nl ^^HlMk' ^^■fe^-*^''

^^E

^^^^^H^^^^^K^ sH^^HH^^&'t^ ^^^^^H^^^^^Ib^ ^

A^^

^S^^

^^^p^

1 jn& ^^^^ ^iHk ^

^Mk ^K, t, $

HARVARD UNIVERSITY.

LIBRARY

MUSEUM OF COMPARATIVE ZOOLOGY.

^IM

Tchojm

UL.

tu/vuey

L,m% — ';Ui7ULooa^xf-i'7/1^03

)

PROCEEDINGS

OF THE

AMERICAN PHILOSOPHICAL- SOCIETY

/■l''M;lli(;f ,1,!/,',

HELD AT PHILADELPHIA

FOR

PROMOTING USEFUL KNOWLEDGE.

VOL. XLI.

JANUARY TO DECEMBER.
1902.

^ PHILADELPHIA :
THE AMERICAN PHILOSOPHICAL SOCIETY.
1902.

i^'

IV

PROCEEDINGS

OF THE

AMERICAN PHILOSOPHICAL SOCIETY

HELD AT PHILADEIPHIJ lOR PROMOTISG USEFUL KNOWLEDGE.

Vol. XLI. Januahy— April, 1902. Ko. 168.

CONTENTS.

PAGB

Stated Meeting, January 3 3

Results obtained from a Search for the Type of Noctua LinD., and
Conclusions as to Types of Huebnerian noctuid Genera repre-
sented in the Korth American Fauna. By A. Radcltfpe

Grote 4

A Modern Delaware Tale. By J. Dyneley Prince 30

Stated Meeting, January 17 34

Stated Meeting, February 7 35

Stated Meeting, February 21 35

Stated Meeting, March 7 36

Stated Meeting, March 21 36

General Meeting, April 3, 4 and 5. . .'. 36

Tho Embryology of a Brachiopod, ^erebratulina septentrionalis

Couthouy (with plates)P^y Edwin G. Conklin 41

The Spermatogenesis of Oniscus asellus Linn., with Especial Refer-
ence to the History of the Chromatin (with plates) ^L!;(By M.
Louise Nichols TT

philadelphia i

The American Philosophical Society,

104 South Fifth Street.

1902.

It is requested that all correspondence be addressed

To THE Secretaries of the

AMEEICA.N PHILOSOPHICAL SOCIETY,
104 South Fifth Street,

Philadelphia, U. S. A.

Members will please communicate to the Secretaries any
inaccuracy in name or address as given on the wrapper of this
number.

It is requested that the receipt of this number of the
Proceedings be acknowledged to the Secretaries.

Members who have not as yet sent their photographs to the
Society will confer a favor by so doing ; cabinet size preferred.

]UN 6 1902

PROCEEDINGS

^.x^ERICAN PHILOSOPHICAL SOCIETY

HELD AT PniLADELPIII\ FOR PROMOTING USEFUL KNOWLEDGE.

Vol. XLI. January, 1902. No. 168.

Stated Meeting^ January 3^ 1902.

Curator Lyman in the Chair.

Present, 6 members.

Prof. Dana C. Munro, a newly elected member, was pre-
sented to the Chair, and took his seat in the Society.

The list of donations to the Library was laid on the table
and thanks Avere ordered for them.

The decease of Mr. Clarence King, at Phoenix, Ariz., on
December 24, 1901, aged 60 years, was announced.

Prof. A. Kadcliffe Grrote presented a paper entitled " Re-
sults Obtained from a Search for the Type of Noctua Linn.,
and Conclusions as to Types of Hubnerian Noctuid Genera
Represented in the North American Fauna."

Prof. J. Dyneley Prince presented a paper entitled " A
Modem Delaware Tale."

Messrs. Joseph Willcox, Joseph C. Fraley and Patterson
Du Bois, the Judges of the annual election for Officers and
Councillors, reported that the same had been held on this day,
between the hours of 2 and 5 in the afternoon, there being
present sixty duly qualified members, and that the following-
named persons were elected, according to the laws, regulations
and ordinances of the Society, to be the officers for the ensu-
ing year :

•i GROTE—SEARCPI FOR THE TYPE OF NOCTUA LINN". [Jan. 3,

President.

Isaac J. Wistar.

Vice-Presidents.
Coleman Sellers, Isaac J. Wistar, George F, Barker.

Secretaries.

I. Minis Hays, EdwinJ^G. Conldin, Arthur W, Goodspeed,

Morris Jastrow, Jr

Treasurer.

Horace Jayne.

Curators.

Charles L. Doolittle, William P. Wilson, William B. Scott.

Councillors to serve for three years.

George F. Edmunds, James T. Mitchell, Albert H. Smyth,
Jcseph Wharton.

RESULTS OBTAINED FROM A SEARCH FOR THE TYPE
OF NOCTUA LINN., AND CONCLUSIONS AS TO TYPES
OF HUEBNERIAN NOCTUID GENERA REPRESENTED
IN THE NORTH AMERICAN FAUNA.

BY A. RADCLIFFE GROTE, A.M.

{Read January 3, 1902.)

In view of the preparation of a general Catalogue of North
American Lepidoptera, I have been asked to give the types of
Hiibnerian Noctuid genera. It is essential that systematists state
the type of the generic title they use, and their work will be lasting
in proportion as its literary basis has been proved. The scientific
edifice will stand when the bricks are sound. A catalogue which
employs the true, historically ascertained generic types has the
advantage of possessing a permanent framework, even if later on
the position of the objects designated be altered. And by using
correct names a great advantage is secured to collectors and to lit-
erature. In my studies of the North American Noctuids for the

, 1902.] GROTE — SEARCH FOR THE TYPE OF NOCTUA LINN. 5

past forty years, I have had occasion to investigate the subject.
The results, as to the types of our genera, are given by me in 1874,
in the Bull. Buff. Soc. N. Sciences, and in the two following years in
the Buffalo Check List ; in 1895 ""^ the Abh. Naturw. Verein, Bre-
men, also in the pages of the Entotnologisf s Record, London,
England, Vol. vi, 27 et seq.; in 1900 in the Can. Entomologist,
209 ; also in publications of the Reenter Museum and in these
Proceedings.

In the present paper I have brought together the historical evi-
dence as to the types of certain leading generic titles, often, per-
haps commonly, used in a perverted sense, or given with a wrong
authority. I have also investigated the question of the use of Noc-
tua as a generic title in the Lepidoptera. I could not have attempted
this latter without the kind aid of Mr. Jno. Hartley Durrant, of
Thetford, England. The type here ascertained is pronuba. The
name Noctua is first used by Klein in 1753 ^"^^ ^ genus of MoUusca.
Linne introduced it then, in 1758, into the Lepidoptera in his com-
bined term Phalaena Noctua. Fabricius follows with Noctua as a
generic term in 1775, 177^-77, and claims the authorship. For
those who reject any limitation for the application of the law of
priority, its use in 1753 will prevent its being later employed in a
different group of animals. It was not used in the Birds until
1809 by Savigny, a fact to which Boisduval drew attention in
1829.

In my late List (1895) of the North American Noctuids, I gave
the ascertained types ; what very few corrections have been found
necessary are here made. The concluding portion of this List, em-
bracing the Catocalinae and Hypeninse, is not yet published. The
unemployed terms in the Verzeichniss of Hiibner need not be con-
sidered in the American Catalogue. They may be neglected until
such time when the faunae of Europe and America be so minutely
compared, that subjective opinion can seize upon the smallest char-
acter for generic differentiation. As a rule, Hiibner's genera in the
Verzeichniss are of mixed contents, and I believe all having present
application have been noticed by me.

In conclusion, I must thank Mr. Louis B. Prout, of London,
England, and Mr. J. D. Alfken, of Bremen, for bibliographical
assistance.

6 GROTE — SEARCH FOR THE TYPE OF XOCTUA LTNN. [Jan. 3,

NOCTUA.

LiNNE, SysL Naturcc, ed. x, Holmiae (Salvii), 1758, Phalaena
Noctua.

The '^Phalaense " (496 footnote) are divided into seven groups,
of which the *'Noctu8e" — antennis setaceis, nee pectinatis — form
the second. Linne gives the foot-structure of the larva of his
** Phalaena Noctua" (497 footnote), so it seems reasonable, in a
selection of the type, that this should be sought among the species
whose larvae he described. These are : Phalaena Noctua strix, fagi,
bucephala, humuli, dominula, fuliginosa, iacobaese, quadra (this
would be, however, excluded by Linne's nota bene), pacta, pro-
nuba, gamma (not a "possible type" from Linne's remark — Dur-
rant /. /.), festucae, meticulosa, psi, chi, aceris, umbratica, exsoleta,
verbasci, brassicae, rumicis, oxyacanthae, oleracea, pisi, atriplicis,
praecox, triplasia, pyramidea, typica, delphinii, citrago.

If we date the commencement of our nomenclature from Linne's
tenth edition, the type of '^ Phalaena Noctua" should then be one
of these. Geoffroy makes no use whatever of Phalaena Noctua
or of Noctua, simply using Phalaena with unnamed subdivisions
(Durrant /. /.). The earliest restriction of the species of Phalaena
Noctua brought to my notice is: Poda, Ins. Mus. Grcec, 88-91,
1761. The species there cited from Linne are : Noctua iacobaeae,
quadra (not a "possible type," vide ante), dominula, pacta (Poda,
90: this is not Linne's species, but is nupta Linne, therefore the
name has no effect), pronuba, gamma (not a "possible type "), ex-
clamationis (excluded, since Linne did not describe the larva),
? secalis.

Of these species iacobaeae is made the type of Hipocrita Hiibn.j
1806, dominula of Callimorpha Latr., 1810, and there would re-
main pronuba as the type of Noctua ; exclamationis being conge-
neric with segetum, taken as type of Agrotis Hiibn., 1806, and
secalis being cited with a query. This latter is the same as didyma
Esp., made the type of Apamea Ochs., 1816, through Duponchel,
1829. Before following the subsequent fate of pronuba, we will
examine Linne's own restriction of his term Phalaena Noctua,
which has given rise to the idea that the type of Noctua falls within
the limits of Schrank's genus Catocala, the type of which I have
shown to be fraxini, through Hiibner's restriction in the Ver-
zeichniss. This type covers our modern use of Catocala Schrank,
1802, which should in no case be disturbed. '

1902.] GROTE — SEARCH FOR THE TYPE OF NOCTUA LINN. 7

LiNNE, Mus. Ludov. Ulr. RegincB, Holmi^e, 1764.

In this work Linne gives the following species : Phalaena Noctua
strix, punctigerata, fulvia, ornatrix, heliconia, rubricollis (removed
now to Bombyx, so that this species is excluded), fraxini, pellex.
It is probable, from this restriction, the idea has arisen (communi-
cated to me in letters) that fraxini was the type of Noctua, because
rubricollis and fraxini are the only two of these species included by
Linne in the Fauna Svecica, 1761, as Mr. Durrant writes me.
Linne now, in 1764, excludes rubricollis, thus restricting the type
to fraxini. But, since fraxini was not included by Poda in 1761,
'^ this can be at once disregarded as of no effect."

Crotch, Cist. Ent., i, 61, 1872, writes:

Noctua — N. sponsa Lamark (1801). Cuvier andLatreille (1805)
concur in this, but afterward Latreiile (1810) selected N. pronuba
as his type. With this selection the writer would be here agreed,
and it remains to be seen what has been since done with pronuba.

TRIPHMNA.
1816. OcHSENH., Schm. Eur., iv, 69.

Interjecta, subs'equa, comes (orbona), prosequa, consequa, lino-
grisea, pronuba, fimbria, ianthina (ianthe, domiduca).

1816. HuEBNER, Verzeichniss, 221.

Interjecta, subsequa, comes, consequa, pronuba.

1829. DuPONCHEL, Hist, Nat. Lep. Noct., Tom. iv, Pt. 2, 71.

Gives pronuba as the type of Triphaena. Therefore Noctua
Linn., in the Lepidoptera, and Triphaena Ochs. would be synony-
mous, having same type. Mr. Meyrick (1895) ^^es Triphaena to
the exclusion of this type. And this opens up the question as
to the validity of the genus, which the type-seeker is not called
upon to answer in the first instance. If pronuba, as being type of
Noctua, could not be taken as type of Triphaena, then Mr. Mey-
rick's use of the latter term may be correct. This question does
not seem necessary to answer for the North American Catalogue.

I now follow the use of Noctua by authors subsequent to Linne.

Fabricius, Sy sterna Entomologice, Flensburgi et Lipsiae, 1775.

In this work 122 species are enumerated under Noctua, pp. 590-
619.

b GKOTE — SEARCH FOR THE TYPE OF NOCTUA LINX. [Jan. 3,

FarbiciU3, Genera Insectoruiu. . . . Mantissa specterum, Chilonii.

There is no date on title-page, but the Preface is dated Kiliae,
Dec. 26, 1776. This work is not given by Staudinger and Rebel,
p. xviii, but is cited for viminalis with the date 1777. Fabricius
quotes it, in 1 781, as '^ Gen. Ins. Mant." It contains only six spe-
cies under Noctua, but these are all new and constitute no restric-
tion of those given previously. They are as follows :

(i) p. 282, Noctua bokti. This is Scardia boleti, a Tineid.

(2) p. 282, Noctua virescens. This appears to be the earliest
description of the North American Noctuid Chloridea virescens
Westw. ex Fab. and is neglected in the Washington Catalogue,
1893.

(3) p. 283, Noctua roboris. I cannot find this citation in Stand-
inger and Rebel. Reference is made to Roesel, I, tab. 50, and the
insect there depicted maybe Dryobota roboris B., Cat. I, No. 1821.

(4) p. 283, Noctua monilis. This appears to be the earliest de-
scription of the North American Noctuid Hypsoropha monilis
Hiibn. ex Fab., with a wrong locality, *' Anglia."

(5) p. 283, Noctua lanceolata. The habitat is given as Germany.
I cannot find the citation in Staudinger and Rebel.

(6) p. 284, Noctua viminalis. This is Cleoceris viminalis, re-
ferred incorrectly in the Catalogue, No. 1560, to Bombycia. The
type of Bombycia Hiibn., 1806, is B. or.

Fabricius, Species Insectoruui^ Hamburgi et Kilonii, II, 1781.

In this work 150 species are enumerated under Noctua, pp. 209-
241. The six of the Gen. Ins. Mant. are included.

Fabricius, Mantissa Insectorum, Hafnise, II, 1787.

In this work 309 species are enumerated under Noctua, pp. 135-
184, and those previously described appear to be all carried for-
ward.

In his Genera Insectorum, 1776, Fabricius cites " Phala^na Linn.
Geoff." as equivalent to his genus Noctua, of which he evidently
considers himself the author. Fabricius restricts Phalaena (p. 164,
/. c.') to the Geometrids, using the term in a generic sense and
citing Linn. Geoff, as authority. Following his own precedent he
should here have applied Linne's term Geometra. Linne's '*Pha-
laense," 1758, is evidently employed in a comprehensive sense, em-
bracing all the seven groups : Bombyces, Noctuoe, etc. I have

1902.] GROTE — SEARCH FOR THE TYPE OF NOCTUA LINN. 9

made no search after the type of Noctua, Fabricius. It is evident
he took the name from Linne, whether he credit it to him or not.

OcHSENHEiMER, Schmetterlinge Europa' s. Vol. iv, 1816.

Ochsenheimer has no genus Noctua ; pronuba is included by him
in his genus Triphaena, with other yellow-winged Agrotids, differ-
ing in structure. On page viii, Ochsenheimer cites by its full title
the Tentamen of Hiibner, and says, literally : dieses Blatt kam
mir erst lange nach dem Abdrucke des dritten Bandes zu Gesichte,
daher konnte ich friiher nichts davon aufnehmen. Already in 1876
I have shown that Hagen misquoted Ochsenheimer {vide Buffalo
Check List and Can. Enf.), who in reality borrowed generic names
and ideas from Hiibner's Tentamen and properly gives him credit.
Later writers, who are here so greatly indebted to their predeces-
sors, could profitably take example.

Ochsenheimer's groupings of the Noctuids must be considered
as expressing his idea of their affinities, because on page ix he says
that he only catalogues and describes what he could compare in
nature, not relying upon descriptions or figures, and that his syste-
matic list is at the same time the catalogue of his collection. He
gives no descriptions of his genera, any more than Hiibner in the
Tentamen.

BoiSDUVAL, EtcropcBorum Lepidopterorum Index Methodicus .

Dated on title-page 1829, but the Preface is dated Sept. 30,

1828. The work has priority over Duponchel's volume, March,

1829, or Curtis, May, 1829. '* Noctua mihi," p. 6^, contains
names of some 70 species; Boisduval cites ^'Agrotis et Noctua
Treits." and *' Agrotis et Graphiphora Ochs." as synonymous.
The type of Agrotis Hiibn., 1806, segetum, is included. '' Tri-
phcena Ochs. Treitsch.," p. (iZ, contains 7 species, among them
pronuba, designated by Duponchel as type.

After Fabricius, the responsibility for the use of Noctua mainly
rests with Boisduval. I cannot find that Hiibner ever used the
term in a generic sense.

Boisduval, Genera et Index Methodicus.

Dated on title-page and in Preface 1840.

"• TriphcBna Treits. Boisd." contains 8 species, among them
Duponchel's type.

10 GROTE — SEARCH FOR THE TYPE OF NOOTUA LINN. [Jan. 3,

Opigena Boisd., 1840, monotypic for polygona.

Chersotis Boisd., 1840, with 8 species.

'^ Noctua Treits.," sagittifera and 18 other species.

Spcelotis Boisd., 1840, for augur and 22 other species.

'^ Agrotis Ochs. Tr.," agricola and 36 other species, including
exclamaiionis, designated by Duponchel in 1829 as the type of
Noctua, but erroneously so, since this is taken by Agrotis, 1S06,
being congeneric with segetum. It is also excluded by Durrant as
being unknown in the larval stage to Linne.

Speyer, in the second edition of Dr. Schenckel's Schmetterlings-
sammler.

Undated, Mainz, C. G. Kunze. Has a genus '* Tryphsena," as
used by Ochsenheimer and Boisduval, and employs Opigena for
polygona. In late editions, undated, of his popular book, '' Schmet-
terlingskunde," Speyer continues to use Tryphcena (Triphaena) in
Hiibner's sense, and includes pronuba in its second section. These
authors, therefore, regarded Triph(Ena as a distinct genus from
Agrotis. Since I have not found in the N. Am. Noctuid fauna the
precise structural equivalent of pronuba, it may not be necessary
for the American Catalogue to use either Trtphcsna or Noctua.
Agrotis gilvipennis Grote, referred by me in 1890 to Triphcena,
belongs, I believe, having no specimen at present, to Lampra. It
remains for the systematist to decide what species, other than pro-
nuba, can be taken as type of Triphcena. Duponchel's type, pro-
nuba, can remain, if my view that Noctua is untenable obtains.

Lederer, Noctuiden Eiiropa^ s, Wien, 1857.

Lederer has no genus Noctua, the species here regarded as typi-
cal being referred to one of the sections of Agrotis. Lederer
divides the numerous species of Agrotis primarily upon secondary
sexual characters, the male genitalia. Already, in 1874, I had pro-
posed to divide the species into two chief groups — those species
which had all the tibiae spinose and those in which the middle and
hind tibiae alone are armed (^BulL Buf. S. N. S..,\\). Subsequently,
in the Canadian Entomologist, I proposed a further addition, in-
cluding the genus Carneades. This classification of mine gives
"three principal divisions for the North American species :

Front smooth, fore tibiae unarmed: Epllectra, Lampra Hiibn.
Front smooth, fore tibiae armed : Triphcena C, Agrotis Wxsl^xx.
Front tuberculate, fore tibiae armed : Carneades Grote.

1902.] GROTE — SEARCH FOR THE TYPE OF NOCTUA LIJ^X. 11

Lederer makes, I believe, some structural misstatements. He
gives the male antenn?e of linogrisea as ''pyramidal zahnig."
This species is the type and sole species of Epilectra. Its diagno-
sis should read : Thoracic vestiture scaly ; male antennae simply
brush-like, nearly naked ; fore tibiae unarmed ; front smooth. The
eyes, as in all these structures, naked. Lederer further gives agathina
as having the fore tibiae armed and triangulum unarmed, whereas
the reverse appears to be the case. In depuncta the thoracic vesti-
ture seems scaly, whereas Lederer places it in a section where this
is hairy. Neither Epilectra or Triphcena (Noctua L.) need ap-
parently affect the American Catalogue. The species referred
in the " Revision" to Noctua belong to Amathes. Lederer's
neglect of Hiibner and his uncritical use of several generic names
has increased the confusion, which is the more to be regretted since
his structural observations are usually so valuable.

To sum up : There seems no use in disturbing Duponchel's type,
pronuba, for Triph?ena, until it is settled whether the term Noctua
Linne can be employed. I conclude that the historically indicated
type of Fhalcsna Noctua Linne is pronuba, and that the term
Noctua cannot be used in the Lepidoptera because preoccupied by
Klein in the Mollusca in 1753. The earliest plural form I find,
which could be used, outside of Noctuae, for the family is Apatelae
Hiibner, 1806, and the family type would be Apatela aceris. The
name Agrotidae, H.-S., based on Agrotes Hiibn., 1806, which
latter occurs on the same page, is a more appropriate title for the
whole group in Lederer's sense. Lederer himself gives no scientific
title to the group. In the present case, if we exclude the term
Noctua, there can be no doubt that the leading genera of the group
are : Apatela, Agrotis, Hadena, Cucullia, Plusia and Catocala.
Three of these belong to Schrank, 1802, and three to Hiibner,
1806. Hiibner's names have the preference for a family title,
because he employs also the plural form, with the evident intention
of using them for comprehensive groups, an intention he carries
out ten years later, in 18 16, in the Verzeichniss.

Taking the opposite conclusion, that Noctua Linn, is a valid
generic title, its type htmg pronuba, then the question comes up : Is
profiuba congeneric with Agrotis segetuin ? If so, then Agrotis falls
before Noctua Linn. Meigen (1832) includes 155 species under
Noctua, with Hadena, Orthosia, etc., as subgenera. His subgenus
Noctua contains baja, candelisequa, brunnea, festiva, rhombsidea,

12 GROTE — SEARCH FOR THE TYPE OF NOCTUA LINX. [Jan. 3,

gothica (!), C. nigrum, triangulum, flammatra, musiva, plecta,
punicea. He remarks : der Rlicken hat einen Schopf. In the main
this seems to be the group intended by Prof. J. B. Smith as Noctua,
but it cannot include either pronuba or segetum. Meigen places
the latter correctly under the subgenus Agrotis, but classifies
pronuba under the distinct genus " Tryphaena " section A, which
he characterizes as having the third palpal joint reduced, hardly
noticeable. It does not seem as though subjective opinion would
ever rest content with the reference oi pronuba as congeneric with
segetum, and therefore the question of the genus Noctua need not
affect the North American Catalogue,

At the present time the study of the Noctuids in America is
suffering under the evil duplication of specific names and a reckless
disregard of the historically indicated types of the generic titles.
In this connection may I ask how Noctua comes to be applied to the
group in Prof. Smith's Revision, except by a kind of restriction?
For Linne's original Phalaena Noctua contains insects belonging to
several distinct families and only by some sort of literary precedent
has it come to designate Owlet Moths or Noctuids. The same sort
of historical research, only carried out with more exactness, reveals
the types I must insist upon for certain genera. And, unless it can
be shown, in any special instance, that I have erred (the study has
often proved intricate), it will be clearly to the advantage of science
that my results be adopted in the new N. Am. Catalogue. I now
give here references I have made and the types which they reveal :

MAMESTRA.
1816. OcHSENH., Schm. Em\, iv, 76.

Fisif splendens, oleracea, suasa, aliena, abjecta, chenopodii, albi-
colon, brassicae, furva, persicarise.

1816. HuEBNER, Verz.y 214.

Pisi, unaminis, leucophsea. Under this restriction pisi became
type, since Hlibner's two other species are not included originally.

(March) 1829. Duponchel, Hist. Nat. Lep. Noct.y T. iv, Pt. 2, 71.

Designates brassicae as type, but this restriction of Mamestra is
no longer possible since Hiibner's action in the Verzeichniss.
Hlibner must have taken this generic name from Ochsenheimer,

1902.] GROTE— SEARCH FOR THE TYPE OF NOCCUA LINN. 13

i8i6j hence this part of the Verzeichniss must be of later issue,
probably 1822, but at any rate earlier than Duponchel.

1874. Grote, Bull. Buff. S. N. Sci., 12.

Lists the N. Am. species and takes //>/ as type. This accords in
a general way with the modern definition of Mamestra: Hadenoid
forms with hairy eyes, the non-extruded ovipositor and different
larval habit separating them from Hadena (type cucubali) Schrank
non Lederer (= Dianthoecia Boisd.). I list the North American
species of Dianthoecia, for which name Hadena Schrank must now
be substituted, and give the characters in Rev. Check List, N. Am.
Noct., 1890, 13 (Bremen, Homeyer & Meyer).

HADENA.

1802. Schrank, Fauna Boica, II, 2, 158.

Refers to this genus the species of his families M. and N. These
species are: typica, atriplicis, pisi, oleracea, chenopodii, praecox,
xanthographa, piniperda, deaurata, referred to family M, and meti-
culosa, lucipara, cucubali, referred to family N. One of these
twelve Noctuids must then be the type of the name of Hadena.
According to modern views species i, 2, 8, to and ii are mono-
typic, 3-5 are Mamestrians, 6-7 Agrotids. The contents are much
mixed, referable to nine genera.

to'

1816. OcHSENH,, Schm. Eur., iv, 70.

Excludes all the species of Schrank's family M, but includes all
of N, among his 29 species of Hadena. The mixture is now more
frightful than it was at first. The three original species of Hadena —
meticulosa, lucipara and cucubuli — are, however, included, and one
of these three must now be the type. It is noticeable, however,
although species with hairy and naked eyes are indifferently cited,
that all the species of Dianthoecia are included by Ochsenheimer.

1816. HuEBNER, Verzeichniss, 216.

This part of the Verzeichniss is of later date than Ochsenheimer's
volume. Hiibner includes under his genus Hadena only two of
Schrank's original species, typica and cucubali. The first is ex-
cluded by Ochsenheimer's first restriction in 1816, and moreover
became the type of Naenia Stephens in 1829. Cucubali becomes,
therefore, the type of the genus Hadena, and is to be looked upon

14 GROTE — SEARCH FOR THE TYPE OF NOCTUA LINN. [Jan. a,

as the original "Triibeule." It is unnecessary, having found the
type, to follow the fortunes of Hadena further. It was used im-
properly by Lederer for a large genus of naked-eyed species separ-
able from Mamestra on this character.

1895. Grote, Ent. Record^ vi, 78.

Designates cucubali as type of Hadena, and states that Dian-
thoecia Boisduval, will probably prove synonymous.

XYLENA.
1806. HuEBNER, Tent., i.

Lythoxylea (lithoxylea) sole species and therefore type.

1 81 6. OcHSENH., Schm. Eur., iv, 85.

Vetusta, exoleta, conformis, lapidea, rizolitha, petrificata, con-
spicillaris, patris, spinifera, scolopacina, rurea, hepatica, polyodon,
lateritia, lithoxylea, petroriza, pulla, cassinea, nubeculosa, pinastri
(scabriuscula), rectilinea, ramosa, lithoriza, hyperici, perspicillaris,
platyptera, antyrrhini, linari^, opalina, delphinii. Ochsenheimer
quotes Hiibner and spells the genus as he does, Xylena. This is
the worst of Ochsenheimer's mixtures and, while enlarging Hiib-
ner's genus, the beginning of all subsequent confusion in applying
this generic title. This abuse is still being perpetuated, although I
gave again the type in 1876. Later writers than Ochsenheimer
take out the Lithophanoid forms (Fam. A in part, petrificata, etc.),
and use for them a genus " Xylina Ochs. or Tr.," whereas Ochsen-
heimer has no generic term so spelled. They then reject the Hade-
noid forms (Fam. B in part), which include Hlibner's type lithoxy,
lea, instead of the reverse. Hiibner himself, in the Verzeichniss-
refers lithoxylea to the same group as petrificata, and the truth
seems to be that, perhaps up to Stephens, the generic types I now
give to Xylena and Lithophane were thought congeneric or nearly
allied. The genus Xylophasia Stephens is a synonym of Xylena
Hubner, having the same type.

(1828) 1829. BoiSD., Eur, Lep. Ind. Afeth., ^6.

Cites ''Xylina Tr. and Xylena Ochs.," and suppresses Ochsen-
heimer's reference to Hiibner for the term.

(March) 1829. Duponchel, Hist. Nat. Lep., iv, Pt. 2, 72.
Gives vetusta as type, but this is impossible.

1902.] GROTE — SEARCH FOR THE TYPE OF NOCTUA LINN. 15

1876. Grote, Buff. Check List Noct., 37.

Restores Hiibner's type and spelling, and gives Hadena (Lederer
nee Schrank) as identical. The type of Schrank's genus was not
then ascertained.

I show, in 1874, that the modern genus ''Xylina" must be called
Lithophane Hiibn., 1816, with the type socia (petrificata) — a far
more appropriate name.

The American species referred to Hadena, Lederer nee Schrank,
should be catalogued under the following genera : Xylena Hiibn.
(=Xylophasia Steph.), type lithoxylea; Helioscota Grote, type
miselioides; Oligia Hiibn. (nee Grote, Smith), type strigilis;
Pseudanarta Grote, type flava (crocea) ; Monodes Guen. (rzz Oli-
gia Auct. nee Hiibn.), type nucicolor (paginata). A very good
notice of the species of Monodes will be found in E?itom. Am.,
Vol. V, p. 145, under the name Oligia. It may be said of all
these genera, what is there said of Monodes, that they are not
"strongly characterized." They have in common naked eyes, un-
armed tibiae, smooth clypeus and hadeniform cut of wing. Xylena
may have a strong character in the thoracic shield of the larva.
The species belonging to these genera vary from being robust, hairy
and tufted down to slighter, scaly and smoother forms. To Xylena
belong species like lignicolor, auranticolor, genialis, cristata, vul-
garis, verbascoides, cuculliiformis, hulsti, vultuosa, sputatrix (I do
not acknowledge this to be Walker's dubitans),- devastatrix,
occidens, arctica, violacea, Bridghami, apamiformis, lateritia, suf-
fusea, remissa ; to Helioscota : miselioides, marina, chlorostigma,
mactata, modica, diversicolor. From want of space and material
I do not carry these references further here.

APAMEA.

I proposed at one time to take Ochsenheimer's nictitans as
type of Apamea, it is his first species ; this nictitans is not the Gor-
tyna nictitans L. of Lederer, but is nictitans Esp., a variety of
secalis L.= didyma Esp.== oculea Guen. {Cat. Stand. a?td Rebel, p.
175). My reference was correct, for this species had become type
of Apamea through Duponchel in 1829. The similarity of the
name led me, however, to mistake Ochsenheimer's species for nicti-
tans Bkh. (given by Lederer as of Linne) = chrysographa Hiibn.
{Cat. Stand, and Rebel, p. 186), which latter is the type of Hydrce-
cia Guen., as shown by me in these pages and elsewhere. It is

16 GROTE — SEARCH FOR THE TYPE OF NOCTUA LINN. [Jan. 3,

probable we have N. Am. species congeneric with didyma (secalis
L.), but at this writing I cannot indicate them. Lederer's restric-
tion of Apamea to testacea, which I followed in 1895, should not
be accepted ; this is the true type of Luperina Boisd. (see Grote,
Ca7i. Ent., 1900, 211). Boisduval, in 1829, refers both nictitans
(chrysographa) and nictitans (didyma) to Apamea.

PSEUDANARTA.
187S. Grote, Bull. U. S. Geol. Surv,, 178.
Crocea (flava), sole species given and therefore type.

1882. Grote, New Check List, New York, 27.

Flava, var. crocea, singula, flavidens, aurea. The name, without
citation, is credited to Hy. Edwards, under the mistaken idea,
derived from correspondence, this author had used it. Pseudanarta
was originally proposed by Grote in letters to Hy. Edwards for this
author's Anarta crocea.

1889. J. B. Smith, Ent. Afn., v, 175.

Falcata, aurea, flava (crocea), singula, flavidens. The genus is
credited to Hy. Edwards and the citation: ^' Proc. Cal. Ac. Sci.,
Vol. 6, p. 133, 1875," is supplied. But this page contains the
original description of ^;zd!r/^ crocea, and the name Pseudanarta
does not occur in any of the communications of Hy. Edwards to
the California Academy : "Pacific Coast Lepidoptera, Nos. i to
22," all published. This erroneous citation is twice repeated in
the Washington Catalogue, 148, 1893,

1895. Grote, Abh. Naturw. Ver. Bremen, xiv, 37.

Flava, var. crocea, singula, flavidens. The genus is limited to
these three species ; falcata and aurea are excluded, owing to Prof.
J. B. Smith's remark on their tibial structure in 1893.

COPANARTA.

1895. Grote, Abh. Naturw. Ver. Bremen, xiv, 70.

Aurea, falcata, aterrima ; aurea specified as type.

PLUSIA.
1806. HUEBNER, Te?it., 2.

Chrysitis, sole species and therefore type. This name is errone-
ously given to Ochsenheimer, who however cites Hiibner's Tenta-

1902.] G ROTE — SEARCH FOR THE TYPE OF NOCTUA LINN. 17

men and includes his type. Lederer in 1857 cites Plusia Fabr.,
but I can find no such genus in Fabricius and the name should be
restored to Hiibner. Chrysoptera Latr., 1825, is said to be preoc.
cupied. It is used by Meigen in 1832 for concha, deaurata and
moneta alone.

The names and types of the subgenera of Plusia are given by me
in these Proceedings, 417 (1895). Typical N. Am. species of
Plusia are : derea, dereoides, balluca, metallica (lenzi, scapularis).

GRAPHIPHORA.
1806. HuEBNER, Tent., i.
Gothica, sole species and therefore type.

181 6. OcHSENH., Schm. Eur., iv, dZ.

Ravida and sixteen other species belonging to Agrotis in sensu
Lederer, excluding Hiibner's type, though taking the name from
Tentamen. The confusion now commences in European literature.
The genus is used for Agrotidians, with which gothica was origin-
ally held as allied, until the type is made also the type of Taenio-
campa, Guenee, which must fall.

1 81 6. HuEBNER, Vei'zeichniss, 220.

Has no genus, but a Stirps Graphiphorse, which comprises
numerous genera, mostly of Agrotidians, among them Epi-
sema, which he takes from Ochsenheimer, including gothica.
No -examination had been made then of the structure of
the eyes and legs ; pattern and size seemed at that time
to warrant the juxtaposition of Taeniocampids and Agrotidians
(still difficult to separate, e.g.^ Pachnobia and Metalepsis). But
the original sense of Graphiphora must be restored. Boisduval, in
1829, refers "Graphiphora Ochs." as a synonym of Noctua and
Agrotis, and' includes its type gothica (/. c, 67) as structurally
identical. This proves the accuracy of the statement given above
as to the views prevalent at the beginning of the last century.

1875-76. Grote, Buffalo Check List, 13, 37.

Gives the North American species, referred to Taeniocampa, to
Graphiphora, and designates gothica as type. Repeats this in 1895,
Entom. Record, 29, and last Check List, Abh. Brem. Nat. Ver.,
xiv, and now ''finally" insists.

PROC. AMER. PHILOS. SOC. XLI. 168. B. PRINTED MARCH 18, 1902.

18 GROTE — SEAHCH FOR THE TYPE OF XOCTUA LIXN. [Jan. 3,

XANTHI.\.

1806. HuEBNER, Tentamen, 1.

Fulvago (puleacea), sole species and therefore type.

181 6. OCHSENHEIMER, Schifi. Eur.^ iv, 82.

Luteago and sixteen other species. Cites Hiibner, but includes
his type under Cosmia.. The similar endings of the names of
the yellow autumnal species, in ago, may have helped to
increase the confusion in their application which prevails in early
European literature. Hiibner's erroneous use of '^fulvago" may
have led to his generic title being misapplied. Species of Citria
and Orthosia are constantly referred in America to Xanthia, which
term should be kept in the North American Catalogue for paleacea
alone, specimens of which I described under the name of infumata,
not knowing the European species, now believed to be identical
with our own. Enargia Hlibn. Verz. has paleacea also for^type
and falls before Xanthia.

COSMIA.
1806. HuEBNER, Tentamen, i.
Affinis, sole species and therefore type.

18 [6. OcHSENH., Schm. Eur., iv, 84.

Fulvago (W. V. Hiibner = paleacea), gilvago, abluta, trapezina,
diffinis, affinis and pyralina. Cites Hiibner's Tentamen and in-
cludes his type of Cosmia. The genus should be restored to Hiib-
ner, but has no place in our American Catalogues. Ochsenheimer
corrects Hiibner's erroneous application of '' fulvago."

AMPHIPYRA.

1 8 16. OcHSENH., Schm. Eur., 70.

Tragopoginis, tetra, livida, cinnamomea, pyramidea, perflua,
spectrum.

1829. BoiSD., Eur. Lep. Index Meth , 6Z.

Uses it for the same species. The first six species belonged since
1 806 to Pyrophyla (r. Pyrophila), and the type of Amphipyra is
spectrum. The genus is not represented in America. Our species
belong to Pyrophyla Hubn., 1806, type pyramidea.

1902.] GROTE— SEARCH FOR THE TYPE OF NOCTUA LINX. 19

ACONTIA.

I Si 6. OCHSENH., Schm. Eur., iv, 91.
Malvae, aprica, caloris, titania, Solaris, luctuosa.

1816. HuEBNER, Verzetchniss, 257.
Malvae, sole species and henceforth the type.

1895. Grote, Entom. Record^ 79.

Designates malvse as type through Hiibner's restriction. This
part of Hubner's Verzetchniss is of later issue than Ochsenheimer's
volume, from which Hubner takes such genera as Acromcfa, Ma-
mestra, Triphcena, etc. The genus Acontia should not be used by
the American Catalogue, as it is confined to Europe. Our species
belong to Tarache.

TAR AC HE.

1 81 6. HuEBNER, Verzetchniss, 261.

Caloris (caffraria), Solaris, insolatrix (ined.), aprica, opalina.

1874. Grote, Bull. B. S. N. S., s^.
Designates aprica as type.

ERASTRIA.

1806. HuEBNER, Tenlamen, 2.

Amataria, sole species and therefore type. This is a genus of
Geometrids and the name is erroneously applied by Ochsenheimer
to a genus of Noctuids. Its use should be avoided by every careful
and unprejudiced person in the Noctuids for this very good reason.

, • EUSTROTIA.

18 r 6. HuEBNER, Verzeichniss, 253.

Unca, sole species and therefore type. The North American
Noctuids referred to Erastria belong to this genus, which is used in
the Catalogue of 1874, Bull. Buff. S. N. S., 37, and subsequently.
The change back to Erastria in the Washington Catalogue is inex-
cusable.

EUCLIDIA.

1806. HuEBNER, Tentamen, 2.
Glyphica, sole species and therefore type.

20 PRINCE — A MODERN DELAWARE TALE. [Jan. 3,

l8l6. OCHSENHEIMER, ScJim. Eur., iv^ 96.

Monogramma, glyphica, triqiietra, mi. Cites Hiibner's Tenta-
men for name and includes his type. Ochsenheimer gives no gen-
eric description, and yet he is constantly cited as author. Hiibner's
property should be restored to him.

LITOGNATHA.

1873. Grote, Bull. Buff. Soc. N. S., 85.
Nubilifascia, sole species and therefore type.

1895. Grote, Broc. Am. Bhil. Soc, 429.

Nubilifascia, cribrumalis. This generic name is referred in the
Washington Catalogue to Hormisa Walker, but Walker's original
specimen over this label we saw in 1867, and it was a specimen of
Epizeuxis jemula. This determination is supported by the text of
Walker's description of the genus Hormisa, which agrees with Epi-
zeuxis and absolutely contradicts Litognatha. Litognatha should
be restored.

ZANCLOGNATHA.

1857. Lederer, iVi?^/. ^?^r., 211.

Tarsiplumalis, tarsicrinalis and others.
1895. Grote, Broc. Am. Bhil. Soc, 424.

Tarsiplumalis, tarsipennalis and others. Tarsiplumalis may be
taken as type, as stated in Buffalo Bulletin, 1874.

RoEMER Museum, November, 1901.

A MODERN DELAWARE TALE.

BY J. DYNELEY PRINCE, PH.D.

{Read January 3, 1902.)

The chief differences between the two ancient dialects of the
Lenape, viz., the Unami-Unalachtigo and the Minsi, have been
pointed out by the late Dr. Brinton {The Lendpe and their Legends,
pp. 9 iff.). Both these varieties of Delaware speech are still in use
in a modern form — the Unami-Unalachtigo by the descendants of
the Delawares who now occupy lands in Indian Territory, in the

1902.] PRINCE — A MODERN DELAWARE TALE. 21

Muskogee Agency of the Cherokee Nation, and the Minsi by about
three hundred Indians in Ontario, Canada, viz., one hundred at
Munceytown, one hundred at Moraviantown, the seat of a Moravian
mission, and the same number at Hagersville, on the Six Nations'
(Iroquois) Reserve. There are also a few Minsis at New Westfield,
near Ottawa, Kansas, most of whom are under the charge of the
Moravian Church.^

The following witchcraft story in the modern Minsi was sent to
me, with other MS. material, by Mr. Nelles Montour, Chief of the
Minsis at Hagersville, Ont., a well-educated Indian who writes his
own language with great clearness. Like all Indian scribes, how-
ever. Chief Montour writes syllabically, separating the syllables of
his texts and not the words, a process which makes a correct edition
of his MSS. extremely difficult. For example, in the following tale
in II. ^ Montour wrote keer/i keeth gta, as three distinct syllables.
This resolves itself under analysis mto kee?'hkee th'q'ta * by the fire.'
His translation also is in many instances so free as rather to
obscure the true meaning of the original. Thus, in IV. % he
renders chee quack leetahhawa dulwihkawawh ' I am a greater man
than he.' The correct translation is undoubtedly ' Do not think
about it ; I will overcome him.' Then, too, the not always uniform,
cumbrous English system of spelling followed by Montour, in com-
mon with those of his tribe who are members of the Church of Eng-
land, makes an accurate analysis of his texts doubly trying. The
English values of the consonants probably do not reproduce the
Indian sounds with great exactness, as may be seen from Montour's
constant use of the spelling quack 'what,' which clearly should be
written queq (see below on III. ''), as well as from his consistent
omission of the n prefix of the first person before g and before the
intercalary -d-, as in gutauch^ I. " ; diV/ioom, III. ;^, etc. The Mora-
vian Minsis still use the much more appropriate German system of
phonetics.

The analysis of the following tale has been made chiefly by
means of the Old Delaware materials left by the German Moravian
missionaries of the eighteenth century, tabulated in a convenient
form by Dr. Brinton in his Lendpe-EngUsh JDictionary.'^ In cases

1 These details were furnished by Chief Nelles Montour, of Hagersville, Ont.,
and by Mr. Dew M. Wisdom, formerly Indian Agent at Muskogee, I. T.

^ A Lenape- English Dictionary, by Daniel J. Brinton, A.M., M.D., and Rev.
Albert Seqaqkind Anthony, Philadelphia, i888. The material is drawn from a
MS. dictionary preserved in the Moravian archives at Bethlehem, Pa.

22 PRINCE — A MODERN DELAWARE TALE. [Jau. 3,

where the Minsi deviates greatly from the mixed UnamiUnalach"
tigo dialect, in which the missionaries wrote, I have had recourse to
the vocabularies of the cognate Abenaki and Ojibwe languages,'
which have given fairly satisfactory aid in every instance save one
(in V. ^). The chief phonetic variation between Montour's dialect
and the language of the Moravian missionaries is the appearance of
ih (soft, as in 'this') in Minsi as representing s in Unami-Una-
lachtigo ; cp. wsheetha for O. D. ■w' schiessa * his uncle,' the end-
ing -multhoo for O. D. -inallsiuy etc. Brinton asserts {Diet., p. vi)
that this s in O. D. was due to the fact that the Germans were
unable to distinguish the soft th, which they accordingly represented
by s. Thus Anthony, Brinton's native authority, states (Diet., p.
T15) that the common word for ' boy ' in his language is skahenso,
which appears in Montour's text in the form thkuhinthoowh, IV. *,
representing the actual pronunciation. Furthermore, in the letter
from Chief Gottlieb Tobias {Len. Legends, p. 88) we find the form
lichsoagan ' language,' which Montour would write leerhthoowawgun.
In other words, those Indians who read the language according to
the German system lisp the s.

In the following modern Minsi text these important points with
regard to the pronunciation should be noted : i. Medial and final
h is never an aspirate, but merely a pause. 2. The combination ng
is pronounced like ng in 'king.' 3. The combination rh is a
deep guttural gh. Actual r has not existed in Lenape since the
days of the early Swedish colony in Pennsylvania and New Jersey.
It is now represented by /as in modern Abenaki (/= ancient r).
4. W before another consonant is pronounced, as in Passamaquoddy,
with a short unclear vowel following it, similar to the Hebrew
SKva mobile. 5. Wh is a guttural combination composed oi w -^
kh. 6. The apostrophe (') indicates a very short u. 7. The
vowels are to be pronounced exactly as in English.'' The O. D.
words are written entirely according to the German system. The
Abenaki vowels are pronounced as in Italian, except the o, which
has the sound of the French nasal on. The sign ' indicates a soft
guttural voice-stop similar to the Arabic medial He. The vowels

3 The Abenaki material is drawn from a dictionary of tlie modern dialect now
in course of preparation by myself, and the Ojibwe words are taken from Baraga's
Dictionary of the Otchipwe Language, Circinnati, 1853.

♦Cp. Prince, "Notes on the Modern Minsi Delaware Dialect," Atner, Journal
of Philol., xxi, pp. 295-302.

^2.] PRINCE— A MODERN DELAWARE TALE. 23

in the Ojibwe words have the Italian and the consonants the Eng-
lish values.

The subject matter of Montour's tale is interesting, dealing as it
does with cannibalism, a vice which was unknown amorg the Algon-
quin tribes, except in the case of wizards. In this particular
story it should be noticed that the spirit Muttontoe (the Abenaki
Madahodo * Devil ') desires to devour an aged man. This maybe a
survival of the primitive time when it was actually the custom to eat
the old people, apparently in order to get rid of them, as has been
the case until quite recently among the Tierra del Fuego tribes. It
is at least curious that the Muttontoe desires to eat the elderly
rather than the young man, who would be a better subject for mere
cannibalism. It is also very striking that the uncle becomes sick
first and thus incapacitated. This would seem to indicate a
survival of some archaic idea, concealed here under the veil of a
witchcraft superstition, that the old man was the proper prey for the
man-eater. This tale seems to embody a different principle from
that shown in the Passamaquoddy account of two wizards who
retired to an isolated island (Grand Manan) to devour the body
of a man.^ In the latter instance, the cannibalism was of the
ordinary sacramental character, viz., the cannibals hoped to
absorb some of their victim's mental qualities by devouring his
flesh. It is not impossible that the custom of eating grandparents
and other aged incapable persons might have had for its basis a
similar sacramental idea — i. e., that the old people, by entering the
bodies of their descendants, should live again and at the same time
impart to the younger cannibals some of the nature of the aged
victims.

As literature in modern Delaware is so rare, I have given a care-
ful philological analysis of Montour's tale, so far as my imperfect
knowledge of the language has permitted.

A Youth and His Uncle.

WiTHKEELNO WAUK WSHEETHA.

1. * Weekwaum lawee kohpe
weekena withkeelno wauk wshee-
tha mahji kihkweelno wrhalin
neepnumo. ^ Tah lickee wshee-
tha weenamulthoo, oonjeemawuh

5 See Prince, Prog. Amer. Philos. Soc, xxxviii, pp. 182, 184, nr. v.

1. * In a wigwam in the midst
of the forest lived a youth and his
uncle of many summers. ^ Once
upon a time the old man was
taken ill (and) called his nephew

24

PKINCE — A MODERN DEL AWAKE TALE.

[Jan. 3,

wlunquathitha aleh-mawmjeenah
kihkloolaut. " Withkeelno lawa-
lindum, leetahah : ''gutauch
wlutchawha jeeth. ^ Noolihtoo-
mich mihtqueenootee wauk kpu-
heekun waukitch nooshwuhtoo-
nich uhpeeyuhk nahtau aleenaw-
qtheet. ' ' ® Waupungeek andah-
keshihtootah mihtqueenootee
meelaun. Wsheethaha wlalin-
dumoo wekwulup laulpuksho.
' Nulhuh-nuh wtuhlaun wsheetha
ahpeewuyuhpeenang. ^ Waupiin-
geeka weenumultheet ithpeen-
urhka aleet " klithtuh." "" Wti-
lawul withkeelno: ''ah wan itch
pawhji ; cheepeenawqthoo wauk
ahkonjauptoona kweeshulooq-
kich, shuqk chee weeshulooq-
koowih ; muthkuneetahaul ; pa-
woich andah-laweetpihkalik an-
dah-wam-quack-kaweet. ' '

II. ^ Nulnuh peethkahkeek an-
dah-mahji - keeshmeettheeteetah,
withkeelno awuthee tindawing
Imutawpoowh, pahtoon tah nij
alak nih aleetpihkahk. ^ Weerh-
kawa quack konjwah wuhkoong;
ahwan cheepeenawqthoo wcherh-
akahlaun keerhkee th'q'ta :
" " Ugh," owh, ^' baum konjah-
wan nhukee ; nmihwa linno.
Ktuhaulaw ksheeth ; naulaw ;
Ugh, kweeshathee." ** With-
keelno mutahkawh weelno, shuqk
wun keemoorh konjahwan wee-
shulooko nawkawh. * Nul muth-
kuneetaha neepahwooh ; owh : —
" Mawhah geesh-keeshajpinah-

to say to him his last words.
"The young man grieved (and)
thought thus: — ''I will make
everything comfortable for my
uncle. ^ I will construct a bas-
ket with (lit. and) a lid, and I
will put in it all kinds of
downs." ®On the morrow,
when he had finished the basket,
he presented it. His uncle was
pleased and received it weeping
(/. e., with gratitude). ^ He then
placed his uncle in the soft
downy bed. ^ On the morrow,
the sick man stretched out his
hand which meant " attention."
^ He told the youth (then) :—
" Some one is coming at whose
terrible appearance and condi-
tion thou shalt be terrified, but
fear not ; take courage. He
comes in the midnight hour
when all things are sleeping."

11. " On that same night, after
they had eaten, the youth sat on
the opposite side of the fire,
awaiting the outcome of that
night. ^Suddenly there was
something overhead and a cer-
tain terrible-looking being
dropped down by the fire :
''"Ugh," said he, ''I myself
am here ; I eat man ; thou lov-
est thine uncle ; I want him ;
Ugh, thou fearest me . " "* The
youth had fought with wild ani-
mals (?), but this wizard, as he
must be, frightened him for a
while. ''Then, summoning his
courage, he stood up on his feet

1902]

PRIXCE — A MODERN" DELAWARE TALE.

wa." Ovvh yohquh : — '' Law-
peewhich baum ; keeshajpina-
witch. ' ' ^ Nul ktithpihlaun aleen-
qahtang.

III. " Nul withkeelno Imutah-
poowh lawpeewh wtilawul whu-
kee yul : — ^ ** Kalahaat checpah-
wan. Shurhke kalahnickulooq-
kich jeeth. Quackwichha dil-
noom? Dulmitheemich ahlih-
wthihkawk, tauthrha ahvvana-
wah." *= Withkeelno uhloomth-
oowh, shuqk wtilawul wsheethul:
— " Lawpeewhich baum."

IV. ^Aloorhwat quack, 5nh
weekwaum thkuhinthoowh pat-
chihkcheewh ; owh : — " Taunha
wtindin ksheeth ?" ^ (Mawsha-
lindum) Mawsheelahwahkoo
almawsheel warn wawihtoon ay-
lackwloowheen. '^ Wauk uhloom-
thoowh wauk lawpeewh moorh-
kum weekwaum ahwawhlihkoo
shawa wninahko wtil-sheewa-
lindumoo weenawqthowh. ^ Nul
warn wtilauch mookahwaun.
Wtil wturhquon cheepahwan.
Shawa wninootumin wuh linno
nunrhat Muttuntoe. * Nul wtil-
awul withkeelno : '' Chee quack
leetahhawa dulwihkawah. Ktilil
yoonich ktilnumin wauk ktilooh-
moolin wanjich ahloowhweekah-
wut."

and said : — '' I cannot have him
ready." Again said (the wiz-
ard) : — " I shall come here once
more ; let him be ready " (then).
^Then he leapt up through the
smoke-hole.

III. ''The youth sat down
again and spoke thus with him-
self:—" " Truly he is awful. It,
must be that my uncle shall
leave me. What am I to do ?
I will go toward the setting sun.
(Perhaps) I may find people
(there)." "The young man
(then) departed, but he said to
his uncle: — *' I shall come
again."

IV. " After journeying a little,
he came to a wigwam (where) a
small boy came out (and) said :
— ' * How is it with thine uncle ?' ^
^(The traveler) thought it
strange : — " Can one so odd
looking know all about our con-
dition ?' ' ° And he went on, and
again he found a wigwam where
there was a wizard, who at once
saw that he (the traveler) was in
trouble ; that he looked sad.
■^Then the youth explained all
to him. He described to him
the terrible being. Immediately
that man knew that this was
Muttontoe (the evil spirit). ® So
he said to the young man : —
''Do not think about it, I will
overcome him. I will tell thee
what thou shalt do, and I will
explain to thee how to overcome
him."

26

PKINCE — A MODERN DELAWARE TALE.

[Jan. 3,

V. * Withkeelno andah-wam-
loohmoonda uliloomthoowh. An-
dah-nuhpahtah, wama wtilauch
mookuhwaun wsheethul. ^ Nul
andah-keeshmeetthihteet, wtul-
wachpeen alningich keesha-wam-
cheekhung neethkak. " Wsheeth-
ul wtuhlaun nakah wtupeenang
wauk wluqknuhaun waupah-
thauni alpookwuhk andauch
pookwuheeng, warheetawshta
nakah wsheethul wtupeenang.
"^Nul wtilahmooltheen wtilkee-
shich uhloowhweekwaun. ® Nul
ninandpeethkahk lawpee chee-
pawaun lawinda wcheerhakah-
laun : '* Ugh,dupih,neecheepah-
waun konjawan ; keeshajpe."
^Nul andah-tahwining kpuhee-
kun, pajkcheewh withkeelno
cheepeenawqthoo uhj althith-
poocheengwat uhpee. ^Wiyoh
mawhaul linnapa weeshauth-
oowh uhloomihlawh.

VI. * Withkeelno wauk wshee-
thul nulowhwee ayahpoowhuk.

V. • After the youth had been
shown all, he departed. When
he returned, the young man told
all to his uncle. ^Then after
they had eaten, ? ? ? ? he swept
up all the dirt. "^ He put his
uncle in his (the youth's) bed,
and covered him with a white
blanket with a peep-hole in it,
and he lay down on his uncle's
bed. ^ Then he felt that some-
thing strengthened him (with
power) to overcome. ®In the
dead hour of night, the hideous
monster again dropped down in
the middle (of the wigwam).
'' Ugh !" (he said) ^^ am here.
I am a monster. Be ready."
^Then when he opened the lid
(of the basket bed), the young
man, looking terrible, stepped
out completely covered with
feathers. ^ That man-eater be-
came frightened (and) departed
(through the smoke-hole).

VI. * The youth and his uncle
are (still) living (there) con-
tentedly.

Philological Commentary.

I. * Weekwaum (A.® wigwoni) ' house, dwelling ' from V week.
Cp. Weekena * they dwell, inhabit' (A. tu'wigino), of which week-
zvaum is the cognate accus. : — ' they inhabit a house.' Note the
use of the present tense in narration to denote past relation. Lawee
* in the midst of = O. D. lawi and A. nowi (reduplicated nano-
wiwi) in the middle. See V.^ Kohpe * forest ' is undoubtedly

^ A. = Abenaki; O. D. stands for Old Delaware, the mixed language of the
missionaries.

AJP. ^ Amer, Journal of Philology.

1902.J PRIISrCE — A MODERN DELAWARE TALE. 27

cognate with A. k'piwi ' 'n\ the woods.' Withkeelno 'a young
man/ composed of withkee, A. uski, Oj. oshki * yo\xx\g' and Itnno
* man.' See on IV. '^. IVauk 'and,' written woak in O. D."'
Wsheetha ' his uncle ' =r O. D. schiess 'uncle'; A. nzasis 'my
mother-in-law's brother'; Oj. nijishe 'my uncle.' Seel.'', ^, but
III. % wsheethul with obviative -/. Mahji ' already ' = O. D.
metschi and A. majimiwi ' oXwd^ys' ; cp. Oj. aji 'already.' Kihk-
weelno 'old man,' from kihkwee ; cp. O. D. kikey -\- linno ' man.'
Wrhalin ' many '; cp. O. D. chweli. Ncepnuino ' summer ' = O.D.
nipen ; A. 7iiben ; Oj. nibin.

I. ^ Tah lickee = O. D. ^a/i likhique ' once upon a time.' Mon-
tour had written wrongly ian lickee here. O. D. likhique ' now,
about this time.' Weenamulthoo (O. D. winamallsin, A. akuamalsi)
'he feels sick.' Oo7ijeeniawuh 'he calls him'; cp. O. D. wunt-
schiman 'he summons him,' composed of wuntshi 'from' and
Vma ' call' ; so A. uwikwimon 'he calls him,' where the last part
of the stem is identical with the Minsi. Wliinquathitha ' his
nephew '= O.D. limk 'nephew.' Aleh-mawmjeenah-kihkloolaut.
Aleh ' in order that '; mawDijeenah = O. D. inamtschitsch ' for the
last time ' (A. iiiomjessald) ; kihkloolaut is a reduplicated participle,
3 p. anim. 'bespeaks' ixomV klooL See Prince, AJP., xxi, p.
298, on this stem and cp. A. kalolomuk ' one speaks.'

1," Lawalindut?i, cp. O. D. uschuwelendam 'he is grieved.'
Leetahah 'he thinks' =0. D. litchen ; A. alidahomuk 'one
thinks.' Gutauch for tigutauch ' I will' make ' (it), with n- pref. of
I p. and -ch sign of the future (A. -ji), Wlutchawha ' so that it
pleases him.' The first element \% wule- ^ good^,' 'pleasing' (A.
wuli). Jeeth 'my uncle' for njeeth=^0. D. tischiess. Montour
always leaves off the n prefix of the first person before a consonant ;
cp. below III. ^ ; dilnoom dulmeetheemich.

\. ^ Noolihtoomich 'I will make it'; «= 'I'; i p. prefix;
oolihtoo ' make '; w is the sign of inanimate ; ich = fut. ending. Cp.
A. noliionji ' I will make it.' Mihtqueeiiooiee^i^O. D. (Zeisberger)
micMquinotees (dim.) ' a basket, something made of sticks '; cp. A.
w'mi ^kwtonakwdno ' they pry it open with sticks.' Kpuheekun =
O. D. kpahikan 'cover, lid'; 'something to shut;' cp. O.D.

7 A. = Abenaki ; O. D. stands for Old Delaware, the mixed language of the
missionaries.

AJP. = Amer. Journal of Philology.

28 PRIXCE — A MODERN DELAWARE TALE, [Tan. 3,

/^/d!^^/ 'shut the door '; A. kbaha imv. oi k ad ho jnuk ' owe shuts.*
The subst. ending -eekun = A. -higan, as mpask-higan ' gun,' lit. ' a
shooter ' (also Passa. -htg'n, as in wighign ' book '). IVaukitch =
wauk -f- itch, fut. ending. Cp. A. ta ' and ' + fut. ending -//.
Nooshwuhtoomich * I will put it in,' with inan. -;;/ and fut. ich.
This stem may be cognate with Oj. moshki ' fill,' as in fi'inoslikinadon
'1 fill it' (inan.). Uhpeeyuhk ' iedX\vex^'' ; cp. N.^Ahpee (?).
Nahtau probably means ' down,' the soft under feathers (?). Aleen-
awqtheet 'of all sorts' is a participle; cp. O. D. elinaquoi ^ \\v\s
or that. '

I. " Waupungeek 'on the morrow' = O. D. woapank, Oj. wa-
bang 'to-morrow'; cp. A. woban 'daybreak.' See I. ^ waupun-
geeka. Andah is an inseparable prefix = O. D. enda ' when *
(rel.). It is probably cogn. with Oj. anindi ' where ?' Keshiiootah,
a parte. 3 p. 'he making it ' (inan.). Cp. O. D. gischiton ' he
makes it '; A. ngizito?t ' I make it.' Meelaun * he gives it to him ';
cp. A. w'mildn. The ending -ha in wsheethaha seems to be a par-
ticle of asseverative force, as in quackwich-ha, III. ^. Wlalindumoo
'he was pleased with it,' from wule 'good' and -Itndum, as in
lawalindiniy I. ^ Wekwulup ' he received it '; cp. A. w^wikwnernen
'he took it.' The stem is Vz^///^. The ending -up is the sign of
the imperfect; A. -ob ; Fenoh. -pan, Laulpuksho ' he weeps,' from
lep ; cp. O. D. lepakgik ' those who weep '; lepakawagan ' weeping.'

I. ^ Nulhuh-nuh 'then'; cp. O. D. nail ' 2X last.' The first ele-
ment nul here = the resumptive nul, as in II. \ III. *. Wluhlaun
' he puts him ' is the animate form of the same stem as O. D. hatton
inan. Ahpeewtiyuhpeetiang ' in the feather-bed '; see above on I. **,
and cp. Oj. apishimon 'abed, anything to lie upon.' This word
seems to contain the stems ahpee ' feather ' and uhpee ' sit, lie '; cp.
V. \

I. ^ Waupungeeka with temporal ending -a ' when,' as in A.
paiodida 'when they came.' See above on I. ^ IVeenumultheet,
parte. 3 p. 'he is sick'; cp. I.^. IthpeenurJika 'he stretches out his
hand ' = O. D. schipinachgen^ from nachk ' hand.' The first part
of the O. D. iox'cw schipi \^ cogn. with A. siba-liljawi 'stretch out
thy hand.' Aleet ' that which is '; al^ rel. particle -j- eet = parti-
ciple 3 p. of verb 'to be'; cp. A. ali-a'it. Klithtuh 'hearken' =
O. D. gli stain.

I. '' Wtilawul ' he says to him '; tv pref. of 3 p. -f infixed / be-
fore a stem beginning with a vowel -\- il ' say ' -|- wul obviative

1902.] PRINCE — A MODERN DELAWARE TALE. 29

ending. Cp. A. ivdi^ldn Mie tells him.' Ahwaniish 'someone'
with ihh fut. ending. With ahwam, cp. O. D. aiiwen 'who?' and
' someone '; also A. aivani ; Penob. aweni. Pawhji * he will
come.' Note that the fut. ending here is -ji as in A. Cp. O. D.
pejaf 'he who comes' and A. wbaidji 'he will come.' See
below pawoich the fut. participle. Cheepeenawqthoo 'one who
looks strange,' from cheepee?i = O. D. ischipin 'strange' and
•awqthoo ' he appears.' Ahkonjauptoona ' one who is ' (?) from
Vkonj 'exist' (?). Kweeshulooqkich 'thou shalt fear it'; cp.
O. D. wischassin ' he is afraid.' See below on II. ^ . Shuqk ' but '
= O. D. schuk. Chee weeshulooqkoowih ' fear thou not !' Chee =
neg. prohibitive particle, as in IV. ^ For weeshul see above. The
neg. ending here is -oowih. Aluthkuneeiahaul ' be brave.' The
stem muthkun is probably cogn. with Oj. songl- 'brave,' as \Visongi-
deewin ^ co\ir2igQ.^ The Minsi ending -^d'/^/?^/// undoubtedly con-
tains the stem seen in leetahah ' he thinks '; cp. I. % IV. ^, and see
on II. ^ Pawoich 'he will come'; fut, psLVtic'iple pawoi'f -\- c/i.
See above pawhji. Andah-laweetpihkahk * when it was midnight '
= O. D. lawitpikat. It is a comb, of lawi ' midst ' and pihkahk
'night'; cp. A, nowitebakak 'midnight.' Andah-wam-quack-
kaweet. Andah 'when '; quack 'thing'; also 'what?' (cp. O. D.
keco ? A. kagid ?'). It should be written queq and not quack.
Kaweet^ ptc. 3 p. ' it, he sleeps '; cp. A. kawi ; Oj. nin gawingwash
' I fall into a deep sleep ' For this whole sentence, cp. A. : — tdtii
adoji mziwi kagui kawit (in A. we usually find the recipr. form as
kawold^ wak ' they are asleep '.).

II. ^ Nulnuh ; see on I. \ Peethkahkeek ' it was night '; cp. O. D.
pisgeu ' dark '; pisgeep ' it was night ' (^-eep = sign of the past) ;
cp. A. pesgid' bakak 'it is dark.' Mahji 'already'; see on I. *.
Keeshmeettheeteeiah ' they had eaten '; parte. 3 p. pi. Keesh =
sign of perfect ; meetthee ' eat ' -|- ieet, ending of 3 p. pi. parte. ;
ah = temporal ending as in waupungeeka, I. ^. In A. kizi-
mitsihidit 'after they had eaten '; cp. O. D. mizin 'one eats ' and
mizewagan and miistiwagan ' food ' (the last form from Zeisberger).
Montour renders here freely ' after the evening meal,' but this
would necessitate the use of the word ulakunipoagan 'supper.'
Awuihee 'opposite.' Tindawing, loc. 'at the fire,' from tindey
'fire.' Lmuiawpoowh 'he sat'; cp. O. D. wulumachdappiji ^\\q

8 Note that/ in O. D. has the value of consonantal^!'.

30 PRINCE — A MODERN DELAWARE TALE. [Jan. 3,

sits with his legs in front of him ' — /. e.^ on the ground. The last
part of this combination contains the same root as that seen in
ahpeewuyuhpeen (I/) ^ bed ' and A. abi ^ sit.' Pahtoon ' he waits ';
cp. O. D. pehawah, pehowen * he waits.' Tah Miow,' the same
element seen in taimha, IV. *. NiJ *that' with fut. sign. Alak
' which is '; al = rel. particle -\- ak ^ p. ptc. ending inan. JViVi
aleet * that which is '; cp. I. ^ and laweetpihkahk, I. ^.

II. ^ Weerhkawa * suddenly'; cp. O. D. wiechgawotschi 'unex-
pectedly.' Quack konjwah luuhkoong ' something there was above.'
With wuhkoong cp. O. D. hokunk, probably cogn. with A. agudat
'above.' Ahwan, see on I.'' ; cheepeenawqthoo, see on I. ^.
Wcherhakahlaun ' he jumps down ' = O. D. loaktschehellen.
Keerhkee tJi q ta 'by the fire.' With keerhkee cp. O. D. giechgi
'near, by' and with th'q'ta 'fire' cp. A. skweda ; Passa. skwut ;
Oj. isJikote. This seems to be a pure Minsi expression. Tindey is
the Unami word ; see II. ^.

II. " Owh ' he said '; cp. Oj. iwa ' he says.' Baum ' here,
hither'; see also II. ^ Is this cogn. with Oj. oma 'here'? Kon-
jahwan ' I am '; parte, i p. sg. See below on II. ^. Nhukee^ lit.
' my body ' = ' I myself; cp. III. ^, whukee ' himself.' In O. D.
hakey is 'body'; cp. A. nhaga 'my body,' but it is not used to
denote the emphatic pronoun. In Oj., however, we find niiaw
'myself; lit. ' my body.^ The A. pronoun nia ' I ' maybe cogn.
with this. Nmihwa 'I eat'; cp. A. n' dwwo 'I eat him.' In A.
mitsi =z^ Q2it' in general, as 'a meal,' but mowd means rather
'devour.' Linno 'man,' the same stem contained in lendpe 'a
male creature'; see Prince AJP., xxi, p. 298 n. 1. Ktuhaulaw
' thou lovest him '; naiilaw (we expect rather ntuhaulaw /) 'I want
him ' = O. D. ahoalan ' love '; cp. Prince, op. cit., p. 299. Kwee-
shathee ' thou fearest me.' Note ending of i p. -ee.

11. ^ Mutahkawah ' he fought with ' = O. D. machtagen, perhaps
cogn. with A. miga^kamuk 'one fights.' IVeelno (?) 'wild
animals '; so Montour, but I cannot find the stem. Wun demonstr.
'that'; cp. A. wa. Keemoorh 'wizard,' probably = O. D.
kemocliwen ' one who steals away something secretly.' Koiijahwan
pane. 3 p. 'as he was '; see above II. ". Weeshidooko ' he scares
him.' See above on I. *", II. ^ Nawkawh = O. D. nakewi 'a
little while '; cp. A. tCmakaiwi.

II. ® iV/// is used as a resumptive exactly like Passa. ////, which
occurs so often at the beginning of a sentence. It is a demonstra-

1902.] PRINCE — A MODERN DELAWARE TALE. 31

tive originally. Cp. nulhuhnuh I. \ and nulnuh II. \ Mitihkunee-
tahah ' summoning his courage.' See on I. ^. It has the temporal
ending here -ah, as in waupungeeka I. ^. Neepahwoowh ' he stood
erect ' = O. D. nipachton. Maivhah ' not ' ; cp. O. D. maita
' not.' Geesh ' I can ' for ngeesh. Keeshajpinahwavf'iih. neg. end-
ing -wa (cp. IV. " and I. ^) from keeshajpin ' be ready '; cp. O. D.
gischhatton 'be ready/ also the form keeshajpifiaivitch, 3 p. imv.,
* let him be ready. ' A. has kizojo ' he is ready. ' Yohqiih ' now again '
= O. D. yucke, used as a sort of resumptive. Lawpeewhich, a
comb, oi lawpee * again' 2Ci\A peewhich *I shall come' (for iipee-
whicli) ; cp. III. *. Baum ' hither '; see II. ^

II. ^ Kiithpihlaun ' he jumps up '; cp. wcherhakahlaiin ' he jumps
down/ II.'', 2.Y\A uhloo7?iihlawh Mie goes up/ V. \ The ending
'ihlawh seems to mean * jump.' Aleenqahtang (loc. -ang) ' through
the smoke-hole.' It is probably connected with O. D. linquechin
Mook.'

III. "" Lmutahpoowh, see on II. '. Lawpeewh 'again ' = O. D.
lappi. See on II. ^ lawpeewhich. Wtilawul, see on I. *". Whukee,
see on II. \ Yul, pi. o{ yun (inan.) ; cp. A. ytilil ' these/ pi. of
yi7 ' this ' (inan.).

III. "^ Kalahaat ' truly ' is a comb, of kalah = O. D. kehella
'verily, yes'; Penob. kehela, and aat the ptc. of 'to be.' The
literal translation is ' true it is.' A. kalaato ' verily ' is an exact
equivalent oi kalahaat. Cheepahwan 'one who looks horrible';
cp. O. D. tschipilen ' it is awful.' See V. ^ Shurhke ' certainly '
= O. D. schachachki ' surely.' Nickulooqki-ch 'he will leave me,'
from Vnickul = O. D. nukaian ' forsake ' -}- /-, ending of the i p.
as in kweeshathee, II. \ For jeeth, see I. \ Quackwich-ha ;
quack with fut. ending + the particle -ha (see on I. '.). The w-
ending in quackwich shows that this word must really be pro-
nounced quackw (so Anthony in Len. Diet, ; queq under kolkii).
Dil7too7n ' I do it ' for ndilnoom (?). Dulmeetheemich ' I will go,'
for ndul-, from aal (see Le?i. Diet., under V aan ' go '). The past
of this verb is ahloomthoowh ' he went,' III. % V. ^ Ahlih-wihih-
kawk; ahlih, rel. particle as A. ali 'where' + wthihkawk vf'iih
loc. ending -k = O. D. wsigau 'sunset.' Taiithrha 'I (shall)
find ' (?). Ahwanah 'people/ really 'someone,' from ahwa7i (see
on I. ^).

III. •= Wsheethul ' his uncle ' is obviative with characteristic
ending -uL In I. % " and \ Montour has written wsheetha (?).

82 "PRINCE — A MODERN DELAWARE TALE. I Jan. 3,

IV. " Aloorhwat ^ he traveling,' participle ; cp. O. D. miss-ochwen
* he walks about.' Quack must mean 'somewhat.' Yih^ dem.
pron., piobably 'a certain.' Thkuhinthoowh *a small boy' =
O. D. and Unami skahenso (see Len. Diet., p. 115). Patchih-
kcheewh 'there came forth,' from O. D. ktschin 'go out '; see on
V. ^ Taunha wtindin ^ K. toni wdain 'how is he?' Ksheeth
see II. "=.

IV. ^ Mawshalindum and mawsheelawahkoo appear to be alterna-
tive synonyms. The first is written in parentheses in Montour's
MS. Almawsheel probably means ' that (al = rel. particle) one so
strange.' Warn 'all '; see I. ^. Wawihfoon ' he knows it' (inan.);
cp. A. fC wawawinowd ' I know him.' Aylackzvloowheen ' our con-
dition'; aylack = O. D. e/ek 'as it is'j wloowheen 'our being
thus.' I have translated it in the 3 p. for the sake of the English.

IV. " Lawpeewh ' again '; see on II. " and III. \ Moorhkum
' he found ' = O. D. mochganien. Ahwawhlihkoo probably ' there
was a wizard ' (so Montour). Shawa = O. D. schawl 'at once ';
occurs also IV. ^. Wniiiahko ' he knows '; cp. wnitiootumin, IV. '^,
and Prince, op. ciL, p. 298. Wtil-sheewalindiwioo 'he feels sad ';
wtil-, pref. 3 p. (A. wdelH-)\ sheewa ' sad ' (O. D. schiwamallsm
' he feels grieved '); lindumoo, the ending denoting a state of
mind ; cp. I. %^. WeetiawqtJwwh ' he looks sad,' from ween, same
stem as in weenamulthoo, I. ^, -j- awqthowh 'he looks,' as in
cheepeenawqthoo, I. ^; II. ^.

IV. ^ Warn, see I. ^ ; IV. ^. Wtilauch seems to be a fut. ' he will
tell him '; see also in V. *. It is probably used here vividly.
Mookuhwaun appears to be a synonym of withkeelno ' youth.' Wtil-
wturhquon ' he describes to him,' from wtil-, pref 3 p. + v wturh
-j- qiion, ending 3 p. sing, (see Prince, op. clt., p. 298). Wnlnootu-
mln, 3 p. sing. inan. with def -In, as in A. n^wajonem awlkhlgan
'I have a book,' but n'wajonemen azvlkhlgan 'I have the book.'
Wuh llnno ' that man.' With wuh, cp. A. wa 'that.' Nunrhat is
probably a participial formation as shown by -at. Muitontoe must
be connected with O. D. mattonheen ' he curses ' and -to, the same
ending seen in Manltto ' Spirit.' It is clearly a cognitive of A.
madahodo * evil spirit.'

IV. ^ Chee quack leetahhawa ' don't think anything about it,' not
translated at all by Montour. Composed of chee, prohib. ' dont '
(cp. I. **) -\- quack ' anything ' -|- leetahah ' think ' (occurs also

1902.] PRINCE — A MODERN DELAWARE TALE. 33

I. ^). Note the neg. ending -wa, as in II. ". Dulwihkawawh (for
ndul-) * I will overcome him '; cp. ahloowhweekahwut * the way to
overcome him.' The stem is seen in O. D. allowat ' sirongy
mighty.' Ktiitl ' I tell thee '; cp. A. kdi'iel, both from V~il.
Yoonich r=z yoon * this ' -j- ich (fut.) used here as relative * what.'
Ktilnumin * thou shalt do it '; see III. ^. Wauk ktiloohinoolin^ dind.
I will explain it to thee.' The k- prefix = 'thee'; the ending
-ool = I p. ' I ' -f the def. -in. See Prince, op. ctt., p. 299.
Wanjich = O. D. wentschi ' for, in order that ' with fut. -ch.

V. * Andah warn loohmoonda * when he had shown him ail ' (not
translated correctly by Montour); from O. D. allohumassin *he
shows it.' Uhloointhoowh Mie departed'; also 111.^,°; IV. ".
Note the lack of subjects here which must be supplied by the con-
text. I have avoided this by a passive periphrasis. With nuhpah-
tah ' return,' cp. O. D. apatschin.

V. ^ Andah keeshmeeithihteet, so in II. *. Wtulwachpeen alningich
I cannot translate. Montour's MS. is confused at this point.
Keesha, sign of perfect, as geesh in II. ®; warn 'all'; cheekhungj
from same stem as O. D. tschikhammen 'he sweeps it.' Neethkak
' dirt ' = O. D. niskeu. The last part of this stem -eethk^ O. D.
isk seems to be cogn. with Oj. aj-ishki ' mud.'

V. ^ Nakah wtupeenang ' on his bed '; nakah ' on '; wtupeenang
from ahpee ' bed ' (cp. I. ^) with pref. 3 p. ze/' with infixed / before
a vowel. Wluqknuhaun 'he covers him'; cp. O. D. metiach-
quohemen ' he covers it ' and Oj. pada-givanawa ; the common
stem evidently being V kwena. Waupahihauni = O. D. woapach-
saney ' white blanket.' Alpookwuhk andauch pookwuheengy lit. ' he
made a hole there in a hole '; cp. O. V). pquihillen. Andauch =
undach. Warheetawshta^ probably ' he lies down.'

V. ^ Wtilamooltheen 'he feels '; with -mooltheen, cp. I. ^. Wtil-
keeshich 'he will make him'; cp. O. D. gisch 'make' — /. <?., 'he
feels someone making him (giving him power) to overcome.'
Uhlowhweekwaun ; cp. IV. ^.

V. * Nin andpeethkahk, see on II. *. Lawinda ' in the midst ';
cp. I. ^ Dupih ' I am (here) ' from uhpee = O. D. achpin ' be in
a place '; A. abi ' sit.' Keeshajpe ' be ready ' (imv.), see on II. ^

V. ^ Andah'tahwining ' when he opened ' = O. D. tauwmznum-
men ' he opens it.' Kpuheekun^ see on I. ^. Fajkcheewh ' there

PKOC. AMER. PHILOS. 800. XLI. 168. C. PRINTED MAY 5, 1902.

34 MINUTES. [Jan. 17>

came forth'; eg. patchihkcheewh,lY. "". UhJ {}). Althithpoocheen-
gwat uhpee ' he is covered with feathers '; see I. ^ uhpeeyuhk.

V. « U^tyoh, demonstr. pron. Mawhaul Hnnapa ' he who eats
man ' ; cp. n'7nihwa ' I devour,' II. ". On hnnapa from linno
' man ' and -ape * a male ' par excellence ; the race name of the
Delawares, see Prince, op. cii., 295, n. 1. Weeshauthoowh, see I. ^ ;
II. \ Uhloomihlawh * he jumped up '; cp. kiithpihlaun, II. \

Yl.^ Nulowhee 'well, happily.' Ayahpoowhuk * they dwell'
from V ahp; A. abi 'sit.'

Stated Meeting., January 17., 1902.

President Wistar in the Chair.

Present, 11 members.

General Wistar, in taking the Chair, returned thanks for the
honor done him in election to the Presidency of the Society,
and ofiered some remarks concerning the future welfare of
the Society.

The list of donations to the Library was laid on the table,
and thanks were ordered for them.

The decease of the following members was announced :

Cornelius Petrus Tiele, Ph.D., D.C.L., at Leyden, on Jan-
uary 11, 1902, aged 71 years.

Philip P. Sharpies, at West Chester, Pa., on January 15,
1902, aged 91.

Prof. Alpheus Hyatt, at Cambridge, Mass., on January 15,
1902, aged 63.

The Standing Committees for the ensuing year were
chosen, as follows :

Finance. — Philip C. Garrett, William Y. McKean, Joel
Cook.

Hall. — Joseph M. Wilson, Harold Goodwin, John Marshall.

Publication. — Henry Carey Baird, Patterson DuBois, Joseph
Willcox, Amos P. Brown, William H. Furness, 3d.

Library. — George F. Barker, Albert H. Smyth, J. G.
Kosengarten, Edwin G. Conklin, K. C. H. Brock.

The meeting was adjourned by the presiding officer.

1902.] MINUTES. 35

Stated Meeting, February 7, 1902.

President Wistar in the Chair.

Present, 10 members.

Hon. James T. Mitchell, on behalf of the Committee on
Historical Documents, reported that arrangements had been
made for the publication in full of the original journals of
Lewis and Clark.

The following were elected officers to fill vacancies :

Vice-President^ Prof. Samuel P. Langley.

Councilor, Prof. Ira Remson.

The Society was adjourned by the President.

Stated Meeting, February 21, 1902.

Mr. Benjamin Smith Lyman in the Chair.

Present, 3 members.

Letters were read from Prof. Samuel P. Langley, acknowl-
edging his election to the Yice-Presidency, and from Presi-
dent Ira Remson acknowledging his election as a Councilor.
A communication was received from the Congr^s Interna-
tional des Orientalistes de Hanoi, announcing the opening of
an International Exposition, and of a Congress of Orientahsts
in connection with it, at Hanoi in November next, and ask-
ing the Society's cooperation.

The list of donations to the Library was laid on the table,
and thanks were ordered for them.

The meeting was adjourned by the presiding member.

86 MINUTES. [April 3, 4, 5,

Stated Meeting, March 7, 1902.
President Wistar in the Chair.
Present, 25 members.

A letter was received from the Committee formed to
arrange for the XIII International Congress of Orientalists,
to be opened at Hamburg, on September 4, 1902, inviting
this Society to send a special delegate to the Congress, and
on motion the President was authorized to appoint a delegate
to represent the Society.

The list of donations to the Librar}^ was laid on the table,
and thanks were ordered for them.

The decease was announced at Philadelphia, on March 2,
of Francis W. Lewis, M.D., aged 76 years.

The meeting was adjourned by the President.

Stated Meeting, March 21, 1902,

President Wistar in the Chair.

Present, 9 members.

A letter was read from the Secretary of the Kobel Com-
mittees of the Eoyal Academy of Science at Stockholm,
^nclosino; the Code of Statutes of the Nobel Foundation.

General Meeting, April 3, 4, and 6, 1902.

Present, 115 members.

April 3. — Morning Session, 10 A.M.

President Wistar in the Chair.

The President delivered an Address of Welcome.

The Secretaries presented a communication from the Ad-

1902.] MINUTES. 87

visory Committee in Astronomy of the Carnegie Institution
(Prof. E. C. Pickering, Chairman), inviting suggestions
regarding investigations in astronomy, which should be
aided by the Carnegie Institution.

The following papers were read :

" Origin of the Oligocene and Miocene Deposits of the
Great Plains," by Prof. John B. Hatcher, of Pittsburg.

" The Upper Cretaceous and Lower Tertiary Section of
Central Montana," by Prof. W. B. Scott, for Mr. Earl Doug-
lass, of Princeton.

" On South American Mammals," by Prof. William B.
Scott, of Princeton.

" The Mammals of Pennsylvania and New Jersey," by
Mr. Samuel N. Ehoads, of Audubon, N. J.

" The Identity of the Whalebone Whales of the Western
North Atlantic," by Dr. Frederick W. True, of Washington.

Afternoon Session, 2 P.M.
President Wistar in the Chair.

The following papers were read :

' ' On the Molluscan Fauna of the Patagonian Formation, ' '
by Prof. W. B. Scott, for Dr. H. von Ihering, of Sao
Paulo, Brazil.

' ' A Comparison between the Ancient and Eecent Mollus-
can Fauna of JSTew England," by Prof. Edward S. Morse, of
Salem, Mass.

" Distribution of Fresh- water Decapods and its bearing
upon Ancient Geography," by Prof. Arnold E. Ortmann,
Ph.D., of Princeton.

' ' Systematic Geography, ' ' by Prof. William Morris Davis,
of Cambridge, Mass.

" On Drift Casks in the Arctic Ocean," by Mr. Henry G.
Bryant, of Philadelphia.

" On the Magnetic Properties of Mckel," by Mr. Joseph
Wharton, of Philadelphia.

38 MINUTES. [April 8, 4, 5,

Evening Session, 8 P.M.

The following papers were read :

" The Eelation of the American University to Science,"
by President Henry S. Pritchett, of Boston.

" The Advancement of Knowledge by the Aid of the
Carnegie Institution," by President Daniel C. Gilman, of
Baltimore.

April 4.— Morning Session, 10 A.M.
Vice-President Langley in the Chair.

The following papers were read :

" Results of Observations with the Zenith Telescope at
the Sayre Observatory," by Prof. Charles L. Doolittle, of
Philadelphia.

"On a New Method of Transiting Stars," by Prof.
Monroe B. Snyder, of Philadelphia.

" On the Evolution of Martian Topography," by Mr.
Percival Lowell, of Flagstaff, Ariz.

" Historical Investigation of the Supposed Changes in the
Color of Sirius since the Epoch of the Greeks and Romans,"
by T. J. J. See, Ph.D., of Washington.

" Recent Progress in the Lunar Theory," by Prof. Ernest
W. Brown, F.R.S., of Haverford, Pa.

" On the Spectra of Gases at High Temperature," by Prof.
John Trowbridge, of Cambridge, Mass.

Executive Session, 12.40 P.M.
President Wistar in the Chair.

Pending nominations were read, and the candidates for
membership were balloted for, and the Secretaries reported
the election of the following :

Residents of the United States —

John A. Brashear, Sc.D., Allegheny, Pa.

1902,1 MINUTES. 39

Andrew Carnegie, LL.D., New York.

Prof. William B. Clark, Baltimore.

Prof. Hermann Collitz, Ph.D., Bryn Mawr.

Grove K. Gilbert, Washington.

President Arthur Twining Hadley, New Haven.

Prof. George B. Hale, Williams Bay, Wis.

Prof. Paul Haupt, Baltimore.

Prof. Albert Abraham Michelson, Sc.D. (Cantab), Chicago.

C. Hart Merriam, Washington.

Prof. Theodore William Eichards, Cambridge, Mass.

Prof. Felix E. Schelling, Ph.D., Philadelphia.

Prof. Eobert Henry Thurston, Ithaca.

Benjamin Chew Tilghman, Philadelphia.

Prof. Robert S. Woodward, New York.

Foreign Residents —

Antoine- Henri Becquerel, Paris, France.

Jean-Gaston Darboux, Paris, France.

Sir Michael Foster, F.R.S., D.C.L., Cambridge, Eng.

Prof. G. Johnstone Stoney, F.R.S., London, Eng.

Prof. Silvanus P. Thompson, F.R.S., London, Eng.

Afternoon Session, 2 P.M.
President Wistar in the Chair.

The following papers were read :

"^Is Scientific Naturalism Fatalism ? A one-minute
paper" by Prof. William Keith Brooks, of Baltimore.

'* On Dichotoma, a New Genus of Hydroid Jelly-Fish,"
by Prof. William Keith Brooks, of Baltimore.

'^ On Some Equations Pertaining to the Propagation of Heat
in^an Infinite Medium," by Prof. A. Stanley Mackenzie, of
Bryn Mawr, Pa.

" On the Law of Magnetic Hysteresis," by Prof. M. I.
Pupin, of New York.

" On the Continuity of Protoplasm," by Prof. Henry
Kraemer, of Philadelphia.

40 MINUTES. [April 3, 4, 5,

'' The Embryology of a Brachiopod," by Prof. Edwin
Grant Conklin, of Philadelphia.

" Relationship of the Gordiacea," by Prof. Thomas H.
Montgomery, Jr., of Philadelphia.

"The Spermatogenesis of Oniscus Asellus, Linn., with
Especial Reference to the History of the Chromatin," by
Prof. E. G. Conklin, for M. Louise Mchols, Ph.D., of Phila-
delphia.

" The International Catalogue of Scientific Literature,"
by Cyrus Adler, Ph.D. , of Washington.

April 5. — Morning Session, 10 A.M.
Yice-President Sellers in the Chair.

The following papers were presented :

" Experiments on Cytolysis," by Prof. Simon Flexner, of
Philadelphia.

" A Classification of Economies," by Prof. Lindley Miller
Keasbey, of Bryn Mawr, Pa.

" On Osteitis Deformans," by Prof. James C. Wilson, of
Philadelphia.

' ' The Influence of Acute Alcoholic Intoxication upon Cer-
tain Factors Involved in the Phenomena of Hsemotolysis and
Bacteriolysis," by Prof. A. C. Abbott, of Philadelphia.

" Blindness from Congenital Malformation of the Skull,"
by Charles A. Oliver, M.D., of Philadelphia.

" The Isthmian Canals," by Prof. Lewis M. Haupt, of
Philadelphia.

" Race Elements in American Civilization (illustrated by
German Examples)," by Prof. M. D. Learned, of Philadelphia.

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 41

THE EMBRYOLOGY OF A BRACHIOPOD,
TEREBRATULINA SEPTENTRIONALIS Couthouy.

EDWIN G. COHKLIN, PH.D.

(FROM THE ZOOLOGICAL LABORATORY OF THE UNIVERSITY OF PENNSYLVANIA.)

Plates I-X.
[Read April 4, igo2.)

I. Introduction.

I. Historical — Although Brachiopoda are chiefly notable be-
cause of their great abundance in past geological periods, their
relationships to other groups of animals are still so obscure as to
make them objects of great interest to the general morphologist.
At different times and by different investigators they have been
variously regarded as allied to MoUusca, Polyzoa, Annelida,
Chaetognatha and Phoronis, while others have regarded them as a
distinct phylum of the animal kingdom. Even at the present time
there is no uniformly accepted view as to their relationships, all of
the above affinities (excepting possibly the first) being maintained
by different authors. Of the two morphological methods of inves-
tigating relationships, viz., Comparative Anatomy and Comparative
Embryology, the former has been applied to this group of animals
in a number of noteworthy cases. Not to mention the large num-
ber of older and less important works on this group, the comprehen-
sive studies of Albany Hancock ('58) and the excellent researches
of Lacaze-Duthiers ('61), which are still models of accuracy, the
extended labors of Davidson [^Zd-^Z) and most recently the
series of splendid contributions by Blochmann ('92 and 1900) have
made us as well acquainted with the anatomy of the brachiopods as
we are with the anatomy of most other invertebrates.

The case stands far differently with the embryology of this
group. But two writers have as yet attempted to deal with the
entire embryology of a brachiopod, and both of these studies were
made without the employment of serial sections or modern micro-
scopical and micro-technical aids.

Neglecting the isolated observations of Fritz Miiller ('60 and '61)
of a free-swimming larval brachiopod, and the more extended but
still very fragmentary observations of Lacaze-Duthiers ('61) on the
development of Thecidium, the credit of having made the first
study of the entire development of a brachiopod belongs to the

42 OONKLIN — EMBRYOLOGY OF A BRACHIOPOD. I April 4,

American naturalist, Prof. E. S. Morse ('71- 73)- How thorough
and complete this work was I shall have occasion to remark in the
further course of this paper ; but done as it was at a time before
good microtomes and imbedding means were invented, and long
before serial sections were thought of, it could not but leave much
of the internal structure of the embryo undetermined, especially as
the eggs and embryos of the form studied (^Terebratulina sep-
tentrionalis) are quite small and opaque. Nevertheless Morse's
work stands to-day as one of the two most important works on the
embryology of the brachiopods. The other work referred to is the
later but more detailed and comparative ** Observations on the
Development of Brachiopods," by the great Russian zoologist,
Alexander Kowalevsky (1874). Kowalevsky's work, which was
published in Russian, remained practically unknown to those not
acquainted with that tongue until 1883, when Oehlert and Deniker
published an excellent abstract of it. In this work Kowalevsky
describes his observations on the development of four species —
Argiope {^Ciste/la) 7ieapolitana, Thecidium mediterraneum, Terebrat'
ula fni?W7' and Terebratulina caput-serpentis; only a few observations
were made on the development of the two last- mentioned species,
but his work on Cistella and Thecidium was detailed in character
and nearly complete so far as the stages of development are con-
cerned. Although Kowalevsky employed isolated sections to a
limited extent in his work and also shows certain details of internal
structure in many figures of entire embryos, yet his work of neces-
sity left many important problems of structure unsolved.

In 1879 Prof- ^ • K. Brooks discovered the free- swimming larvae
of Linguia {Glotlidid) pyramidata and described in detail the
structure and further development of these larvae up to the adult
condition. This work, although dealing only with the larval stages
and metamorphosis, is still the most complete extant on the devel-
opment of the Ecardines, the most primitive group of the brachio-
pods. With characteristic insight Brooks has used his many
important discoveries on the later development of Glottidia in an
extremely valuable discussion of the systematic position of the
brachiopods.

The small portion of Shipley's (1883) paper on Argiope {Cistella)
which treats of the development of that form adds little to the
much more extensive work of Kowalevsky on that animal. His
principal contribution consists in his determination of the fact that

1902.] CONKLIN — EMliRYOLOGY OF A BRACHIOPOD. 43

the so-called ''segments " of the larva are not true segments, as
Kowalevsky supposed, but are mere folds in the body wall.

The papers by Beecher ('91, '92, '93) on the development of
brachiopods deal almost entirely with the developmental changes
which occur in the shell and not with the general embryology.
Beecher has proposed a very interesting and important classification
of the brachiopods based on the developmental characters of the
shell; since however the present work deals only with the early
embryology, we need not further consider Beecher's work here.

2. Material, — For the material which has formed the basis of
this study I desire at the outset to express my profound obligations
to my friend Dr. Edward G. Gardiner, of Wood's Holl, Mass. Dr.
Gardiner had collected the material (which consists of about thirty
different stages in the early embryology of Terebratuli7ia septeritri-
onalisy forming a fairly complete series from the unsegmented egg
up to the beginning of the metamorphosis) at Eastport, Me., during
the early summer of 1894. For various reasons he was prevented
from making an immediate study of this material, and when in the
summer of 1898 in conversation with him I expressed my desire to
study the cell lineage of a brachiopod, he graciously offered me the
material which he had collected with the request that I should use
it in any way I might see fit. I soon found that it would be impos-
sible to work out the cell lineage, not only because of a lack of
sufficient number of cleavage stages, but also and chiefly because
of the great difficulties which the material itself offered ; the eggs
were quite opaque and, except in a few cases, it was impossible to
render the nuclei visible in preparations of the entire egg ; the
cleavage was almost entirely equal and I was unable to find any
constant landmarks which might be used in orientation, and finally
the cleavage was found to be more or less irregular and inconstant.
I was compelled therefore to abandon the plan to study the cell
lineage of Terebratulina and the material was laid aside, until a few
months ago I found opportunity to again take up this subject with
the view of working out the early development of this interesting
animal in as great detail as the material would allow.

3. Methods.— K\\ the material was, I believe, preserved in Per-
enyi's fluid, and while the general form and size of the embryo
as a whole, and also of its constituent cells and nuclei, has been
faithfully preserved, every trace of the cilia, which according to
Morse ('71-73) cover the surface and line the alimentary tract and

44 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

body cavity, has entirely disappeared. The method of fixation,
together with the long residence of the material in alcohol, ren-
dered it difficult to stain. The picro-haematoxylin, which I have
used with such good results in other cases, was of little value here,
and I have found that in the preparation of entire eggs or embryos
the most useful stain is borax carmine, while in the preparation of
sections iron haematoxylin has given the best results. Both entire
preparations and serial sections were mounted in balsam and studied
and drawn under an immersion lens (Zeiss. Apochromat. 3 mm.,
Comp. Occ. 4).

Perhaps I may be pardoned a word- in defense of the rather large
use of surface views and optical sections which I have made in this
paper. This has not been due to the fact that I have made few
serial sections, for I have made and studied serial sections of many
hundreds of embryos, but because with material which is at all
favorable the orientation of the embryo and the interrelation of
its various parts can be more safely and satisfactorily determined
from the study of whole embryos than by means of serial sections ;
and this is especially true where it is possible to use an immersion
lens in the study of entire preparations. Further, more points of
structure can be shown in a single figure of this kind than in dozens
of figures of serial sections. Of course, serial sections must always
be used in connection with the study of entire preparations, and in
the present paper all the details of internal structure which are
shown in the surface views and optical sections have been confirmed
again and again by serial sections. Any one accustomed to the
study of both entire preparations and serial sections knows that few
things are more deceptive than the latter when not checked by a
study of the former, while the publication of whole series of sec-
tions contributes more to the pride of the author and the income
of the illustrator than to the edification of the reader.

II. The Egg and its Cleavage.

Morse ('71) has described the method of egg laying, and has
called attention to the fact that the mature eggs are usually kidney-
shaped, though they vary considerably in shape and size. None of
the unsegmented eggs which I have examined are kidney-shaped ;
they are slightly elliptical, being about 160 /x in the longest diame-
ter and 144/^. in the shortest. This elongation of the egg in one

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 45

axis is probably the precursor of the first cleavage at right angles to
that axis. Morse ('73) also mentions the presence of a " distinct
granular membrane, the ovishell, which is formed while the eggs
are still in the perivisceral cavity." I have found this membrane in
all early stages (Figs, i-io), though I have not been able to recog-
nize it after about the twenty-cell stage. From the fact that it lies
close to the periphery of each cell, following all of its individual
curvatures (Figs. 2-10), I am convinced that it is not a distinct
membrane but only a peripheral layer of clear protoplasm (ecto-
plasmic layer of Harmer). There is no protoplasmic area in the
egg distinct from the yolk, though there is a slight aggregation of
protoplasm around the nuclei, but the entire egg is densely packed
with small yolk granules which jender it opaque.

I have not attempted to study the nuclear phenomena of the
maturation and cleavage since the material is very unfavorable for
such work. Two polar bodies are formed, the first of which soon after
divides (Figs, i and 2). These polar bodies do not remain attached
to the egg after the sixteen-cell stage, and in some eggs they appear
to pass into the cleavage cavity, though in such cases it is difficult
to distinguish between polar bodies and small spherules which are
cut ofif from the inner ends of the cleavage cells, and which contain
protoplasm and yolk but no nuclei (Figs. 11 and 37). At the stage
when the gastrulation begins these spherules are found in consider-
able numbers in the cleavage cavity (Fig. 37) ; they disappear in
later stages. Similar spherules have been observed by Caldwell
('85) in Phoronis.^

The first cleavage is meridional and divides the egg into two
slightly unequal blastomeres (Fig. 2) ; the second cleavage is also
approximately meridional and divides each of the blastomeres
equally ; as a result of this cleavage four blastomeres are formed,
two of which are somewhat smaller than the other two (Fig. 4). A
polar furrow is present (Figs. 4 and 7) which, taken in connection
with the overlapping of certain cells (Fig. 3), indicates that in
some eggs at least the cleavage is of a spiral type. The third
cleavage is equatorial and leads to the formation of eight blasto-
meres, all of which are nearly equal in size (Figs. 5, 6, 7) ; in some
eggs the four cells at the animal pole lie just above those at the
vegetal pole (Fig. 6) ; in others they have rotated through various

1 Quite recently Ideka (1901) has fully described these spherules in Phoronis ;
he calls them pCasmic corpuscles.

46 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

angles (Figs. 5 and 7). In the eight-cell stage a considerable
cleavage cavity appears with openings at the animal and vegetal
poles (Figs. 6 and 7).

The cleavage forms just described and represented in Figs. 1-7
are among the most regular ones observed ; others are irregular and
unequal from the first. One of the most common variations is pro-
duced by very unequal divisions, the chief mass of the egg seeming
to bud off small cells, usually at the animal pole (Fig. 8). Later
stages in which one or two of the blastomeres are much larger than
the others, and in which the cleavage cavity is smaller than usual,
are not infrequently found. Such eggs appear to give rise to nor-
mal blastulae, in which all the cells are of equal size, by the more
rapid division of these larger blastomeres.

The eight-cell stage gives rise to the sixteen-cell stage by the
meridional division of each of its blastomeres. Fig. 7 shows each
blastomere of the eight-cell stage indented at its periphery, pre-
paratory to this division, which occurs simultaneously in all of the
cells. The sixteen cells shown in Fig. 9 and in optical section in
Fig. 10 are all of approximately the same size. Except for the
occasional presence of the polar bodies at this stage it would be
impossible to distinguish the animal from the vegetal pole. The
cleavage cavity is now larger and it no longer opens widely to the
exterior.

In subsequent cleavage stages division does not take place simul-
taneously in all of the cells ; this is shown, for example, by Fig. 1 1, in
which twenty cells are present, some of which are considerably
larger than the others. In the eggs represented in Fig. 1 2 about
forty-eight cells are present and some of these are larger than
others, indicating that with them division has been delayed. The
egg shown in Fig. 12 has been flattened by the cover glass, so that
its apparent diameter is greater than normal ; at the same time the
blastomeres are separated from one another in an abnormal manner.
At all stages the blastomeres are but loosely joined together and they
break apart at the slightest pressure. In the later development I
have found many embryos which are about one-half or one-quarter
the size of the normal embryo, and it seems likely that such
embryos have developed from isolated blastomeres of the two- or
four-cell stage.

After this brief description of the cleavage, I think it will be
quite apparent to everyone that it would be extremely difficult, if

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 47

not quite impossible, to work out the cell lineage of Terebratulina,
even with an abundance of most favorably preserved material.
With the material at my disposal such work was wholly out of the
question. In the blastula, even at the time when gastrulation
begins, one is struck by the great uniformity in size and quality of
all the cells. I have found it quite impossible to distinguish any
difference between the cells which invaginate and those which do
not until after the gastrulation is well advanced.

III. Gastrulation and Formation of Body Layers
AND Cavities.

Gastrulation occurs by typical invagination, and at the time
when the infolding begins there is no difference in the cells at the
two poles (Figs. 13 and 37). The infolding continues until the
inner layer comes into contact with the outer one and the blasto-
coel is entirely occluded (Fig. 14 et seq.). During this process
there is a decided change in the character of the cells of the inner
layer ; they become very much shorter and henceforth are cuboid
or rounded in shape ; the cells of the outer layer remain columnar
in shape and are very long, so that the ectoderm is quite thick.
Large deeply staining granules are found at the free ends of all the
cells, both those which are invaginated and those which are not,
and in the invaginated cells these granules are so dense that they
frequently obscure the nuclei and cell boundaries. In the ectoderm
these granules lie on the outer side of the nuclei (Fig. 37 et seq.
Fg), while the inner ends of the cells are left free from gran-
ules and nuclei and hence are very clear. I suggest that these
granules may be associated with the cilia, which in life cover the
embryo and line the enteron and coelom (see Morse, '73).

Almost as soon as the inner layer comes into contact with <-he
outer one — /. <?., when the infolding is complete — the innermost por-
tion of the archenteron becomes slightly constricted from that
portion lying nearer the blastopore. This constriction is deepest
anteriorly, least marked posteriorly, while it is about equal on the
left and right sides of the archenteron (Figs. 14, 15, 16, 38). On
the anterior side this constriction develops into a partition wall,
which grows downward and backward into the archenteron, shutting
off the enteron above from the enteroccel below (Figs. 16, 20,
42fz and 42^). So far as I am able to determine this partition wall

48 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4.

is always but one cell thick (Figs. 40-42), though in the earlier
stages of its development it represents a fold in the wall of the
archenteron (Figs. 42^^-42^). The backward growth of this par-
tition wall continues until the enteron is entirely separated from
the enterocoel save for a narrow slit-like communication at the
posterior end (Figs. 20, 24).

While the enteron is thus being separated from the enterocoel
the blastopore is gradually closing and the whole embryo is becom-
ing flattened in a dorso-ventral direction and elongated antero-
posteriorly. The blastopore is at first a circular opening ; it then
becomes narrowed from side to side and apparently elongated
antero-posteriorly (Figs. 17-21). The blastopore groove thus
formed is shallow posteriorly and deepest at its anterior end where
it opens into the enterocoel and enteron (Fig. 17). This groove
continues to grow narrower and to be filled up at its posterior end
until it becomes a mere slit, opening by a small pore near its
anterior end into the enterocoel (Figs. 21, 22). Finally this pore
also closes (Fig. 24) and the enterocoel and enteron are completely
shut off from the exterior, though still communicating with each
other by a narrow opening in the region posterior to where the
blastopore closed (Figs. 24 and 46, 47). The blastopore groove
persists for some time after the pore has closed but ultimately dis-
appears, though a depression is left at the anterior end of this
groove which becomes a part of the anterior mantle furrow ; it is
probable that at this very point the oesophageal invagination occurs
at a stage after the fixation of the larva (see p. 56).

In stages in which the blastopore is still circular the enterocoel
is but little larger than the enteron (Figs. 14 and 38). In looking
at an entire egg of this stage from the oral side one sees two cavi-
ties of about the same diameter, one above the other, which com-
municate with each other by a wide opening ; the cavity nearest
the blastopore is the enterocoel, the one nearest the aboral side the
enteron. In an older stage (Fig. 17) in which the blastopore has
begun to narrow one still sees that these two cavities are of nearly
the same diameter. As the enteron becomes separated from the
enterocoel, however, the latter becomes much more extensive than
the former, and an oral view of an embryo at this stage shows the
enterocoel lying on the oral side of the enteron and entirely
surrounding it except on the aboral side (Fig. 21, also optical sec-
tions, Figs. 19 and 20). This rapid enlargement of the enterocoel

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 49

is due in large part to the change in shape of the embryo, which
becomes flattened in a dorso-ventral direction and enlarged in its
other axes. Associated with the growth of the enteroccel is an
important change in the character of its bounding cells;- at first
they form a quite regular, cuboid or columnar epithelium (Figs.
14-16 and 40-42^), but as the enterocoel increases in size the
epithelium becomes less regular, particularly at the anterior end,
and here many mesenchyme cells come to lie in the cavity of the
enteroccel (Figs. 20, 21, 42^^, 43). Later such mesenchyme cells
are found generally throughout most of the coelom. The cells
bounding the enteron remain cuboid or columnar throughout the
development.

With the flattening of the embryo and the closure of the blasto-
pore, the ventral wall of the enteron is brought into contact with
the ectoderm at the place where the blastopore closes (Figs. 24
and 44-47), and consequently the enterocoel is here divided into
right and left cavities, which however still communicate with each
other at the anterior end and open into the enteron posteriorly
(consult Figs. 43-47 which are transverse sections of an embryo of
the stage shown in Fig. 24, Fig. 43 being the most anterior section
drawn and Fig. 47 the most posterior). Very soon after this stage
the communication between the enteron and the enterocoel is com-
pletely cut off and the definitive coelom is thus formed, consisting
of two sacs, still opening into each other anteriorly and posteriorly
but separated throughout the middle region of the embryo (consult
Figs. 48-52 which are cross sections, in order from the anterior to
the posterior region, of an embryo of about the stage shown in
Fig. 28).

The gastrulation and formation of body cavities in brachiopods
has been observed heretofore only by Kowalevsky. A comparison
of the method of gastrulation and coelom formation in Terebratu-
/I'na with. Kowalevsky's observations on Cistella and other brachio-
pods reveals many resemblances and some interesting differences.
Kowalevsky found that the gastrula was formed by invagination in
Cistella, Terebratula and probably Terebratidina ; by delamination
or ingression va Thecidium. In all cases, however, he describes
the coelom as arising as two lateral pouches from the archenteron in
the same manner as in Sagitfa, viz., by the folding into the arch-
enteron of two lateral partitions. In this way the archenteron is
divided into three portions, a median one which becomes the

PROC. AMER. PHILOS. SCO. XLI. 168. D. PRINTED MAY 5, 1902.

50 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

enteron and two lateral ones which form the coelom. It is diffi-
cult to determine from Kowalevsky's figures from which portion
of the archenteron these lateral partitions arise, but there can be
no doubt that they are regarded as folds of the archenteric wall
(see his Fig. 6), nor that they divide the archenteron into three
cavities, the enteron and two coelom sacs. In both of these
respects Cistella is very unlike Terebratulina ; in the latter the
archenteron is first divided into two cavities and not three (the
enterocoel is for a long time unpaired), and the single partition wall
by which this division is brought about consists of a single layer of
cells and not a plication of the archenteric wall (though in its
earliest stages this partition wall probably represents such a plica-
tion. Fig. 42«). The former of these differences is perhaps not
so great as would at first thought appear, being principally due
to the fact that in Terebratulina the -enteron occupies but a small
part of the archenteron, and hence the partition wall which
separates it from the enterocoel leaves the latter a large unpaired
cavity, whereas in Cistella the division of the archenteron is
more nearly equal and when completed separates two lateral
enterocoel pouches from the median enteron. But one cannot
overlook the fact that according to Kowalevsky two partition
walls are formed in Cistella, whereas but a single one is found in
Terebratulina. Moreover these partitions are lateral in position
according to Kowalevsky, whereas in Terebratulina the single par-
tition grows out from the anterior wall and merely curves around on
to the lateral walls of the archenteron (Fig. i6). The difference
in the structure of this partition in Cistella and Terebratulina is
also important \ in the former it consists of a double layer of cells,
in the latter of a single layer ; in Cistella the wall of the enterocoel
sacs lying next to the enteron becomes the splanchnic layer of the
mesoblast, in Terebratulina the splanchnic mesoblast is derived
from mesenchyme cells. Before attempting to explain these im-
portant differences between Cistella and Terebratulina in the
formation of the coelom it would be well to know that they actually
exist, and it seems highly desirable that the embryology of Cistella
should be reinvestigated with the aid of modern histological
methods.

The differences between Terebratulina and Sagitta in the mode
of forming the coelom are of interest since they remove one im-
portant argument for the supposed relationship between Brachio-
poda and Chaetognatha.

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 51

IV. Orientation of the Embryo and Establishment
OF Definitive Axes.

It is well known that previous investigators of the embryology of
the brachiopods have found the proper orientation of the embryo
one of their most perplexing problems. The complete closure of
the blastopore at a relatively early stage and before many land-
marks have appeared by which its position relative to definitive
organs could be determined, as well as the fact that both the oral
and aboral sides of the embryo are very similar, has been chiefly
responsible for this uncertainty of orientation. However if one
studies carefully whole embryos of the critical stage when the last
remnant of the blastopore is disappearing, the proper orientation of
the embryo of a brachiopod is no more difficult than is that of any
other animal.

In the stage shown in Figs. 23 and 24 the antero-posterior axis
of the embryo is well defined, while the remnants of the blastopore
are still present. The enlarged end of the embryo (to the right
in Fig. 24) is anterior and gives rise to the head, while the
posterior end (to the left in the figure) is narrowed and gives rise
to the peduncle. The point where the blastopore closed (Fig. 24,
Br) lies near the middle of the ventral side, while the blastopore
groove runs backward almost to the posterior end of the embryo.
Directly opposite the blastopore is a groove which runs transversely
across the dorsal side of the embryo ; this is the dorsal mantle
groove, and the prominent ridge anterior to it is the dorsal mantle
fold (Fig. 24, Md). If now Fig. 24 be compared with Figs. 20
and 16 it will be seen that the blastopore occupies the ventral-pos-
terior region of the embryo, and that the anterior pole of the
embryo is rounded while the posterior pole is pointed, the embryo
being flattened on its postero-dorsal side. The prominent ridge
opposite the blastopore in Figs. 16 and 20 corresponds with the
dorsal mantle fold in Fig. 24. A comparison of these three
figures further shows that the axis connecting the middle of the
blastopore v/ith the apex of the gastrula invagination (<?. g., E \v\
Fig. 14) is ultimately bent on itself through an angle of more than
90°. It is difficult to say whether this bending of the gastrula
axis is chiefly due to the forward shifting of the blastopore on the
ventral side or to the forward shifting of the apex of the gastrula
invagination, since there are no points in the embryo which may be

52 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

considered as fixed in position. It is highly probable however that
both movements occur and that both the apical pole and the blas-
topore pole are shifted forwards, while the middle of the gastrula
axis is shifted backwards, the gastrula axis thus being doubled on
itself exactly as is the case in Turbellaria, Annelida and Mollusca.
Since the polar bodies have long since disappeared it is impossible
to locate exactly on the embryo the point which corresponds to the
animal pole of the egg. It is probable, however, that this point
lies on the ectoderm directly opposite the apex of the gastrula
invagination, and therefore anterior to the ridge which develops
into the dorsal mantle fold. In the forward shifting of the apex of
the gastrula invagination it is highly probable that this point is also
shifted forward and continues to lie opposite the apex of the enteron.
If this be true the animal pole of the egg coincides very nearly
with the point where the line from Ce in Figs. 20 and 24 touches
the ectoderm.

The bending of the gastrula axis which has just been described
shows that Terebratulina, like most bilateral animals, belongs to
the group designated by Hatschek (^^^^ Heteraxonia, and by
Goette ('82) Hypogastric forms. Goette has divided bilateral
animals into two groups: (i) the pleurogastric, in which the chief
axis of the egg becomes the chief axis of the larva or adult, as in
Sagittadccidi the echinoderms, and (2) the hypogastric, in which one
of the ''cross axes" of the egg becomes the chief axis of the
larva or adult, as in worms, mollusks and arthropods. There can be
no doubt that Terebratulina should be classed among the hypogas-
tric forms, and if it be true, which however seems questionable,
that Sagitta and the echinoderms belong to the pleurogastric type,
it shows a very important difference between the embryo of the
brachiopod and of the chaetognath.

V. Development and Organization of the Larva.

There is of course no natural Ime of demarcation between the
embryo and the larva, but for the sake of convenience we shall
designate those stages which precede the closure of the blastopore
as embryonic, while those which extend from the closure of the
blastopore to the end of the free-swimming life we shall call lar-
val stages.

The flattening of the embryo in the dorso-ventral axis and its

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 53

elongation antero-posteriorly have already been described. At the
same time the embryo becomes wider at its anterior end and nar-
rower posteriorly. The dorsal mantle groove and fold appear and
the fold extends over on to the ventral side of the larva as a slight
swelling on each side of the midline (Figs. 25 and 26). These
are the halves of the ventral mantle fold and at first they are sepa-
rated in the mid-ventral line by the blastopore groove (Fig. 26, Bf)^
while the blastopore remnant {^Br) lies at the anterior end of this
groove. Very soon after this stage the halves of the ventral mantle
fold fuse with each other, thus obliterating the last trace of the
blastopore groove, while a deep notch on the anterior side of the
ventral mantle fold (Fig. 29, O) represents the place at which the
blastopore remnant was last seen.

I. The Mafitle J*old is at this stage a ring-Hke prominence which
extends all the way around the larva (Figs. 29, 30, 31). This ring
passes obliquely around the larva, being nearer the anterior end
on the dorsal side and nearer the posterior on the ventral side
(Fig. 31). Two mantle furrows are now plainly distinguishable,
one in front of and the other behind the mantle fold. The anterior
furrow is deepest on the ventral side, while the posterior one is
deepest on the dorsal side (Figs. 29-31), With the appearance of
the mantle, bounded anteriorly and posteriorly by these constric-
tions, the mantle furrows, three regions may be recognized in the
larva, viz., the cephalic region, in front of the anterior mantle
furrow, the mantle, between the anterior and posterior furrows,
and the peduncular region, behind the posterior mantle furrow.

These constrictions, which I have called the anterior and pos-
terior mantle furrows, continue to grow deeper but at no time do
they form true septa which divide the ccelom. The regions which
they separate are not, therefore, segments, as Kowalevsky supposed.
The larva is at this, and all other stages which I have studied,
unsegmented, and the appearance of segmentation is due merely
to the formation of the mantle from the middle region of the
body.

The mantle becomes a very prominent ring around the body, and
then its free edge is turned backwards until it surrounds the pedun-
cular regions on all sides (Figs. 32-36). A space is left between
the mantle and the peduncle which is the peduncular chamber
(Figs. 34 and 36, PC). This chamber is a little deeper and wider
on the dorsal than on the ventral side, which is due to the fact

54 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

that the posterior mantle furrow is deeper and is farther forward on
the dorsal than on the ventral side (Fig. 31, Fp).

When the mantle has reached the limit of its backward growth it
entirely covers the peduncle, the extremity of which nearly fills the
circular opening of the mantle chamber (Figs. 34 and 36). This
is the oldest stage which I have had opportunity to study. Morse
C 11) has observed in detail the transformation of this larva into the
adult. He figures a great many stages showing the manner in
which the mantle is turned forward over the cephalic region, its
free edge being directed forward and its originally internal surface
becoming external. This happens after the larva has become fixed
by the tip of its peduncle, and it is evident from Morse's figures
and descriptions that the stage shown in my Figs. 34-36 is one of
the last stages in the free-swimming life.

2. The Cephalic Regmi lies in front of the anterior mantle fold
and is nearly hemispherical in shape, being however somewhat
variable in form (consult Figs. 32-36), which is probably due to
the fact that it is extremely contractile, as Morse has observed. At
its anterior end and slightly toward the dorsal side is a shallow
depression, the apical sense plate, which bears a tuft of long cilia
in life (see Morse). The enlarged end of the enteron as well as a
portion of the mesoderm and coelom extend into the cephalic
region.

3. The Pedunculai' Region is cylindrical in shape and is con-
tracted near its posterior end. This contraction is due to the fact
that the coelom and mesoderm terminate abruptly some distance
in front of the end of the peduncle (Figs. 34 and 36), and it is
certainly not to be taken as constituting a fourth region of the
larva, as Shipley ('82) suggests in the case of Cistella. The endo-
derm is continued as a solid cord of cells nearly to the end of the
peduncle.

4. The Ectoderm covering the larva is unusually thick, though
consisting of but a single layer of cells ; these cells are however
extremely long. Their inner ends are clear and free from nuclei
and granules, so that on first examination a clear zone seems to
separate the outer from the inner layer (Figs. 38 et seq.). Only on
the anterior side of the mantle fold does the ectoderm become
cuboid or squamous, while over the cephalic and peduncular regions
it is particularly thick.

5. SetcB Sacs. — High columnar cells line the peduncular chamber,

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 55

and at four places (two median and two lateral) on the dorsal side
of the mantle the epithelium is invaginated to form the setae sacs
(Figs. 34, 36, 56-64, SS). No traces of setae are preserved in the
material which I have examined, but there can be no doubt from
Morse's ('73) account that these invaginations give rise to the
lateral and median bundles of temporary setae.

6. Sense Plates and Gafiglia. — A little toward the dorsal side of
the apex of the cephalic lobe is a depression in the ectoderm, and
in this region the cells are deeply pigmented, especially at their
free borders. This is the apical sense plate {Scheitelplatte) (Figs.
26, 28, 29, 31, 32, 34, 35, 36, 56, 57, 58, CG), and in life bears
a tuft of long cilia (see Morse, '73), though no trace of these is left
in the material which I have examined. At the base of the cells of
this sense plate ganglion cells are cut off from the epithelial cells,
but continue to lie in the ectoderm (Figs. 28, 36, 56, 57). These
ganglion cells are small and I have been unable to observe their
further development, but there seems no reason to doubt that they
represent the cerebral ganglion.

A similar sense plate and ganglion is formed on the midventral
line immediately posterior to the place where the blastopore rem-
nant closed and in the region where the blastopore lips fused along
the mid-line (Figs. 29, 31, 35, 58, 61, 62, 6'6^). This is the ventral
sense plate, and the cells of this plate are pigmented as are those
of the apical plate ; I think it probable that they bear a tuft of long
cilia in life, although no one has observed this feature as yet. As
in the case of the apical plate, ganglion cells are cut off from the
basal ends of the epithelium of the ventral plate, and here again
ihere seems every reason to believe that these ganglion cells become
the subcesophageal ganglion. The oesophagus has not formed in
the oldest larva which I have been able to study, but a slight invag-
ination of the ectoderm immediately anterior to the ventral sense
plate probably represents the earliest step in the formation of the
oesophagus (Fig. 58, C?). Heretofore no observations have been made
on the early development of the nervous system. Neither Morse
nor Kowalevsky observed any stages in the formation of the ner-
vous system. Shipley has observed in the head region of Cistella
a small clump of cells without granules, which he suggests may be a
nerve ganglion ; his Fig. 35 shows however that it lies entirely
within the mesoderm, and it cannot therefore be a ganglion.
Brooks has described in detail the nervous system of the larva of

56 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

Glottidia, but this system is already well formed in the earliest
larva which he observed.

7. Enteron. — The cavity of the enteron is flask-shaped, the
enlarged end lying in the cephalic region while the pointed ex-
tremity extends into the peduncle. In early larval stages the
transverse diameter of the enteron is greater than its dorso-ventral
diameter (Figs. 25 and 44-50) ; in later stages it becomes circular
in cross section (Figs. 53-55), while in still later stages it becomes
compressed laterally so th'at its greater diameter is in a dorso-ven-
tral direction (Figs. 33, 35, 56-60). In the oldest stages which I
have observed a diverticulum from the anterior end of the enteron
grows out toward the ventral side, and at the same time an invagi-
nation of the ectoderm appears in the anterior mantle furrow, at the
very point where the blastopore remnant disappeared, and grows in
toward this diverticulum (Figs. 35 and 58). I presume that this is
the first step in the formation of the oesophagus. During the whole
of the larval period the enteron has no opening to the exterior.
According to Morse ('73; the mouth is formed late in the meta-
morphosis, and while he does not describe the place or manner of
its formation his Fig. 90 shows it at the anterior extremity of the
young brachiopod. Kowalevsky (' 74) also has described and figured
what he considers to be the formation of the oesophagus at the
anterior end of the cephalic lobe. From what I have observed I
feel confident that the mouth is first formed on the ventral side, in
the region of the anterior mantle furrow, and if it is later found at
the anterior end of the young brachiopod, its change in position
must have been brought about by secondary changes. Kowalevsky
expresses some doubts as to whether the invagination observed by
him at the anterior end of the cephalic lobe is really the oesopha-
gus, and from its location I would suggest that it is the apical sense
plate and cerebral ganglion rather than the oesophagus. My obser-
vation as to the location of the stomodaeal invagination thus con-
firms Heider's ('93) theoretical suggestion and brings the brachiopod
larva into close relationship with the trochophore.

The cells bounding the lumen of the enteron are cuboid in early
stages and columnar in later ones (compare Figs. 43-50 with Figs.
56-60). These cells, enclosing a minute lumen, extend through
the entire peduncular region (Figs. 2i^, 57, 58, 61-64).

From its earliest formation the enteron is in contact on its dorsal

1902.] CONKLIX — EMBRYOLOGY OF A BRACHIOPOD. 57

side with the ectoderm, while laterally and on its ventral side it is
bounded by mesoderm (Figs. 43-52, 58 and 61-63).

8. Coslom. — For a long time after the division of the ccelom into
right and left cavities by the flattening of the embryo and the
closure of the blastopore these sacs communicate with each other
both anteriorly and posteriorly (Figs. 43-47, 48-52, 53-55). In
still later stages these communications are closed by the practical
elimination of the coelom in the cephalic and peduncular regions
through the proliferation of mesenchyme cells (Figs. 56-63).
The coelom sacs, which are at first of nearly the same size
both anteriorly and posteriorly (Fig. 25), become much con-
stricted in the peduncular region while they still remain
large in the head and mantle regions (Fig. 26). In the latter
region they then become lobulated, often showing a trefoil condi-
tion (Figs. 27, 28), and with the further development of the dorsal
and ventral mantle folds a lobe of the coelom is sent into each of
these folds (Figs. 29-33). ^^^ the same time the coelom in the
cephalic and peduncular regions grows smaller, while that in the
mantle grows larger. Finally almost the entire coelom is con-
tained in the mantle, the portion in the head and peduncle being
very small (Figs. 2>^ and 56-63). The posterior limits of the
peduncular coelom is marked by a narrowing of the peduncle,
which probably represents the fourth ** segment" of Shipley. The
coelom however is never segmented though it may be constricted in
certain places. The constrictions shown in Figs. 27-31 are quite
constant in position and are connected with the extension of the
coelom into the mantle lobes, but they never coincide in position
with the superficial constrictions of the body (mantle furrows). In
a few abnormal larvae of the stage shown in Fig. 25 I have found
each coelom sac partially divided by mesenchyme cells into three
cavities. That these divisions, however, have no real importance
is shown by the fact that their number differs in different larvae and
is sometimes different on opposite sides of the same larva.

In the early larval stages the enteron is in close contact with the
ectoderm on the dorsal side, while a collection of mesoderm cells
on the ventral side of the enteron separates the two coelom
sacs and may be considered the rudiment of a ventral mesentery
(Figs. 49, 50).

In later stages the coelom is almost entirely obliterated, except
in the mantle, and consequently the enteron is surrounded by

58 CONKLIN— EMBRYOLOGY OF A BRACHIOPOD. [April 4,

a dense mass of mesoderm cells, except on the dorsal side, where
it is still in contact with the ectoderm (Figs. 56-64). At this
stage therefore there can be no mesenteries since there is practically
no coelom.

In early stages of larval life the mesoderm cells are mesenchy-
matous in the anterior region of the body and epithelial in the
posterior regions {^cf. Figs. 43-46). In the later stages the meso-
derm cells of the posterior regions become more mesenchyme-like
(Figs. 48-52 and 53-55), while in still later stages they become
densely packed and pigmented and it is impossible to distinguish
their cell boundaries (Figs. 56-64).

VI. General Considerations and Conclusions.

Although I am not one of those who expect to find phylogeny
'' writ large " in ontogeny, yet it may be worth while to point out
the bearings of the development of Terebratulina on the supposed
relationships of brachiopods. Since my own studies cover only
the embryonic and larval periods, I shall of course limit to those
periods the comparison of Terebratulina with other forms. Within
these periods we may compare (i) the cleavage, (2) mesoderm and
coelom formation, (3) orientation of embryo and larva, (4) the gen-
eral morphology of the larva.

I. The Cleavage. — As has been said already, there is no evidence
that the cleavage of Terebratulina resembles that of mollusks or
annelids. It is now known that in a great many annelids and
leeches and in all groups of mollusks except the cephalopods the
cleavage is of a certain determinate or morphogenetic (Child,
1900) type. The principal characteristics of this type of cleavage
are that the ectoderm is segregated in three quartettes of cells, that
the greater part of the mesoderm appears in one cell (4d) of the
fourth quartette, and that the remaining cells of the fourth quartette
together with the basal cells (macromeres) constitute the endoderm,
and finally that the elongation of the embryo takes place by the
teleoblastic cleavage of certain cells in the ectoderm and mesoderm
(first and second somatoblasts, 2d and 4d) and possibly also in the
endoderm (endodermic derivations of 4d). In addition to these
general characteristics of the cleavage of annelids and mollusks
there are other characteristics less general in application, such as
the derivation of the prototroch, the stomodaeum, the cerebral

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 59

ganglia, etc., from certain cells of the ectoderm. Wilson ('99) has
shown that many of the general characteristics mentioned above
are applicable also to the cleavage of the Turbellaria.

With regard to some of these general characteristics it is im-
possible at present either to affirm or deny their presence in Tere-
bratulina. For example, owing to the difficulty of identifying
individual cells I cannot say whether or not the ectoderm is segre-
gated in three quartettes. In fact I am wholly unable to recognize
quartettes at all after the eight-cell stage. It is certain, however,
that the mesoderm is not formed by teleoblastic growth from a sin-
gle cell and that the embryo does not grow in length by the teleo-
blastic cleavage of two somablasts. Furthermore the cleavage of
Terebratulina shows no resemblance to any type of determinate or
morphogenetic cleavage which has yet been described, whether
among annelids, moUusks, turbellarians, nematodes or ascidians.
On the other hand it does resemble in some respects the indeter-
minate cleavage of echinoderm.s, Phoronis and ectoproctous Bryozoa.

2, Mesoderm and Cosio?n. — The gastrulation occurs by typical
invagination ; however, this method of gastrulation is found in
almost every great group of animals, and therefore no phylogenetic
significance can be attributed to it. In the formation of the coelom
however the case is somewhat different. The method of mesoderm
and coelom formation in Tei-ebratulina is totally unlike that which
is found in annelids, mollusks, platyhelminths, nematodes and
arthropods, while it shows certain resemblances to chaetognaths and
echinoderms. A more detailed comparison shows however that
even these resemblances are not very close.

In echinoderms the enterocoel is formed at the inner end of the
archenteron, while the enteron arises from that portion of the arch-
enteron nearest the blastopore; in brachiopods the enteron is
formed from the inner end of the archenteron, while the enterocoel
arises from that part of the archenteron which in echinoderms
gives rise to the enteron. It is evident therefore that no real
resemblance exists between echinoderms and brachiopods in this
respect.

In chaetognaths the method of coelom formation is more like that
in brachiopods— in fact Ko\valevsky supposed that the two were
identical— and yet there are important differences here also. In
Sagitta, according to both Kowalevsky ('71) and Hertwig ('80),
two bilateral folds of the archenteric wall grow into the archen-

60 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

teron from its apex, thus dividing the cavity into a median one, the
enteron and two lateral ones, the enterocoels. The stomodseum
is said to open into the apical end of the enteron, directly opposite
the blastopore, which has however closed at an earlier period. In
Terebratulina, on the other hand, the enteron is separated from the
single enteroccel by one partition which grows out from the anterior
wall of the archenteron and divides the latter into a dorsal cavity,
the enteron, and a ventral one, the enteroccel ; only later, by the
closure of the blastopore and the flattening of the embryo, does the
enteroccel become divided through its middle region into right and
left cavities, which, however, continue for some time to communi-
cate with each other both anteriorly and posteriorly. Moreover the
stomodaeum in Terebratulina is formed in the position of the for-
mer blastopore and not on the opposite side of the embryo. There-
fore, although there are certain general resemblances between the
two, I cannot regard the coelom formation in chaetognaths and
brachiopods as being more than analogous processes, and as such
devoid of phylogenetic significance.

In Phoroms, according to Caldwell ('85), two pairs of coelomic
cavities are formed by a modified type of enteroccel formation,
which however bears no resemblance to that in Terebratulina.
The anterior one of these cavities gives rise to the cavities of
the epistome and lophophore, the posterior one to the chief body
cavity.

Masterman's (1900) work on the development of Phoronis fur-
nishes the most complete account of the early development of this
interesting form which has yet been given. ^ In an earlier work
('97) he found that there were three separate and distinct coelomic
cavities in the larva, a preoral or epistomal cavity, a collar or lopho-
phoral cavity and a trunk cavity. In his later paper he describes
the origin of these cavities ; the first of these arises as a median
outgrowth from the anterior side of the archenteron, the other two
arise as paired masses of solid mesoderm cells in which cavities
appear later. In the matter of the formation of the anterior or
procoelomic cavity there is much resemblance between Phoronis and
Terebratulina^ but in the latter animal the mesocoelomic and meta-
coelomic cavities are entirely absent.

1 Since the appearance of Ideka's (1901) work this statement is no longer
true. Ideka has given by all odds the most complete account of the embryology
of Phoronis yet published. (See postscript, p. 70.)

1902.J CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 61

Very little is certainly known of the formation of the mesoderm
and coelom among the Bryozoa. In the Ectoprocta the larval form
is usually solid, the coelom and enteron having undergone extensive
if not complete degeneration, while in those forms in which the
coelom is still preserved its method of origin is highly peculiar.
For example, in the Phylactolaemata the central cavity of the em-
bryo is generally considered to be the ccelom, although at the stage
at which it appears there is neither endoderm nor enteric cavity in
the embryo (see Korschelt and Heider's Text-Book).

Among the Endoprocta the mesoderm arises in Pedicellina,
according to Hatschek, from two pole cells which appear at the
posterior edge of the blastopore ; these cells by repeated divisions
give rise to two short mesoderm bands, and from these bands meso-
derm cells arise which fill the space between the ectoderm and the
endoderm. It appears therefore that no direct comparison can be
made between Terebratidina and the Bryozoa in the matter of the
formation of mesoderm and coelom,

3. Orientation of Embryo and Larva. — As has been pointed out
already the relation of the chief axis of the gastrula to the chief
axis of the larva is the same as is found in all Heteraxonia (Hat-
schek) or Hypogastric forms (Goette). The animal pole of the
tgg and the apical pole of the gastrula become the cephalic pole
{Hirnfeld) of the larva, while the blastopore comes to lie on the
ventral side. Such a relation of the embryonic and larval or adult
axes is of very general occurrence, being found at least in all Tro-
chozoa (Hatschek). Moreover in having a blastopore which
becomes narrow from side to side and then closes from behind for-
ward, and also in the formation of the stomodaeum at the place
where the blastopore closed, Terebratidina agrees with a large
number of bilateral animals belonging to widely different phyla.
These characters are so general, therefore, as to be of little value in
determining the affinities of the brachiopods. Regarding the
apical sense plate as anterior and the suboesophageal sense plate as
ventral in position, it follows that the peduncle is posterior and the
mantle folds dorsal and ventral ; consequently even after the meta-
morphosis the peduncle is posterior, and the valves which are
formed by the mantle lolds are dorsal and ventral, while the opening
of the valves is anterior. This orientation is the commonly
accepted one and is directly opposed to Caldwell's ('82) remark-
able views, according to which both valves and the peduncle are
ventral in position.

62 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

4. General Morphology of Larva. — The resemblances between
the early embryos of Terebratiilina and those of other meta-
zoa are so general in character that they afford little assistance in
determining the affinities of the brachiopods. We must therefore
rely largely upon the structure of the larva and of the adult for the
solution of this problem.

(a) Comparison with Trochophore.

Among the chief characteristics of the trochophore larva, as
enumerated by Hatschek {^ZZy p. 307), are the following :

I. (i) Bilateral symmetry, (2) mouth on ventral side, (3)

anus near posterior end, (4) shape ovoid.

II. (5) Apical tuft of cilia, (6) preoral ciliated band

(Trochus), (7) postoral ciliated band (Cingulum),
(8) adoral ciliated zone, (9) ventral ciliated furrow be-
tween mouth and anus, (10) small cilia over general
surface of larva.

III. (11) Epithelial nervous system, (12) apical plate {Schei-
telplatte), ganglion and sense organs, (13) oesophageal
nerves and buccal (ventral) ganglion, (14) ventral (and
sometimes dorsal) longitudinal nerves.

IV. (15) Alimentary canal (oesophagus, stomach and intes-
tine) horseshoe-shaped and ciliated throughout, (16)
stomach retort-shaped, (17) intestine reaches to poste-
rior end of body.

V. (18) Mesoderm partly mesenchymatous, partly epithel-

ial, (19) mesenchyme gives rise to branched connective
tissue cells and thread-like or branched muscle cells,
(20) ventral and dorsal longitudinal muscle pairs, (21)
preoral and postoral ring muscles, (22) dilatators and
constrictors of oesophagus and intestine, (23) meso-
thelium gives rise to the paired protonephridium, which
is a longitudinal ciliated tube closed at the anterior end
by terminal cells, and opening posteriorly on the ven-
tral side in front of the anus, (24) paired coelomic sacs
at the posterior end.
Of these characteristics, numbers 5, 6, 7, 8, 11, 12, 13, 15, 16,
17 and 23 are undoubtedly the most important, and all of these
except the last are found also in brachiopod larvae. All inves-
tigators of the embryology of brachiopods have described the
apical tuft of cilia (5) ; the preoral ciliated band (6) is probably

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 63

repr^ented in the brachiopod larva by the circle of longer cilia
found at the base of the preoral lobe (see Kowalevsky's figures of
Argtope \bx\^); the postoral ciliated band (7) is probably repre-
sented by the mantle, which attains a much greater development in
the brachiopod larva than in the ordinary trochophore ; the adoral
ciliated furrow (8) is represented by the anterior mantle furrow ;
the fact that adult brachiopods have an epithelial nervous system
(11) has long been known, and I have observed the beginnings of
such a nervous system in the larva of Terebratulina ; I have also
observed (p. 55) the apical plate and ganglion (12) and the ventral
plate and ganglion (13) in Terebratulina ; only the beginnings of
the oesophageal invagination are shown in the oldest larva which I
have studied, but it is evident from its position that after it joins
the stomach the alimentary canal will be horseshoe-shaped (15) ;
the stomach in Terebraiulina larvae is retort-shaped (16) and the
intestine reaches to the posterior end of the larva (17).

In addition to these more important characteristics the larvae of
Cistella, Thecidium or Terebraiulina agree with the trochophore in.
the possession of the following characteristics : Bilateral symmetry
(i), blastopore and anlage of mouth on ventral side (2), ventral
blastopore groove (9) and general ciliation of body (10), mesoderm
partly mesenchymatous, partly epithelial (18) ; finally ventral and
dorsal longitudinal muscles (20) are present in the peduncle of
Cistella. These brachiopod larvae also agree with many chaetopod
larvae in the possession of ectodermal seta sacs and provisional
setse.

The points in which the larvae of testicardinate brachiopods
differ most from the trochophore are in the absence of mouth and
anus and the lack of a protonephridium. In the Ecardines how-
ever both mouth and anus are present during larval life, and in all
brachiopods a single pair of nephridia appears after the larval
period. The absence of these larval structures therefore indicates
a retardation or less perfect development of the larval brachiopod
as compared with the typical trochophore. I believe therefore
that the brachiopod larva belongs unquestionably to the trocho-
phore type.

(b) Cofnparison with Actinotrocha.

The larva of Fhoroms, while showing many peculiarities, bears
a most decided resemblance to the trochophore. Among its prin-
cipal characteristics may be enumerated the following :

64 CONKLIN — EMBRYOLOGY OF A BRACHTOPOD. [April 4,

I. (i) There are three sections of the body: (a) the pre-

oral lobe, (b) the postoral section (collar) which carries
tentacles, and (c) the posterior or anal section (trunk);
(2) the preoral lobe probably represents the umbrella of
the trochophore ; ( 3) the cilia at its margin probably
correspond to the preoral ciliated band (Trochus), while
(4; the postoral ciliated zone (collar) which carries the
tentacles probably corresponds to the postoral band
(Cingulum) of the trochophore; (5) this postoral cili-
ated zone (Cingulum) runs obliquely around the body,
being further posterior on the ventral than on the
dorsal side ; (^6) tentacles appear near the ventral mid-
line and fresh pairs are added dorsally.

II. (7) There is an epithelial nervous system, (8) an apical

plate, ganglion and (in some species) eye spots, and
(9) an oesophageal commissure.

III. (10) The coelom is composed of an anterior unpaired
cavity and two pairs of cavities posterior to this (Mas-
terman);^ (i 1} the anterior coelom sac arises as an entero-
coel, the posterior paired ones as schizocoels ; (12) there

; is one pair of protonephridia, which end blindly inter-

nally in connection with excretory cells.

IV. (13) There is a ventral invagination posterior to the
zone of tentacles and a peculiar metamorphosis by the
evagination of this invagination; (14) during metamor-
phosis the tentacles turn forward, and (15) the anus
comes to lie on the dorsal side of the mouth, the intes-
tine thus forming a loop.

Comparing now the larvae of brachiopods with the Actinotrocha
we find that, in addition to the general resemblances to the trocho-
phore which both show, there are the following special resem-
blances between the two : (i) In both brachiopods and Actinotro-
cha the postoral ciliated zone (Cingulum) is greatly enlarged and
runs obliquely around the body, being farther posterior on the ven-
tral than on the dorsal side. (2) In both cases this forms' the
mantle or lophophore, though the tentacles or cirri which are borne
upon it appear much earlier in Actinotrocha than in the brachiopod
larva. (3) Brooks has shown that in Lingula the ventral pair of

See postscript, p. 70.

1902. CONKLIN — EMBRYOLOGY OF A BRACIIIOI'OD. 65

tentacles appears first and that successive pairs of tentacles are
added dorsally, exactly as in Actinotrocha. (4) In the meta-
morphosis the mantle (lophophore) is turned forward over the pre-
oral lobe in exactly the same way in both cases. These are
extremely important resemblances, and in themselves lend support
to the view that Phoronis and the Brachiopoda are closely related.'

On the other hand, according to Masterman's ('97 and 1900)
recent work on Actinotrocha, there are certain important respects
in which Actinotrocha differs decidedly from the brachiopod larva :
(i) The ccelom consists of an anterior unpaired cavity and of two
pairs of cavities, one of which lies in the lophophore and the other
in the trunk region. The anterior unpaired cavity somewhat
resembles in position and method of origin the anterior portion of
the enterocoel of Terebratulina, but the lophophoral and trunk
cavities of Actinotrocha differ from the mantle and peduncular
coelom of Terebratulitia in that the latter are a part of the entero-
coel and are never completely separated from one another, whereas
in Actinotrocha they arise as schizocoels and are always separate.
(2) Actinotrocha also has rudiments, at least of a second pair of
nephridia. (3) It also has two endodermal outgrowths from the
anterior portion of the enteron, which are composed of large vacu-
olated cells, and are homologized by Masterman with the notochord
of the Hemichorda.^

I have had no opportunity of studying the later stages in the
development of the brachiopod, in which alone the two last-men-
tioned structures might be looked for, and cannot therefore deter-
mine whether there are real differences between the brachiopod and
Phoronis in these respects. With regard to the differences shown
by the coelom, one must bear in mind the fact that in the brachio-
pod larva the coelom almost entirely disappears, except in the man-
tle, and a segmentation of the coelom in later stages could not
therefore be observed, even if it had at one time existed in the
ancestors of the brachiopods. There can be no doubt however
that in Terebratulina the entire coelom arises from a single entero-
coel, in which respect there is a decided difference between the
brachiopod and Phoronis. The resemblances mentioned above

1 The presence of " plasmic corpuscles " (Ideka, 1901) in the blastoccel of both
forms is another interesting resemblance (see p. 45).

2 See postscript, p. 70.

PROG. AMER. PHILOS. SOC. XLI. 168. E. PRINTED MAY 6, 1903.

66 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

however are so important and extend to such details that I am
inclined to accept the view that Phoronis and the Brachiopoda are
related, and to look to future work on the development of both of
these groups to harmonize the apparent differences between them.

(c) Comparison with Larval Folyzoa,

Brooks in particular has emphasized the resemblance between the
larvae of Polyzoa and Brachiopoda, basing this comparison, how-
ever, rather upon the external characters in which both resemble
the trochophore than upon a detailed comparison of internal
structure.

Ectoprocta. — It is extremely difficult to compare larval brachio-
pods with larval ectoprocts, owing to the great variety of forms
presented by the latter, their many secondary characters, and the
conflicting accounts of their structures and homologies which have
been given by various authors. There is some reason for believing
however that the ectoproct larva belongs to the trochophore type,
and that the following parts of the two may be homologous: (i)
The retractile disk may correspond (at least in part) to the apical
plate, (2) the corona in part to the trochus, (3) the sucker to
the trunk of the trochophore, or to the ventral evagination of
Actinotrocha. Furthermore one may trace a certain resemblance
between the invaginated sucker of Bugula and Lepralia and the
peduncle and mantle of Terebrafidiiia. In both cases attachment
takes place by the peduncle, while the covering folds (mantle in the
case of brachiopods) are turned forward as the peduncle is protruded.
However the degeneration and modification of structures, both in
the larval stages and in the metamorphosis, are so extreme that any
attempt at the present time to trace homologies between larval
Ectoprocta and other forms must be accompanied by a lively
imagination and a ready facility in guessing.

There is good evidence in the degeneration of the intestine and
coelom of the ectoproct larva, and in the general degeneration
which accompanies its metamorphosis, that we are dealing with a
highly modified type of development, which is little likely to throw
light upon the affinities of the Polyzoa. However the resemblances
between the adult Polyzoa and Phoronis and the Brachiopoda is
such as to warrant the conclusion that these groups are at least
remotely related to one another.

Enfoprocta. — Among larval entoprocts there are few, if any,
undoubted homologies with either the trochophore, the actinotroch,

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 67

or the brachiopod larva. It is possible that the ciliated disk of
Pedicellina and Loxosoma is homologous with the retractile disk of
the ectoproct larva and with the apical plate of the trochophore,
and that the margin of the vestibule (ciliated ring) in the former
corresponds to the trochus of the latter, but these possible homol-
ogies are too hypothetical to be affirmed with any degree of
assurance.

5. Conclusions. — Neglecting the older views as to the affinities of
the brachiopods with lamellibranchiate Mollusca, which were
founded merely upon superficial resemblances, we find that within
recent times the brachiopods have been associated, at different
times and by different authors, with Chaetopoda, Polyzoa, Chce-
tognatha and Phoronis.

Both Morse ('73') and Kowalevsky ('74) independently reached
the conclusion that the brachiopods are chaetopod annelids. Morse
says in summing up his work on the subject ('73', p. 57): "We
must regard the brachiopods as ancient cephalized chcetopods, while
Serpula, Amphitrite, Sabella, Protula and others may be regarded
as modern (later) cephalized chcBtopods''; and Kowalevsky ('74)
maintained that the brachiopods ought to be considered simply as
an order of the annelids, which present at least as many resem-
blances to the chaetopods as do the leeches.

Morse has enumerated twenty-four characteristics in which brachi-
opods resemble more or less closely Vermes, sedentary annelids and
Gephyreans. Kowalevsky also names a considerable number of
points in which brachiopods resemble chaetopods. Some of these
features are not actually characteristic of the brachiopods, as, for
example, the segmentation of the larva ; others are of such a gen-
eral character as to apply to almost all Bilateralia, as Brooks has
shown, while still others represent real resemblances betv/een the
brachiopod larva and the trochophore. The trochophore larva
however is of such wide occurrence among bilateral animals, that
the mere classification of the brachiopods among the Trochozoa
throws no light upon the nearer affinities of this group.

Huxley, Lankester, Claus and others have regarded the brachio-
pods as more or less closely related to the Polyzoa, and Brooks in
particular has held that the two groups belong to the same phylum
and class. ''The organization of the Lingula larva," he says,
"shows that it is not merely like a Polyzoon, but that it actually is
one; as much so as the hydra stage of an Hydro-Medusa is a

68 CONKLIX — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

Hydra, or the tailed larva of Botryllus is an Appendicularia, and
more so than a tadpole is an urodellan Batrachian." This close
relationship he bases largely upon the external resemblances be-
tween the larvae of Thecidium and various Polyzoon larvae. It
seems to me that some of these resemblances are real homologies,
but on the other hand the differences between these larvae, as well
as between the adults of these two groups, are so great that it
would be inadvisable to place them together in the same class ;
though I believe they should be placed in the same phylum.
Moreover it seems to me that Brooks' view, that the Polyzoa are
the ancestral form of which the Brachiopoda are a specialization,
is just the reverse of the real relationship ; larval as well as adult
Brachiopoda show less specialization and certainly less degeneration
than the Polyzoa.

The resemblances of the brachiopod larva to the MoUuscan
veliger, upon which Brooks lays emphasis, are in the main the same
as the resemblances to the trochophore, the veliger and trochophore
belonging to the same type of larva.

The idea that the brachiopods are related to the chsetognaths,
which was suggested by Blitschli and Hertwig ('80) and maintained
by van Bemmelen ('83), has little more in its favor than the sup-
posed resemblance in the method of formation of the coelom and
in certain histological details.

So far as the formation of the coelom is concerned, I have
already pointed out the fact that in Terebratulina it forms in a very
•different manner from what obtains in Sagitta, and as for the histo-
logical resemblances they are by no means confined to the two
groups in question. On the other hand there are so many im-
portant differences between the two groups, both in their embryology
and in their adult structure, that one could as well maintain the
•affinity of the Brachiopoda with Echinodermata, Enteropneusta or
Chordata, as with Chsetognatha.

Caldwell ('82) first pointed out in detail the resemblances be-
tween Fhoronis and the Brachiopoda. In this paper he has urged
*'an entirely new view of the homologies of the body surfaces in
Brachiopoda." He regards the Brachiopoda as fixed by their ven-
tral surface, and both valves of the shell as ventral in position, the
peduncle of the brachiopod corresponding to the ventral invagina-
tion of Actinotrocha. While there are some facts which may be
urged in favor of this view there are many which may be used

1002.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 69

against it. The fact that in both Phoronis and Lingula the intes-
tine forms a loop, the anus opening near the mouth, and that fixa-
tion takes place by the posterior extremity, has led to Caldwell's
view as to the homologies of the body surfaces in the brachiopods.
On the other hand the ventral mantle fold of Terebratulina forms
directly across the region where the blastopore lips fused and imme-
diately posterior to the place where the blastopore remnant closed and
where the mouth later appears. Upon the anterior face of this fold
the suboesophageal sense plate and ganglion appear; there can be
no doubt therefore that this fold is ventral in position. The dorsal
mantle fold appears at a very early stage (Figs. i6 and 20) on the
apical side of the gastrula and just posterior to the chief gastrula
axis ; it is impossible therefore that it should be considered as ven-
tral in position. Furthermore the mantle folds of the brachiopod
correspond to the zone which bears the tentacles in Actinotrocha (col-
lar, Masterman) and not to the margins of the ventral invagination ;
and since the mantle folds surround the body posterior to the
mouth, both of them cannot be ventral in position. Whether the pe-
duncle is ventral or not cannot perhaps be determined with certainty
until we know the embryology of a brachiopod in which the anus
and terminal portion of the intestine are present. In Lingula, as
is well known, the anus opens near the mouth and on the left side ;
in Crania it is terminal in position, and the embryology of either
of these forms should throw light on this question as to the mor-
phological position and homologies of the peduncle. Brooks'
work on Lingula deals only with stages in which the anus and the
intestinal loop are already present, and one cannot therefore tell at
what point relative to the blastopore the anus appears and how the
loop is formed. It is certain however that the ventral invagination
and remarkable metamorphosis of Actinotrocha are ccenogenetic
rather than phylogenetic characteristics, and that parallel phe-
nomena need not be expected in other groups of animals. Further-
more it is certain that the peduncle in Terebratulina is derived
from that portion of the gastrula which is posterior to the blasto-
pore ; I do not see therefore how it can be regarded as ventral in
position. But however this problem of the relation of the pedun-
cle of the brachiopod and the ventral invagination of Actinotrocha
may be decided by future work, it seems to me that the affinities of
Phoronis and Brachiopoda are well established. I agree therefore
in the main with the views of Caldwell, Lang and Blochman, and

70 GONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

more particularly with the position of Heider, as to the affinities of
the brachiopods. The relationship between Phoronis, Bryozoa
and Brachiopoda seems to me sufficiently close to justify the placing
of them in the same phylum, though not in the same class, as Lang
has done.

Postscript.

Since this paper was written I have seen Ideka's (1901) very
important contribution on the '* Development, Structure and Meta-
morphosis of Actinotrocha." Ideka's work is in all respects the
most thorough and extensive which has yet been done on the devel-
opment of Actinotrocha, and in many very important points he
differs decidedly from Masterman. Some of the differences be-
tween Actinotrocha and the brachiopod larva, which are pointed
out on p. 65, disappear in the light of this work. For example,
Ideka finds that there is but one complete septum in the body, that
between the collar and the trunk, while the cavities of the preoral
lobe and collar are in communication through a very incomplete sep-
tum. Furthermore Ideka finds no trace of a second pair of neph-
ridia or of a ** proboscis pore," such as Masterman described,
while the two '* chorda" diverticula of Masterman (Diplochorda)
are represented by a single unpaired diverticulum in the Japanese
species. Whether this is a glandular or skeletal structure is left an
open question.

With the exception then of the single septum between the collar
and the trunk regions there are no important differences between
Actinotrocha and the brachiopod larva. This septum occupies a
position in Actinotrocha corresponding to the posterior mantle
furrow of the brachiopod larva, and it would be interesting to
know whether, in stages of the brachiopod larva later than those
which I have studied, any trace of a septum can be found in this
position.

Literature List.

Beech ER, C. E. ('91). Development of the Brachiopoda. Am Jour. Set.

Vol. 41.
Beecher, C. E. ('92). Development of the Brachiopoda. Am. Jour. Sci.,

Vol. 44.
Beechkr, C. E. ('93). The Development of Terebratulina obsoleta. Trans.

Conn^ Acad., Vol. 9.
Blochmann, F. ('92 and 1900). Untersuchungen ilber dm Ban der Brachio'

poden. Jena.

1902.1 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 71

Brooks, W. K. ('78). The Development of Lingula and the Systematic Posi-
tion of the Brachiopoda. Johns Hopkins University, Chesapeake Zoologi-
cal Laboratory, Scientific Results of the Session of 1S7S.

Caldwell, W. H. ('82). Preliminary Note on the Structure, Development and
Affinities of Phoronis. Proc. Roy. Soc. London, Vol. 34.

Caldwell, \V. H. ('85). Blastopore, Mesoderm and Metameric Segmentation.
Quart. Jour. Mic. Sci., Vol. 25.

Child, C. M. (1900). The Early Development of Arenicola and Sternaspis.
Arch. Entzvick. Mech., Bd. 9.

Davidson, T. ('86, "^'j, '88). A Monograph of Recent Brachiopoda. Trans.
Linncean Soc. London, 2d Series, Vol. 4,

Hancock, A. ('58). On the Organization of the Brachiopoda. Trans. Roy.
Soc. London, Vol. 148.

HatsCHEK, B. ('88). Lehrbuch der Zoologie. Jena.

Heider, K. ('93). Chapters on Phoronis, Ectoprocta and Brachiopoda in Kor-
schelt and Heider's Lehrbuch der vergleichenden Entwicklungsgeschichte
aer -ii-irbellosen Thiere, Drittes Heft. English Translation of Vol. II by
Matilda Barnard ('99).

Hertwig, O. and R. ('80). Die Coelomtheorie.

IDEKA, IwAji (19GI). Observations on the Development, Structure and Meta-
morphosis of Actinotrocha. Jour. Coll. Sci. Ltnp. Univ. Tokio, Vol. 13.

KowALEVSKY, A. ('71). Embryologische Studien an Wurmern und Arthro-
poden. Mein. VAcad. Itup des. Sci. de St. Petersbourg, 7th Series, T. 14,

KowALEVSKY, A. ('73)- On the Development of the Brachiopoda. Cong.
Russian Naturalists, Kassan, 1873; Sec. Anat. Phys. Zool. and Comp.
Anat. (Russian Ref. by Hoyer in Jahresber. ueber die Fortschritte der
Ajiat. 2ind Phys., Bd. 2, 1873).

KowALEVSKY, A. ('74). On the Development of the Brachiopoda. Lzvyest.
imp. Obshch. Lynbit. Estestv. Anthrop, i. Ethnog., Vol. 14. Moscow.
(Russian. Abstract by Alex. Agassiz in Am. Jour. Sci., '74, and by Oehlert
et Denicker in Arch, de Zool. exp., 2d Series, T. 1, ^Zt,.)

Lacaze-Duthiers, H. de ('6i). Histoire de la Thecidie. Ann. Sci. JVat .
4th Sen, T. 15.

Lang, A. ('88). Lehrbuch der vergleichenden Anatomie, Heft 1. Jena.
English Trans, by Henry M. and Matilda Bernard ('91).

Masterman, a. T. ('97). On the Diplochorda. Quart. Jour. Mic. Sci.,
Vol, 40.

Masterman, A. T. (1900).

Morse, E. S. ('71). On the Early Stages of Terebratulina septentrionalis.
A/e?n. Boston Soc. Nat. Hist., Vol. 2.

Morse, E. S. ('73). On the Embryology of Terebratulina. Mem. Post. Soc.
Nat. Hist., Vol. 2.

Morse, E. S. ('73^). On the Systematic Position of the Brachiopoda. Proc.
Boston Soc. Nat. Hist.,Yo\. 15.

Mueller, F. ('60). Beschreibung einer Brachiopodenlarve. Milller's Archiv
fur Anat. und Phys.

72 CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

Mueller, F. ('6i). Die Brachiopodenlarve von St. Catharina, Zweiter Beitrag.

Archiv filr Naturgesch.
Shipley, Arthur E. ('83). On the Structure and Development of Argiope.

Mittheil. ZooL Stat. Neapel, Bd. 4.
Van Bemmelen, J. F. ('83). Untersuchungen liber den anatomischen und his-

tologischen Bau der Brachiopoda Testicardines. Jenaische Zeit. f. Natur-

wiss, Bd. 16.
Wilson, E. B. ('99;. Cell Lineage and Ancestral Reminiscence. IVcods Noll

Biological Lectures for 1898.

Reference Letters.

Be, Blastoccel.

Bf, Blastopore groove.

Bp, Blastopore.

Br, Blastopore remnant.

C, Coelom.

Ce, Enteroccel.

Cc, Cephalic Coelom.

Cm, Mantle Coelom.

Cp, Peduncular Coelom.

Cent, Cephalic and Mantle Coelom.

Cmd, Dorsal extension of Mantle Coelom.

Cmv, Ventral extension of Mantle Coelom.

CG, Apical Sense Plate and Cerebral Ganglion.

D, Dorsal.

DA, Dorsal-Anterior.

E, Enteron.

Em, Egg Membrane (ectoplasmic layer).

Fa, Anterior Mantle Furrow.

Fp, Posterior Mantle Furrow.

Gc, Gastrocoel.

M, Mantle.

iMd, Dorsal Mantle Fold.

Mv, Ventral Mantle Fold.

mc, Mesenchyme.

O, Point where blastopore closed and oesophageal invagination appears.

P, Peduncle.

Pb, Polar Bodies.

PC, Peduncular Chamber.

Pg, Dark Staining Granules.

SG, Ventral Sense Plate and Sub-oesophageal Ganglion.

SS, Setae Sacs.

V, Ventral.

VP, Ventral-Posterior.

1902.] CONKLIN — EMBRYOLOGY OF A BRACHIOPOD. 73

Description of Figures.

All the figures illustrating this paper were drawn with Camera Lucida at the
stage level under a Zeiss Apochromat. Homog. Immers. Obj. 3mm. Comps. Occ.
4. In the process of reproduction they have been reduced about one-third.

Plate I.

Fig. I. One-cell stage ; polar body in process of being formed; egg elliptical ;
ectoplasmic layer surrounds egg and polar body.

Fig. 2. Two-cell stage; one cell larger than the other; three polar bodies
present.

Fig. 3. Transitional stage between two-cell and four-cell stages ; showing over-
lapping of certain cells and « spiral " character of cleavage.

Fig. 4. Four-cell stage, two cells (at right) larger than the other two; proto-
plasmic areas surrounding nuclei shown ; two polar bodies lie in the polar
furrow.

Fig. 5. Eight-cell stage, apical view; three polar bodies at animal pole.

Fig. 6. Eight-cell stage, side view ; one polar body at animal pole.

Plate II.

Fig. 7. Eight-cell stage, apical view; each cell indented at periphery; polar

furrows at right angles to each other at opposite poles.
Fig. 8. Seven-cell stage ; irregular cleavage ; animal pole indicated by three

polar bodies.
Fig. 9. Sixteen-cell-stage, apical view ; two polar bodies at animal pole.
Fig. 10. Sixteen-cell stage, optical section, showing blastocoel, ectoplasmic layer

{^Em') and polar body or yolk spherule between two of the cells.
Fig. II. Tiventy-cell stage; some of the cells elongated and probably dividing

yolk spherules within blastoccel.
Fig. 12, About forty-eight-call stage; egg flattened by pressure and the blasto-

meres partially separated.

Plate III.

Fig. 13. Early invagination stage, optical section.

Fig. 14. Gastrulation completed; blastoccel obliterated; gastroccel partially

divided into enteron {E) and enteroccel (6V).
Fig. 15. Optical section of older embryo, viewed from posterior; enteron still

further constricted from enteroccel.
Fig. 16. Optical section of embryo of same stage as preceding, lateral view,

showing anterior extension of enteroccel and partition wall growing down

on anterior side between enteron and enteroccel.

74 COJ^KLIN— EMBRYOLOGY OF A BRACHIOPOD. [April 4.

Fig. 17. Oral view of embryo, showing elongated blastopore opening at its
anterior end into the enterocoel and enteron; the enterocoel but little larger
than the enteron.

Fig. 18. Antero- ventral view of an embryo of about the same stage as the
preceding.

Plate IV.

Fig. 19. Optical section of older embryo, viewed from posterior; showing the
enteron separated from the enterocoel.

Fig. 20. Optical section of an embryo of same stage as preceding, lateral
view ; showing the enteron almost entirely separated from the enterocoel.

Fig. 21. Oral view of an embryo of about the same stage as the preceding; the
blastopore narrower than in preceding stages ; the enteron completely
cut off from the enterocoel except in the region of the blastopore ; mesen-
chyme cells {mc) line the anterior part of the enterocoel.

Fig. 22. Oral view of an older embryo in which the blastopore has closed to a
narrow groove except for a small opening near its anterior end ; mesenchyme
cells are abundant in the anterior and posterior parts of the enterocoel ; man-
tie folds show at the sides of the embryo.

Fig. 23. Aboral view of an embryo of the same stage as the preceding, showing
dorsal mantle fold {M).

Fig. 24. Optical longitudinal section of an embryo in which the blastopore has
completely closed, leaving however on the ventral surface the blastopore
remnant {Br) and groove; the dorsal mantle fold {Md) and furrow are
shown on the dorsal side ; the enteron still communicates with the enterocoel
at its posterior end.

Plate V.

Fig. 25. Dorsal view of a larva in which cephalic, mantle and peduncular
regions are well defined.

Fig. 36. Ventral view of an older larva, showing the ventral mantle folds
meeting m the region of the blastopore groove [^Bf); the blastopore rem-
nant {Br) lies in a notch on the anterior side of the mantle fold.

Fig. 27. Dorsal view of a slightly older larva, showing the increased promi-
nence of the mantle and the lobulation of the coelom.

Fig. 28. Optical longitudinal section of a larva of the same stage as the pre-
ceding, showing the dorsal and ventral mantle folds {Md axidi. AIv) ; apical
sense plate and ganglion ; enteron and coelom.

Fig. 29. Ventral view of an older larva, showing the ventral mantle folds fused
in the midline ; the anterior and posterior mantle furrows are shown as
shaded lines ; in the anterior mantle furrow the place at which the blastopore
remnant disappeared and where the oesophageal invagination will occur is
marked ( O) ; the apical and ventral sense plates ( CG and SG).

1902.] CONKLIN— EMBRYOLOGY OF A BRACHIOPOD. 75

Fig. 30. Dorsal view of a larva of the same stage as the preceding, showing the
dorsal mantle fold {Md); the anterior and posterior mantle furrows; the
union between the two halves of the peduncular ccelom.

Plate VI.

Fig. 31. Lateral view of a larva of the same stage as the preceding, showing

anterior and posterior mantle folds {Fa and Fp)^ apical and ventral sense

plates ( CG and SG), enteron and ccelom.
Fig. 32. Dorsal view of older larva, showing increased prominence of mantle

and deep constriction of anterior mantle furrow.
Fig' ZZ' I^orsal view of an older larva, showing the mantle in process of

growing back over the peduncle ; the ccelom is almost entirely confined to

the mantle.
Fig. 34. Dorsal view of one of the oldest free-swimming larvae ; the mantle has

almost entirely covered the peduncle.
Fig. 35. Lateral view of a larva of the same stage as the preceding, showing

apical and ventral sense plates ( CG and SG), dorso-ventral extension of

enteron and greater width of mantle chamber on dorsal side.
Fig. 36. Optical section in longitudinal frontal plane of a larva of the same

stage as the preceding, showing apical sense plate and cerebral ganglion ;

mantle chamber and setae sacs ; cephalic, peduncular and mantle ccelom

[Cc, Cp, Cf?i).

Plate VII.

Fig. 37. Section of an embryo of the stage shown in Fig. 13, showing dark

staining granules in the outer ends of the cells and yolk spherules in the

blastocoel.
Fig. 38. Section of an embryo of the stage shown in Fig. 14.
Figs. 39-42. Four transverse sections of an eaabryo of the stage shown in Fig.

20 ; Fig. 39 being the most posterior and Fig. 42 the most anterior section

drawn.
Figs. 42r7 and 42^. Longitudinal sections through an embryo of about the same

stage as is shown in Figs. 39-42, showing the formation of the septum which

separates the enteron from the enterocoel.

Plate VIII.

Figs. 43-47. Five transverse sections of a larva of the stage shown in Fig.

24; Fig. 43 being the most anterior and Fig. 47 the most posterior section

drawn.
Figs. 48-52. Five transverse sections of a larva of the stage shown in Fig.

29; Fig. 48 being the most anterior and Fig. 52 the most posterior section

drawn.

76 CONKLIX — EMBRYOLOGY OF A BRACHIOPOD. [April 4,

Plate IX.

Figs. 53-55. Three oblique sections from the dorsal anterior (DA) to the ven-
tral posterior ( VP) region of a larva of the stage shown in Fig. 31 (the
sections are nearly in the plane of the reference line from £ in Fig. 31).

Figs. 56, 57. Two longitudinal frontal sections of a larva of the same stage as
the preceding, Fig. 56 being ventral to Fig. 57.

Fig. 58. Longitudinal sagittal section of a larva of the same stage as the pre-
ceding; the section passes through both the apical and the ventral sense
plates (CG and SG).

Plate X.

Figs. 59-64. Six transverse sections of a larva of the stage shown in Figs.
34-36; Fig. 59 being the most anterior and Fig. 64 the most posterior sec-
tion drawn. Fig. 59 passes through the cephalic region ; Fig. 60 lies just
behind the anterior mantle furrow ; Figs. 61-64 are through the mantle and
peduncle.

1902.] NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINX. 77

THE SPERMATOGENESIS OF ONISCUS ASELLUS LINN.»

WITH ESPECIAL REFERENCE TO THE

HISTORY OF THE CHROMATIN.'

BY M. LOUISE NICHOLS.

(Plates XI-XVIII.)

{Read April j^, 1902.)

This study was begun in the month ot February, 1899, i" order
to ascertain the mode of origin of the peculiar spermatozoa of the
land Isopods. I have now completed, so far as I am able at the
present time, the investigation undertaken for that purpose. Before
entering, however, upon a description of my observations, I wish,
at the close of a work which has proved both interesting and
instructive, to express my gratitude to my instructors. Prof. E. G.
Conklin and Prof. Thomas H. Montgomery, Jr., for the inspiration
and the many valuable suggestions which have aided me toward its
completion. To the latter I am particularly indebted for his help-
ful criticism concerning the earlier stages of the spermatogenesis.

Methods.

The material was fixed either in Flemming's fluid, Hermann's fluid
or in Gilson's fluid (acetic-nitric sublimate). It was stained for
the most part with iron haematoxylin, but for purposes of compari-
son also with saffranin and malachite green (Wilcox) (1895),
saffranin and gentian violet, Delafield's haematoxylin and Bordeaux
red, and with the Biondi-Ehrlich triple stain. The study of the
spermatozoa was also pursued by teasing apart the vas deferens
with needles, staining the fresh material with haematoxylin or with
acetic-methyl green, and mounting in glycerin. Permanent mounts
were also made of fixed and stained material. I consider Wilcox's
double stain with saffranin and malachite green to be a valuable
one, for the reason that it can be used with good effect on material
fixed in Flemming's fluid. It gives in reality a triple stain, for in
successful preparations the cytoplasmic structures stain green,
active chromatin, centrosomes and true nucleoli red, while resting
chromatin takes a purple color. Its chief disadvantage is that it
will in time fade.

iThe species was determined by means of the works of Budde-Lund (1885)
•and of Richardson (1900). The species is also known as 0. murarius, Cuv.

2 A thesis for the degree of Ph.D. at the University of Pennsylvania.

78 NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. [April 4,

I. Structure of the Male Reproductive Organs.

The male reproductive organs of the land Isopods are paired and
lie on either side of the tubular intestine, occupying almost the
entire length of the thoracic region.

The testis consists of three narrow lobes, which are attached to
the body musculature by slight strands of tissue. These lobes are
distinct from each other and open successively into the anterior
expanded portion of the vas deferens (Fig. i). Posteriorly the
vas deferens narrows to a more slender tube, which joins its fellow
of the opposite side and opens through the penis, which is unpaired
and is said by Gerstaecker (1882) to be an outgrowth from the
seventh thoracic segment. It is enclosed by the modified internal
lamellae of the first abdominal appendages (Fig. i).

Sections of the vas deferens (Fig. 3) show its expanded portion
to be lined with cells of large size, which possess prominent
spherical nuclei. The nucleus is sometimes surrounded by a clear
space, varying somewhat in size. The chromatin is in the form of
closely crowded granules. Between these are sometimes other
granules, which with the iron haematoxylin stain are less deeply
colored, and with Bordeaux red and Delafield's hsematoxylin take
a red tint. The periphery of the latter is usually darker in color.
The surrounding cytoplasm is filled with particles of a rounded
shape, which take cytoplasmic stains (Fig. 4a). In one prepara-
tion the cytoplasm of these cells was filled with particles, not
rounded in shape but thread-like, and taking a very dark stain with
iron hsematoxylin. The chromatin consisted of granules of varying
size, which appeared lighter in the centre and possessed a darker
margin (Fig. 4^). I do not know whether there is any connection
between the particles within the nucleus and those without ; the
subject might possibly repay further research. The appearance of
these cells suggests strongly that they have a secretory function ; no
doubt the fluid which bathes the spermatozoa is produced by them.
They are more abundant at those places where the follicles open
into the vas deferens and grow more scarce in the region where the
narrow portion of the vas commences (Fig. 3). Between them are
to be seen nuclei of smaller size, whose chromatin is not so
distinctly and regularly granular. These lie in a cytoplasmic reti-
culum of a coarse mesh, without well-defined cell boundaries and
containing no granules. This tissue apparently forms a supporting

1902.] NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. 79

membrane for the secretory cells. It is continuous with the layer
of cells which line the narrow portion of the vas deferens and is
similar to it in structure (Fig. 3). The narrow portion of the vas,
as a rule, is covered externally by dark pigment, thus forming a
marked contrast in the fresh state to the milk-white walls of the
anterior portion (Fig. i). Between the pigment layer and the
lining cells, delicate muscle-fibres are occasionally discernible.

The three lobes of the testis are seen in section to be three folli-
cles (Fig. 2). Each follicle is covered by a thin membrane which
is provided with delicate muscle-fibres (Fig. 5W, /). The margins
of the follicle are occupied by large nuclei of unsymmetrical outline,
containing irregular blocks of chromatin interspersed with finer
granules. Cell boundaries between these nuclei are not visible.
They can sometimes be seen to be undergoing amitotic division, of
a character similar to that described by vom Rath (1891) for
Astacus (Fig. 5/, c). In ibllicles of a certain stage of development
these nuclei, as will be explained later, are subject to degeneration.

The strands of tissue, by means of which the follicles are
suspended from the body wall, are made up of cells which also
divide amitotically and which are similiar in appearance to the
follicle nuclei, inasmuch as their outlines are irregular, but the
blocks of chromatin are of larger size and the nuclei are separated
from each other by distinct cell walls (Fig. 6).

The interior of the follicle, except during the migration of the
follicle cells, is occupied entirely by the germ cells, which are in
differing stages of development in the three follicles of one side.
Corresponding follicles of opposite sides of the body contain, how-
ever, germ cells which have developed to very nearly the same
degree.

Fig. 2 shows, in a typical case, the comparative degrees of devel-
opment to which the cells of the three follicles have attained. Each
follicle may be divided into two principal regions of growth, com-
posed of cells of different generations and of different degrees of
development. Thus, in the most posterior of the follicles (a), the
apical third is occupied exclusively by spermatogonia, some of
which can be seen in mitosis ; the basal region, on the other hand,
by spermatids in a not very advanced stage. Follicle cells occur
on the outside of the follicle, being especially abundant in the basal
region. In the adjacent follicle (^), the apical two-thirds is occu-
pied by cells in the synapsis stage, the remaining portion by sper-

80 XICHOLS— SPERMATOGENESIS OXISCUS ASELLUS LINN. [April 4,

matids in a stage of development later than that of follicle {a).
Along the margin of the follicle are found scattered small groups of
spermatagonia (Fig. 2, spg). The third and most anterior follicle
{c) contains chiefly spermatocytes in a late prophase. Groups of
spermatogonia similiar to those of follicle (^b) are here also found
scattered along the margin and nearly filling the extreme apical
portion. The follicle cells in the basal region are undergoing not
only active amitotic division, but to a certain extent degeneration
(Fig. 5). Their active multiplication or fragmentation causes them
to crowd in toward the axis of the follicle.

From a comparison of the extent of these growth regions in the
three follicles, the developmental cycle may be conceived somew^hat
as follows. The spermatozoa, when fully formed, are forced into
the vas deferens. Since they have no motion of their own, this is
probably caused by the contraction of the muscle layer of the folli-
cle, perhaps assisted by the pressure of the growing cells in the
apical region. During this process, the spermatogonia in the apical
portion of the follicle divide and come to fill the space left vacant
by the discharged sperm. The rate at which this replacement takes
place and the comparative development of the cells in the two
regions may vary in different follicles, for I have preparations in
which few or no spermatozoa are seen — in other words, most of them
had probably been discharged, and at the same time the replacing
spermatogonia are scattered and itw in number. In others, as is
shown in the diagram (Fig. 2a), the spermatozoa, in an early stage
of development, occupy the basal portion of the follicle, while the
apical portion is packed with spermatogonia. The majority of the
spermatogonia thus filling up the follicle proceed in their develop-
ment, while the remainder form the groups of cells along the margin
of the follicle already described in follicles {b) and (^), and which
are destined later to again supply a new generation of cells. The
spermatids also proceed in development and are forced into the vas
deferens. A condition like that represented in diagram {b) (Fig. 2)
thus arises — the basal region filled with spermatozoa in a late stage
about to pass into the vas deferens and the apical region with, cells
which have progressed as far as the synapsis stage. Later, the
spermatozoa having been completely discharged, the cells of the
apical region come to occupy the basal part of the follicle, being
now less compactly pressed together (Fig. 2c). Their development
progresses until, having become mature spermatozoa, they pass into

1902.] NICHOLS— SPERMATOGENESIS ONISCUS ASELLUS LINN. 81

the vas deferens, the spermatogonia again fill the apical region, and
the cycle is repeated.

The invasion of the follicle cells begins, as a rule, when the germ
cells are in an advanced prophase and may continue later. Many
of the germ cells likewise degenerate, and they, together with the
follicle cells, form a disintegrated mass in which the spermatids lie.
In young follicles, which have not as yet matured sperm, the basal
region is filled with follicle cells, the apical region with spermato-
gonia. This is sometimes true also of older follicles which have
recently discharged the sperm.

It will thus be seen that a series of stages, illustrating the com-
plete history of the changes through which the germ cell pass, can
be obtained only by an examination of numerous testes. Dupli-
cates are often obtained and some of the stages occur very infre-
quently, probably owing to a greater rapidity of development at
certain periods.

This study was begun in the latter part of February. In March
or April, according to the rigor of the weather, the land Isopods
in the vicinity of Philadelphia commence to breed. The breeding
season continues during the summer months. There are, in a single
year, several cycles of development of the reproductive elements ;
the exact number I have not determined. It is therefore pos-
sible, at almost any time of year, by examination of a sufficient
number of individuals, to procure a complete series of develop-
mental stages.

II. Spermatogenesis.
I. Spermatogonia.

The resting spermatogonia are distinguished from the follicle
cells by their smaller size, the distinctness of the cell walls, and by
the fact that in their nuclei the chromatin masses are of smaller
size and show indications of an arrangement into a network (Fig.
\\d). They possess a prominent true nucleolus of more or less
rounded form. Some cells contain one or more smaller nucleoli.

It is impossible to determine the exact number of spermatogonic
divisions. They are probably numerous, since it must require a
considerable number of divisions of the spermatogonia remaining
in the follicles to fill the space left vacant by the discharged sperm.
The cells vary somewhat in size. When the apical region of the
follicle is filled with spermatogonia the individual cells are small,
but when the follicle is not well filled and the spermatogonia are

PROC. AMER. PHILOS. SOC. XLI. 168. F. PRINTED MAY 7, 1902.

82 NICHOLS— SPERMATOGENESIS ONISCUS ASELLUS LINN. [April 4,

beginning the task of producing a new generation, individual cells
often equal the spermatocytes in size (Figs. lo and 13). In the
cytoplasm are occasionally seen irregular masses of a dull brown
tint (yolk ?), but neither sphere substance nor centrosomes are
apparent in the resting cell.

In nuclei preparing to divide, the chromatin is seen to be
arranged in the form of slender, elongated threads, which, so far as
I have been able to discover, in no case form a continuous spireme
(Fig. 7). In the cytoplasm surrounding a nucleus of this character
are visible two minute black specks joined by a delicate thready
presumably the centrosome undergoing division. The nuclear
membrane at this stage begins to fade. Figs. 8, 9 and 10 show
stages immediately succeeding the stage shown in Fig. 7. The
threads have become shorter and thicker, the nuclear membrane has
entirely disappeared, and the centrosomes have become more widely
separated. The amount of segmentation of the thread varies in
different cells. In Fig. iib is shown a nucleus in which very little
segmentation has taken place, although the thread is considerably
thicker than that shown in Fig. 7. The linin threads joining the
chromosomes are of extreme delicacy and difficult to discover.
Occasionally, however, (Fig. 9) fine fibres may be seen stretching
from one chromatin thread to the next, The shortened and thick-
ened chromosomes then arrange themselves into an equatorial plate
(Fig. 12). The appearance of the plate, both in side and in pole
view, is irregular. The division of the chromosome into chromo-
meres and their longitudinal division is visible only in very thin
sections, which have been stained with iron haematoxylin and
rather strongly decolorized (Fig. 14). The centrosomes and
spindle-fibres of the spermatogonic, mitotic figure are not quite so
prominent as those of the spermatocytic divisions. The same is true
of the polar radiations. Central spindle-fibres are apparently entirely
lacking. After splitting of the chromosomes the halves diverge, in
the manner of the two legs of a pair of compasses, the divergence
commencing at one end, while at the other end the two halves
remain in contact (Fig. 15).

A still later anaphase is shown in Fig. 17. The chromosomes
have become massed together, the spindle-fibres are beginning to
disappear and the centrosomes are almost lost to sight. The con-
striction of the cell body, observable to a slight degree at this
stage, becomes more marked and a membrane comes to separate
the two daughter cells (Fig. 19).

1902.] NICHOLS— SPERMATOGENESIS ONISCUS ASELLU3 LINN. 83

The reconstruction of the nucleus consists of the breaking up of
the chromosomes into fine granules, which are connected by linin
threads of great delicacy, and in the development of a nuclear
membrane (Figs. 19, 20, 21). The change in chemical composi-
tion of the chromatin is indicated in sections stained with saffranin
and gentian violet by a gradual change in color from red to blue.
As the cell body constricts slight thickenings are discoverable on
the connective spindle-fibres in the equator (Fig. 18), which, as the
constriction proceeds, grow fewer in number and more conspicuous
in size until they are finally reduced to a single large swelling, from
which radiate the spindle-fibres, by this time grown faint (Fig. 19).
At a stage a little later than the one just described I have occa-
sionally seen a small black body wedged in the angle between the
daughter cells (Fig. 20). Its appearance is similar to the " Flem-
mingscher K5rper " described by Hoffmann (1898) for Limax
maximus (see his Figs. 31, 32, 33) and strikingly like that of the
rabbit described by von Winiwarter (1900, Figs. 9 and 10).

2. Growth Period.

The anaphase of the last spermatogonic division is decidedly
different from that just described. The chromatin threads lie
massed together and entangled near the centre of the cell (synap-
sis). They are surrounded by a clear space bridged over by
slender acromatic fibres, which connect the chromatin threads
with a narrow layer of cytoplasm lying close to the cell wall. No
trace of centrosome or sphere substance (idiozome) is discoverable
(Fig. 22). The chromosomes now spread apart, although still
connected by strands of linin. They are seen to be for the most
part V-shaped. The chromatin granules are rather irregularly dis-
tributed, being frequently massed together in lumps (Fig. 24).

In a thin section of a cell at a stage slightly later than this there
appeared a minute black dot, surrounded by a vaguely defined
area, slightly more dense than the rest of the cytoplasm (Fig. 23).
I hesitate to attach importance to this, as it occurred in very few
cases.

The threads now elongate, and during this process the granules
of which they are made up divide, so that the thread becomes
longitudinally split. The granules apparently do not divide simul-
taneously. Even in the same thread some of them show division,
while others remain entire (Fig. 25). The split is to be seen with

84 NICHOLS— SPERMATOGENESIS ONISCUS ASELLUS LINN. [AprU 4,

the greatest clearness in sections stained with iron haematoxylin
and strongly decolorized. The chromosomes are very irregularly
distributed, only occasionally a part of them, six or seven, may be
grouped with reference to a central point. Of the entire number
of chromosomes present it is difficult to be certain, owing to
the fact that they overlie each other so closely. The number,
however, is certainly, less than that present in the spermatogonia
and not greater than sixteen (Fig. 26). The reduction in the
number of chromosomes, therefore, apparently takes place at this
stage, and the V-shape so prevalent is due to the approximation of
two chromosomes to form a single bivalent one. The place of
union is frequently covered by chromatin, but a connection of
linin can sometimes be discovered (Fig. 27). This figure also
shows the varying angle at which the univalent chromosomes may
approach each other. Occasionally they may even form a complete
ring.

The threads become more and more attenuated (Fig. 28), and
finally by anastomosis are transferred into the nuclear reticulum of
the resting spermatocyte (Fig. 30). During the elongation of the
chromosomes the chromatin granules divide and redivide (Figs.
33-28), so that they become very numerous, and as the elongation
progresses the longitudinal split becomes less easily discoverable,
until in the resting cell it can no longer be made out. Cells are
sometimes seen in which, just before the formation of the nuclear
membrane, the network lies to one side, being connected by slight
strands of linin with the surrounding cytoplasm (Fig. 31).

The fact that the chromosomes remain distinct until just before
the formation of the nuclear membrane points to a maintenance of
their individuality in the resting cell.

The nuclear membrane appears to form as a condensation of
achromatic substance, upon which later appear granules staining
deep blue with haematoxylin (Fig. 29).

A peculiar fact with reference to the last spermatogonic division
has struck my attention and I have been unable to explain it very
satisfactorily. It will be seen from Fig. 2 that nearly all the cells
in the apical portion of follicle (^b) are in the synapsis stage. It
might be supposed from this that sections would be obtained of
follicles filled apically with the spindles of the last spermatogonic
division. Such a condition, however, I have never found, although
I have examined a large number of testes at different seasons of the

1902.] NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. 85

year. The karyokinetic figures of the spermatogonia are always
scattered and it is impossible to distinguish between the early and
late ones.

5'. The Maturation Divisions.

In preparing for the first maturation division the meshes of the
nuclear network become coarser, the granules more distinct and
aggregated into separate threads, joined together by linin (Figs.
32-36). The manner of their origin again lends support to the
view concerning their individuality in the resting cell. A still
greater condensation of the granules leads to a shortening and
thickening of the chromosomes (Figs. 37 and 38), the final result
of which is the production of sixteen compact masses of chromatin,
still connected by linin threads (Fig. 44). Condensation does not
proceed at an equal rate in all the chromosomes of a nucleus.
Fig. 45<^ shows a small portion of a nucleus in which lie side by
side two chromosomes, in one of which the final dumbbell-shape is
almost completed, while in the other the condensation of the chro-
matin is but little advanced. These sixteen masses are of various
forms. Some are dumbbell-shaped, two spheres of chromatin
joined by linin ; some are crescent-shaped and still others are more
or less complete rings (Figs. 39-45). The different forms may
occur in the same nucleus, but apparently without constancy in the
ratio of relative frequency of occurrence. The dumbbell-shape,
straight or slightly curved, is abundant, some cells containing no
complete rings (Fig. 41). Other cells contain a comparatively
large number of rings or crescents (Figs. 39 and 40).

Two main types may be distinguished among the chromosomes
according to their structure and mode of origin — />., (i) those
in which the bivalent chromosome consists of two univalent chro-
mosomes lying end to end, as in those having the dumbbell-shape,
and (2) those in which the univalent chromosomes lie side by side,
as in those arising through a ring or narrow V-shape. A form in-
termediate between these is represented by those having a crescent-
shape. The different types and their probable mode of origin are
shown in the diagram (Fig. (>Zay d, c). It is interesting to note
that these types can be distinguished in the synapsis stage (Fig. 27),
although they are here not so well marked as in the prophases of
the first maturation division.

In cells stained with iron haematoxylin, which have been strongly

86 NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. [April 4,

decolorized, a longitudinal split is evident and likewise a division
of the chromosome into chromomeres. If the chromosome is of
the second type and seen from above, two of the chromomeres will
be seen longitudinally split (Fig. 46; cf. Fig. 53). An end view of
a chromosome of the first type shows simply a single chromomere
longitudinally split (Fig. 59).

Linin connections between the chromosomes are much more evi-
dent than in the spermatogonia, and they can be seen to extend
from the sides as well as the ends of the chromosomes.

With regard to the origin of the first maturation spindle-fibres it
is difficult to be certain, but they appear to arise, at least in part,
from within the nucleus. The centrosome is not evident until a
rather late prophase (Figs. 39, 40, 43). In many cases it lies
within a more densely staining mass of cytoplasm of ill-defined out-
line applied close to the nucleus (sphere substance, idiozome of
Meves) (1898) (Figs. 39^, 43). This is not, however, invariably
the case, as may be seen from Figs. 40, 39^^, where the centrosomes
lie freely in the cytoplasm. Fig. 38^ perhaps represents an early
stage in the development of the sphere substance. In the two
adjacent cells (Fig. 38^ and 38^:) are shown rounded bodies of a
dull tint lying within clear vacuoles. I met these in but one
preparation and am unable satisfactorily to interpret them. ^The
division of the centrosome and the formation of the spindle is
shown in Figs. 46, 47, 48, 52. The centrosomes and spindle-fibres,
as well as the polar radiations, are more prominent than in the
spermatogonic spindles. During this time the sphere substance
disappears.

In the equatorial plate the chromosomes become arranged with
the longitudinal split parallel to the axis of the spindle in the case
of chromosomes of the first type, but at right angles to it or nearly
so in the case of chromosomes of the second type (Figs. 49, 50 and
53). In Figs. 55 and 56 are represented pole views of both types
of chromosomes. It may be gathered from these, as well as from
the figures of the prophases, that chromosomes of the second type
are not nearly so numerous as those of the first nor so numerous as
those of the intermediate type.

From what has been said with regard to the origin of the chro-
mosomes, it will be seen that in the metaphase the bivalent chromo-
somes are separated into their univalent components, and conse-
quently \}ci^ first division is one of reduction.

1902.] NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. 87

A well-marked mid-body is visible in the late anaphase (Fig.
6ia). The interzonal fibres are sharply constricted and oftentimes
the nuclei completely separated before a cell wall makes its appear-
ance. In stages like this a noticeable bending of the fibres is often
observed. This is slightly evident in Fig. 61 a.

Apparently the plane of the second spermatocytic division is to
be at right angles to the first, if Figs. 6 id; and 61^ are interpreted
as early stages in the formation of the equatorial plate of the second
spermatocyte.

The equatorial plate of the second spermatocytic division is
shown in lateral view in Fig. 62. The length of the chromosomes
is less than that of the chromosomes of the first spermatocytic
division. The question as to whether the second division is actu-
ally equational is difficult to decide. The chromosomes of the
first maturation figure, consisting of a double row of four granules,
are separated by karyokinesis into halves, and each half contains a
double row of two granules (Fig. 58). It thus has the appearance,
although only the appearance, of a true tetrad. It will be seen
that some of these daughter chromosomes have a length equal to
their width, whereas in others the length is slightly greater than
the width. If we turn to the fully-formed spindle of the second
division (Figs. 62, 6$) we find similar phenomena. It might be
argued from these appearances that the second division is also
reducing. In view, however, of the weight of evidence in favor of
both methods of division (equation and reduction) being necessary
to the maturation of the sexual cells among the Arthropods, I hesi-
tate to accept this interpretation without further corroborative
evidence. When the length of the chromosomes is equal to their
breadth, it is obviously as impossible to decide here concerning the
plane of division as in the case of the true tetrads of the Copepods,
Canthocamptus^ Heterocope or Diaptomus. If the length is greater,
as in the anaphase, the appearance might be referred to the elonga-
tion of the mother chromosome (Figs. 49, 51, 53), some of the
daughter chromosomes not having recovered from the stretching
apart of the chromatin in the metaphase. The apparently greater
length of some of the chromosomes in the spindle of the second
division (Fig. d^) may be explained by the assumption that some
of the chromosomes commence to divide earlier than others, and
consequently become elongated, an assumption which is not with-
out parallel in the first spermatocytic and especially in the sperma-

88 NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. [April 4,

togonic divisions (Fig. 15). In Fig. 61, a stage intermediate
between 58 and 62, some of the chromosomes likewise appear of
greater length than others. It might be supposed that the longer
ones represent the side view, the shorter ones the end view, of the
chromosomes. This need not, however, necessarily be the case,
for the chromosomes vary amongst themselves in size (Fig. 58 and
previous figures). It is possible, too, that in some cases the chro-
mosomes are seen slightly foreshortened and that their true dimen-
sions do not appear in the figure. I feel it, therefore, impossible
to ascertain with the desired degree of certainty the plane of the
second spermatocytic division.

In the late anaphase (Fig. 66) the chromosomes are more or less
indistinguishably massed together. On each of the interzonal
fibres in the equator is a minute swelling. These become reduced
in number (Fig. 67).^

4. Metamorphosis of the Spermatids.

The chromosomes spread apart, a nuclear membrane is developed
and the daughter cells become the spermatids. The gradual con-
version of the chromosomes into a fine reticulum is illustrated in
Figs. 69 and 70.

The nucleus now commences to elongate at one end (Fig. 72),
and this continues until the entire nucleus is transformed into a
shape somewhat like that of a narrow flask (Fig. 74). The nuclear
network is extremely delicate and takes the iron hsematoxylin stain
more faintly than previously. In cross section (Fig. 74^^; numer-
ous fine dots appear interspersed with clear areas (vacuoles). This
vacuolated appearance is sometimes evident at an earlier stage
(Fig. 71).

During the transformation of the nuclei the cell boundaries have
entirely disappeared and the nuclei lie in a common mass of cyto-
plasm. Several of them become associated together, and their
extremities, elongated into slender threads, are surrounded by a
clear, homogeneous, well-defined area of cytoplasm, while the
more or less contorted bodies of the nuclei still lie in an undefined
mass of cytoplasm (Fig. 77^).

A cytoplasmic thread of extreme delicacy can be traced from the

^ During the examination of the foregoing stages I have seen nothing similar
to the accessory chromosome (chromatin nucleolus) of insects, as described by
Montgomery (1898) and Paulmier (1899).

1902.] NICHOLS— SPERMATOGENESIS 0NISCUSA8ELLUS LINN. 89

slender extremity of each nucleus for some distance into the clear,
homogeneous area (Fig. 77^). At this stage also there can be
clearly seen in the undefined mass of cytoplasm a bundle of fibres,
which run in between the nuclei, but which cannot be seen to have
any connection with them. I have a preparation of a stage, earlier
than that just described, stained with hsematoxylin and Bordeaux
red, in which these striations appear, near the margin of the follicle
(Fig. 73). So early a development of the fibres is rather unusual.
The fibres are here apparently incomplete and not massed together
as they later are. On account of their indistinctness it is difficult
to say whether or not they are independent of the nuclei. At first
sight it might appear as if they were continuous, but it is impossi-
ble to state definitely that this is so because of the impracticability
of tracing a single fibre for any great distance.

The further changes in the nuclei consist in their gradual elonga-
tion into filaments, in which the network has entirely disappeared
and which have acquired the power to take a vivid and homogene-
ous stain. Their free ends, at first divergent, gradually approach
each other and finally come to lie close together (Figs. 77-79, 85
and 86). In regions of the follicle where the cells are closely
crowded together the nucleus is often seen to be bent or coiled
upon itself (Fig. S^).

There is at first a small quantity of cytoplasm around the nuclei,
but as they increase in length this disappears. The cytoplasmic
fibres also increase in length at the expense of the surrounding
cytoplasm. Their length, indeed, becomes truly marvelous, many
times exceeding that of the nuclei. They crowd in between the
follicle cells (Fig. 2) and in cross sections of the follicle can be
seen in great numbers around the margin. From the anterior end
of the bundle is developed a slender flagellum (Figs. 85, 86). The
entire bundle has the appearance at first sight of a single sper-
matozoon, and such I thought it before having studied its develop-
ment.

The term ''spermatophore " has been applied by Gilson to the
bundle. This term, however, has been used by Grobben and others
to designate an envelope secreted by the cells of the vas deferens
(in the Decapods) and surrounding a mass of spermatozoa. It
does not, therefore, seem applicable to the bundle of spermatozoa
found in the Oniscidse. Ballowitz applies the term *'spermozeugma"
to a large bundle of double spermatozoa found in the vas deferens

90 NICHOLS— SPERMATOGENESIS ONISCUS ASELLUS LINN. [April 4,

of the Dytiscid, Coly?nbetes striatus. These adhere together after
having reached maturity. Their structure and mode of origin is,
therefore, not the same as that of the bundles of Oniscus. The
term "compound spermatozoon" has been suggested to me, but
the word spermatozoon might carry with it certain implications
with regard to behavior in fertilization. I prefer, therefore, to use
the term sperm colony, at least until a better one offers itself.
Gilson uses this term also, although not so generally as the word
'* spermatophore."

The number of nuclei entering a colony varies within rather
wide limits. I have counted as few as six and also as many as four-
teen. In cross sections stained with saffranin and malachite green,
they are seen as red bodies surrounding a central mass of green
dots, the sections of the cytoplasmic fibrils (Fig. 80). The red
dots diminish in size toward the anterior end of the bundle, and
at one point can be seen merging directly into delicate green
threads (Fig. 80^). At the extreme anterior end of the bundle the
delicate green threads alone will be cut (Fig. %oa). It might be
supposed that the bundle of cytoplasmic fibres previously described
are the tails of the spermatozoa. If they are really the tails of the
spermatozoa, one would expect to find them at some place con-
nected with the nuclei, or with the delicate fibres which can be
demonstrated to be continuous with the nuclei. A comparison of
sections obtained at different levels seems to leave but two alterna-
tives : either the long bundle of cytoplasmic fibres stops abruptly
before the anterior end of the colony is reached, or the connection
is of so tenuous a character as to escape observation. In structures
of such minuteness the latter might easily be the case.

A point bearing on this matter, and therefore of interest to deter-
mine, is the number of cytoplasmic fibres as compared with the
number of nuclei. Attempts to determine this might be made in
two ways. The mature sperm colonies taken from the vas and
teased apart might be examined and an attempt made to count the
fibres at the frayed end of the bundle, or one might try to count
the number as seen in cross sections. By either method it is difii-
cult to be sure of an accurate count, for in the frayed ends of col-
onies one or more of the fibres may adhere together. In cross
sections the fibres appear as minute dots, as a rule, closely crowded
together. Occasionally they may be more loosely distributed.
Fig. Z\b represents a cell of this sort in which the num.ber of cyto-

1902.] NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. 91

plasmic fibres equals that of the nuclei. I cannot be certain that
this is invariably the case. With the iron ha^matoxylin stain the
bundle of cytoplasmic fibres stains deeply, like the nucleus, and it
is therefore impossible to distinguish between them in cross section
where both appear. The delicate fibril previously mentioned,
which joins the nucleus, stains faintly and can therefore be distin-
guished from the nucleus. In Figs. 8i and 82 cross sections of
sperm colonies at slightly different stages of development, colored
with this stain, are compared. In both images may be seen similar
to that of Fig. 80 — /. e., a circle of dots merging into faintly stain-
ing fibres. Sometimes the latter have a granular or beaded struc-
ture (Fig. 8i«). These are sections near the anterior end of the
colony, and here again the central circle of dots, representing the
posterior cytoplasmic fibies, is lacking. Fig. 82/^ represents a sec-
tion which I interpret as having been cut slightly posterior to Fig.
82^. The tail fibres here begin to appear.

A comparison of the two stages illustrates the gradual dwindling
of the cytoplasm which surrounds the bundle. It will be remem-
bered that shortly after the complete reconstruction of the spermatid
nucleus, cell boundaries disappear and the nuclei lie in a common
plasma. When, however, the nuclei come to be associated in
groups, the cytoplasm again becomes sharply defined and in cross
sections an appearance like that of separate cells is obtained (Fig.
80). The cytoplasm in the anterior region becomes comparatively
homogeneous and the nuclei often lie in a central clear space (Fig.
81^). More posteriorly it breaks up and assumes a granular
appearance (Fig. Zic), while still farther back the fibrillar bundles
lie isolated, with vague remnants of cytoplasm between them (Fig.
81). In Fig. 82 the diameter of the colony is less and the cyto-
plasm surrounding the fibres decidedly less extensive.

The sperm colonies when mature, or nearly so, are forced into
the vas deferens, probably by contractions of the muscle layer of
the follicle. In the vas they are surrounded by a fluid secreted by
the large cells which form its lining, and which causes them readily
to adhere to needles or forceps. The mature colony has the appear-
ance shown in Figs. 85, Zd and 87. I have not heen able to isolate
a single colony entire, for in teasing the long fibres are almost
invariably torn. I have been able to trace them for a considerable
distance, however, and can state that they are exceedingly long.
The filamentous nuclei are invariably partially frayed from the

92 NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. [AprU 4,

sheath and often entirely torn from it, lying twisted and contorted
at some distance from the sheath. According to Hermann, 1883
(2), the spermatozoa of the Isopods retain their immobility in the
oviduct of the female. The function of the extraordinarily long
fibres, if the spermatozoa remain motionless, is to me a matter of
great perplexity. It becomes still more puzzling if, as my prepa-
rations seem to indicate, there is no direct connection between
them and the nuclei. Their function and their true relation to the
nuclei might possibly be elucidated by a study of their behavior
in fertilization, a study in which I hope to engage at some future
time.

5. T/ie Nucleolus.

In the resting spermatogonia the nucleolus is present as a rounded
or oval body, staining pink with the eosin of the Biondi-Ehrlich
stain and red with saffranin. When the mitotic figure is fully
formed it is, as a rule, no longer visible, nor is it seen in the pro-
phase immediately preceding. The newly constructed daughter
nuclei likewise show no trace of it (Figs. 10, 14, 20). Possibly it
may consist of metabolic products developed in the resting cell and
quickly dissolving during or before mitosis. In the synapsis stage,
subsequent to the last division of the spermatogonia, the nucleolus
is, however, clearly visible, lying to one side of the tangled mass of
chromatin threads.

In the very earliest synapsis of which I have sections it is not
discernible (Fig. 22), but as the threads elongate and separate it
becomes evident. It continues to be present throughout the syn-
apsis and is finally enclosed within the nucleus of the resting sper-
matocyte by the development of the nuclear membrane (Figs. 23,
26, 28, 29 and 30). Throughout the prophases of the first sper-
matocyte it is still to be seen within the nucleus (Figs. 32, 33 and
43^), and after dissolution of the nuclear membrane and formation
of the mitotic figure it is cast off to one side of the spindle, where
it persists for some time (Figs. 47, 51, 52, 55, 61, 65-67 and 6()F).
With saffranin and malachite green the nucleolus is very evident,
coloring bright red, while the chromatin of the resting cell is pur-
ple. With iron haematoxylin it is not so readily distinguished, but
with the Biondi-Ehrlich stain it can be seen as a pink body lying
to one side of the spindle.

1902.] NICHOLS— SPER:\rATOGENESIS ONISCUS ASELLUS LINN. 93

6. Summary.

The main results of this study may now be briefly summarized as
follows :

(i) The spermatogonia chromosomes are joined together in
pairs in the synapsis to form sixteen bivalent chromosomes. They
may be joined {a) in an approximately straight line, {I)) to form a
more or less narrow V, or {c) into a more or less complete ring
(Figs. 26, 27).

(2) A longitudinal splitting of the chromatin threads takes place
at this stage (Figs. 25^, by c).

(3) The distinctness maintained by the chromosomes up to the
formation of the nuclear network of the resting spermatocyte, and
the manner of origin of the spermatocytic chromosomes from it,
lends support to the theory of their individuality in the resting
nucleus (Figs. 28 and 32).

(4) In the structure and mode of origin of the bivalent sper-
matocytic chromosomes two main types may be distinguished :
(a) The component chromosomes lie end to end, or {b) they lie
side by side (Figs, d'^a, b, c).

(5) Inasmuch as univalent chromosomes are separated, the first
maturation division is reductional (Figs. 48-59).

(6) Sphere substance (idiozome) is not observable, except for a
short time during the prophases of the first spermatocyte (Figs.
39 and 43).

(7) The nucleolus of the spermatogonia disappears shortly after
dissolution of the nuclear membrane, while that of the spermato-
cytes, first discovered in the synapsis, persists throughout the
divisions (Figs. 8-10, 47, 26, 29, 2iZ, 47, 48, 51^ 52, 55^ S^, 60a,
61, 65-67, 69).

(8) The spermatids become associated in groups to form colonies
of nuclei lying in a common plasma (Figs. 73-75).

(9) Within the latter arise bundles of fibres of great length,
whose connection with the nuclei, if actual, is very slight and
occurs very late, as well as single fibres of greater delicacy which
are continuous with the nuclei (Figs. 76-83).

(10) The mature sperm colony consists of a variable number of
filamentous nuclei contained, together with the bundle of cyto-
plasmic fibres, in a tenuous sheath which is flagellate at its anterior
extremity (Figs. 84-86).

94 NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. [April 4

III. Critical Review of the Literature on Crustacean
Spermatogenesis since 1878.

I. Spermatozoa.

a. Review.
Decapoda.

1878. Grobben in his valuable work investigates principally the
form of the Decapod spermatozoa and their transformations from
the immature to the mature state, as well as the nature of the case
(spermatophore) in which they are enclosed. With regard to the
spermatozoon of Astacus fluviatilis, he states that the head develops
from a structure arising near the nucleus, while the nucleus itself
disintegrates. He gives also a review of the literature on Crus-
tacean spermatozoa up to that time, which therefore need not be
repeated here.

1883 (i). Herrmann describes the spermatozoa of the Podop-
thalmia, chiefly the Macrura and Brachyura. The study of the
development, he says, shows a series of transitory forms which
enable us to seize clearly the bonds of relationship existing between
the different adult forms. The transitional forms of some resemble
the complete forms of others.

1884. Nussbaum (Astacus fluviatilis) considers the change of the
spermatid into the spermatozoon. He traces the gradual condensa-
tion and transformation of the nucleus from spermatid to sper-
matozoon, and the transformation of a large body lying in the
cytoplasm into the peculiar ''kopfkappe" of the mature sperma-
tozoon (see his Figs. 53-68). He regards the nucleus as the head
of the spermatozoon.

1885. Sabatier published a short article on the spermatogenesis
of the Decapod Crustacea, principally Astacus.

1886. Gilson describes the spermatozoa of a considerable num-
ber of Decapod species, among others Astacus fluviatilis. The
structure of the spermatozoon of the latter he delineates more fully
than either of his predecessors. The nucleus he shows to be
present and saucer-like in shape. It is covered by a layer of pro-
toplasm which is extended laterally into pseudopodic processes.
From the centre of the protoplasmic layer sometimes arises a pro-
tuberance, to which he gives the name ''globule achromatique.''
The nucleus surmounts a bladder-like vesicle often perforated at the
opposite pole. Into this from the centre of the concavity of the
nucleus projects what he calls "la tigelle."

1902.] NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. 95

1895. Auerbach compares the spermatozoon of Astacus fluviatilis
with those of other Crustacea, Insects and Vertebrates, with a view
to discovering homologies of head, apex, middle-piece and tail.
The cyanophilous, saucer-shaped nucleus corresponds to the head of
more highly developed spermatozoa, its pole therefore to the ante-
rior end of a flagellate spermatozoon and the surrounding proto-
plasm to the sheath of the head. The ''globule achromatique " of
Gilson is the anlage of the apex. The " tigelle " of Gilson, which
Auerbach found to be erythrophilous, he regards as the anlage of
the middle-piece. In the genera Pagurus, Eupagurus, Clibanarius
and Ethusa the '' tigelle " is prolonged into what Auerbach regards
as a rudimentary tail. The bladder-like vesicle is perhaps a kind of
'* Schwanzkappe," possibly comparable with the sheath sometimes
surrounding the place of origin of the tail in immature vertebrate
spermatozoa. The extremity regarded by Grobben as the head
would, according to Auerbach's interpretation, be the tail end.
For a more detailed account of the Decapod spermatozoa, of which
that of Astacus may be taken as a type, the reader is referred to
the works cited above.

Stomatopoda, Schizopoda, Amphipoda.

1885. Gilson, in his excellent and very comprehensive work, de-
scribes also the spermatozoa of the Stomatopod Squilla, the Schizo-
pod Mysis and the Amphipod Gammarus. The whip-like sperma-
tozoon of Mysis is strikingly similar in shape to that of the
Isopods. That of Gammarus is flagellate and that of Squilla
vesicular.

Isopoda.

1883. Herrmann studied among the Isopoda, Ligea, Idotea and
Sphaeroma. His description is unaccompanied by figures and is
difficult to comprehend. The spermatic filaments, he says, are
united in numbers varying from eighty to one hundred. The
bundles are found lying amongst the cells which line the walls of
the tube. He did not find isolated spermatozoa, except in the
oviduct of the female, where they retain their habitual form
and immobility. The large cell of the vas deferens he con-
e-iders as homologues of ovarian cells and calls them ''ovules
males."

1884-1886. Gilson (Oniscus asellus). Groups of six spermatids

96 NICHOLS — SPERMATOGENESIS OXISCUS ASELLUS LINN. LApriU,

C spermatoblaste ") were observed surrounding a protoplasmic
stem and their origin referred to the small cells in the apical por-
tion of the caecum. The structure of the nuclei and the changes
in them and in the surrounding protoplasm, by which the mass is
converted into the mature ''spermatophore," are described at
some length and illustrated with numerous figures. The name
^'spermatophore " is applied for the following reasons: " Les
cellules spermatozoides sont done contenus dans un etui resistant
derivant de la differentation du protoplasm, c'est-a-dire dans une
production particuliere, on pourrait done appliquer aux faisceaux la
denomination de spermatophore." The name '^plasmodium
parietal " is applied to the follicle cells and the surrounding proto-
plasm, and to it is ascribed the function of taking part in the
formation of the tails, thus reinforcing the insufficiency of the pro-
toplasm of the germ cells. The tails of the spermatozoa are thus
thought to arise in the plasma and to attach themselves to the
nuclei ^' vers le haut." The exact level is not determined. The
form of the spermatozoa is compared to that of a whip, the long
tail representing the handle and the nucleus the lash. This would
seem to indicate that the tail is conceived as being attached to the
nucleus at its upper extremity. The entire bundle is said to
measure 0.15^0^ mm.

The sheath (etui) enclosing the spermatozoa is most evident at
the anterior end. The apparent absence of protoplasm around the
filamentous nuclei is explained as perhaps due not to degeneration
or absorption of the protoplasm, but to a condensation and fusion
with the nucleus, perhaps applying itself so closely to the filament
that an effect of refraction communicates to it the same coloration
This hypothesis is based on results obtained by treating the flagellae
with nuclear solvents. When submitted to the action of potassium
carbonate in concentrated solution or strong hydrochloric acid for
several days the filaments become scalariform ; a skeleton formed
of little chambers is seen which communicate with each other, and
which were previously filled with the nuclear substance. The char-
acteristic frayed appearance of the bundles is thought to be due to
artificial rupture.

The nuclear flagellae are said to grow considerably after having
attained their distinctive form. From the figures given to show
this (Figs. 329 and 330, PL VIII), it seems probable that this
appearance may be due to a portion of the filaments having been
broken off by teasing.

1902.] NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. 97

The large cells lining the vas deferens are described and also
the smaller cells between them. The latter are believed to arise
from the larger ones by segmentation. The function of the large
cells is said to be the secretion of the fluid which bathes the sper-
matozoa. The nucleus of these cells is figured as a network of
great regularity.

The mature colony of Asellus, as figured by Gilson in Vol. 2 of
La Cellule, PI. X, Figs. 385-395, agrees with that of Oniscus in
general appearance. The spermatozoa in the bundle, however, are
more numerous and much less compactly bound together. Asso-
ciated with them in their development is a large cell (''noyau
femelle "). The tail is shown to be distinctly continuous with the
nucleus. The granular mass surrounding the nucleus at its free end
is said to consist of caryoplasm and the remains of the nuclear
membrane. Its formation is shown in Figs. 387-393.

A few figures are also given of Idotea.

1886. Wielowieyski, in a short paragraph concerning Asellus,
states his opinion that the ''noyau femelle" of Gilson is an artifi-
cial product, caused by the confluence of the protoplasmic mass with
one of the large cells on the margin of the testicle.

Cirrepedli.

1886. Gilson figures the spermatozoa of Lepas anatifera and
Balanas perforatus. They are flagellate, the nucleus a slender
thread occupying the anterior end.

1894. Ballowitz, K., studied Balanas improvisus Darw. and Lepas
anatifera L. He makes the astonishing statement that the head is
demonstrable as a distinct structure neither by its form nor by its
staining reaction. He mentions the work of Nussbaum (1890) on
a Californian Cirrepede (Pollicipes polymerus) in which the head
is described.

Copepoda.

1895. Steuer gives a figure to show the spermatozoa of the
marine Copepod, Sapphirina gemma. They are flagellate, shaped
somewhat like a javelin. He mentions the spermatozoa of the
Calanidse as being of spherical shape.

Osiracoda.

1886. Stuhlmann. The spermatozoa of the Cypridae are de-
scribed as having at first the shape of a ribbon, through the length

PROC. AMER. PHILOS. SOC. XLI. 168. G. PRINTED MAY 8, 1932

98 NICHOLS— SPERMATOGENESIS ONISCUS ASELLUS LINN. [April 4,

of which the nucleus runs as a thread. They are stated to increase
in size through the assimilation of a secretion of the vas deferens.
They then become spirally twisted while in a certain limited section
of the vas deferens, presumably by a motion of their own. This is
said to be caused by a fibre running spirally the length of the
spermatozoon. The mature spermatozoon has the spirally twisted
structure of a rope of tow. It contains a twisted central fibre, not
visible externally, and the entire structure is surrounded by a
hyaline sheath. The spermatozoa are nearly motionless while in
the body of the male, but become extremely active in the recep-
taculum seminis of the female. This is said to be due to the loss
of the hyaline sheath. The curious fact is noted that the sperma-
tozoa coming from the right side of the animal are twisted to the
left and vice versa.

1889. Mliller discovered in the spermatid of Ostracoda one or
two '' Nebenkerne." These form a *' Schwanzstuck " which s^rows
very long and is of complicated structure. Through the middle of
the tail runs the central fibre, at or near one end of which the
nucleus is located. The spiral twisting is referred to the contrac-
tion of the middle one of the three threads which surround the
central fibre. He does not agree with the opinion of Stuhlmann
concerning the inhibitive function of the sheath while in the body
of the male.

Phyllopoda.

1885. Zacharias describes the results of his observations and ex-
periments on the spermatozoa of the Phyllopod, Polyphemus, which
he shows to be capable of amoeboid movements.

b. Commentary.

The Crustacea as a class show an astonishing variety in the form
of the male reproductive elements. Knowledge of their intimate
structure is of course at present too incomplete to enable us to discuss
at any great length the homologies existing between them. But a
rough classification of them according to their external appearance
would place the bell-shaped or vesicular form characteristic of the
Decapods in one group and the form found in the Isopods, Gam-
marus, Mysis and Balanus, with more or less elongated nucleus and
tail of varying length, in another. The extremely peculiar form of
the spermatozoon of the Ostracoda might perhaps be referred to the

1902.] NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. 99

latter group. It is possible, and I advance it simply as a tentative
hypothesis needing corroboration, that these strikingly dissimilar
forms have arisen from a primitive one, simple and amoeboid in
character like that of Polyphemus.

The ingenious series of homologies drawn by Auerbach between
the head, tail, apex and middle-piece of the spermatozoa of Verte-
brates and Insects and similar structures in Astacus appears plausi-
ble. Since, however, the location of the centrosome and the
sphere substance remains undetermined, the homologies cannot be
said to be in all respects established. A more detailed and thor-
ough examination of the spermatozoa of the Crustacea, especially
of their behavior in fertilization, might extend these homologies.
If the spermatozoon of Oniscus be compared with the type most
frequently occurring in animals, the part immediately adjacent to
the nucleus, the delicate fibril shown in Figs. 77^, 79, corresponds
in location to the middle-piece. Whether this is in reality the
habitation of the centrosome might be discovered through a study
of its fate during fertilization. My observations on the spermato-
genesis throw no light on the question.

The Isopods are unique among the Crustacea in the formation of
colonies of spermatozoa of a nature so close that they appear as
units. Concerning their origin in Oniscus, I can confirm M.
Gilson's statement that the formation of the bundle takes place in
a Plasmodium, cell boundaries being for a time entirely absent, and
with the main outlines of his account of the changes taking place
in the development of the spermatids into the mature colony I am
thoroughly in accord.

The number of nuclei entering into a bundle, according to my
observations, is not invariably six, but may vary within considera-
ble limits. The number of cytoplasmic fibres is assumed by M.
Gilson to be equal to the nuclei, but in his Fig. 328, PI. VIII, they
are shown to be more numerous. As has been already said, I have
been unable to convince myself of a direct continuity between these
fibres and the nuclei. In his Fig. 320 (an immature spermato-
phore) the cytoplasmic fibres may be traced directly to the nuclei.
I have, however, riot been able to obtain images of equal clearness
from my preparations. Nor have I obtained anything at all similar
to the rings or vacuoles, shown in Gilson's Figs. 328, 329 and 330,
near the anterior end of the bundle. In Sphaeroma serratum, Gilson
states, the continuance of head and tail is very evident, forming an

100 XICHOLS — SPERMATOGENESIS OXISCUS ASELLUS LIXN. [April 4.

Open and regular ring. The close relationship of nucleus and cyto-
plasmic fibre in Oniscus is shown only in Fig. 320. In Figs. 323
and 326 they are represented as discontinuous. In Fig. 334
(Asellus) the fibres are pictured as arising independently of the
nuclei, although it is shown in later figures that they eventually
become attached. If the follicle nuclei and the surrounding proto-
plasm take part in the formation of the tails, it is only, in my opin-
ion, in so far as they become converted into the substance of the
germ cells.

In attempting to reconcile the fact of the direct continuity of
head and tail, shown by Gilson so clearly in Asellus and stated by
him to be present in Sphaeroma, with the lack of demonstrable con-
nection in Oniscus, it occurred to me that the condition in Oniscus
might represent a different phase in the evolution of the Isopod
spermatozoon. Either the connection, at one time evident, between
the nucleus and the unusually long tail may have grown so slight as
to be no longer recognizable, or, if the spermatozoon of Oniscus
for any reason is to be looked upon as the more primitive form, it
may be that the connection, which will later in the course of evolu-
tion become more marked, is as yet but little developed. Although
in the present state of our knowledge both alternatives may perhaps
be considered open, the former seems to me far more plausible, for
not only are the land Isopods in other structural peculiarities to be
regarded as more specialized than Asellus, but the sperm colony
itself in Asellus is less compact and less completely developed as a
unit. The obscurity of this point serves to emphasize the desira-
bility of further study of the Crustacean spermatozoa and the estab-
lishment of accurate homologies between them.

The " noyau femelle" of Asellus is, in my opinion, to be regarded
as homologous with the follicle cells of Oniscus. I am inclined to
doubt the correctness of M. Gilson's conclusions as to the origin of
the small cells of the vas deferens of Oniscus from the larger ones
by segmentation, and, although I have not devoted much time to
the elucidation of the point, I think it more probable that the re-
verse is true, for I have seen the small cells segmenting, but never
the large ones.

1902.] NICHOLS — SPERMATOGENESIS OXISCUS ASELLUS LINN. lOl

2. The Earlier Stages in the Development of the Germ Cells

IN Crustacea, with Especial Reference to the

Problem of Reduction.

a. Review.
Decapoda.

1878. Grobben gives almost no figures of the earlier stages and
does not consider the subject in detail.

1884. Nussbaum (Astacus fluviatilis) does not distinguish between
spermatogonia and spermatocytes. Five figures of mitoses are given
in which the chromosomes are shown to be spherical at the begin-
ning of the metaphase, but they soon elongate to a rod-like shape.

1885. Carnoy studied among the Decapods, Astacus fluviatilis,
Crangon vulgaris and several species of Brachyura and Anomura.
In no case are more than thirteen figures given. It is impossible to
determine in every case the generation to which the cells belong.
The mode of origin of the chromosomes is not fully traced, and it
is impossible to determine with accuracy, therefore, anything with
regard to the question of reduction. In the case of Astacus, as far
as can be judged from the figures given (Figs. 246^, b, c, d, e and
f), the division is transverse. The mitosis figured occurred in

August, and, according to vom Rath, it is from this month until
December that the final divisions of the spermatogonia and those
of tlie spermatocytes take place. A transverse spermatogonic
division is improbable. The chromosomes are shown to arise,
however, through the shortening and thickening of rods, resulting
from the breaking up of the nuclear network. The transverse
division, if it be such, is therefore probably that of the first sperma-
tocyte. The same is perhaps true of Crangon cataphractus (Figs.
247 and 248). Of peculiar interest is the constitution of the chro-
mosomes of Crangon cataphractus, as shown in Figs. 249^, b, c, d,
PI. VII. According to these a chromosome in longitudinal view
consists of a double row of from three to five granules. A recon-
struction of the chromosome from these figures leads to the concep-
tion of a rod split longitudinally several times.

Cytoplasmic Structures. — A dense mass, lying within the cyto-
plasm during the prophases and migrating to the poles of the spin-
dle as it is formed, is shown for Crangon. No centrosome is
figured as lying within this mass, to which the name *'Nebenkern"
is given. The same name is applied to a body lying in the cyto-

102 NICHOLS — SPERMATOGENTESIS ONISCUS ASELLUS LINN. [April 4,

plasm in Astacus. This body, however, seems not to be affected
by mitosis, but lies passively to one side. In the vicinity of the
poles are, however, numerous granules (''corpuscles polaires") (Fig.
246/, PL VII). The " Nebenkern" of Crangon, according to the
description, behaves like the substance designated idiozome by
Meves. The "corpuscles polaires" of Astacus may be of a similar
nature. For the other forms studied no bodies of any kind lying
in the cytoplasm are shown. The substance seems to be unusually
prominent in Crangon and Astacus, The cells of both are of large
size.

1 89 1, vom Rath settled the question of amitotic division of the
germ cells of Astacus in the negative. He states that a minority of
the spermatogonia undergo no change at first, but give rise by
mitosis to new spermatogonia after the discharge of the ripe sper-
matozoa. He mentions a case of regeneration of an entire follicle
from a single spermatogonium. With the first appearance of the
spermatids the follicle cells (''Randkerne") commence to grow in
size and divide amitotically. The direct division apparently takes
place by a sharp breaking apart of the portions of the nuclei, re-
sembling a slicing. Degeneration of the nuclei follows. At the
point of transition between follicle and duct there is often an extra-
ordinary growth of cells by amitosis. The results of his research
are interpreted by vom Rath to mean that two kinds of cells have
arisen from indifferent epithelium, one dividing mitotically, the
other amitotically.

Is op o da.

1884. Gilson states that it is only at certain seasons of the year
that the spermatogenesis of these animals can be studied with
profit. In the case of Oniscus asellus, from July to November is
the most favorable season for obtaining preparation of what he
calls the first stage ("premiere etape"). In the case of Asellus
aquaticus it is later — about the month of February.

Oniscus asellus. — The cells filling the apical end of the caeca
(spermatogonia) are mentioned, and the opinion is expressed that
they constitute a reserve mass destined to replace by proliferation
the elements organized in the lower part of the tube and later
evacuated. Karyokinesis in these cells (spermatogonia) was ob-
served but once, and the stages intervening between them and the
spermatocytes were not discovered.

1902.] NICHOLS— SPERMATOGENESIS 0XI3CUS ASELLUS LIXX. 103

The condition of the lower part of the tube is thus described :
**I1 y aurait dans les c?ecums testiculaires des Oniscus une sort de
Plasmodium contenant une grande nombre de noyaux et entourant
une masse centrale formee d'elements spermatiques en formation.
Ce fait est si etrange qu'on n'ose a peine I'accepter." The amitotic
division of nuclei occupying the lower portion of the follicles and
referred by Gilson to the germ cells is probably that of follicle
cells, for they are described as occupying the periphery of the tube
in its median portion.

1885, Carnoy makes the following statement concerning the
Isopods (pp. 222, 223) : " Chez I'Oniscus asellus, au moment de la
plus grande activite cellulaire preludant a la formation des sperma-
tozoides, on ne rencontre pour ainsi dire que des noyaux en voie
d'etranglement ou de division acinetique. Les figures caryocine-
tiques y font le plus souvent defaut. Depuis trois ans nous n'en
avons rencontre que deux, une couronne equatoriale et une couronne
polaire qui sont reproduite dans la PL VI, Fig. 227; et cependant
nos observations ont ete nombreuses et pratiquees a toutes les
epoques de I'annee.

*'Nous avons constate les memes phenomenes sur plusieurs ani-
maux du meme groupe, sur les Idotea en particulier. La division
directe est tres frequente chez ces derniers, et s'y fait normalement.
Nous n'y avons point remarque de caryocinese; mais nous devons
ajouter que nos observations sur ces Crustaces bien que fait seri-
eusement ont ete beaucoup moins nombreuses que sur Oniscus.
Chose remarquable, chez les Idotea la multinuclearite des grandes
cellules qui vont se transformer en autant de faisceaux de spermato-
zo'ides est due exclusivement a la segmentation du noyau primi-
tive. Ces faits sont d'autant plus singuliers que dans un genre
voisin, le genre Armadillo, les figures caryocinetiques sont fre-
quents ; tandis que les cas de division directe y sont beaucoup plus
rares." I have examined testes of Armadillo and also of Porcellio
and find that they do not differ greatly from Oniscus as to the
manner and frequency of the divisions.

Copepoda.

1890, 1892. The work of Hacker on the eggs of Cyclops has
been corrected by the later research of Riickert and need not,
therefore, be mentioned here.

1892. Ishikawa gives a figure of the testis of a Copepod cut

104 NICHOLS— SPERMATOGENESIS ONISCUS ASELLUS LIXX. [ApiiU.

longitudinally, showing it to be divided into regions called by him
formative, growing and ripening zones. The formative region cor-
responds in Oniscus to the reserve groups of spermatogonia, the
growing region to the apical part of the follicle and the ripening
zone to the basal part of the follicle. Ishikawa's conclusions con-
cerning reduction have not been substantiated by recent research.

1894. Riickert. This well-known paper concerns the ovogenesis
of the Copepods, Cyclops strenuus, Heterocope and Diaptomus.

In Cyclops the number of chromosomes is 22-24. The germinal
vesicle shows double threads of chromatin, a longitudinal split
having occurred at an early period. At the beginning of matura-
tion these contract to double rods, whose number is the reduced
one and which have, moreover, become transversely split. As the
spindle is formed the chromosomes come to lie in the equator, with
the longitudinal split at right angles to the axis of the spindle.
The first division is thus equational. In the second division the
chromosomes are separated along the transverse split, and this
division is therefore reducing.

In Heterocope and Diaptomus open rings are formed which,
through condensation, become the tetrads. The plane of the first
division is not so easily determined for these Copepods. In the
opinion of Riickert the first maturation division of Diaptomus is
equational.

1895. Hacker studied the ovogenesis of the Copepod, CajitJio-
ca7nptus. The reduced number of chromosomes is twelve. There
are apparently two divisions of the ovogonia. The last division is
followed (i) by a transverse breaking apart of a doubly split thread
and a shortening and thickening of the segments so that twelve
double rods are produced. Some of these are transversely split
Or (2) the last division of the ovogonia is followed by a condensa
tion and longitudinal division of the thread as a whole and a sub
sequent breaking apart of the thread into twelve double rods
These become transversely split and form chromosomes correspond
ing to the tetrads of the first mode. In either mode the changes
follow immediately upon the last division of the ovogonium, and no
true reticulum is formed in the germinal vesicle. Since the width
of the chromosomes is equal to their length, it is impossible to
settle the question as to the order in which the longitudinal and
transverse divisions occur.

1895. vom Rath describes the ovogenesis of marine Copepods

1902.] NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. 105

mentioned by him in his earlier works on Gryllotalpa and Sala-
viandra. He studied the genera Eucluvta, Euca/anus, Anomalocera
and Fleuromma. He calls attention to the differences that may
exist between the ovogenesis of different species of marine Cope-
pods and between the ovogenesis and the spermatogenesis of the
same species. His conclusions on the subject of reduction agree
substantially with those of Riickert. Particularly in the case of
Euchceta marina and Eucala?ius attenuatus is the aspect of the first
maturation figure similar to that of Cyclops. Here, too, the divi-
sion seems to be equational.

Ostracoda.

1898. Woltereck describes a well-marked synapsis zone in the
ovary of a parthenogenetic Cyprid. He rejects, as not applying to
the object which he studied, the theories of Moore, Brauer and
Hacker concerning the relation of the synapsis to the last ovogonic
division and to the processes of reduction and maturation. "Von
*Reduktion,' " he says, " ist nicht die geringste Andeutung vorhan-
den, von der Reifungstheilung sind die Eier noch durch eine lange
Phase getrennt, in der das Chromatin kaum sichtbar ist and gegen
die Auffassung als Dispirem die excentrisch Zusammenballe bei
deutlich vorhandenem Nucleolus, sowie das Vorhandensein aller
Uebergange aus einem lockeren, hellen Fadenknauel in die Synapsis
und aus ihr in die segmentirten Chromosome."

Phyllopoda.

1892. Brauer thus summarizes his results on the ovogenesis of
Branchipus: "Die Beobachtungen, welche ich bei Branchipus
gewonnen habe, zeigen nun folgendes Bild :

"i. Keimblaschen : durch Quertheilung entstehen 6 Schleifen ;
eine neue Quertheilung erhoht ihre Zahl auf 12. Dann folgt eine
doppelte Langspaltung. Resultat : 12 viertheilige Chromosomen
bil den die Aquatorialplatte der ersten Richtungsspindle " (p. 53).
In describing the Figs. 8 and 9, Taf. I, upon which he bases this
conclusion, he says: "Ich will gern zugeben, dass diese Beobach-
tung schwierig sind und eine Taiischung moglich ist, doch muss
ich vorheben, dass ich kein Bild gesehen haben, welches eine
Vermehrung der 12 Faden durch eine Quertheilung auf 24 zwei-
theilige auch nur andeutete und spatere Verklebung von je zwei-
theilige zeigte. Solche Bilder, welche ganz ahnlich aussehen

106 NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. [April 4,

mlissten wie das in Fig. i dargestellte, waren mir, glaube ich, nich
entgangen."

1893. Brauer. The study of the closely related Phyllopod
Artemia was undertaken by the same author with the object of
ascertaining whether reduction took place in parthenogeneticaily
developing eggs.

The number of chromosomes in the germinal vesicle is eighty-
four, and their structure is quadripartite, /. e., each consists of four
spheres. In the first maturation division two of these spheres are
separated from the others. After this has taken place the matura-
tion may proceed in two different ways. The second polar body
may be formed and the elements of the dyad separated, or there
may be an abortive attempt to form the second polar body, the
chromatin, however, remaining undivided and the elements of the
dyad not separated.

Cleavage and further development of the ^gg may take place in
both of the above cases. In the first case it is necessary for this
that the second polar body be drawn back into the &gg, where it
acts as would a male pronucleus. In the second case the nucleus
left within the tgg after the formation of the first polar body, be-
comes the cleavage nucleus. In the first case the somatic number
of chromosomes is 168, in the second case 84.

It thus appears that the tetrads of the germinal vesicle are
bivalent chromosomes and that the actual reduction may or may
not take place.

1893. Moore published the results of his studies on the reproduc-
tive elements in Apus and Branchipus. With regard to Branchipus,
the chief stress of the paper is laid upon the relation between
karyokinesis and protoplasmic structure, the author believing ** that
the divisional phenomena of these cells are intimately related to a
protoplasmic structure, which might be fitly described as ' Schaum-
plasma,' and one of the initial physical impulses toward meta-
morphosis is a fusion of some of the intra-nuclear globules ; and a
considerable portion of the complicated karyokinetic figures^ with
their centrosomes, pseudosomes and dictyosomes, appear to be the
logical as well as the actual consequence of the continuance of this
process."

The question of reduction is not entered upon in much detail.
From the nucleus of the resting spermatocyte, however, are shown
to arise ten chromosomes of dumbbell-shape. These become

1902.] NICHOLS— SPERMATOGENESIS ONISCUS ASELLUS LINN. 107

arranged in the equatorial plate with the transverse constriction in
the plane of the equator. This division consequently is apparently
reducing. No longitudinal split is shown and the second sper-
matocytic division is very inadequately worked out.

b. Cofnmentary.

Although agreeing with many points in the description of Gilson
concerning the metamorphosis of the spermatids of Oniscus, my
observations do not entirely coincide with his account of the earlier
stages. The statement defining the most favorable season for ob-
taining preparations of the first stage does not hold true for the
locality of Philadelphia, for I have sectioned material collected
during every month of the year, except December and January, and
have not found one monlh to be preferred over another with regard
to the abundance of any particular stage.

I feel sure that the function of replacing the evacuated elements
which he ascribes to the spermatogonia is the true one, but that
their multiplication takes place by direct division I am unable to
believe. On the one hand the weight of the evidence of modern
research is against the occurrence normally of amitotic division in
the germ cells. Moreover the work of vom Rath on Astacus creates
a strong probability that the phenomena are similar in Oniscus. I
have never seen amitotic division in the germ cells of Oniscus, and
believe that the error arose from a failure to distinguish between
the germ cells and the follicle cells. I cannot help a feeling of
surprise that mitosis should have been so infrequently seen both by
M. Gilson and his colleague, M. Carnoy. It is true that the
mitoses of the spermatogonia are scattered, and occasionally no
spindles at all will be met with in a follicle, but by cutting a
sufficient number of sections cell division will be abundantly
seen.

With regard to the question of reduction in the Crustacea, my re-
sults, much to my own surprise, do not coincide with those obtained
by Riickert and vom Rath in the Copepods. The case in Cyclops
is so clear that it seems to admit of no doubt, and its very clearness
makes it probable that the divisions take place in a similar manner
in a form so closely allied as Canthocamptus. The figures given by
Hacker of this object do not, however, conclusively prove this to be
the case, since the tetrads are cubical in shape, the length no greater

108 NICHOLS— SPERMATOGENESIS ONISCUS ASELLUS LINN. [ApriU,

than the width. Indeed Hacker himself says of this object that it
is not adapted to the solution of the problem of reduction. The
like may be said of Artemia.

With Brauer's results on Branchipus, those obtained with Oniscus
likewise do not agree. The double longitudinal split claimed by
Brauer for the chromosomes of Branchipus is not shown in the
figures with the clearness that might be desired. An oblique view
of an elongated chromosome in Fig. 8 shows it to be split longitu-
dinally, not twice but only once. In the absence of direct evidence
to the contrary, the Figs. 8 and 9 might be explained equally well
on the assumption that the twelve tetrads represent two univalent
chromosomes longitudinally split and joined end to end.

In Oniscus, inasmuch as the first division separates two originally
distinct chromosomes and the second presumably divides the chro-
matin longitudinally, the manner of reduction resembles that of
Insecta as described by Henking (i890-'92), Paulmier (1899) and
Montgomery (1898, '99).

If my interpretation of the method of reduction in Oniscus be
correct, and that of Riickert concerning reduction in Cyclops be
equally so, it becomes clear that the cell generation in which the
true reduction takes place need not be the same for all members of
a given class of animals. The order in which the reduction and
equation divisions take place is, therefore, relatively unimportant ;
the significant thing, so far as our knowledge at the present day
goes, appears to be that in the Arthropods both divisions should
take place. Further research alone can show whether the apparent
cases of transverse division in the first spermatocytes of Astacus,
Crangon and Branchipus are really such. To the future must also
be left the question as to which method of reduction, the Copepod
or the Isopod type, is the rule among Crustacea.

M. Louise Nichols.

January 10, 190 1.

1902.] NICHOLS — SPERMATOCxEXESIS ONISCUS ASELLUS LINN. 109

Explanation o.^ the Plates.

All of the figures, with the exception of i, 2 and 68, are camera drawings
made at the level of the microscope stage, and all except 3, 4a, 5, 6, 84-86, were
drawn with a Zeiss homogeneous immersion objective ^^, ocular No. 6, tube
length 100 mm. In those marked * the chromosomes are not all shown.

Plate
XI

Plate
XII

Plate
XIII

1. Free-hand drawing to illustrate the male reproductive organs of

one side, a, d, c, lobes of the testis ; v, vas deferens; /, penis;
X, suspending tissue.

2. Lobes of the testis in longitudinal section (semi-diagrammatic), a,

d, c, as before ; spg.y spermatogonia ; / c, follicle cell.

3. Longitudinal section of the vas deferens (Zeiss ocular 4, obj. AA).

d, c, lobes of the testis.
\a. Small portion of the wall of the anterior region of the vas (oc. 6

obj. D).
4b. Secretory cell from the anterior region of the vas.

r 5.

6.
7-

*8 *c

Plate
XIV

Small portion oi a testis lobe in longitudinal section. /. c, follicle
cells; m. /., muscular layer (oc. 6, obj. D).

Cells from the suspending tissue (^cf. Fig. i, x) (oc. 6, obj. D).

(^) Spermatogonium in an early spireme stage; ncl, nucleolus.
Centrosomes beginning to divide, {b) Resting spermatogonium
with large masses of chromatin, probably beginning to degen-
erate.
9, 10. Later spireme stages.

11. (a and f) Resting spermatogonia. {b) Spireme beginning to
segmeni.

1 2. Equatorial plate in side view.

13. («) Equatorial plate in pole view. (*3) Spermatocytic prophase,
171. /., muscle layer.

14. Equatorial plate in side view, showing the longitudinal split in
the chromosomes.

*I5. Metaphase.

16, Approximate pole view of a stage similar to 15.

17, Anaphase.

18, Late anaphase. Mid-body.

*I9, *20. Reconstruction of the daughter nuclei. In 20 the mid body
has migrated to the periphery.

*2i. {a and b) Reconstruction of the daughter nuclei, (t) Degen-
erating spermatogonium.

22, 23-28. Synapsis. 26. 1-16, chromosomes.

27. Different forms of the chromosomes in the synapsis.

29. Formation of the nuclear membrane.

30. Resting spermatocyte, spg., spermatogonium.

31. Irregular arrangement of the nuclear network, occasionally seen
juit before the formation of the nuclear membrane.

110 XICHOLS— SPERMAT0EGXESI3 ONISCL^rf ASELLUS LIXN. [April ,

Plate
XV

Plate
XVI

Plate
XVII

f 32> *33-47- Prophases of the first spermatocyte. 44. 1-16, chromo

somes.
46 and 47. Strongly decolorized sections showing the longitudinal

split of the chromosomes. Divergence of the centrosomes.
48. {a) Side view of the equatorial plate of the first spermatocyte.

{b) Anaphase of the first spermatocyte.
49-53. Equatorial plate of the first spermatocyte in side view.
54. Chromosomes of the first spermatocyte, showing the longitudinal

split.
55,56. Pole views of the equatorial plate of the first spermatocyte.
57. Slightly oblique view of the same.

58.

59.
60.
61.
62.

63.
64.
65.
66,
68.

69,
70.

71.

72.

73-

Anaphase (side view).
Anaphase (pole view).
Anaphase (tangential sections).
Telophases.

Side view of the equatorial plate of the second spermatocyte.
The same more strongly decolorized.
Pole view of the same.
Metaphase.
67. Telophases.

{a and b') Mode of formation of the two main types of chromo-
some in the first spermatocyte, {c) Intermediate form.
70. Reconstruction of the nucleus of the spermatid.
Disappearance of the cell boundaries, x, remains of degenerated
cells.

Variation in the appearance of the spermatid nucleus.
Commencing elongation of the spermatid nucleus.
Group of spermatids from near the margin of the testis lobe. Ap-
pearance of cytoplasmic striaticns (hematoxylin and Bordeaux
red).
79. Further development of the spermatids.

Stage succeeding 72. (<?) Longitudinal; {b) transverse section
(iron hematoxylin).

Longitudinal section (iron hematoxylin).

Longitudinal section (Biondi-Ehrlich). The middle figure alone
is complete anteriorly.

{a) Longitudinal; (i^) oblique section (iron hematoxylin).
79, Nearly malura sperm colonies in incomplete longitudinal sec-
tion (hematoxylin and Bordeaux red).

Cross sections at different levels of nearly mature sperm colonies.
{a) anterior to the nuclear region ; (^b, c and d') nuclear region ;
((?), posterior to nuclear region (saffranin and malachite green,
black ■=^ red, gray r=r green).

1902.] NICHOLS— SPERMATOGENESIS ONISCUS ASELLUS LINN. Ill

8i. Cross sections of colonies of about the same stage (iron hoema-
toxylin).

82. Cross sections of colonies at a later stage (iron hcematoxylin).
Plate 83. Group of spermatids with convoluted nuclei. Cytoplasm of the
XVIII I individual cell still evident (iron haematoxylin).

84. Mature sperm colony (Delafield's hcematoxylin) (oc. 6, obj. D).

85. The same (hsematoxylin and Bordeaux red).

86. The same (oc. 4, obj. AA).

Literature List.
Note. — Those papers marked * I have not had an opportunity of examining.
1895. Auerbach, L. Spermatologische Mittheilungen : 72 Jahr. Schles.
Gesell. f. Vaterl. Cult.-Zool. bot. Sect., pp. 30-34.

1895. Ballowitz, E. Die Doppelspermatozoa der Dytisciden : Z. w. Z,, XLV.

1894. Ballowitz, K. Zur Kenntniss der Samenkorper der Arthropoden : Int.
Monat. Anat. u. Phys., il.

1892. Brauer, A. Das Ei von Branchipus Griibii von der Bildung bis zur
Ablage : Abh, preuss. Akad. Wiss.

1893. Brauer, A. Zur Kenntniss der Reifung des parthenogenetisch sich ent-
wickelnden Eies von Artemia salina. A. m. A., XLIII.

1885. Budde-Lund, G. Crustacea Isopoda Terrestna.

1885. Carnoy, J. B. La cytodierese chez les Arthropodes: La Cellule, Vol. I.
*i883. Friedrich, H. Geschlechtsverhaltnisse der Onisciden : Zeit. Natur-
wiss.

1882. Gerstaecker, A. Bronn's Klassen u. Ord. des Thierreichs Fiinfter Band,
II Abth.

1884-86. Gilson, G. Spermatogenese chez les Arthropodes: La Cellule,
Vols. I and II.

1878. Grobben, C. Beitrage zur Kenntniss der mannlichen Geschlechtsorgane
der Dekapoden: Arb. Zool. Inst. Wien.

1890. Hacker, V. Ueber die Reifungsvorgange bei Cyclops : Z. A., pp.

551-55^-

1892. Hacker, V. Die Eibildung bei Cyclops und Canthocamptus : Zool.

Jahr., V.

1895. Hacker, V. Die Vorstadien der Eireifung : A. m. A., XLV.
i890-'92. Henking, H. Untersuchung iiber die ersten Entwicklungsvorgange

in den Eiern der Insekten : Z. w. Z., XLIX, LI, LIV.

1883. Herrmann, G. Sur la Spermatogenese des Crustaces podopthalmes. C.
R. de I'Akad. des Sciences, 97, p. 958.

1883. Herrmann, G. Sur la Spermatogenese des Crustaces edriopthalmes. C.
R. de I'Akad. des Sciences, 97, p. 1009.

1898. Hoffmann, R. W. Ueber Zellplatten und Zellplattenrudimente : Z. w.
Z., LXIII.

1892. Ishikawa, M. Studies on the Reproductive Elements : Jour. Coll. Sci.
Imp. Univ. Japan, Vol. V.

*iSg$. Mari, M. Sopra la regenerazione delle spermatogonio nei crostacei
Decapodi : Bull. Soc. Ent. Ital. Anno 26, pp. 396-407.

112 NICHOLS — SPERMATOGENESIS ONISCUS ASELLUS LINN. [April 4,

*i895. Mari, M. Canateri delle cellula seminali nel Granchio di Fiume.
Contribuzione alia Istologia del vesticolo del Decapodi brachiuri. Ibid. Anno
27, pp. 3-10.

1898. Meves, F. Zelltheilung : Merkel u. Bonnet Erg., Anat. Hefte.

1898. Montgomery, T. H., Jr. The Spermatogenesis of Pentatoma, etc.
Zool. Tahrb., XII.

1899. Montgomery, T. H., Jr. Chromatin Reduction in the Hemiptera, a
Correction. Zool, Anz., Bd. XXII, No. 580.

1893. Moore, J, E. S. Some Points in the Origin of the Reproductive Ele-
ments in Apus and Branchipus. Q. J. Mic. Sc, XXXV.

1889. Miiller, G. W. Die Spermatogenese der Ostracoden : Zool Jahrb. 3.
Anat. Hefte., p. 677.

1884. Nussbaum, M. Ueber Veranderungen der Geschlechtsproducte bis zur
Eifurchung: A. m. A., Bd. 23. (Astacus.)

*i890. Nussbaum, M. Studien an californischen Cirrhipedien Bonn.

1899. Paulmier, F. C. The Spermatogenesis of Anasa tristis : Am. J. M.,
XV. Suppl.

1891. vom Rath, O. Ueber Bedeutung der amitotische Kerntheilung von
Ploden : Z. A., 14, pp. 342-363. (Astacus.)

1895. ^OT^ Rath, O. Neue Beitrage zur Frage der Chromatinreduktion in
den Samen-und Eireife : A. m. A., 46, p. 168.

1900. Richardson, Harriet. Synopsis of North American Invertebrates :
Amer. Nat., April, p. 303.

1894. Ruckert,J. Die chromatinreduktion bei der Reifung der sexualzellen •
Merkel u. Bonnet Ergeb. d. Anat. u, Entwick., Anat. Hefte. 2re Abt. Ill, pp.
544-548.

1894. Riickert, J. Zur Eireifung bei Copepoden : Merkel u. Bonnet, Anat.
Hefte, I Abt., XII Heft.

1885. Sabatier, A. Sur la spermatogenese des Crustaces Decapodes. C. R.,
100, p. 391.

*i892. Sabatier, A. De la spermatogenese chez les Crustaces Decapodes.
Nac. Inst. d. Zool. d. Montpellier, S. 2, Mem. 3.

1895. Steuer, A. Zoologische Ergebnisse VI. Sapphirinendes Mittelmeeres
und der Adria, etc. : Denkschr. Akad. Wien, Math. Nat. C. 62.

1886. Stuhlmann, F. Beitrage zur Anatomic der inneren mannlichen Ge-
schlechtsorgane und zur Spermatogenese der Cypriden. Z. w. Z., 44, p. 536.

1886. Wielowieyski, H. de. Spermatogenese des Arthropodes : Archiv.
Slavs de Biologie, T. II, p. 33.

1895. Wilcox, E. V. Spermatogenesis of Caloptenus and Cicada. Bull. Mus,
Comp. Zool. Harvard College, Vol. XXVII, No. i.

1900. von Winiwarter, H. Le Corpuscle Intermediaire et le nombre des
Chromosomes du Lapin : Arch, de Biol., T. 16.

1898. Woltereck, R. Zur Bildung und Entwicklung des Ostrakoden-Eies :
Z. w. Z., 64, p. 596.

1885. Zacharias, O. Ueber die amaboiden Bewegungen der Spermatozoen
von Polyphemus pediculus : Z. w. Z., XLI.

Magellanic Premium.

FOUNDED IN 1786, BY

JOHN HYACINTH DE MAGELLAN,
OF London.

1902,
The American Philosophical Society,

Held at Philadelphia, for Promoting Useful Knowledge

ANXOLWCES THAT IN

DECEMBER, 1902,

IT WILL AWARD ITS

MAGELLANIC GOLD MEDAL

to the author of the best discovery, or most useful invention, relating to Navi-
gation, Astronomy, or Natural Philosophy (mere natural history
only excepted) under the following conditions :

1. The candidate shall, on or before November i, 1902 deliver, free
of postage or other charges, his discovery, invention or improvement,
addressed to the President of the American Philosophical Society. No.
104 South Fifth Street, Philadelphia, U. S. A., and shall distinguish his
performance by some motto, device, or other signature. With his dis-
covery, invention, or improvement, he shall also send a sealed letter
contaming the same motto, device, or signature, and subscribed with the
real name and place of residence of the author.

2. Persons of any nation, sect or denomination whatever, shall be
admitted as candidates for this premium.

3. No discovery, invention or improvement shall be entitled to this
premium, which hath been already published, or for which the author
hath been publicly rewarded elsewhere.

4. The candidate shall communicate his discovery, invention or im-
provement, either in the Enghsh, French, German, or Latin language.

5. A full account of the crowned subject shall be published by the
Society, as soon as maybe after the adjudication, either in a separate
publication, or in the next succeeding volume of their Transactions, or
in both.

6. The premium shall consist of an oval plate of solid standard gold
of the value of ten guineas, suitably inscribed, with the seal of the Society
annexed to the medal by a ribbon.

All correspondence in relation hereto should be addressed
To THE Secretaries of the

American Philosophical Society,

No. 104 South Fifth Street,

PHILADELPHIA, U. S. A.

TRANSACTIONS

OF THE

American Philosophical Society

Held at Philadelphia,
^■" For Promoting Useful Knowledge.

Vol. XX, Nezv Series. Part /, 4to, pp. 62, with j Plates.
Lately Published.

CONTEJNTS.

Art, I. — The History of the Pelycosauria, with a Descrip-
tion of the Genus Dimetrodon Cope. By G.
Baur and E. C. Case.

Paid II. ^to, pp. ij/j, with 5 Plates,
yust Published.

CONTENTS.

Art. II. — The Chronological Distribution of the Elasmo-
branchs. By O. P. Hay.

Art. III. — Results of Observation with the Zenith Telescope
of the Sayre Observatory from January 19, 1894,
to August 19, 1895. By Charles L. DooLixgrLE.

Art. IV. — A Study of the Chromosomes of the Germ Cells
of Metazoa. By Thomas H. Montgomery, Jr.
Ph.D.

SUBSCRIPTION— FIVE DOLLARS PER VOLUME.

SEPARATE PARTS ARE NOT SOLD.

Address THE LIBRARIAN OF THE

American Philosophical Society,

No. 104 South Fifth Street,

Philadelphia, U. S, A.

^Vf-"^

PROCEEDINGS

A^xMERICAN PHILOSOPHICAL SOCIETY

HELD AT PHILADELPHIA FOE PROMOTING CSEFUL KNOWLEDGI.

Vol. XLI. April, 1902. No. 169

CONTENTS.

PAGB

Origin of the Oligocene and Miocene Deposits of the Great Pla5ns.

By J. B. Hatcher 113

On the Molluscan Fauna of the Patagonian Tertiary (with plate).

By H. YON Ihering 132

Spectra of Gases at High Temperatures. By Prof. Johj^ Trow-
bridge 138

The Influence of Alcoholic Intoxication upon Certain Factors Con-
cerned in the Phenomena of Haemolysis and Bacteriolysis.
A Preliminary Note. By A. C. Abbott and D. H. Beegey. 140

A Brief Account of the Disease Known as Osteitis Deformans.

By Prof. J. C. Wilson, M.D 143

Is Scientific Naturalism Fatalism? A One-minute Paper. By

William Keith Brooks 145

A Classification of Economies. By Prof. Lindley M. Keasbey. 146

Drift Casks in the Arctic Ocean. By Henry G. Bryant 154

Blindness from Congenital Malformation of the Skull (with plate).

By Charles A. Oliver, A.M., M.D y. 161

On the Continuity of Protoplasm (with plates). By Henry

Kraemer, Ph.D 174

On Some Equations Pertaining to the Propagation of Heat in an

Infinite Medium (with plates) ^^ By A. Stanley Mackenzie. 181

A New Method of Transiting Stars. By Monroe B. Snydee 200

^philadelphia :

The American Philosophical Society,

104 South Fifth Street.

1902.

It is requested that all correspondence be addressed

To THE Secretaries of the

AMEBIC A.]Sr PHILOSOPHICAL SOCIETY,
104 South Fifth Street,

Philadelphia, U. S. A.

Members will please communicate to the Secretaries any
inaccuracy in name or address as given on the wrapper of this
number.

It is requested that the receipt of this number of the
Proceedings be acknowledged to the Secretaries.

Members who have not as yet sent their photographs to the
Society will confer a favor by so doing ; cabinet size preferred.

U.'^

PROCEEDINGS

OF THE

AMERICAN PHILOSOPHICAL SOCIETY

HELD AT PHILADELPHIA FOR PROMOTING USEFUL EKOIVLEDGE.

Vol. XLI. April, 1902. No. 169.

ORIGIN OF THE OLIGOCENE AND MIOCENE DEPOSITS
OF THE GREAT PLAINS.

BY J. B. HATCHER.
{Read April 3, 1902.)

Skirting the base of the Rocky Mountains and covering the* sur-
face of the plains for some two or three hundred miles to the east-
ward is a series of Tertiary clays and sandstones with a combined
maximum thickness of over 1700 feet. This extends from the Rio
Grande in southern Texas to and beyond the northern limits of
the Black Hills in South Dakota, and covers the greater portion of
the plains of eastern New Mexico and Colorado, southeastern
Wyoming and western Texas, Oklahoma, Kansas, Nebraska and
South Dakota. Within this 1700 feet of Tertiary deposits there are
a number of different horizons, which are usually quite distinct
both faunally and lithologically. The more important of these
were long ago differentiated and given appropriate names by Hay-
den, Leidy, Cope, and others. If we exclude the Equus beds and
certain other deposits at the top, of Pliocene and Pleistocene age,
and which do not fall within the limits of this paper, this entire
series of rocks has been considered to belong to two formations,
the White River, or Oligocene, and the Loup Fork, or Miocene.
The White River, so named from a stream in northwestern Nebraska
and southwestern South Dakota, where it is particularly well repre-
sented, is the lowermost, and therefore the older of these two
formations. It has a maximum thickness of about 700 feet and
consists for the most part of very fine and usually unlaminated clays,
with frequent lenses of sandstones which in places become so coarse

PROC. AMER. PHILOS. SOC. XLI. 169. H. PRINTED MAY 24, 1902.

114 HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. [Aprils^

as to appear as conglomerates. Less frequently there are strata of
limestone. These are usually only an inch or two in thickness,
though occasionally attaining to as much as a foot. They are
always of quite limited extent laterally.

The White River formation has been subdivided into the Titan-
otherium and Oreodon beds, the former at the base, the latter at
the top of the series. The Titanotherium beds have a maximum
thickness of about two hundred feet, and are composed of very fine,
white, reddish- or greenish-colored clays with numerous lenses of
sandstones and conglomerates, not faunally distinguishable, how-
ever, from the clays. The Oreodon beds, with apparently slight
local unconformities, immediately overlie the Titanotherium beds.
They have a maximum thickness of five hundred feet and consist
of brown or pinkish-colored clays, banded but usually not lamin-
ated except at one or two horizons where distinct lamination is
plainly visible. The clays of the Oreodon beds are interrupted by
sandstone lenses, though less frequently than are those of the
Titanotherium beds, and the sandstones of the upper series are
usually of a much finer grain than are those of the lower. Toward
the bottom of the Oreodon beds in the Bad Lands of South Dakota,
there is a series of sandstone lenses known as the Metamynodon
sandstones. These sandstones are faunally distinct from the sur-
rounding clays. At the top of the Oreodon beds in the same
region, these sandstone lenses are replaced by a second series very
similar lithologically to the first, but quite distinct faunally. These
upper sandstone lenses have been called the Protoceras sandstones.
Their fauna differs not only from that of the lower Metamynodon
sandstones, but from that of the adjoining clays as well. While
the Metamynodon and Protoceras sandstones are faunally quite
distinct, both from one another and from the adjoining and under-
lying clays, sandstones and conglomerates of the Oreodon and
Titanotherium beds, they are, in so far as is at present known, of
extremely local distribution. At present neither of these two series
of sandstones has been recognized outside of a very limited area in
the South Dakota Bad Lands. Here they appear as lenses marking
the course of an ancient river channel, that in Oligocene times
crossed these plains in a direction almost at right angles to the
present courses of the Cheyenne and White Rivers, now the two
principal streams of this immediate region. On the same horizon
with the Protoceras sandstones and contemporaneous with them in

1002.] HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. 115

origin, there is in the South Dakota Bad Lands a series of pinkish-
colored nodular clays. These clays are faunally quite distinct, both
from the adjoining sandstones and the underlying clays of the lower
Oreodon beds. Unlike the Matamynodon and Protoceras sand-
stones, this upper series of clays is of wide distribution and has
been recognized in South Dakota, Nebraska, Wyoming and Colo-
rado. Dr. W. D. Matthew, in his most excellent memoir on the
Fossil Mammals of the Tertiary of Northeaster7i Colorado, has very
appropriately named these clays the Leptauchenia Beds, from a
genus of fossil mammals occurring abundantly in them.

From the above remarks it will readily appear thai the White
River formation may be separated faunally into three sub-equal
primary divisions. These are, commencing with the lowermost,
as fellows :

1. The Titanotherium Beds, consisting of 200 feet of fine, white
or greenish-colored clays with numerous intercalated lenses of
sandstones and conglomerates, the latter not faunally distinguishable
from the clays.

2. The Oreodon Beds, consisting of 300 feet of pinkish-colored,
banded and frequently nodular but usually unlaminated clays, with
less frequent lenses of finer sandstones, faunally distinct and known
as the Metamynodon sandstones.

3. The Leptauchenia Beds, consisting of 200 feet of pinkish-
colored, often nodular and banded, but unlaminated clays, includ-
ing the Proteceras sandstones above referred to.

The Loup Fork was the name given by Cope to a series of sand-
stones and clays well represented in western Nebraska and Kansas.
This formation has since been found to have a very wide distribu-
tion, and to extend almost uninterruptedly all along the eastern base
of the Rockies from Mexico to the Missouri River. It attains its
greatest development in southeastern Wyoming and northwestern
Nebraska, where it has a thickness of more than 1500 feet.

The sediments of the Loup Fork formation have not been so
thoroughly studied as those of the White River, and their faunal
and lithological characters are consequently less perfectly known.
The latest attempt at a differentiation of the various horizons within
the Loup Fork is that of Darton. Chiefly by lithological characters
he has divided the Loup Fork of northwestern Nebraska into three
divisions. Commencin.i{ below these are :

IK) HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. [April 3,

1. The Gering Sandstones. — These consist of some 200 feet of

laminated, massive and cross-bedded sandstones, found either
conformably or unconformably overlying the White River
formation at various localities in western and northwestern
Nebraska. They are well shown at the mouth of Monroe
Creek canon, some five miles north of Harrison, Sioux
County, Nebraska. Few fossils have been found in these
sandstones.

2. The Arikaree Sandstones. — These consist of some 500 feet or

more of light- gray, soft, massive sandstones, everywhere
characterized by numerous, flattened, horizontally columnar,
hard, dark-gray concretions. These concretions have an
average vertical thickness of about one foot ; they are fre-
quently several yards in width and often several hundred feet
in length. They have a general northwesterly and south-
easterly trend. The Arikaree sandstones are especially well
developed in the northern face of Pine Ridge, in Sioux
County, Nebraska, and Converse County, Wyoming. In
this region these sandstones may be conveniently subdivided
into an upper and lower series, easily distinguishable both by
faunal and lithological characters. These subdivisions in
the Arikaree will be referred to and fully described later.

3. The Ogalalla Fonnation. — This consists of a series of calcar-

eous grits, loose brown sands and clays with occasional coarse
conglomerates, the whole attaining to an aggregate maximum
thickness of 300 feet. It is the equivalent of the Goodtiight
(Palo Duro) beds of Texas and Kansas, and is especially
well developed in western Nebraska and Kansas, between the
Platte and Arkansas Rivers. It is usually referred to the
Pliocene, but a portion, or all of it, may yet prove to belong
to the Miocene.
Returning to the Arikaree formation, I have already remarked
that in Sioux County, Nebraska, and Converse County, Wyoming,
it is lithologically and faunally divisible into two easily distinguish-
able horizons. Commencing below, these may be named and
characterized as follows :

I. The Monroe Creek Beds. — These are well shown in the north-
ern face of Pine Ridge, at the mouth of the Monroe Creek
canon, five miles north of Harrison, Nebraska, where they
overlie the Gering sandstones, and are composed of some

1902.1 HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. 117

300 feet of very light-colored, fine-grained, not very hard,
but firm and massive sandstones. On account of their
usually barren nature they have been neglected by collectors,
and very little is known concerning their fauna beyond the
fact that toward the top they contain Promerycochoerus.
They decrease in thickness very rapidly to the eastward and
increase to the westward.
2. The Harrison Beds. — These are well shown in the bluffs of
all the small streams that head near the summit of Pine
Ridge, in the vicinity of Harrison, Nebraska. They are
also known to cover a considerable area to the east, west and
south of that village, extending well into the State of
Wyoming. They are composed of about 200 feet of fine-
grained, rather incoherent sandstones, permeated by great
numbers of siliceous tubes arranged vertically rather than
horizontally. They are further characterized by the presence,
often in the greatest abundance, of those peculiar and inter-
esting, but as yet not well understood, fossils known as
Dsemonelix, and by a considerable variety of fossil mammals
belonging to characteristic Miocene genera. They imme-
diately and conformably overlie the Monroe Creek beds and
pass insensibly into them. Above these come :
The Nebraska Beds, of Scott. — These consist of a series of buff-
colored sandstones of varying degrees of hardness and un-
known thickness, with occasional lasers of siliceous (not
calcareous) grits, which protrude as hard, indurated or shelv-
ing masses from the underlying and overlying softer materials.
These beds are rich in vertebrate fossils, such genera as
Cosoryx, Protolabis, Cyclopidius and Merycochoerus pre-
dominating. They are represented at various localities along
the Niobrara River, south of Harrison, Nebraska, where
they are of unknown thickness and immediately overlie the
Harrison beds. Toward the south they pass beneath the
Ogalalla formation.
According to the above classification all the Miocene deposits of
this region are referred to the Loup Fork, notwithstanding their
great thickness and, in certain localities at least, their apparent con-
formity with the underlying Oligocene deposits, and without regard
for the fact that throughout the lowermost 500 to 1500 feet of these
sediments there is as yet practically no paleontological evidence as

118 HATCHER — OLIGOCENE AND MIOCENE DEPOSITS, [Aprils,

to their exact age and correlation. Considering the absence of
such direct paleontological evidence, it may be just as well to con-
tinue to refer this entire series to the Loup Fork; but I believe it
more probable that the Gering sandstones, and perhaps a portion at
least of the overlying Monroe Creek beds, will prove eventually to
belong to the John Day rather than the Loup Fork. The maximum
thickness of these two formations in Converse County, Wyoming,
can hardly be less than 1500 feet, and almost nothing is known of
the fauna of this entire series. Although for the most part quite
barren of fossils, it would seem that somewhere throughout its great
vertical and lateral extent there must be fossiliferous horizons, and
that within these representatives of the John Day fauna will yet be
found. The paucity of these beds as compared with the great
wealth of fossils in the underlying and overlying deposits, have
heretofore caused them to be almost totally neglected by collectors.
I believe a better classification of these beds would be obtained by
making Darton's Arikaree coordinate with the Loup Fork, includ-
ing within it the Gering sandstones and Monroe Creek beds, cor-
relating it provisionally with the John Day.

The following table is submitted as expressing the present author's
views as to the proper classification of the Oligocene and Miocene
deposits of this region. It is based on our present knowledge of
the faunal and lithological characters of the various horizons as
they have been determined, chiefly in northwestern Nebraska and
southwestern South Dakota, where these deposits are best repre-
sented and have been most thoroughly studied.

Table of Oligocene and Miocene Formations of Western

Plains.
f Goodnight = Palo Duro = Ogalalla.

(Nebraska = Upper Deep River,
Harrison z= Hiatus between Lower^and Upper Deep
River.

{Monroe Creek = Upper John Day and Lower Deep
River.
^ Gering Sandstones r= Lower John Day.

p Leptauchenia Clays, including Protoceras Sand-
I stones.

Oligocene = White River. I Oreodon Clays, including Metamynodon Sand-
I stones.

[ Titanotherium Sandstones and Clays.

1902.] HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. 119

The writer is well aware that the above correlation of the Gering
and Monroe Creek sandstones is open to criticism, as being at
present inadequately substantiated by direct paleontological evi-
dence. However, it should be remembered that on the other hand
there are no direct paleontological evidences against such correla-
tion, and that since sedimentation seems to have been continuous
at certain localities in this region, from the base of the White River
to the top of the Loup Fork, the John Day should be represented
somewhere in the series, and that the lithological sequence, as well
as the faunas of the overlying and underlying rocks, point to the
Gering and Monroe Creek sandstones as the logical representatives
of the John Day formation in this region.

Origin of the Deposits.

Until very recently the sediments of this entire series of deposits
have been very generally considered as of lacustrine origin, and the
boundaries of these supposed great Oligocene and Miocene lakes
have been set forth in text-books and scattered papers, and espe-
cially in the classroom lectures on the subject at our various uni-
versities, with a preciseness only surpassed by that of the modern
geographer when dealing with existing lakes.

The earlier writers, including David Dale Owen, King, Hayden,
Leidy, Cope, Marsh and others, were always accustomed to speak
of these deposits as lacustrine, and they are at present so considered
l)y many authorities. Recently, however, their lacustrine origin
has been rejected, at least partially, by a considerable number of
competent observers, several of whom have had most excellent
opportunities for studying them. This is especially true of the
upper or Loup Fork series of deposits, which has now come very
generally to be considered as of combined lacustrine, fluviatile,
flood-plain and seolian origin, instead of as having been laid down
over the bottom of a great and continuous body of water, as was
formerly supposed.

With regard to the origin of the underlying White River series,
however, it has been different ; and with a few exceptions these
deposits are still regarded as of lacustrine origin. Dr. W. D.
Matthew, in an article entitled ''Is the White River Tertiary an
^olian Formation?" published in the American Naturalist, for
May, 1899, was the first to seriously question the lacustrine origin
of these deposits. In his " Fossil Mammals of the Tertiary of

120 HATCHER— OLIGOCENE AND MIOCENE DEPOSITS. [AprU 3,

Northeastern Colorado," published as Part VII of Vol. I of the
Me7fioirs of the Atfierican Museum of Natural History , Dr. Mat-
thew has set forth additional facts in favor of his seolian theory as
to the origin of the deposits, which, if not furnishing conclusive
evidence as to the correctness of his theory, at least make it very
clear that the lacustrine theory is alone unable to explain many well-
known facts relating to the nature of these deposits and the distri-
bution, condition and nature of the animal remains found in them.
W. D. Johnson, in his paper on *' The High Plains and their Utili-
zation," published in the Twenty-first Annual Report of the United
States Geological Survey, has entirely ignored the lacustrine theory
of the origin of any of the Tertiary deposits of the plains, holding
that they are of fluviatile and flood-plain origin, while Dr. J. C.
Merriam, in a recent paper on "The Geology of the John Day
Basin," rejects the lacustrine theory of the origin of those deposits,
which had previously remained unquestioned. The above are the
leading authorities among those who have questioned the lake
theory as to the origin of these beds. On the other side the lacus-
trine origin of the rocks of the White River series, at least, has
been maintained by Todd, Scott, Darton and others, though
none of these authorities have thought it worth while to support
their contentions by the production of any considerable direct or
indirect evidence bearing on the case. Like the earlier writers they
have, almost without exception, set forth their views as if they were
well-established facts and beyond question or criticism. The fol-
lowing quotation from Scott is a fair example. In speaking of the
Oligocene series, on page 507 of his Introduction to Geology, he
says : '^ But in the interior regions are extensive fresh-water deposits
which clearly should be referred to it and which form the White
River stage. The largest body of water of this time occupied
northeastern Colorado, southeastern Wyoming, much of western
Nebraska and South Dakota." But the limits of this supposed
Oligocene lake have lately been greatly extended by Darton, who
has contended that it covered all of eastern and central Wyoming,
and a considerable portion of Montana and North Dakota'as well ;
so that one is at a loss to understand where lived the terrestrial
mammals and reptiles whose remains are now found in such abund-
ance in the deposits.

The lacustrine theory had its origin in the until recently univers-
ally accepted idea that all sedimentary rocks showing stratification;

1902.] HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. 121

or bedding were deposited in either marine or fresh waters. More-
over the color-bands exhibited, more especially by the clays of the
White River series, have been very generally mistaken for examples
of stratification and lamination, while true lamination in the clays
of this series is rare and usually of very limited extent both verti-
cally and horizontally.

Dr. Matthew, in his Mejnoir already referred to, has set forth in
very clear and concise language the principal stratigraphic and
paleontologic evidences against the lacustrine theory as observed by
him for these deposits in northeastern Colorado. It will be the
chief purpose of the succeeding pages of this paper to extend these
observations into southeastern Wyoming and western Nebraska and
South Dakota, and to record some additional facts relating to the
stratigraphy, paleontology and paleobotany of the beds, with espe-
cial reference to their bearing upon the origin and mode of depo-
sition of the latter.

Matthew has already called attention to the physical and topo-
graphical difficulticF, as well as to the lack of terraces and of certain
stratigraphical characters which should exist if these deposits had
their origin in a body of fresh water of a size comparable with that
outlined by Scott. These difficulties, already serious, are only
augmented by the increased dimensions of this lake proposed by
Darton. If we confine it, however, to the much more restricted
limits given by Dr. Scott we still have a lake of very considerable
dimensions, greatly exceeding in size those of any fresh-water lake
of modern times, with no barrier to the east or south to retain its
waters, without recognizable terraces about its shores, and with a
distribution of materials and of remains of fresh water, and of
terrestrial plants and animals which are at least difficult, if not
impossible, of explanation by the assumption of the presence of a
great lake.

Character of the Materials in the White River Series.

We have already observed, while discussing the classification of
the White River beds, the presence in them of frequent lenses of
sandstones and conglomerates. These sandstone and conglomerate
lenses are not arranged concentrically at varying altitudes about the
margins of this supposed great lake, but extend as greatly elongated
and narrow lenses far out into the very centre of the region which
this lake has been supposed to have occupied. They occur at all

122 HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. [Aprils,

horizons, show frequent examples of cross-bedding and their irregu-
lar course, as well as the spheroidal shape of the pebbles they con-
tain/ and the increased fineness of the materials of which they are
composed as one proceeds from the margin toward the interior, are
all characters strongly suggesting that they were deposited in river
channels. Moreover, the materials of these sandstone and con-
glomerate lenses are not only coarser about the western borders of
the beds, but the lenses are far more numerous in that region.
Toward the interior these lenses converge and unite without spread-
ing out laterally, so that in the region lying east of the Black Hills
in South Dakota, at a distance of fifty to seventy-five miles from
the mountains, the sandstones are finer, less frequent and are sepa-
rated by greater areas of fine clays, just as the streams of the present
day unite and become fewer in number as we proceed farther from
their sources.

The Metamynodon and Protoceras sandstones, as well as certain
intermediate and underlying sandstones, present many evidences,
like those just enumerated, which strongly suggest that they were
deposited in river channels. Taking the Protoceras sandstones as
the most favorable example, owing to the greater extent to which
they have been exposed by the subsequent erosion of the overlying
sediments, they are seen to extend as a series of narrow elongated
lenses from the summit of the Cheyenne and White River divide
for several miles to the southward of the last-mentioned stream,
where they pass beneath more recent deposits. Throughout their
entire extent they exhibit frequent examples of cross-bedding,
while the sands become finer and the channels fewer in number and
broader and deeper as ones goes southward toward and across
White River. That they have been removed by erosion over con-
siderable areas lying between their present limits and the Black
Hills is evident. At the summit of the Cheyenne and White River
divide there are several of these sandstone lenses at approximately
the same horizon. These bear many evidences of having been
deposited in the channels of small streams or rivers pertaining to a
single drainage system, which had its source somewhere in the

^A conglomerate accumulated by a running stream can usually be distin-
guished very readily from one formed on the beach of a lake or sea, by the shape
of the contained pebbles. In the first instance the pebbles have been reduced to
irregular spheroids by the rolling motion to which they have been submitted by
the current. In the second they are more generally flattened disks.

1902.] HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. 123

present region of the Black Hills and was tributary to a much
larger river coming from the southwest. These sandstone lenses
appear to converge and unite as one proceeds toward White River,
like the tributaries of recent streams. I am at a loss to understand
how these greatly elongated sandstone lenses, confined laterally to
at most only a few hundred yards in breadth, and necessitating the
presence of strong currents, could have been deposited in the bot-
tom of a great lake. For they appear not only to extend quite
across the entire region which this lake has been supposed to have
occupied, but these or very similar sandstones are found at inter-
vals throughout the entire vertical and lateral extent of the beds,
although as one recedes eastward from the western border they
become less frequent and of finer grain. Such difficulties as those
just mentioned, together with others to be referred to later, long
ago demonstrated to the present writer the untenable nature of
the lake theory as to the origin of these deposits.

If these beds had their origin in a great lake it may very natu-
rally be asked. Where are the remains of the aquatic fauna which a
lake of such dimensions may very reasonably be supposed to have
contained ? The reply has been made, and will be forthcoming
from advocates of the lake theory, that the waters of this great lake
were of such a saline or alkaline nature that it was incapable of
supporting life. Hence the absence of the remains of aquatic
animals. But I shall show presently that such bodies of water as
did exist in this region during the deposition of these beds were
not only not of such a nature, but that they were eminently fitted
for the support of aquatic life and did in fact support such life, both
plant and animal, in great abundance.

Again, if a lake deposit, how did the remains of terrestrial mam-
mals and reptiles receive their present distribution throughout these
beds? It has been maintained by advocates of the lake theory
that the fine-grained, banded clays were deposited in the deep and
quiet waters of the lake and the sandstone and conglomerate lenses
along the shores and about the mouths of tributary streams, while
the preservation and distribution of the remains of terrestrial mam-
mals and reptiles was accomplished by the drifting about in the
lake of dead carcasses brought down by the tributary streams. Such
a theory requires conditions which are not only quite unreasonable
but unparalleled elsewhere, both in the deposits of the lakes and
seas of the present day and those of past geological epochs. Fur-

124 HATCHER— OLIGOCENE AND MIOCENE DEPOSITS. lAprilS,

thermore it not only does not account for but is actually opposed
to the present distribution of the fossils. If, as we are told, the
fine clays were deposited over the deeper and quieter waters of the
lake and the sandstones and conglomerates about the mouths of
rivers and along the shores, why, I may ask, is it that the former
contain absolutely by thousands the remains of giant land-tortoises,
while these if not entirely absent are conspicuously rare in the
sandstones, while the few examples of Trionyx, an aquatic turtle,
have, in so far as I know, all been recovered from the sandstones ?
I have myself collected a number of these latter from the sand-
stones. If the land-tortoises were brought into the lake by the
rivers, ought we not to expect that their remains would be found in
at least as great an abundance in the sandstones as in the clays ?
Again, while it is quite possible to conceive of even a huge animal
of such elephantine size as was Titanotherium as having met death
by drowning or otherwise in or near some stream, where the dead
body inflated by gases would be carried out by the current into the
waters of the lake to sink later, allowing the bones to be preserved
in the clays at the bottom, it is difficult to understand how such
examples could be other than exceptional, and it is totally incapable
of explaining the present distribution and abundance of such bones.
In such a case as that just supposed it seems quite probable that
once decomposition had proceeded far enough to weaken the body
walls sufficiently to permit of the escape of the confined gases, the
carcass would sink to the bottom and the bones of the skeleton be
preserved in approximately their normal position relative to one
another, just as are the skeletons of marine reptiles in the chalk
beds of western Kansas or at Lyme Regis in England. If this
were the case we should expect to find complete skeletons at least
fairly common, but they are in fact exceptionally rare, and for
every even approximately complete skeleton to be found there are
scores of isolated skulls and other bones. Taking Titanotherium as
an example, I have myself collected nearly two hundred skulls of
this animal, while the number of fairly complete skeletons at pres-
ent known may be counted on the fingers of one hand. What
is true of this animal applies likewise to the others found in
the beds.

But, it will be asked, if the lake theory is so objectionable, why
do you not offer a better ? Such has already been done by
Matthew, and it is the purpose of the present paper to support

1902.] HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. 125

in a somewhat modified form the theory advanced by him,
extending his observations and adding certain additional facts
observed by the present writer and bearing directly upon the
subject.

Matthew's theory as to the origin of these deposits may best be
described in his own language. Speaking of the conditions attend-
ing the deposition of these beds, he says : " The nature of the
organic remains, where such have been found, seem to definitely
negative the idea of any vast lake, and to favor less the theory of a
series of lagoons and swamps than that of a broad, open and com-
paratively level plain, with shallow, probably wooded, rivers
meandering over parts of it and deposits partly or chiefly brought
by rivers, but in large part redistributed over the higher sodded
grassland by the agency of the wind." With most of the
principal features of this theory as applied by its author the present
writer is in accord. I believe, however, that the materials on the
whole partake more of the nature of fluviatile and flood-plain
deposits than of those characteristic of prairie loess.

Paleontological Evidences.

The distribution, state of preservation, nature and character of
the animal and plant remains found in the clays and sandstones, as
well as the distribution of the latter, absolutely preclude the possi-
bility of their having been deposited in a vast lake and favor the
presence of streams meandering through low, broad, level, open or
wooded valleys subjected in part at least to frequent inundations,
conditions very similar to those at present prevailing in the interior
of South America, about the headwaters of the Orinoco, the
Amazon and the Paraguay and Parana Rivers.

Now it is evident that if such conditions prevailed in this region
during the deposition of the White River beds there should remain
certain evidences concerning it, such as filled-in river channels and
small lagoons with their characteristic deposits and remains of the
animal and vegetable life peculiar to each. Moreover some indica-
tion at least of the forests should remain and be found somewhere
in this vast region. With these and many other points constantly
in mind the writer passed a considerable portion of the seasons of
1900 and 1 90 1 in exploring these deposits. Particular attention
was given to ascertaining whether or not they contained an aquatic
fauna and flora. The sandstone lenses were especially examined

126 HATCHER — OLIGOCENE AND MIOCENE DEPOSITS, [AprU 3,

with reference to this, for whether the deposits as a whole were of
lacustrine origin or not, there could be little doubt as to the
aqueous origin of the sandstones. Though for the most part
remarkably barren of aquatic life, remains of Trionyx, fishes and
crocodiles were found, and in one locality the casts of unios were
observed in great numbers. A search in the clays of the Titano-
therium and overlying Oreodon beds was rewarded with greater
success, for numerous thin layers of limestone, varying in thickness
from a fraction of an inch to a foot or more and always of limited
areal extent, were discovered at many horizons rich in the remains
cf fresh-water plants and mollusca, such characteristically shallow-
water forms as Chara, Limnaea, Physa and Planorbis occurring in
the greatest abundance. I have submitted these mollusca to Drs.
Dall, Pilsbry and Stanton, and all have assured me that they belong
to species inhabiting swamps and small ponds, and could not have
lived in the midst of a great lake ; while Dr. Knowlton, who has
examined the plants, finds in great abundance the stems and seeds
of Chara, which, as all know, is distinctly an inhabitant of small
springs, shallow ponds and brooks. The presence of these thin
limestone layers with such characteristically swamp plants and
mollusca as are Chara and Physa at various horizons throughout the
White River series, and in the very midst of the region which was
supposed to have been occupied by a great lake, and intercalated
with the clays which advocates of the lake theory maintain were
deposited in the deep and quiet waters, would appear to preclude
the possibility of the existence of such a lake in White River
times. Moreover remains of forests were found at several places
and at different horizons throughout these beds. At various locali-
ties in the Hat Creek basin in Sioux County, Nebraska, I discov-
ered remains of the silicified trunks of trees and seeds belonging
especially to Hickoria and Celtis. These were found at various
horizons from the middle Titanotherium beds to the very top of the
Loup Fork. And in South Dakota, some twelve miles north of
White River, opposite the mouth of Corn Creek, I discovered the
remains of a no inconsiderable forest. Here in the upper Titano-
therium beds and lower Oreodon beds there occur, actually by
hundreds, the silicified stumps and partially decayed trunks of trees,
weathering out of the fine clays of these deposits. It was notice-
able that only the knots and lower stumps had been preserved.
Nothing like complete trunks were to be observed, and the entire

1902.] HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. 127

aspect was that of the remains of a dead and decayed forest on the
margin of some stream, where only the less destructible knots and
stumps would endure sufficiently long to be finally covered up and
preserved. In this same region there were discernible certain
strata which seemed to indicate that during the deposition of these
beds there had been at several horizons an accumulation of vegeta-
ble mould or humus, and on Dry Creek, some five miles northeast
of Chadron, in Dawes County, Nebraska, I observed near the base
of the Oreodon beds a stratum of some two feet of dark-colored
humus, clearly indicating that this region had not been occupied
by a great lake while this stratum was being deposited.

The advocates of the lake theory have always maintained that
the fine clays of the Oreodon and Titanotherium beds were
deposited in the deep and quiet waters of the lake, explaining the
absence of the remains of an aquatic fauna, such as a lake of so
great dimensions might in all reason be expected to maintain, on
the theory that this lake was of such a saline or alkaline nature as
to render its waters uninhabitable by crocodiles, turtles and fresh-
water fishes. But I have shown that the remains of such animals
do occur, though sparsely, wherever there is evidence of sufficient
water to maintain them. The character and abundance of the
mollusca and aquatic plants found in the thin limestone lenses
throughout the clays show that such bodies of water as were pres-
ent, although limited in area, were eminently well adapted to fresh-
water life. The great abundance of land-tortoises in the clays and
their almost complete absence in the sandstones is very strong if
not positive evidence that the former were not deposited in the
bottom of a great lake, for I do not believe that any one will
assume to explain the present distribution of the remains of these
land-turtles on the lake theory. After a careful consideration of
the materials composing the White River deposits and the distribu-
tion and character of the fossils throughout the sandstones, con-
glomerates, clays and limestones, the present writer believes that
the sandstones, conglomerates and a portion of the clays were
deposited in river channels, while the limestone lenses, so rich in
the remains of aquatic plants and mollusks, originated in shallow
ponds and lakes scattered over the higher table-lands and the broad
flood-plains of the rivers, where for the most part the finer clays
were deposited by occasional inundations and through the agency
of the winds. Such a theory as to the origin of the White River

128 HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. [April 3,

beds appears to the writer not only to be in harmony with all the
observed facts, but moreover the conditions which it presupposes
are paralleled by existing conditions in other parts of the earth's
surface.

The following description of the conditions at present prevailing
about the sources of the Parana and Paraguay Rivers in central
South America has been furnished me by Mr. H. H. Smith,
who has spent several years in that region and has had exceptional
advantages for studying the physical conditions that obtain about
the headwaters of these streams and their tributaries. He says :

*' Ascending the River Paraguay from Asuncion, the river hugs
the higher lands of the eastern or Paraguay side or is separated from
them by strips of alluvium. On the western or Chaco side the
ground is always low and flat, hardly above reach of the annual
freshets and proportionally a little lower toward the north.
During the rains water covers large spaces of these flatlands, but
it does not come from the river and is gradually drained away
after the rains cease. Above the mouths of the A^ermejo the
Chaco bank is at first covered with low forest ; farther north
great areas have a scattered growth of Caranda palms with grass
beneath, but with no other vegetation. The Chaco plains extend
far inland to the table-land of Bolivia, which is said to fall abruptly
to the plain.

*^ At latitude 21° 26' 40'' S. the river flows through a narrow pass,
the Fecho dos Morros, between two rocky hills. The hill on the
eastern side is connected by high ground with the Brazilian table-
land. That on the western side appears from the river to form one
of a number of isolated hills which rise from the Chaco plain. It
may be, however, that there is rocky ground extending westward to
the Bolivian table-land, and perhaps connected with the Corumba
hills. This region, however, is practically unexplored, and nothing
definite can be said about it. If there is a connection with the
Bolivian highland, the basin of the upper Paraguay is completely
enclosed like a lake, with only the narrow outlet at the Fecho dos
Morros. If the hill on the western side is isolated, it is probably
one of a chain which extends inland and imperfectly closes the
Paraguay basin on this side.

*' Above the Fecho dos Morros the character of the vegetation
changes ; the Caranda palms disappear ; there is left only open
grassland, with lines of bushes here and there and often a thin

1902.] HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. 129

fringe of forest on the river bank. The river at the flood season
covers these lands almost entirely. It must be remembered that the
upper Paraguay rises about thirty feet annually.

" All the flatlands above the Fecho dos Morros to Villa Maria —
over four hundred miles in a direct line — are subject to river floods,
and these are deepest toward the north. The width of the flood-
plain at the mouth of the Sao Louren^o can hardly be less than one
hundred and fifty miles from the rocky lands on the east to the base
of the Serra dos Dourados. The whole region is a labyrinth of
lakes, ponds, swamps, channels and islands in a grassy plain, the
only forest being near the river. I had a fine view of this plain
from the foothills of the Dourados ; even the flood-plains of the
Amazon cannot compare with it in its tangle of land, water and
marshes. Only the most experienced canoeman can thread his way
through it ; generally travelers trust to the Guato Indians, who are
the only inhabitants of the region and literally live in canoes.
Castelnau was lost there and only found the river channel with
great difficulty. We were lost or partly lost for a few hours, though
we had three experienced hands.

'^ This is the region called Lake Xaraes, or Charaes, by the old
explorers ; Brazilians called it the Pantanaes, literally The Marshes.
Even at low water at least one-fourth of it is flooded : when the
river is at its highest the whole plain is a vast lake covered with
floating grass and weeds ; it is possible to pass almost straight
across it in a canoe, though with great difficulty. Only a few
islands remain here and there ; jaguars, deer and ether animals
take refuge on them, and they are favorite hunting grounds of
the Guatos.

'^ The rainy season is from October to April, the heaviest rains
being toward the last; the small rivers from the highlands are
flooded in March and April, and pour their waters over the flood-
plain. But it takes a long time for these waters to spread over
the plain. Consequently the highest waters on the plain are
in July and August. Then they gradually drain away through
the Fecho dos Morros, and the lowest waters are found about
February.

"The eastern and northern sides of the flood-plains are bordered
by low rocky lands which extend for a few miles inland : then they
rise precipitously 1500 or 1800 feet to the Brazilian plateau. The

PROC. AMER. PHILOS. 80C. XLI. 169. I. PRINTED JUNE 10, 1903

130 HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. [Aprils,

line of precipices forming the plateau may be traced from Villa
Maria to near Miranda.

'' On the Dourados side the case is different. Long bays of the
flood-plain run back among the hills and often contain lakes of
considerable size. The Dourados chain itself is narrow and on the
other side are more floodlands, the region called ''Ceo e Terra " by
the Brazilians. The Brazilian-Bolivian Boundary Commission tried
to pass over this land but had to turn back.

" The Paraguayan flood-plains are connected with the Ceo e Terra
region by several strips which cut the Dourados chain. Hence the
Dourados are like a chain of islands.

" A narrow neck of rocky but low land divides the Paraguayan
flood- plain from similar plains in the Guapore ; those on the
Guapore are continuous to the junction of the Madeira where
there are rapids ; beyond that a flood-plain extends to the Amazon,
broadening out. The Amazonian plain is connected in much the
same way with the Orinoco. The Orinoco, Amazon and Paraguay
are like each other in their flood-plains, which broaden out as we
ascend the rivers.

"■ Except for a single break at the Fecho dos Morros (which may
not be a continuous wall) a great plain stretches from Villa Maria
to Rosario and beyond. The Xaraes, Chaco and Pampa differ
only in their relations to the river floods. The pampas are above
reach of the floods \ the Chaco plains are also above the floods, but
so low that the water drains off slowly ; the Xaraes are covered at
high water. The Paraguay and its continuation, the Parana, cut
deeper and deeper into the plains as they flow southward,
hence the difl'erences in physical features, which are more apparent
than real."

From the above description it will be seen that the flood-plain
of the Paraguay at the mouth of the Sao Louren^o has a width of
one hundred and fifty miles and that it broadens as we ascend the
river. It is a well-known fact, as stated by Mr. Smith, that the
flood-plains of the upper Paraguay, Amazon and Orinoco Rivers
are confluent, and that a vast region about the headwaters of those
streams possesses physical conditions in every way similar to those
just described as obtaining over the flood-plains of the upper
Paraguay. Here it appears to the present writer we have a region
of equal or greater area than that occupied by the Oligocene and
Miocene deposits of our Western plains, with all the physical

1902.] HATCHER — OLIGOCENE AND MIOCENE DEPOSITS. 131

conditions necessary for the deposition and present distribution of
the sandstones, clays and conglomerates, together with the preser-
vation of remains of the faunas characteristic of each.

Many have noticed and Prof. J. E. Todd has recorded the
presence of great deposits of bones at various localities in the
White River beds. He describes them as literally covering the
ground in places where they have weathered out over areas
frequently of more than an acre in extent. It is not only difficult,
but I think impossible to account for these accumulations of bones
of terrestrial animals at the bottom and in the very middle of a
great lake. Since the surrounding clays are usually almost destitute
of bones, it is difficult to understand how the dead carcasses of so
many animals were driven or drawn as by a magnet to so limited an
area. Accepting the other theory, however, we have seen how
during the rainy season the deer, tapirs and other animals are
driven to the islands over the flood-plains of the great South
American rivers. Since in exceptionally high freshets the lower of
these islands become submerged it is not difficult to understand how
great numbers of these animals must annually perish, and indeed it
is a well-known fact that frequently great numbers of them are
caught on low islands and, driven by the rising waters to more
limited confines, they are finally all drowned when the island
becomes entirely submerged. To such or similar conditions the
great deposits of bones in the Oligocene and Miocene deposits of
the West may owe their origin. I have frequently observed these
deposits, though not covering so great an area as that recorded by
Todd, and I have always without exception noted that in the
Oligocene beds they occurred in the very fine clays, while in the
upper or Miocene deposits they occur in the finer sandstones.
Although bones are fairly abundant in the sandstones of all these
series of beds, I never observed these extremely rich deposits in the
coarse sandstones or conglomerates.

The above facts, together with those brought forward by Dr.
Matthew, have driven me, contrary to my earlier opinion, to reject
the theory of a great lake and accept that of small lakes, flood-
plains, river channels and higher grass-covered pampas as the
conditions prevailing over this region in Oligocene and Miocene
times.

Carnegie Museum, March j/, igo2.

132 VON IHERING — MOLLUSCAN FAUNA OF PATAGONIA. [Aprils,

ON THE MOLLUSCAN FAUNA OF THE PATAGONIAN

TERTIARY.

BY H. VON IHERING.
(Plate XIX.)

{Read April 3, 1902.)

During the last ten years the exploration of the Patagonian and
Argentine Tertiary has been very actively prosecuted, but the
results of the new studies have not always represented genuine
progress.

This refers particularly to the deposits of Entre Rios, which
Alessandri regarded as Eocene from his studies of Selachian teeth,
while A. Smith Woodward, reexamining the same material, came
to the conclusion that this formation is Miocene or Pliocene. The
study of the Mollusca of the Entre Rios beds led me to the opinion
that they are Miocene, while Borchert, in view of the large propor-
tion of recent species in this fauna, refers the formation to the
Pliocene.

Having at my disposal one of the best collections of marine
shells from the Brazilian and Argentinian coasts, I have carefully
examined Borchert' s work. This author has had access to a much
richer collection of Entre Rios moUusks than I myself, in which are
represented well-preserved examples of some species — e. g., Cardium
magnum Born — of which I have seen only casts. This circumstance
does not, however, entirely explain the divergence of our opinions,
which is rather due to a number of incorrect determinations by
Borchert. The small shell which he believes to be Dione purpurata
Lam. is not referable to that species. The opinion that the young
shell differs in outline from that of the adult is refuted by the large
series of specimens in my collection. It is not reasonable to attach
much importance to such young shells, especially if represented
only by a single specimen.

'* Cryptogramma brasiliana Gm." is not this species at all, but a
different and extinct species, characterized by its numerous coarse
concentric ribs, as also by numerous radiating strice. Beside these
two species, the following are certainly erroneously determined :

1902.J VOX IHERING — MOLLUSCAN FAUNA OF PATAGONIA. 133

Lithodomus patagonicus d'Orb. Solecurius platensis d'Orb.

Modiola iulipa Lam. Marginella prumivi Gm.

Nucula piielchana d'Orb. Columbella acuta Stearns.
Tivela argentina Sowb.

Of the nineteen species which the author regards as still existing
at least nine are incorrectly determined. Admitting that the
remaining ten species are accurately determined, the proportion of
living species in the Parana formation is as lo : 60, or 17 per cent.

Borchert's attempt to prove that the Parana formation is Pliocene
must be regarded as a failure ; it is opposed to the opinion of
d'Orbigny, Philippi and von Ihering, all of whom were well
acquainted with South American marine MoUusca, both recent and
fossil. We must continue to regard the Parana formation as
Miocene, while the Pliocene of Argentina is represented in the south
by the Cape Fairweather beds, and in the north by the Tehuelche
formation. Of the latter the new collection of Carlos Ameghino
contains an instructive representation, with many new species of
Pectinidce, Carditidce and Veneridce.

From the marine deposits in the upper part of the Pampean
fortnation I have a relatively large collection. As the species
are all still living, I am obliged at present to consider the " Pam-
peano superior" of Ameghino as Pleistocene, It is interesting to
observe that in this formation are two different horizons, the older
of which contains Ostrea arborea Ch., Purpura hcBtnastoma L.,
and other species now common on the Brazilian coast, but which
are wanting in the later horizon, where they are replaced by Ostrea
puelchana d'Orb. and other Patagonian species.

With regard to the Patagonian formation, many new forms are
contained in a large collection sent to me this year by Dr. Floren-
tino Ameghino, and gathered by his brother Carlos in the years
1899 and 1900. This great collection contains representatives of
three different faunas, but the greater part of it comes from the
Patagonian formation. As I have studied many important collec-
tions from this formation, and as my friend Dr. A. E. Ortmann has
also discovered many new forms in the collections of the Princeton
Expedition, it was very surprising to find a great number of new
and interesting forms in this new collection, I reserve the descrip-
tion of these new species until Dr Ortmann's report has been pub-
lished, and describe here only two of the most striking new species.

134 VON IHERING — MOLLUSC AN FAUNA OF PATAGONIA. [April a-

Nautilus caroli-ameghinoi, sp. nov.

(Plate XIX, Figs, i, 2.)

Nautilus testa suborbiculari, imperforata, leevi, suturis simplicibus,
siphone interno, in fundo foss^ latae situ.

The shell, which is filled with matrix and not very well preserved,
is of suborbicular shape, subcompressed and much enlarged toward
the aperture. The outer or peripheral part of the shell is rounded,
smooth and with simple sutures. There is no umbilicus, but a pit
at the central point of origin of the outer lip of the body-chamber.
In this groove, the wall of which covers the umbilicus, there is,
opposite to the origin of the outer lip, a crista, the prolongation of
which passes into the dorsal wall of the body-chamber. The siphon
is placed at the bottom of a large and deep hollow, which has an
internal situation — that is to say, nearer to the dorsal than to the
ventral or external margin of the septum.

Measurements.

Major diameter 74 mm.

Minor « 58 "

Body-chamber, length 42 *'

*< « breadth, central 34 "

" " « in the middle 29 '«

I take great pleasure in dedicating this new species, which is the
first representative of the class Cephalopoda from the Tertiary of
Patagonia, to Mr. Carlos Ameghino, whose excellent work in the
geological exploration of Patagonia I appreciate very highly.

Locality : Golfo de San Jorge, Cabo Tres Puntas.

Formation : '^ Patagonico medio."

Neoinoceramus ameghinoi, gen. et sp. nov.

(Plate XIX, Figs. 3, 4.)

Testa solida, compressa, oblique-ovata, antice posticeque paullulum
hiante, concentrice laminato-costata, laminis altis subreflexis
distantibus, interstitiis loevibus ; auriculis anticis magnis,
posticis minoribus ; area ligamentale lata brevi.

The shell is large, solid, very obliquely ovate, slightly compressed
and gaping on both sides. The anterior lateral margin is convex,
passing gradually into the arched ventral margin ; the posterior

1902.] VON IHERING— ^[OLLUSCAN FAUNA OF PATAGONIA. 135

margin is concave below the posterior auricle, then becoming
convex, passes gradually into the arched ventral margin ; the
ventral extremity is obliquely produced and convex. The outer
surface is provided with numerous concentric ribs, which are three
to four millimetres high and slightly reflected along the free
margin ; the number of these lamellae amounts to eighteen in the
ventral half of the shell ; the upper or dorsal half is somewhat
defective in the central part. The distance between the ribs is, in
general, equal to their height, but there is some irregularity in their
size and spacing. Between the ribs the surface of the shell is
smooth.

The ligamental area, which is thirty millimetres in length, is
destroyed in the central part, as is also the umbo, the situation of
which must have been nearly central. The posterior auricle is
small, but well developed ; the anterior is broken away, but must
have been much larger than the posterior. The lateral remnants of
the ligamental area are strong, eight millimetres broad and
obliquely striated by narrow grooves, separated by small ribs, which
are the direct continuation of the ribs of the outer surface. It is a
remarkable fact that the concentric ribs of the outer surface do not
converge toward the umbo, but toward the dorsal margin of the
ligamentous area. The inside of the shell shows the simple pallial
line, which is distant seven millimetres from the anterior lateral
margin and twenty millimetres from the ventral margin. Behind
the posterior auricle the shell slopes abruptly toward the margin,
while at the anterior margin the transition is gradual. It is impos-
sible to recognize the muscular impressions.

Measui^ements.

Length, from anterior auricle to posterior ventral

extremity 95 mm.

Breadih 60 "

Diameter of half-shell 22 "

Another specimen, represented only by the ventral extremity,
must have had a length of at least 150 millimetres.

Locality : Golfo de San Jorge, east of Punta Nova.

Formation : Lower part of the Patagonian.

It is not easy to define the systematic position of this species,
because the umbo and the central part of the ligamentous area are
wanting. The multivincular ligament, the oblique-ovate shape and

136 vox IHERIXG — MOLLUSCAX FAUNA OF PATAGOXIA. [Aprils,

the concentric ribs indicate a relationship to the genus Inoceratnus,
from which however it is distinguished by the short and broad liga-
mental area and the well-developed auricles of the dorsal margin.
I regard the species therefore as the representative of a new genus,
of which I offer the following diagnosis.

Neoinoceramus, gen. nov.

Genus Aviculidarum testa sequivalvi, oblique-ovata, biauriculata,
concentrice costata, cardine recto, crasso brevi oblique-sulcato.

I believe this genus to be nearly allied to Inoceramus, the species
of which are exclusively Mesozoic. Although, because of the
incomplete preservation of the specimen described, the systematic
position of the genus is not definitely fixed, there can be no doubt
that this species represents one of the most remarkable discoveries
of Mr. Carlos Ameghino's later expeditions.

I have much pleasure in dedicating this exceedingly interesting
species to my friend Dr. Florentine Ameghino, not only in appre-
ciation of his excellent palseontological work, but also as an
acknowledgment of the liberality with which he has confided to me
the study of the invertebrates of his collection, permitting the types
to remain in the Museu Paulista, which for this reason possesses the
finest existing collection of Patagonian invertebrates. Of these two
new species, one is the first representative of the Cephalopoda from
the Tertiary of Patagonia, the other is a new type of Pelecypoda,
nearly allied to the Mesozoic genus Inocera?nus. The collection is
also rich in Corals and Echinoderms ; among the latter, the study
of which I have entrusted to Mr. Loriol le Fort, are also Crinoidea.

With reference to the paper of Mr. Hatcher, I have examined
the question of the significance of the Patagonian and Suprapato-
gonian beds, which Mr. Hatcher regards as only different facies of
a single formation. The fact that some species are common to both
horizons and that the number of such common species increases
with the progress of investigation, induced me for a time to agree
with Mr. Hatcher's opinion. It was therefore of importance for me
to reexamine the question with reference to the new material,
which was not derived from Santa Cruz, but from northern Pata-
gonia. The result was not favorable to Mr. Hatcher's views. This
may be seen from my paper on " The History of the Argentine

1902.] TOX IHERIXG— MOLLUSCAN FAUNA OF PATAGONIA. 137

Oysters," which will soon be published in the Communicaciones del
Miisco Nacional de Buenos Aires. Like the species of Ostrea,
those of Struthiolaria are also characteristic fossils of the Patagon-
ian and Suprapatagonian beds, which represent two different sections
in one great formation.

Of great interest in this new collection are the fossils from the
Pyroiherium beds, collected on the Rio Chico, a tributary of the
Rio Chubut, and the Golfo San Jorge. Among the mollusks are
characteristic Gryphcea concors Ih., and G. pyrotheriorum Ih., and
Ostrea ameghinoi. Of species characteristic of the Patagonian
formation only two occur: Cardiia patagonica wSow. var., and
Rhynchonella plicigera Ih. Among the new species may be men-
tioned Bouchardia patagonica^ Turriiella 7?ialaspina, Struthiolajia
striatissima and Rostellaria cossmanni. The last-named species is
a representative of a genus which does not occur in the other
Patagonian Tertiary formations. In this collection there are
neither existing nor Mesozoic species, and I therefore believe the
Pyrothei'i2un beds to be Eocene, while Florentino Ameghino regards
them as Cretaceous.

The general results given in my paper in the Revista do Museu
Paulista, Vol. II (abstract in English, p. 372 ff.), have not been
essentially changed, either by my later investigations or by those of
Dr. Ortniann. It is therefore singular that Mr. Pfeffer has, in the
past year, repeated his erroneous theory as to the existence of a uni-
form Eocene marine fauna. Even if we leave out of account the
Eocene formations of Patagonia, Chile and New Zealand, we must
consider such elementary faunistic facts as the distribution of the
Nummulites, which in the northern hemisphere extend from North
America to Europe and Asia as far as the Sunda Islands, while they
are wanting in the southern hemisphere. These facts cannot be
explained by supposed differences of temperature, but only by
geographical modifications, for the study of which a knowledge of
the Tertiary Mollusca offers one of the most important means.

Sao Paulo, Brazil, October 28, 1901.

138 TROWBRIDGE — GASES AT HIGH TEMPERATURES. [April 4.

SPECTRA OF GASES AT HIGH TEMPERATURES.

BY PROF. JOHN TROWBRIDGE, OF CAMBRIDGE, MASS.

{^Read April 4, 1902.)

It seems to me highly appropriate that I should speak in Phila-
delphia, the home of Benjamin Franklin, on my researches in elec-
tricity, and that I should bring to the attention of scientific men
here for the first time some remarkable results in the science in
which Franklin was a pioneer.

In the Jefferson Physical Laboratory of Harvard University
there is a Franklin electrical machine, which was ordered for the
College by Franklin when he was one of the Commissioners in
Paris. One can with great labor produce by means of it a thin
spark perhaps one inch in length. In the same laboratory I have
a storage battery of twenty thousand cells which, with suitable
transformers, will generate a spark six and one-half feet in length,
at a voltage of over six million.

In this practical age, especially in America, one is immediately
asked, '' What is the use of this great spark? " Probably a similar
question was asked Franklin in regard to his smaller manifestations
of electricity, and I shall ask you to reflect upon the developments
of electricity since his time — the telegraph, the telephone, the
lighting of cities, the trolley, the X-rays — and answer for me. You
will remember, too, that Franklin, fearing ridicule, which we can
charitably think generally arises from lack of imagination, tried his
kite experiment in secret. I have not hesitated to build the largest
electrical plant at present in existence for the scientific study of
electricity, feeling sure that I could reach an unexplored field ; and
I hope that some of my results which I shall communicate to you
will be considered of scientific importance, and will show that I
have reached such a field. In the first place, Franklin would see
in a spark six feet in length a veritable flash of lightning, brought
out of the skies into a laboratory where it can be studied at all
times and under almost any imposed conditions. I have discovered
that these long sparks do not encounter, so to speak, any greater
resistance in passing through the air than sparks one inch in length.
The entire current used in propelling the electric cars in this city
can fjass along the path opened by these long sparks without suffer-
ing hardly an appreciable diminution. A rarified hole seems to be

1902.] TROWBRIDGE — GASES AT HIGH TEMPERATURES. 139

bored, so to speak, in the air, tlirough which, by means of water
vapor, what we call electricity passes with a loud explosion. I wish
to emphasize this fact in speaking of the scientific results which I
have reached with this large electrical plant. I believe that I have
proved that water vapor is essential for the passage of electricity
through the air or gases. Just as a certain degree of moisture is
necessary for chemical reactions, so is water vapor essential for the
discharge of electricity through gases. I believe that we have never
been able to obtain a perfectly dry gas ; and if we should succeed
in the future, such a gas would be a perfect electrical insulator.

Since the time of Franklin, the subject of spectrum analysis has
been developed. He could study electricity only by means of his
eyes. With the spectroscope, however, we now see instead of a
blinding flash of white light, lights of many colors — in other
words, a spectrum extending from red light to violet light, traversed
by many bright lines which are due to the vibrations of the mole-
cules of the components of the air. These molecules are invisible to
us until revealed by electricity. The large storage battery I have
had constructed enables us to explore a new field in electrochem-
istry, revealed by the motions of the smallest particles of matter in
the world ; particles which are everywhere about us, but are only
evident when agitated by a discharge of electricity. I can surely
claim to have subjected gases to the highest temperature that has
been hitherto reached with this interesting result, that the spectra
of oxygen, hydrogen, nitrogen, the main components of the air,
contain the same spectrum, which is that of water vapor. By
modification of the strengtli of the discharges, one can pass from
the blue spectrum of argon to the red spectrum of this gas, which
was discovered by Lord Rayleigh, even in tubes filled with hydro-
gen. This result is accomplished by a powerful dissociation of
the small amount of air which is always present in glass tubes,
even when great care is taken in preparing the hydrogen. I have
obtained many such singular dissociations in hydrogen tubes which
have been unsuspected.

Another important fact has been revealed by the passage of pow-
erful discharges through glass tubes filled with rarified gases. I
have discovered a rate of molecular vibration to which the photo-
graphic plate is apparently inactive. All gases give bright lines in
their spectra, and consequently these bright lines are dark lines on
the photographic negative. I have discovered dark lines in the

140 ABBOTT AND BERGEY — ALCOHOLIC INTOXICATION. [Aprils,

spectra of gases which give, therefore, bright lines on the negative;
that is, they do not change the silver salt. This discovery, I think,
is of great importance, for it shows that there are rates of vibration
to which the photographic plate does not respond. It is imperfect
in science as well as in art, and does not give a complete history of
the stars, the temperatures of which are probably much higher even
than those which I have reached. These dark lines are not due to
what is called solarization or to absorption. The solar spectrum is
thus probably far more complex even than we have supposed. This
new field of what may be called destructive dissociation of gases in
which I am working, promises to lead to many important results in
the new science of electrochemistry.

[Prof. Trowbridge projected some lantern slides of the spectra
of gases obtained with the discharges from the large storage battery,
which showed the universal spectrum of water vapor and the re-
markable dark lines of which he had spoken. — The Secretaries.]

THE INFLUENCE OF ALCOHOLIC INTOXICATION UPON
CERTAIN FACTORS CONCERNED IN THE PHENOM-
ENA OF HEMOLYSIS AND BACTERIOLYSIS.

A PRELIMINARY NOTE.

BY A. C. ABBOTT AND D. H. BERGEY.

(from the laboratory of HYGIENK, university of PENNSYLVANIA.)

(^Read April 5, 1902.)

In 1896 one of us (A. C. A.) published the results of an investi-
gation upon the influence of alcoholic intoxication on resistance to
infection.^ In that paper attention was directed to the fact that
the susceptibility of rabbits to certain types of infection was
markedly increased through the influence of prolonged alcoholic
intoxication. These results have been fully confirn^ed by others.'^

At the time the results were published no fully satisfactory expla-
nation of the mechanism of this phenomenon was available, though
several suggestions were offered, viz., the reduced resistance may
be referable to the local action of the alcohol upon the gastric mu-

1 See Journal of Exp. Med., 1896, Vol. i.

2 See Laitinen, Acta Societatis Scientiarwn Fennicce, Tom. xxix, No, 7>
I900; also Zeit. f. Hyg. u. Infeklionskrankheiten, 1900, Band 34, S. 206.

1902,1 ABBOTT AND BERGEY — ALCOHOLIC INTOXICATION. 141

cous membrane, thereby impairing the nutrition of the animal to
such an extent as to create conditions analogous to starvation, a
state in which susceptibility is also seen to be increased ; or, to a
diminution in the alkalinity of the blood through the acids result-
ing from the oxidation of the alcohol — such reduction in alkalinity,
though slight, has since been shown by Laitinen to occur ; or, to
the remote action of the alcohol on the nervous system. The value
of neither of these hypotheses was, however, susceptible of ready
determination, so that the matter rested there for a time.

During the past three or four years a series of brilliant investiga-
tions, especially by Bordet, Buchner, Metschnikoff, Ehrlich and
Morgenroth and their associates, upon certain physiological phe-
nomena peculiar to the blood and other fluids of the body, have
acquainted us with many hitherto obscure and unknown phases of
the subject. One of these newly discovered blood reactions seemed
especially adapted to the solution, in part at least, of our problem.

It has been demonstrated by the investigators named that an
animal may be rendered immune from the intoxicating effects of
the blood of another species ; that when such immunity is estab-
lished the blood serum of the immune animal rapidly and com-
pletely dissolves the erythrocytes of the alien blood, even when
mixed with them in a test tube (haemolysis); that if such immune
serum be heated for thirty minutes to 55°-56° C. it loses its hasmo-
lytic power ; and that the power of haemolysis is at once restored
to the heated serum by the addition of a few drops of serum from
a normal mammal. These reactions are believed to occur through
the agency of two bodies present in the serum — the one a body re-
sistant to low degrees of heat, a "receptor" or ''intermediary"
body ; ^ the other a complementary something, perhaps a ferment,
common to all mammalian serums, that is destroyed by heat. The
*' receptor" or "intermediary body" is conceived to have the
property of fixing the invading cells (in this case the blood cells of
another species) on the one hand, and the complementary, ferment-
like body on the other, bringing and holding them together in a
way most favorable to the destructive action of the ferment upon
the invading cell. The destruction of bacteria by the fluids of the
body is thought to take place in an analogous manner, it being as-
sumed that in the blood are "receptors" having the property of

1 Synonyms— Anticorps hemoljtique, Substance preventive, Immune Korper,
Amboceptor, Philocytase, Desmon, Copula, Substance sensibilisatrice, Fixateur.

142 ABBOTT AND BERGEY — ALCOHOLIC INTOXICATION. [Aprils,

fixing, on the one hand, bacteria, and on the other a *' comple-
ment" having the power to destroy such bacteria, the relation of
receptor to bacteria and to complement being in both cases specific.

The question under consideration by us was :

'*Will the sera of animals under the influence of alcohol for
varying lengths of time, but otherwise normal, restore to a heated
immune serum its haemolytic activity in the same way as is done by
the normal sera of non-alcoholized animals? "

If it will, then the action of alcohol upon the animal organism
is plainly not evidenced through a reduction in the amount of the
complementary substance so necessary to normal resistance and to
immunity. If it will not, then the reverse must be the case.

Should the serum of animals under the influence of alcohol prove
to be poorer in haemolytic ^' complement " than that of animals not
so treated, then there is some justification for the belief that the re-
duction of resistance to bacterial infections, noted in our work of
1896, may be due to the suppression (in part or in whole) of a '' com-
plementary " ''proteolytic ferment " (?) that constitutes one of
the natural defenses of the body against the invasion of infective
bacteria. Without discussing our results in detail, it suffices to say
that we found in a number of animals daily intoxicated for a period
of about three weeks, the amount of '' complement " in their sera
to be from fifteen to twenty-five per cent, less than that of normal
sera, as determined by the power to ^' reactivate " a heated immune
serum — /. e., to restore to it its haemolytic properties, a result that
we regard as of fundamental importance in explaining (in part at
least) the results of investigations made in 1896.

In the course of this work a number of important collateral
questions arose, the most significant of which being as to whether
the effect noted by us could be interpreted as a general reduction
of all complementary substances^ in the blood, or as only a reduc-
tion of a single complement specifically concerned in the phenom-
enon of haemolysis ; but as their solution is as yet only in the initial
stages, it is scarcely necessary to introduce them at this time.

1 It is believed by Erhlich and Morgenroth and their associates that the blood
contains a multiplicity of complementary elements, each one of which is specifi-
cally related to particular receptors and to particular irritants and intoxicants;
while Buchner, Bordet, Metschnikoff and their adherents contend that the com-
plement, designated by Buchner and Bordet as ** alexine " and by Metschnikoff
as "cytase," is a single substance possessed of heterogeneous affinities.

1902.] WILSOX — OSTEITIS DEFORMANS. 143

A BRIEF ACCOUNT OF THE DISEASE KNOWN AS
OSTEITIS DEFORMANS.

BY PROF. J. C. WILSON, M.D.
{Read April 5, 1902.)

It will, I trust, be acceptable to the Society if I communicate
some facts in regard to a rare disease of the bones.

This affection was first described by Sir James Paget, in the
Transactions of the Royal Afedical and Chirurgical Society of Lon-
don, in 1877, under the title '^ A Form of Chronic Inflammation of
Bones — Osteitis Deformans." To the five cases which formed the
basis of that communication, Paget was able to add in 1890
eighteen further instances of the disease which he had studied.
Other cases have been observed in Great Britain ; in America up to
the present time eleven cases have been reported ; a number in
France, and a few elsewhere on the Continent of Europe.

It is, however, probable that osteitis deformans is much more
common than the number of the published cases would indicate.
In the preliminary program of the Association of American Physi-
cians just issued two new cases are announced. The fact that the
disease remained long undescribed and is now so seldom recognized,
is due not so much to the infrequency of its occurrence as to the
trifling subjective symptomiS which attend it or their complete
absence, its insidious development and slow progress, and the im-
munity of the bones of the hands and feet. The sufferer from
osteitis deformans may develop advanced changes in the skeleton
before the deformities attract his attention or that of his friends.

The deformities in some instances affect only a limited number
of the bones, more commonly most of them. In the fully developed
disease they are usually symmetrical to a remarkable degree.

They consist in the following changes in the skeleton:

Thickening of the bones of the skull and an alteration in its
shape. The calvarium becomes flattened, the brow broad, the
parietal regions prominent. The general circumference is increased
so that the patient has to wear a larger cap than formerly. The
bones of the face remain unchanged, so that the facies assume a
triangular outline, the base being at the brow, the apex at the chin.

The spine becomes stiffened and curved. There is marked
cervico- dorsal kyphosis, with compensating lordosis of the lower

144 WILSON— OSTEITIS DEFORMANS. [April 5,

dorsal and lumbar spine. In consequence of this change in the
spinal column the head is carried forward and lower than normal,
and the height of the patient is reduced — a reduction much in-
creased by the curvature of the bones of the lower extremities and
amounting in some of the cases to six or seven inches.

The clavicles are prominent and thickened, the chest short and
narrow, the abdomen short and broad and the pelvis wide and low.

Associated with these changes are marked deformities of the long
bones of the extremities. The humerus is thickened and enlarged ;
its surface is irregular, and the shaft is markedly curved, the con-
cavity presenting toward the flexor surface. The ulna and radius
show similar deformities and are strongly bent and twisted. The
bones of the lower extremities are deformed and bent in a like
manner. The femur, tibia and fibula are bent outward and forward.

In fully developed cases the patients bear a curious resemblance
to each other. The diminution in stature causes the arms to appear
disproportionately long — like those of the anthropoid apes.

The disease usually makes its appearance in middle life and is
mostly unattended by subjective symptoms, although in some cases
rheumatoid pains have been present at the outset. It has no con-
stant relation to any particular visceral or nervous pathological
process, nor to malignant disease as was at one time thought. I
have called attention to the high grade of muscular atrophy present
in well developed cases.

Paget, whose name has been given to the disease and whose
original description remains the best that has thus far appeared,
regarded the changes in the bones as inflammatory, and Butlin's
account of the histological changes lends support to this opinion.

The process consists of a progressive absorption of bone tissue
which becomes porous and rarified ; the coincident formation of
new bone, which remains for a time uncalcified so that abnormal
curvatures develop, and finally dense calcification of the subperiosteal
layers of the overgrown and deformed bones. The marrow under-
goes fibrous changes. The pathological changes have been espe-
cially studied by Butlin, von Recklinghausen, Stilling and Packard
Steele and Kirkbride.

The etiology of the disease is involved in complete obscurity.
To state that it is due to trophic derangements is a mere general
restatement of the facts.

Tiie hypertrophic changes in the bones of an extremity, which

1902] BROOKS— IS SCIENTIFIC NATURALISM FATALISM ? 145

have been shown by Schiff, Vulpian and Philipeaux to follow the
section of the nerve supply, cannot be regarded as an analogous
process and are not invariable.

Two views suggest themselves : Osteitis deformans may be due to

1. Infection by some organism, to the action of which bone
tissue is especially liable ; or,

2. To the default of some physiological principle which nor-
mally regulates and limits the growth of bone.

Either of these views may serve as a working hypothesis for
investigations into the cause of the disease.

This affection has points of similarity with osteomalacia, leonti-
asis ossea, acromegaly, gigantism, arthritis deformans and rickets,
but differs from them all in essential particulars.

No treatment has been of any service in arresting the progress of
the disease.

15 SCIENTIFIC NATURALISM FATALISM?

A ONE-MINUTE PAPER.
BY WILLIAM KEITH BROOKS.

i^Read April 4, 1902.)

Berkeley pointed out long ago that all the phenomena in nature
may be expressed in terms of motion. The progress of science is
teaching us this truth, and is thus bringing us to a point of view
which Hume has indicated in these words: *'The necessity of any
action, either of body or of mind, is not in the object which ex-
hibits the action, but in the spectator."

Scientific predictions are based upon our well-founded confidence
that the order which we have discovered in nature in the past will
continue in the future ; but physical analysis neither answers nor
asks why nature should be orderly, or what has made it so. For its
purposes, the notions of agency and efficiency and causation are
irrelevant and useless, because the notion of necessity is something
that we ourselves project into nature and not anything that we find
in nature.

If we agree with Hume, as I think we must, does not his state-
ment carry with it, as its complement and counterpart, a declara-
tion to this effect : Freedom in willing and doing, if there be such

PROC. AMER. PHILOS. SCO. XLT. 169. J. PRINTED JUNE 10, 1902.

146 KEASBEY — A CLASSIFICATION OF ECONOMIES. [Aprils,

freedom, is not in the spectator who considers the action, but in
the agent?

Is our failure to find proof of freedom in our bodily machinery
and its activity anything more than we should look for if freedom
is not in the spectator, so far forth as he is merely a spectator and
not a participant ?

If the certainty of scientific predictions does not imply necessity,
and if freedom in willing and doing is not in the spectator, are we
not led to agree with Berkeley, that '^ certain and necessary are
very different, there being nothing in the former notion which im-
plies constraint, and which may not consist with a man's being
accountable for his actions " ?

If physical necessity is not in nature, but in the spectator ; if
freedom is not in the spectator, but in the agent ; if the certainty
of scientific predictions does not imply constraint ; — does not the
controversy about necessity and freedom come to an end for the
man of science ? Does science afford any ground for controversy ?

A CLASSIFICATION OF ECONOMIES.

BY PROF. LINDLEY M. KEASBEY.
{Read April 5, IQ02,)

Economics has to do with the weal relation between life and the
environment. From life, on the one hand, emanates demand for
well-being; from the environment, on the other hand, is derived
the supply of useful things or goods that minister to well-being.
In the last instance, therefore, the weal relation between life and
the environment is a relation between demand and supply. Now,
demand and supply are connected — made to meet, as economists
say— by the utilization of natural resources. The object of this
process is to derive from the outer world the qualities requisite to
fulfill the demands of well-being, or, more precisely, to convert the
potential utilities inherent in the environment into actual utilities.
Thus, in its simplest sense, an economy may be defined as a system of
activities whereby the potential utilities inherent in the environment
are through utilization converted into actual utilities.

The very existence of life implies some such system of activities ;

1902.] KEASBEY — A CLASSIFICATION OF ECONOMIES. 147

wherever the essential weal relation is established between life and
the environment, there the process of utilization is operative. In
its widest extension, therefore, the term economy can be applied
over the whole range of evolution, from the lowest to the highest
orders of animate existence. Furthermore, cursory comparison
shows that with the developm.ent of life the process of utilization
becomes more and more complicated. Thus, regarded from the
utilitarian point of view, evolution exhibits a succession of econo-
mies increasing in complexity.

It is out of the question, of course, to elaborate this long series
in detail. As a matter of fact, no hard and fast distinctions can be
established between the several orders of economies, since in each
instance the more complex proceed, as it were, by insensible steps
out of the simpler, leaving no appreciable spaces between through
which lines of demarcation may be drawn. Nevertheless, if we
confine ourselves to generalities and content ourselves with obvious
distinctions, it is possible to establish the general order of economic
development and characterize the several types of economies.

For convenience' sake biologists still distinguish between plant
life, animal life and human life, what though they are well aware
that the laws of organic evolution to which the three orders of life
are subjected are essentially the same. It is possible to establish
a corresponding series in the order of economic development, but
we must not lose sight of the fact that the differences to be noted
are merely differences of degree and in no sense distinctions in
kind. This, then, is the primary purpose of the present paper : to
indicate the types of economies characteristic of plant life, animal
life and human life respectively. It will be seen, when this series is
established, that the human economy differs far more from the
economies of the lower orders of life, than the economies of plant
and animal life differ from each other. Though evidently an elabo-
ration of the preceding types, the human economy is in certain
respects so different as practically to constitute a separate system.
Having shown this, to be the case, I shall devote the remaining por-
tion of my paper to establishing the human economy upon its
higher plane.

In the first place, in order to establish the required series of
economies, it is necessary to adopt a canon of distinction. To this
end I would suggest that characteristic types of economies can be
distinguished from each other in two ways : subjectively, according

148 KEASBEY — A CLASSIFICATION OF ECONOMIES. [Aprils,

to the incentive leading to utilization ; and, objectively, according
to the means employed in the process.

Applying this canon of distinction in the first place to the sim-
pler systems of activities, it is possible to establish two types of
economies — the automatic and the instinctive — characteristic
respectively of the plant and animal worlds.

Under the automatic system the stimulus inciting utilization is in-
voluntary, and as this is the case, the means employed in the pro-
cess are necessarily natural organs that act without the intervention
of the will. Thus plants, for example, as well as some of the lower
orders of animals, assimilate the life-sustaining elements inherent in
their immediate environment by simple reflex action, involving no
conscious effort on their part.

Under the instinctive system, on the other hand, the impulse
leading to utilization is voluntary, and as this is the case, the means
employed in the process consist for the most part of natural organs
that act in obedience to the will. Thus, as opposed to plants, ani-
mals may be said to be urged by their appetites to utilize natural
resources. It is instinct in their case that induces economic activity.
That is to say, the higher animals as a rule are impelled by their
natural desires of self and kind preservation to acquire such pro-
ducts of their local environment as go to gratify their own appetites
and provide for the preservation of their progeny. And as nature
has provided them for the most part with the natural organs neces-
sary to gratify their desires, little or no ingenuity is necessary to
this end.

The most complicated economy is that characteristic of human
life. In contradistinction to the foregoing, this highly complex
system may be designated as the rational economy. Right early
in the course of their development, human beings appear to have
become imbued with an intelligent purpose to meliorate their mate-
rial condition and so raise the standard of life of themselves and
their associates. And not being physically equipped by nature to
realize their economic ideals, far back in the course of their career
they began to exercise ingenuity in the manufacture of artificial
instruments of utilization. Thus, to distinguish the human economy
from that characteristic of the animal orders, it may be said: under
the rational system the motive making for utilization is purposive,
and the means employed in the process consist for the most part of
artificial implements manufactured for the purpose.

1902.1 KEASBEY — A CLASSIFICATION OF ECONOMIES. 149

Having applied our canon of distinction over the whole range of
economic development, there appear to be three fundamental types
of economies, the automatic, the instinctive and the rational, char-
acteristic respectively of plant, animal and human life. In the
automatic economy the stimulus exciting utilization is spontaneous,
and the means employed in the process consist of natural organs
that act without the intervention of the will. In the instinctive
economy the impulse leading to utilization is voluntary, and the
means employed in the process consist for the most part of natural
organs that act in obedience to the will. In the rational economy
the motive making for utilization is purposive, and the means em-
ployed in the process consist for the most part of artificial imple-
ments manufactured for the purpose.

The foregoing classification gives a general idea of the order of
economic development, and enables us to distinguish superficially
between the three fundamental types of economies. The distinc-
tion between the automatic and the instinctive systems, it will be
noticed, is not nearly so marked as that between these simpler sys-
tems, on the one hand, and the highly complex human economy on
the other. Indeed, if Professor Loeb is right in regarding instinc-
tive action as essentially the same as reflex action, the separation of
the instinctive economy from the automatic economy must betaken
to express simply a superficial distinction, or at most to mark a
minor difference of degree. Rational activities are, however, radi-
cally different from instinctive acts, though here too, no doubt, the
difference is ultimately one of degree. Wherein these latter differ-
ences consist is the task of the psychologist to show. It is enough
for the economist to take cognizance of the facts and establish his dis-
tinctions accordingly. On the face of it, the fact that the human econ-
omy constitutes a rational system evidently places it upon a higher
plane than the economies characteristic of the lower orders of life.
Then, again, regarded from the point of view of economic develop-
ment, a further distinction is discernible in the process of utilization
characteristic of the rational system. In the rational economy utiliza-
tion appears to make for progress ; whereas under the automatic
and instinctive systems utilization seems to be simply conservative.

It is evident enough, as has already been indicated, that with the
development of plant and animal life the process of utilization
becomes more and more complicated, but in all these cases increased
complexity appears to be rather the effect of variation and selection

150 KEASBEY — A CLASSIFICATION OF ECONOMIES. [Aprils,

than the outcome of economic initiative. Thus the instinctive
system, characteristic of the animal world, becomes more and more
complicated as we advance from the lower to the higher orders of
animal life ; but there is nothing to indicate that this increase of
complexity is due to conscious effort on the animal's part. Lamarck,
it is true, attributed appetency to animals and endeavored to prove
that evolution is to a large extent the result of active initiative ;
but modern opinion still inclines to the belief expressed by Darwin
that the process is effected unconsciously, through natural selection.
But it is not necessary at this juncture to go into this abstruse ques-
tion of the relative importance of appetency and variability in the
evolutional process. We are dealing, it will be remembered, merely
with differences of degree, and may accordingly content ourselves
with establishing obvious distinctions. This much at least is evi-
dent from casual observation : if we exclude the development of
the human species from our survey, progress in the economic sense
is not a notion that can properly be applied to the evolution of
animal life, and of course much less to plants. Even the highest
animals, when once adapted to their environment, show no disposi-
tion in their natural state to improve their material condition or
meliorate the lot of their progeny. On the contrary, to the extent
that they remain uninfluenced by selection, animals and their off-
spring appear to be urged by the same appetites, to utilize the same
resources in the same way from generation to generation. The im-
pulse leading to utilization is in their case instinctive, and therefore
more or less rigidly determined along certain definite lines. And
inasmuch as nature has provided them with the means of utilization,
it is not necessary for them to exercise ingenuity in the invention
of artificial instruments. Some animals do, to be sure, manufacture
artificial implements of production — witness, for example, beavers
that build dams, or certain ants that actually cultivate their fields.
Still even in such cases nature supplies the necessary tools, and it
would be difficult to find instances in which animals were led to
improve their productive processes with a view to meliorating their
material condition. Thus, from the fact that the impulse leading
to utilization is in their case instinctive, and from the further fact
that the means employed in the process are for the most part natu-
ral organs that act without the intervention of intelligent foresight
on their part, animals may be said to subsist in a circle. Appetite
impels them in first instance upon their food quest, and the

1902.] KEASBEY — A CLASSIFICATION OF ECONOMIES. 151

nutriment when acquired is assimilated. During the process of
digestion a period of rest or play ensues until the original appetites
are re-aroused, when hunger again sets them in search of subsist-
ence with the same result. The life of the anaconda is the most
striking example of this circular sort of existence, though the
description applies in a less degree to all orders of animals, whose
existence for the most part amounts to a monotonous round of
acquisition and assimilation as long as life lasts, and is afterwards
carried on in much the same way by their offspring. Obviously
there is nothing in such a system to stimulate progress, for the
economic sequence once established is recurrent : demand tends
toward utility, utility leads to utilization, and utilization results in
supply, over and over again.

Turning from the instinctive to the rational economy, the phe-
nomenon of progress becomes immediately apparent. If we
extend our survey to include the activities of mankind, it is evident
enough that utilization is a potent factor of development. Not
that the human species is not subject, like all other animals, to the
process of selection ; by no means — indeed, as ethnology shows,
the human species has in the course of time, through the interaction
of variability and environment and by dint of selection, become
differentiated into a number of ethnic stocks. Only the process of

human development does not appear to stop there. In man's case

and, as far as I can see, in man's case alone — utilization has made
for further progress along economic lines. That is to say : men of
the same descent, who do not differ from each other ethnically to
any appreciable extent, who are to all intents and purpose alike as far
as structure and function are concerned, still exhibit striking differ-
ences in their manner of life. Thus the Frenchman of the prov-
inces and the Frenchman of Paris are ethnically alike, but differ
enormously in their economic activities. And offspring that vary
ever so slightly from their parents in the organic sense very often show
decided increase of economic capacity. For example, the English-
men of to-day are very much like the Englishmen of three hundred
years ago, but in their manner of life they differ widely from their
ancestors. On the other hand, people of diverse ethnic stocks, if
placed under the same economic conditions, soon conform to an
established standard of life and adopt similar ways of living. Our
own country furnishes a striking instance of this. The population of
the United States is recruited from all countries of the world, but

152 KEASBEY — A CLASSIFICATION OF ECONOMIES. [Aprils,

despite this ethnic divergence a distinctly American standard of
life has been established to which all citizens, foreigners and natives
alike, endeavor to conform. Since such are the facts it is evidently
necessary in man's case to draw a sharp distinction between prog-
ress through selection and progress by utilization — between what
may be called ethnic variations and economic distinctions.

Let us examine the situation a little more closely. Looking first
to the subjective side, human beings do not seem to be content, as
most animals are, to consume the same goods day after day, year
after year, and from generation to generation. On the contrary,
man appears to be bent on obtaining variety. The gratification of
one set of desires seems to cause a new series to emerge in the
mind. We imagine we shall be satisfied with what we want, but
acquisition soon convinces us to the contrary — like the boy who
found a watchkey, and on the basis of this possession asked his
father for a watch. In short, the mere fact of acquisition extends
the horizon of our wants and arouses a desire for further acquisi-
tion; or, to put it in economic terminology, the possession of cer-
tain essential goods- stimulates a demand for complementary goods.
Without dvi'-elling on this pyschic phenomenon, so familiar to us
all, it may be stated as a general proposition : human beings
naturally seek variety and strive to extend the scope of their
consumption.

The emergence of new wants in men's minds naturally suggests
a corresponding series of satisfactions ; demand is necessarily cor-
related with supply. Suppose we turn, then, to the objective side
and take the extrinsic factors into account. The moment the con-
ditions of supply are considered, it becomes apparent that man's
desire to extend the scope of his consumption is met by obstacles
arising from the character of the environment. Outer nature
affords a few free goods, it is true, but by no means enough to
satisfy man's expanding wants. For the rest, raw materials must be
transformed into pleasure-giving products by artificial processes.
To this end implements are necessary, since human beings are not
equipped, as most animals are, with the technical means of produc-
tion. Organization is also essential, as it is only through the
systematic division and association of their productive forces that
men are able to provide the requisite variety of goods. Because
his expanding wants outstrip his inherited capacity, to overcome the
obstacles arising between demand and supply, man is accordingly

1902.] KEASBEY — A CLASSIFICATION OF ECONOMIES. 153

required to exercise ingenuity in invention and undertake economy
in organization. Or, to express it more concisely : in order to
extend the scope of their consumption human beings are compelled
to improve their means and methods of production.

Putting two and two together, the situation seems, then, to be
this : man's desire for variety urges him to extend the scope of his
consumption, and in order to extend the scope of his consumption
he is obliged to improve his means and methods of production.
Thus, in contradistinction to the circular sort of existence charac-
teristic of animal life, the course of human progress is upward, so
to speak, along the lines of a spiral. The emergence of elementary
wants in men's minds stimulates invention and organization and
results in the production of goods. The consumption of these
essential goods causes wants for complementary goods to emerge in
the mind, and these new wants in turn stimulate further invention
and organization. Thus new wants call continually for the im-
provement of productive processes, improved productive processes
provide a further variety of goods, which in being consumed cause
still other wants to emerge in the mind that call for further
improvement of productive processes, and so on ; want inducing
satisfaction and satisfaction inducing want almost indefinitely.

Thus in the rational economy the economic sequence is progres-
sive and not merely recurrent as in the instinctive economy. In-
stead of demand tending toward utility, utility leading to utiliza-
tion, and utilization resulting in supply over and over again, as is
the case with most animals, in man's case expanding demand tends
toward the augmentation of utility, the augmentation of utility
leads to increasing utilization and increasing utilization results in
the differentiation of supply.

154 BRYANT — DRIFT CASKS IN THE ARCTIC OCEAN. [Aprils,

DRIFT CASKS IN THE ARCTIC OCEAN.

BY HENRY G. BRYANT.
{Read April 3, 1902.)

Among the many notable sessions of this venerable Society, per-
haps none in recent years have been more interesting than the
*' Nansen " meeting held on the afternoon of October 29, 1897.
It was one of the last occasions on which our late President, Fred-
erick Fraley, occupied the chair. The occasion was noteworthy,
not only by reason of the paper on *' Some of the Scientific Results
of the Fram Expedition," read by the distinguished Norwegian
explorer, but also because of the supplementary discussion which
gave opportunity for Rear Admiral George W. Melville and other
competent authorities to give expression to their views on the
importance of Arctic research and the best methods of prosecuting
it in the future.

In the course of his discourse on '*The Drift of the Jeannette,"
Admiral Melville — after recommending that future attempts to ex-
plore the unknown area should start from the Bering Sea side —
called attention to the fact that much valuable data relating to
circumpolar currents could be obtained by setting adrift in the
waters north of Bering Strait specially constructed casks containing
the requisite records. A certain percentage of these floating mes-
sengers might fairly be expected to survive the perils of the Arctic
pack and eventually be looked for in waters adjacent to Franz
Joseph Land, Spitzbergen or Greenland. In this connection he
remarked: "I do believe, however, from the information we
have gained from the drift of the Jeannette and the Fram, that
vessels of any kind, such as casks or driftwood, will come out by
way of Spitzbergen — though not necessarily across the Pole. The
only reason for sending men in ships is, that they may be observers

to make a daily record of events But for this, I say, a

hundred casks, properly numbered, made after the manner of a
beer keg of twenty gallons capacity, properly hooped, and the ends
extended out to complete a parabolic spindle, would demonstrate
the drift." ^

This idea of studying ocean currents from data obtained from
** bottle messages" is not entirely new, and has, in fact, been em-
ployed by the Hydrographic Office of the U. S. Navy and by

1 " The Drift of the Jeannette," Prog. Am. Phil. Soc, Vol. xxxvi, No. 156.

1902.] BRYANT— DRIFT CASKS IN THE ARCTIC OCEAN. 155

Other agencies for some years past. But in such cases the mes-
sages have been enclosed in an ordinary bottle and have been dis-
tributed along the ordinary routes of ocean travel. But the idea
of investigating circumpolar currents by means of specially con-
structed drift casks originated, I believe, with Admiral Melville,
and the project possesses certain features which will commend it to
that large body of students who are interested in the problems of
oceanography. On another occasion I outlined some of the pre-
liminaries of this experiment, and in this connection I venture to
quote from that statement of the subject : ^

''This proposed method of studying Arctic currents without
endangering human life having been brought to the notice of the
Geographical Society of Philadelphia, that body determined to un-
dertake the project. In view of the exigencies of a long voyage on
the floe ice, special attention was given in the construction of the
casks to shape and strength of materials. Thus, to more readily
escape crushing by the ice, as intimated above, their shape con-
formed to that of a parabolic spindle, while they were made of
heavy oak staves one and one-quarter inches think, encompassed
by iron hoops three sixteenths of an inch thick and two inches wide.
A coating of black ' half stuff ' (pitch and resin mixed) was then
applied. In addition to the preservative qualities of this coating,
the thickness of the wood and metal used is believed to be sufficient
to resist the attrition of the ice and the effects of corrosion during
the long drift. The staves, so tapered as to form the spindle, were
covered on the ends by light galvanized cast-iron caps, held in
place by an iron rod five-eighths of an inch in diameter, extending
the length of the cask and secured by conical nuts at each end. As
above stated, a heavy coating of black water-proofing material was
applied to the casks to guard against corrosion and decay. From
the color used they will be more easily seen, and will also the more
readily sink — under the action of the summer sun— into the body
of the ice and be preserved from destruction by crushing. The
number of each cask was etched into the wood, as well as painted
on the outside. In accordance with the instructions of the origin-
ator of the plan, the vessels must be placed on the heavy floe ice.
If set adrift in open water they would be too much at the mercy of
winds and waves, whereas by being deposited on heavy ice, which

1 " Drift Casks to Determine Arctic Currents," by Henry G. Bryant, Ver-
handlungen des Siebenten Internationalen Geographen-Kongresses, Berlin,
1899, Zweiter Theil, Seite 663.

156 BRYANT — DEIFT CASKS IN THE ARCTIC OCEAN. [Aprils,

is more affected by under currents, they will probably be carried on
a more correct drift. A reinforced bung-hole with bung was pro-
vided, and through this the message bottle was inserted

This latter consisted of a narrow cylindrical tube made of flint
glass, and technically known as an 'ignition tube,' accompanying
which were suitable corks and sealing wax. As an additional pre-
caution, these tubes were in turn enclosed in cases made of maple
wood provided with screw tops.

'' The message paper enclosed in this way was printed on linoleum
paper by a permanent blue-print process, which renders it practi-
cally impervious to salt water. The enclosed message was printed
in the English, Norwegian, German and French languages, and
embodied the following particulars :

"(<:?) Space for name of vessel and master assisting in distribution,
date, number of cask, and latitude and longitude of point where
it was set adrift.

^\b) Directions as to filling in record and sealing up tube.

^'{c) Blank space for insertion of name of finder, date and locality
where cask was picked up.

^'{d) Clause requesting finder to forward message paper to the
nearest United States Consul at his home port, or to send it direct
to the Geographical Society of Philadelphia.

*' Accompanying each consignment of casks was a set of printed
instructions to masters of vessels engaged in their distribution."

In the important and hazardous work of distributing the fifty
casks provided for the experiment, the promoters of the enterprise
have received the assistance of the U. S. Revenue Cutter Bear,
which makes yearly trips to Point Barrow, Alaska, in the interests
of the American whalemen. We have also profited by the cooper-
ation of the Pacific Steam Whaling Co. and of Messrs. Liebes &
Co., of San Francisco, both of whose vessels have assisted in placing
the casks adrift in the far North. The reports of the accomplish-
ment of this preliminary work have come in rather slowly owing to
the length of the whaling voyages and other causes. Thus the
first consignment of casks was shipped from San Francisco as early
as March, 1899, ^^^ ^^^^ others as soon after as opportunity offered ;
and yet, of the thirty-five casks whose distribution has thus far
been reported, intelligence of the last distribution arrived here no
later than December 11 of last year.

Thus only within the last few months has it been possible to
report definitely in regard to the launching of the greater number

1902.] BRYANT — DRIFT CASKS IN THE ARCTIC OCEAN. 157

of the casks, and I have, therefore, availed myself of the present
opportunity to present some details relating to the present status of
the experiment. Tlie directions to masters of vessels having in
charge the distribution embodied the main ideas of the originator
of the plan and recommended ''that special efforts be made to
carry a number of casks north of Bering Strait and thence to the
westward, where a number of them should be set adrift at or near
Herald Island. Then proceeding northward along the eastern edge
of the ice pack until the highest safe latitude is obtained — say lat.

75° N., long. 170° W. from Greenwich At this point final

sets of casks are to be set adrift to demonstrate, if possible, the
currents to the eastward or northward and eastward, if any there
prevail." In examining the reports sent in, I find that these in-
structions have been carried out in a praiseworthy manner. Thus
I find, by plotting the positions indicated, that on August 19 and
21, 1 90 1, the U. S. Revenue Cutter Bear, under Capt. Francis
Tuttle, placed fifteen casks adrift at three different points on the
floe ice north and northeast of Herald Island, making a northing
in one instance of 72° 18'' near the 175th meridian of west longi-
tude.

In September, 1899, Capt. D. T. Tilton, of the S. S. Alexander,
belonging to Messrs. Liebes, placed four casks adrift south and east
of Herald Island, and in the same month Capt. Sherman, of the
Pacific Steam Whaling Co.'s steamer Thrasher, discharged one
cask W. N. W. of Point Barrow, while in September of last year
(1901) the same company's vessel, the Narwhal, succeeded in
launching three casks in three different locations well north and
west of Herald Island. The highest northing yet reported as a
delivery of the casks was attained by the vessel just mentioned on
September 7, 1901, when 73° N. lat. was reached.

Thus we find that twenty-two casks have been successfully
launched at different periods on the great ice pack north and
northeast of Herald Island.

With a view of testing the probable existence of a northeastern
or North American drift through the Parry Archipelago, and along
the route followed by McClure fifty years ago in accomplishing the
Northwest Passage, the whaling captains were requested to distribute
some casks in the region of Banks Land, near the western approach
to the Northwest Passage route. This plan coincided with the
movements of the whaling fleet, the members of which frequently
extend their voyages for considerable distances east of their winter

158 BRYANT — DRIFT CASKS IN THE ARCTIC OCEAN. [Aprils,

rendezvous, Herschel Island. Hence we find that some nine casks
were set adrift off Banks Land in 1899 and 1900 by the steam
whalers Alexander, Thrasher, Narwhal and Beluga. It seems to be
altogether reasonable to assume that quite a large percentage of the
water contributed to this part of the Arctic Ocean by the Macken-
zie River should find its outlet by means of the devious channels
which extend eastward among the islands of the American Archi-
pelago ; but just which route the drift casks will take, or how long it
will take them to reach the whaling grounds in Lancaster Sound, it
is idle to conjecture. Should any number of the casks be recovered
on the Atlantic side, however, the time occupied by them on the
journey between the known termini can be ascertained with some
accuracy, and the resulting data should throw some light on the
speed of the current in question. From the representatives of this
miniature flotilla which were cast adrift in the waters north of
Bering Strait, we may look for more definite results.

It has been known for years that no appreciable amount of water
from the Polar Ocean escaped through the narrow, shallow outlet
of Bering Strait, while the knowledge gained from the drift of the
Jeannette and Fram point to the existence of a well-defined drift
across the circumpolar area to the shores of Franz Joseph Land,
Spitzbergen and East Greenland. The presence of quantities of
Siberian driftwood in the localities named can be explained by no
other intelligent hypothesis, while it is well known that Dr. Nansen
based the theory of his voyage primarily on the finding of the
Jeannette relics on the west coast of Greenland, three years after
the crushing of that vessel in the sea northeast of the New Siberian
Islands. Prince Krapotkin, the distinguished Russian writer, gives
due importance to the Jeannette's voyage as bearing on the solu-
tion of this problem, and commends Nansen for "embodying the
drift of the Jeannette and the East Greenland ice drift in one
mighty current. A formidable ice current, almost as mighty and
of the same length as the Gulf Stream, a current having the same
dominating influence in the life of our globe, has thus been proved
to exist." ^ Those who are interested in this experiment indulge in
the hope that these casks, which have been consigned to the sea
ice near the locality where the Jeannette began her drift, will pur-
sue their voyage across the Polar basin impelled by the same ele-
mental forces which carried the Jeannette so far on her journey,

1" Recent Science," Nineteenth Century^ February, 1897, P- 259.

1902.] BRYANT — DRIFT CASKS IX THE ARCTIC OCEAN. 159

and which subsequently swept the brave little Fram across a great
portion of the unknown area.

From the nature of the case, it is difficult to prophesy the time
that will be required to complete the drift.

The Jeannette was put into the ice in latitude 71° 35' N. and
longitude 175° W. and consumed twenty-two months in making
her zigzag drift of 1300 miles. The provision list signed by
Lieutenant DeLong, and the other articles believed to belong to the
Jeannette, were three years in traversing the distance from the place
where that vessel was crushed in latitude 77° 15' N., longitude
155° E., to the point where they were picked up by the Eskimo,
off Julianhaab, in South Greenland, a distance of 2900 miles. ^

Assuming that the resultant of the drift of these casks will be the
same as that of the Jeannette before she sank, and assuming that
their subsequent drift will be at a rate of speed corresponding to
that of the relics — that is, about 2.6 miles per day of twenty-four
hours — we find that a period of about five years will be required to
bring them to the same locality ; but it is only fair to assume that
a certain percentage of the casks which are carried in this great
current — estimated to be 300 miles in width — will find their way to
the shores of Franz Joseph Land, Spitzbergen or Nova Zembla,
in which event the chances are quite good of their being recovered
at an earlier date by Norwegian walrus hunters or fishermen.

The controlling influence of winds in their relation to the recog-
nized and well-defined ocean currents is a fact accepted by all
meteorologists at the present day. It is said currents are set in
motion by this agency which attain a speed of from three to four
miles per hour. If such is the case where the ordinary ocean sur-
faces are concerned, how much more potent must the impelling
force of the winds be in conditions where countless ice surfaces are
presented to its action. In reading of the drift of the Jeannette,
nothing is more striking than the rapid progress the imprisoned
ship made in the summer months as a result of the influence of the
continuous southeast winds which prevailed. Admiral Melville
alludes to the effect of innumerable hummocks of ice, '' like mil-
lions of sails set to catch the breeze," and states that after each of
these disturbances had subsided a setback drift to the southeast set
in.^ The experiences of the Fram party appear to have been quite
similar ; and these facts would seem to point to some uniform and

1 The Farthest North, Dr. Fridtjof Nansen, Vol. i, p. 19.

2 « The Drift of the Jearnette," Id.

160 BRYANT — DRIFT CASKS IN THE ARCTIC OCEAN. [Aprils,

consistent set of conditions which may be explicable from a meteor-
ological standpoint. The published weather charts show that there
is an area of low barometric pressure where a cyclonic disturbance
takes place whose centre in summer time is well north of Bering
Strait, and which in yielding to seasonal changes drops down at
the approach of winter to a region south of Bering Strait.

Thus in the summer months this centre of disturbance would be
north of the position occupied by the Jeannette, and winds in fol-
lowing their accepted course from west to east would naturally be
drawn in and approach the cyclonic area from the southeast, caus-
ing the continuous gales referred to by Admiral Melville.

Meteorologists also refer to the existence of a centre of cyclonic
disturbance between the 70th and 80th parallels of north latitude,
which, following the general law, progresses with the seasons on a
more or less uniform path from west to east around the Polar basin.
The presence of this moving centre of attraction (if it is accepted
as an existent factor in circumpolar meteorology) must exert a con-
trolling-influence on the winds of this region, and it follows as a
logical sequence that the direction of the ocean currents must be
regulated by the same phenomena. Without claiming any special
knowledge of this branch of the subject, it occurs to me that in
this connection we may find an explanation of the existence of
more or less constant winds at certain times of the year in the re-
gion referred to, and in these phenomena may lie the interpretation
of the reality and constancy of the great Polar current.

With the generous cooperation of the various agencies men-
tioned, the drift-cask experiment has been successfully inaugurated.
It is our intention to bespeak the aid of the U. S. Consuls in
northern Europe likely to come in contact with seafaring people
who may visit the northern waters where these casks may eventually
put in an appearance. It is hoped that such publicity will be given
to the project, that when the time arrives for these inanimate mes-
sengers to appear in waters frequented by men, a certain percentage
of them may be recovered and reported upon.

We look forward with keen interest to the outcome of the present
campaign of the gallant Peary, and to the efforts of the superbly
equipped Baldwin-Ziegler expedition in its attack on the Pole.

To the Norwegian expedition under Sverdrup and the Russian
one under Baron Toll, we also wish a full measure of success. And
we trust all these explorers will return with important contributions
to our knowledge of the far North.

1902.] OLIVER — BLINDNESS FROM MALFORMATION OF SKULL. 161

Should the prize of the centuries be denied to these intrepid
voyagers, however, it may be that some devoted enthusiast will be
moved to attempt to explore the unknown area in an expedition
planned on the lines of the Fram voyage, which after all promises
the greatest chance of success.

In such an event it is hoped that the data secured as a result of
this drift-cask experiment may be found to be a contribution of
some value to the hydrography of the Arctic regions.

BLINDNESS FROM CONGENITAL MALFORMATION
OF THE SKULL.

(Plate XX.)

BY CHARLES A. OLIVER, A.M., M.D.

{Read April 5, 1902.)

Congenital malformations of the skull assert their evil effects
upon the integrity of the tissues of the visual apparatus and its
consequent functioning in definite ways. Should the disturbing
factors be set into activity during intrauterine existence, while the
cranial bones are passing through their primary stages of develop-
ment, the direct effects of such disturbance will be so great that
not only will organic changes appear in the ocular structures, but
coarse associated faults will manifest themselves in the related and
contiguous tissues.

The posterior portion of the cranium is proportionally the
largest during the early stages of development of the skull, the
parietal regions beginning to enlarge at about the eighth week of
intrauterine life, followed soon afterward by the frontal and the
occipital regions.

The newly born cranium is relatively very large in comparison
with the rest of the body. In contrast with the facial portion it
exhibits a predominance of the cerebral part in proportion of seven
to one. The six membranous fontanelles and the fibrous septa
between the adjacent osseous structures continue intracranially with
the dura mater and extend extracranially to form the pericranium,
giving rise to sacs in which bony plates without diploe are situated.
At this period of life there are cartilaginous areas scattered through
the occipital bone, while the presphenoid portion of the sphenoid

PKOC. AMER. PHILOS. SOC. XLI. 169. K. PRINTED JULY 8, 1902.

162 OLIVER — BLINDNESS FROM MALFORMATION OF SKULL. [April 5,

bone fails to exhibit any of the sinuses that are seen in adult life.
The optic foramina are large in size and triangular in shape, having
been obtained by the confluence of the presphenoidal and orbito-
sphenoidal centres. The superciliary ridges and frontal sinuses
are not yet present. The lacrymal bones consist of simple delicate
sheets. As a rule, the nerve foramina occupy sutural points or
positions of ossific centres.

Both the primary and the secondary foramina, particularly the
latter, are disturbed by distortion-processes taking place during
their passage through many complicated bony tunnels before they
escape through the dural sheath, as is primarily done by the former
types.

Minor arrests and perversions of development in the bones of the
upper face are so frequent that they constitute the daily findings of
the scientific ophthalmologist and trained optician. Orbital de-
formities, more especially those of the rim of the orbit, are very
common, and although they have decided effects upon refractive
error and exterior-ocular muscle-equilibrium, they fail to exert but
little, if any, damage upon combined visual functioning when the
resultant functional faults are either orthopedically or radically
corrected. More pronounced osseous deformation, the result of
disturbances of development of the bones of the face, show coarser
signs of fault in the eyeballs and their adnexa ; exhibiting, for
example, monolateral and bilateral stenoses of the nasolachrymal
ducts. In the grosser forms of congenital malformation leading to
antenatal or, later, postnatal blindness (the subject-matter of this
communication), it is probable that the primary changes have taken
place in the notochordal and trabecular regions during the chondral
stages of development of the brain-case. In these types, both
irregular ossification with consequent cranial contraction in one
situation and undue expansion in another, and undue sutural closure
from inflammation of the osteophytic membranes with resultant
thickenings and ridge-like eminences along the osseous junctures,
especially in the basilar series of bones at their asteriorial, inional
and lambdal points of junction, may appear.

The normal morphology of the skull is expressed in three stages.
The brain vesicles are at first enclosed in a thin delicate sac, a part
of which gradually hardens into a fibrous membrane, while the rest
persists to form the dura mater of postnatal life. The second stage
is represented by a partial conversion of the metamorphic tissues into

1902.] OLIVER — BLINDNESS FROM MALFORMATION OF SKULL. 163

cartilage, particularly at the sides and the base of the membranous
cranium. During the third stage, true osseous material obtained
from both the membrane bones and cartilage bones appears, until
finally a more or less completed bony coveri::g containing rem-
nants of chondral matter is obtained.

The occipital bone originates from four centres : the basioccipital,
formed from cartilage at about the seventieth day ; the two exoc-
cipitals, also derived from cartilage a few days later; and the
squamoccipital, composed of two parts, the interparietal and the
supraoccipital, which appear from separate nuclei at about the
eighty-fourth day, and unite in about twenty-four days' time. At
birth all of these parts are connected by cartilaginous strips. They
are not fully fused until the seventh year of postnatal life, the two
exoccipitals and the squamoccipital becoming ankylosed some two
years later.

The sphenoid bone arises from twelve bone nuclei arranged in
pairs, these being divided into two pair for the presphenoidal and
four for the postsphenoidal centres. These centres successively
appear from the fifty-fourth to the ninety-first day of intrauterine
life.

The parietal bones are of interest, as they constitute a great por-
tion of the vault and sides of the skull, and are in direct relation-
ship with some of the most important sutures — the sagittal with its
fellow, the coronal with the frontal, the lambdoidal with the
squamoccipital, and the squamus with the squamal ; the anterior
inferior angle articulating with the sphenoid, and the posterior
inferior angle articulating with the mastoid portion of the petrosal.
As a rule, each parietal bone ossifies from a single earthy spot,
situated in the outer layer of the membranous covering of the cra-
nium, at about the forty-second day of intrauterine existence.

The frontal bone, another important suture-bearing roof bone,
arises from two earthy spots in the external layer of the membranous
covering of the cranium, about a week later than those that are
intended for the parietal bones. These two portions, as a rule,
inite soon after birth by a median suture-line known as the metoptic.
Ankylosis commences at about the second year of postnatal exist-
ence. A portion of the bone helps form a part of the orbits and
has its main connections with the ethmoidal, the lacrymal, the
malar, the superior maxillary, the nasal, the parietal, and the sphe-
noidal bones.

164 OLIYEK — BLINDNESS FEOM MALFORMATION OF SKULL, [April 5,

The epipteric bones, wedged between portions of the frontal, the
parietals, the sphenoid and the temporal bones, are of importance
in this study. They are present from the second year of life to
about the age of adolescence ; they then persist as true ossicles or
help to form new sutures. They are variable in size.

The Wormian bones, that at times are found in great numbers in
the various sutures of the cranial part of the skull, must also be
considered of value in this connection.

The sphenoid bone, the most important and the most irregular
of the basilar bones, is situated in the region of the anterior and
middle fossa. It practically contains all of the foramina and fis-
sures intended for the emergence and the exit of the sensory and
motor nerves, blood vessels and lymph channels connecting the
intracranial and external portions of the visual apparatus. The
middle fossa is the most complicated of the three great depressions
in the floor of the cranial cavity, it containing all of the most im-
portant nerve communications and vascular and lymph channels
that are in association with the optic nerves and eyeballs. The
posterior fossa hold the occipital lobes in their subdivisional cere-
bral fossa, that are situated above the groove that is intended for
the course of the lateral sinus.

It is a well-known fact that cranial asymmetry is almost universal.
Study of the main foramina and fissures of the various orbital cavi-
ties of man exhibit marked variabilities in their relative sizes,
shapes and positions. The average depth of the orbit of the Negro
race, for example, is at least an eighth greater than it is in the orbit
of the Caucasian ; while the early ossification of the septum with
the superior maxilla in the same race produces a normal flattening
of the glabella, with a lateral broadening of the alse of the nose.
Moreover, in this class of subjects the characteristic prognathism of
the race becomes apparent when the individual has passed the
pubertal period, at which time of life an over-development of the
inferior maxillary bone occurs. Here there is type-form of indi-
vidual with a flattened nose, a wide interpupillary distance, a broad,
flat forehead and a projecting malar prominence, that are all so
characteristic of the usual brachycephalic head : here there are
individuals representing one of the principal subspecies of human
life in whom there are probable retentions of some of the most
pronounced features of the quadrumana; a true acceleration, as it
were, passing beyond the Caucasian retardation of embryonic
development.

1902.1 OLIVER — BLINDNESS FROM MALF0R:\[ATI0N OF SKULL. 165

It is not, however, with these minor and relatively undisturbing
types that this communication deals. It is with the grosser forms
of cranial malformation ; those that particularly involve the basil-
ary fossa and their many fissures and foramina ; types which sooner
or later give expression to blindness as one of their most prominent
and characteristic symptoms.

The gross configuration of the skull and the condition of the
various portions of the visual apparatus are so strictly in accord with
one another, that certain forms of cranial asymmetry can, with
almost definite precision, be associated with certain kinds of blind-
ness. Five coarse clinical types of cranial deformation — the well-
known oxycephalic, the scaphocephalic, the leptocephalic, the
trigonocephalic, and what I have elsewhere described as the occipi-
tal or occipito-parietal — may be cited.

The oxycephalic or even the gross hypsicephalic type is char-
acterized by the so called steeple-shaped or dome-like head. It is
dependent upon an improper union of the parietal bones with the
occipital bone, the temporal bones and the sphenoid bone, produc-
ing compensatory over-developments along the sagittal suture and
in the position of the anterior fontanelle. The pterion region with
its anterior lateral fontanelle and later Wormian bone, and the
region of the lambda marking the situation of the posterior fonta-
nelle, with its intervening sutures and angular articulation, are all
too early united and ankylosed, giving rise to corresponding rela-
tive disturbances in the calval portion of the cranium, particularly
along the sutural lines and in the most nearly related fontanelles.

In this type, which may be very slight or of the grossest charac-
ter, as shown in the accompanying reproductions of two undeniable
cases occurring in my public practice at the Philadelphia and Wills'
Hospitals (Plate XX, Figs, i and 2), the visual signs of the disease
vary from the veriest eye symptom to the coarsest ocular expres-
sion, and may first appear at any time during early or middle life.

Case I. — The gross example shown in Fig. i was that of a sixty-two-
year- old negro, who, with a history of an acute attack of convulsive
seizures from fright, occasioned, he asserted, by a fall ^ at one year of
age, had three years later the expression of "pop-eyed" epithetically
applied to him, this pseudonym having since persisted throughout life.
The patient stated, and I one day had a clinical demonstration of the

^ The postnatal fall as a causative factor for the cranial malformation must be
rejected when the congenital stigmata are considered.

166 OLIVER — BLINDNESS FROM MALFORMATION OF SKULL. [April 5,

same, that he had more than once pushed his right eye out between the
lids. Five years before I saw him he accidentally discovered that he
could not see with the left eye. Two and a half years after this the
sight of the right eye began to gradually fail, until at the time of exam-
ination it was found that vision with it was reduced to a faint doubtful
perception of light in an inferior temporal field. The left eye was blind.
The superficial areas of the two orbits were immense. The lids were
large and the palpebral fissures were broad and long. When the posi-
tion of the left eye was gauged so as to have its supposed visual axis
directed straight ahead, the right eye projected two and a quarter milli-
meters forward beyond the superior and the inferior margins of the
orbit, and diverged some thirty degrees out and three degrees down.
When the right eye was placed in the same relative position, the left eye
was found to be almost as greatly diverged and was directed somewhat
more downward.^ Curiously, extraocular motion was very little if at
all disturbed, although palpation showed that the eyeballs were situated
in extremely shallow, almost saiicer-like orbits, the shallowest portions
of the cavities being situated toward the median line. The corneal
epithelium was thickened and the deeper structures of the membrane
were opaque in a couple of places. The pupils were large and the irides
seemed disproportionately sluggish in their various reactions to the
amount of local disturbance. Both lenses presented evidences of dense
secondary degeneration, that of the left eye being so opaque that the
fundus of the organ was invisible. A faint red glare, with the appear-
ance of a few retinal vessels — best seen with a minus spherical lens of
twenty diopters' strength — made it probable that portions of the sec-
ondary ocular lesions were due to a high-grade myopia. Intraocular
tension in each eye was normal. The anterior scleral vessels were not
engorged, and there was not any ciliary tenderness.

The conformation of the skull was typical. The lower jaw, which
was increased in size, was mesognathous, if not prognathous in shape.
The condition of the hands, as seen crossed upon the body, discredited the
belief of any disease of the pituitary body. The bitemporal diameter
of the skull was but thirteen and a half centimeters, and the biparietal
was but one and a quarter centimeters wider. The occipito-frontal diam-
eter equaled eighteen and a half centimeters, while the occipito-mental
was somewhat in excess of twenty-six centimeters. The trachelo-breg-
matic diameter was twenty-three and a half centimeters in length.^

1 The exophthalmus and divergence can be easily differentiated by examina-
tion of the reproduction of the photograph of the case.

2 I am under obligations to Dr. Clarence Van Epps, one of my Residents in
both institutions, for presentation of the copy of the photograph of the first
subject taken by Mr. James F. Wood, of Philadelphia; to Dr. Frederick C.
Krause, one of my former assistants, and now Assistant Ophthalmic Surgeon to

1902.] OLIVER — BLINDNESS FROM MALFORMATION OF SKULL. 167

The second example of the type, in a German, an excellent
illustration of possibly an extreme hypsicephalic skull with a pre-
ternaturally elongated bregmato -mental diameter, is not quite so
rare, I having the opportunity to systematically study four or five
such patients in a total number of some sixty to seventy thousand
cases of ophthalmic disease that I have seen in the combined public
and private practice of myself and others.^

The reproduction of the photograph of the case shown in Plate
XX, Fig. 2 gives a good idea of the general appearances of the head
in profile. In this case the suboccipito-bregmatic circumference
equaled twenty inches, the occipito-frontal circumference was
nineteen and a half inches, and the occipito-mental circumference
equaled twenty-six and a half inches.

Case IT. — The patient, who was born in Germany, was a thirty-five-
year-old farmer. He stated that he had always had a curiously shaped
skull. He had been free from all disease until he was ten years old, at
which time he had had a series of spasms. These convulsions were
associated with a permanent divergence of the eyes and a persistent in-
different vision which was more pronounced in the left eye. Three
weeks before I saw him, he noticed that the sight of his good eye began
to fail, this failure being associated at times with deeply seated orbital
pains on the same side. His habits, he said, were good, and there were
not any signs of gross hereditary or acquired disease. No other mem-
bers of his family "for three generations back had gone blind." His
parents were not blood relations.

Vision with the right eye was reduced to an incorrectible one-eighth
of normal in an excentrically placed field, with its fixation-point situ-
ated far up and in. Color perception for green, red, blue and yellow
was lost. Vision with the left eye was almost gone, there being but one
small area of doubtful at times light-perception situated in an extreme
temporal field as the last remnant of sensory functioning. Intraocular
tension in each eye was normal. The pupil of the left eye, which was
round, was about two millimeters larger than the similarly shaped one
of the right eye. The right iris responded fairly well to light-stimulus
and accommodative efforts, giving rise to rather prompt consensual
reactions of the iris of the almost blind left eye during both of these

St. Christopher's Hospital, in Philadelphia, lor photographing the second case ;
to Dr. William L. Zuill, one of the Assistant Surgeons at Wills' Hospital, for the
craniometric measurements of the second case ; and to Dr. Frank R. Harrison,
of East Liverpool, Ohio, for securing the photograph of the third case.

1 Individuals from two races have been purposely used in the elucidation of
this phase of the subject in order to obtain exceptionally broad standpoints of
observation.

168 OLIVER — BLINDNESS FROM MALFORMATION OF SKULL. [Aprils,

impulses. The left iris was almost immobile to light-stimulus thrown
upon its retina, but responded feebly to forced movement for supposed
accommodation, and gave quite prompt consensual reaction to the iris
of the less affected organ. Gross downward convergence of the two
eyes, by having the patient endeavor to look at his nose tip, rapidly
brought the pupillary areas down to one millimeter each in size.

In spite of a left divergence of about thirty degrees out and slightly
down, the exterior muscles of the two organs seemed to enjoy good
movement. An almost constant lateral nystagmus that increased upon
attempts at near fixation was a prominent symptom.

The patient's eye-grounds were characteristic of consecutive atrophy,
that of the right eye showing evidences of a recent optic neuritis of
postocular type.

Although not hoping for any permanency of result, I gave the patient
the benefit of therapeutically driving more blood through the half-
starved and degenerating neural tissues of the affected optic nerves.
This was done by the internal administration of large and frequently
repeated doses of strychnia, resulting in a temporary betterment.^

The main disturbances upon the visual apparatus in this case,
therefore, which were probably of twofold character — meningitic
and mechanical — were mainly exerted upon the optic nerves at the
optic foramina. Secondary degeneration changes were only too
certain, as later proven by the steady decline of vision in spite of
all constitutional treatment that could be conscientiously and judi-
ciously directed against any supposed dyscrasia.

The scaphocephalic type of cranial malformation exhibits a boat-
shaped form of deformity of the cranium, with an extremely broad
forehead. The deformation is dependent upon a premature union
of the sagittal suture between the medial margins of the parietal
bones. Here the brunt of the disturbance seems to exert itself
upon the median posterior portion of the anterior fossa, the limbus
of the lesser wing of the sphenoid bone, and the anterior medial
portion of the middle fossa. True optic neuritis with consecutive
atrophy; prominent, sightless and divergent eyes; pupils partly
dilated, and irides fixed to light-stimulation, are the most prom.inent
eye-symptoms in such cases. Intelligence is but fair, convulsive
seizures are not infrequent, and a lethal ending from some ordinarily
innocuous disease is most frequently an early event. Rapid and

1 During a portion of my studies of this case the patient attended the public
clinic of my friend, Dr. George C, Harlan, at the Pennsylvania Hospital. Dr.
Harlan's findings and results of treatment coincided with my own.

1902.] OLIVEE — BLINDNESS FROM MALFORMATION OF SKULL. 169

unstable increases of intracranial pressure from ventricular disturb-
ances are frequent, giving rise to repeated optic nerve-head swell-
ings and retinal extravasations.

The head of the leptocephalic type is small. This condition is
caused by a too early union of the fronto-sphenoidal suture between
the alae of the frontal and sphenoidal bones. In this type the in-
tracranial distortions, particularly thofe that affect the foramina and
fissures between the body and the greater and lesser wings of the
sphenoid bone, bring optic nerve atrophy from previous inflamma-
tion, and later palsies of the exterior ocular muscles, into existence
very soon after birth. <

The trigonocephalic or three-cornered type of cranial deformity,
with its small end situated anteriorly, is dependent upon a prema-
ture or improper ossification of the frontal and parietal bones along
the coronal suture, particularly in the region of the bregma or
pterion ; or, at times, it may be due to a fault in osseous ankylosis
of the combined frontal bones along the metoptic suture-line.
Postneuritic atrophy, the principal ocular expression of the disease,
occurring quite early in postnatal life, is apt to appear in the gross
examples of the type.

The rarely seen occipital or occipito-parietal type of cranial
deformity exhibits a flattened curving of the posterior portion of
the cranium. The condition seems to be dependent upon either a
too early syntosis of the occipital suture, especially at the lambdal
region, or an improper union of the medial portion of the lamb-
doidal and postero-inferior part of the sagittal suture in the region
of the posterior fontanelle. Here, in the superior and the posterior
parts of the deepest portion of the intracranial cavity (in the
interparietal parts of the occipital bone above the grooves for the
lateral sinus), the osseous tissues are distorted and flattened. In
certain places this condition is so pronounced, that in some situa-
tions the cerebral fossa are almost annihilated, and the inmost por-
tion of the elevation of the superior longitudinal sinus and falx
cerebri is increased. The most marked ocular signs are almost
wholly sensory in character. Vision in each eye is nearly or entirely
lost. The orbits are shallow, particularly at their postero-mesial
parts. The eyeballs are but slightly proptosed, somewhat enlarged,
and enjoy full freedom of movement. The motor apparatus of the
exterior of the eyes, with the exception of a few minor discrepan-
cies of probable improper nuclear action, is in good working

170 OLIVER — BLINDNESS FROM MALFORMATION OF SKULL. [April 5,

order. The pupils are but slightly, if at all, oversized. The irides
are prompt to light-stimulus, efforts for accommodation, and con-
vergence. The ciliary muscles are active. The eye-grounds, in
every detail of neuronic, vascular, and lymph structure, appear
normal j in fact, the eyeballs, with their entire adnexa, are healthy
and perform their functional duties properly.

This complexus of symptoms, with its absolute blindness and
concomitants of slight globular protrusion, divergence and the
rotary nystagmus as the only ocular signs, constitute a most remark-
able clinical picture. In it is seen a blindness, the proving of which
necessitates a careful study of every possible direct and indirect
ocular detail ; a blindness that, from the ocular signs and associated
conditions, may be assumed as intracranial in type, and most
probable, until autopsy proves to the contrary, mainly cortical in
character.

The accompanying reproduction of a photograph (Plate XX, Fig.
3) of a case recently studied by me and described in full elsewhere,^
gives an excellent idea of the cranial deformation and the peculiar
facial appearances and expression in an American-born type of case
of this character. In this child the optical and receiving portions of
the visual apparatus were apparently perfect. No visual perception,
however, could be evolved in this case, no matter how centrally the
impression reached (surely in this case back to the midbrain).
Cortex sensation was lost ; the discharging station was functionless.^

The cases thus far described exhibit but little, if any, mental
involvement. The grossest of the resultant disturbances are mainly
basilar in character, and in measure affect the vascular channels,
the lymph cavities and the coarse nerve fibrils as they pass through
both the primary and the secondary foramina. Trophic ocular dis-
order soon takes place ; ophthalmic irritation signs and palsies early
appear ; sensory changes in the organs of vision quickly ensue ; and,
sooner or later, the main portions of the receiving, transmitting
and discharging parts of the visual apparatus degenerate and become
useless. Should the main distortions be situated in the anterior
and central portions of the cranial base, producing antero-midbrain
disorder, as in the first illustrative case, the more frequently in-

^ The American Journal of the Medical Sciences, January, 1902.

2 It is probable that cases of the badly termed condition " amaurotic fam-
liy idiocy," with their peculiar lesions in the fundus of each eye, have some such
similar origin.

1902.] OLIVER— BLINDNESS FROM MALFORMATION OF SKULL. 171

volved become the ocular end organs. In this type the most bizarre
motor ophthahiiic signs are commingled in complicated yet definitely
determinate interrelationships. On the contrary, the further back
the coarse osseous changes are found, the greater become the sen-
sory deficiencies of the visual apparatus and the better preserved
remain the organs of vision and their contiguous parts.

In the anterior types the main basal cause of the condition may
be summed as a series of asymmetries of basilar structures, with
coarse anomalies in the various portions of the underlying sphenoi-
dal and contiguous bones.

In the posterior types histological examination reveals cortex and
nuclear changes in the posterior part of the sensory portions of the
visual apparatus. In some such individuals the cellular elements
may have attained a good size, and may have been able to function
most excellently during early postnatal life. This can be under-
stood when it is realized that nearly ninety per cent, of the gross
volume of the brain mass is obtained during the first stage of post-
natal existence ; later, the association fibres and the neural cells
continue to be the main factors of growth. This development, of
course, exerts its influence upon the formative processes taking place
in the osseous cranium.

In the majority of cases of these types there is a true tissue-
sclerosis.

In deformation of the cranium occurring at a very early antenatal
stage, the visual apparatus is more liable to become affected than
any of the other special sense organs. On the contrary, morbific
causes which affect the same apparatus during the later stages of
development of the skull and its contents are not so apt to affect
the organs of vision. It may be also of interest to state that the
sensory portion of the visual apparatus being developed much earlier
than the motor, and not possessing so many separations and ramifi-
cations in midbrain, is better able to withstand coarse pathological
changes than the latter. Statistics and personal observations, how-
ever, have determined that the great majority of congenitally blind
subjects possess malformations of the skull and its appendicular
elements.

Far different are the grosser forms of more generalized cranial
deformation, such as the two great classes, microcephales and ma-
crocephales. Coarser disturbances of sensation, grosser peculiarities
of motion, and increased degrees of trophic condition affect other

172 OLIVER— BLINDNESS FROM MALFORMATION OF SKULL, [April 5,

situations more markedly than they do the visual apparatus. Such
cases always present mental inefficiencies and disturbances from
either gross organic change or deficient development and growth in
the intracranially placed tissues. Circulation of but small quanti-
ties of blood and lymph of poorly nourishing quality through the
distorted and ofttimes inflamed and even contracted tissues, is seen
in so many cases of this coarse type of disease, that it seems no
wonder that cerebral development and growth soon become affected.
Many such subjects are fortunately early victims of convulsive
seizure, mental hebetude, general wasting from ectogenous infection,
and death.

The cerebral alterations in these types are many. Should the case
exhibit mental deficiency, the convolutions are generally gross,
narrow and uncomplicated, while the related gyri are small and
badly developed. Fissural confluences may be present, and not
infrequently the occipital lobes do not extend over the usually too
large cerebellum. These conditions are probably also found in
other forms of genetous idiocy with and without eye lesions.

Some cases of that rare condition, microcephales, from too early
ossification of the cranial sutures with and without idiocy, may
have true microphthalmus as a jmrt of the products of the same
morbid cause ; though functioning power, particularly that for
color-perception, as far as can be scientifically determined, may,
even in minor cases, remain practically undisturbed. As a rule,
the eyeballs of such subjects are relatively well placed, and exterior
ocular muscle action seems good.

In hydrocephales, on the contrary, there are frequent disturbances
of muscle action in and around the eyeball, particularly during
attempted movements of coordination, and when the parts are
brought into association with the related ocular reflexes.

An extremely broad interpupillary distance with a broadening of
the zygomatic arches forms one of the characteristic ophthalmic
features of congenital cretins, whether they be endemic or sporadic
in origin. In this peculiar type of cases sight is generally undis-
turbed, the sensory part of the visual apparatus usually being good*
The visual organs, however, are somewhat differently sized. In
such cases disturbances with the motor portions of the visual appa-
ratus are quite common. The affected individuals are frequently
deficient in hearing and are often unable to enunciate. The size
of the orbits in these cases is unequal. The osseous irregularities.

1902.] OLIVER— BLINDNESS FROM MALFORMATION OF SKULL. 173

however, are greater at the base of the skull. There is always a
marked tendency to cranial asymmetry, the most pronounced
abnormality consisting in a premature ossification of the spheno-
basilar bones. In these cases the distance from the glabellar point
to the occipital foramen is said by some to be quite short ; by
others this shortening is denied. Curiously, such subjects are said
to never shed tears. Investigations, however, especially as to the
•condition of the secretory apparatus in these cases, should be made
before any such dogmatic assertion as this can be hazarded.

It must be remembered that this communication does not deal
with monstrosities such as cyclocephales, in which it is stated there
is a circumscribed impairment of development and growth from
mechanical pressure, exerted in some instances by the amniotic
hood, an increase of intracranial pressure, resulting in rupture of
the early cerebral vesicle, or an arrested development of the anterior
vesicle as one of the results of anomalies in the amnion. This
form of malformation presents several varieties. The first type of
a true cyclopic monstere is that exhibiting the rhinocephalic mal-
formation. Such an individual is represented by a head containing
two more or less completely fused rudimentary eyes in a single orbit,
the nose consisting in a proboscis situated above the orbit When
there is a complete fusion of the orbital cavities and eyeballs with-
out the vestige of a nose or a proboscis, the variety receives the
designative term of cyclocephalus. Should the lower part of the
face be additionally affected and the integument overlying the im-
perfectly developed superior and inferior maxillary bones hang in
folds, the condition is known as stomacephalus.^

The artificial deformation of the skull of the infant in all manner
of fantastical ways, which has been practiced by many tribes
throughout the world before even the time of Hippocrates, is inter-
esting in the fact that although of necessity the three great portions
of the combined visual apparatus — the receiving, the transmitting
and the discharging — must in every instance have been more or less
pressed upon and distorted, yet probably by reason of the distortion

1 These type-forms do not strictly include the nose-headed or ethmocephalic
form of monster, in which there are two eyes and two orbital cavities, the nose
being represented by a proboscis that is provided with either one or two nostrils.
Neither do they include the monkey-headed or cebocephalic variety, in which
there are two orbital cavities and two eyeballs, but not any nose, the intra-
•ocular region being both narrow and flat.

174 KRAEMER — CONTINUITY OF PROTOPLASM. [April 4.

having been gradually accomplished after birth, gross bulbar dis-
turbance, blindness, faulty muscle action, and coarse atrophic dis-
order have not been produced, and hence remain unmentioned as
ordinary consequences in such cases.

Blindness from deprivation (postnatal causes), as in the wide-
world known case of Laura Bridgman, which on autopsy was found
to be associated with optic nerve and optic tract atrophy and thin-
ning of the gray matter of the occipital cortex, is also a subject for
discussion elsewhere.

ON THE CONTINUITY OF PROTOPLASM.

BY HENRY KRAEMER, PH.D.
(Plates XXI and XXH.)
{Read April I^, 190S.)

While Schleiden^ conceived each cell to have an independent
existence, Hofmeister' contended that the protoplasts of contigu-
ous cells are united,, forming a higher unity; that is, one synplast.
In later years both Sachs * and Strasburger ^ have supported the view
of Hofmeister. And even so great an authority as Nageli ® ex-
pressed the view that neighboring plant cells are united by means
of threads of protoplasm in much the same manner as in the sieve
tubes first described by Hartig ® some thirty years before.

In 1878 Thuret and Bornet' first called attention to the fact that
in certain of the Florideae the contents of certain of the cells of
the trichophore and carpogonium are directly connected by means
of pores. Fromann * appears first to have called attention to the
direct connection of protoplasm in the higher plants, in the epider-
mal and parenchyma cells in the leaves of Rhododendron and Dra-
cena. While TangP was preceded by these several investigators,
the establishment of the view that there is a continuity of proto-
plasm is due for the most part to his researches. On treating dry
sections of the endosperm of Strychnos Nux vomica with dilute
iodine solutions, he observed a distinct lamellation of the cell wall
as well as the formation of yellowish striae, which latter he con-
ceived to be plasma threads connecting the different cells. The
appearance thus produced he compares to the structure of the sieve
tubes, but in speaking of the contents of the latter, he states that

1902.] KRAEMER — CONTINUITY OF PROTOPLASM. 175

they can hardly be considered to be in the nature of protoplasm,
and substantiates this statement by quoting from De Bary and
Sachs.

A few years later Gardiner/*" while working in the laboratory of
Sachs on certain sensitive plants, observed by the use of sulphuric
acid or chlor-zinc-iodide and Hofmann's blue or methylene blue,
colored stride in the walls of certain of the cells, which he consid-
ered to be in the nature of threads of protoplasm. A number of
other workers have also considered this subject, using a similar
technique to that of Gardiner, confirming his observations and
extending the number of species showing a continuity of proto-
plasm.

The results obtained by these investigators tend to show that
there are two kinds of continuity of protoplasm, one through open-
ings in the pores which apparently occur in the larger number of
cases, and another in which the threads of protoplasm extend
through walls in which there are no pores. Several investigators "
even go so far as to express the view that probably every cell is
connected with its neighboring cells by protoplasmic threads.

That there is a continuity of protoplasm has become almost a
fundamental principle in botany, it being considered necessary in
the transmission of irritation currents and in the distribution of
protoplasm and such bodies as starch grains and oil globules, intact
and quickly from cell to cell.

While fully cognizant of the plausible arguments which have
been advanced in favor of the continuity of protoplasm, and, fur-
thermore, not desiring to consider the subject theoretically, by the
discussion of certain facts in regard to solution, osmosis, the ascent
of sap, and other physical phenomena that might more favorably
assist the plant in its various functions than a protoplasmic connec-
tion between the cells, the author presents herewith some of the
results of his studies on the structure of the starch grain and cell
wall, in the belief that they will throw some additional light on the
subject under consideration.

Suffice it to say that these results seem to offer a different expla-
nation for the phenomena observed by the investigators already
mentioned, in their studies on the continuity of protoplasm. In
other words, the appearances described by these authors as indicat-
ing a continuity of protoplasm are due to a peculiarity in the
structure of the cell wall, which is made manifest by the reagents

176 KRAEMER — CONTINUITY OF PROTOPLASM. [April 4,

employed and which bears an analogy to the structure of the
starch grain.

In the author's studies on the starch grain, the following obser-
vations have been made :

(i) The illustrations of potato starch in the various text-books
show two kinds of grains, one with the point of growth and the
alternate lamellae light in color, as figured by Sachs (Plate XXI, Fig.
i), and the other with the point of growth and alternate lamellae
dark, as figured by Strasburger (Fig. 2). This appearance, how-
ever, is not due to a difference in the grains, but is brought about
by the manner of focusing on them. In the figure given by Stras-
burger the lamellae are viewed from above, while in the figure of
Sachs the view is from below.

(2) On treating the starch grain with water at different tempera-
tures and a number of reagents,* a radiating crystal-like structure
is observed in the successive layers (Fig. 5). This crystalline
structure appears to be most pronounced in the layers alternating
with the point of growth, and is succeeded by the formation of a
number of clefts or fissures (Figs. 6 and 7). In potato starch these
clefts are more or less feather-like in appearance, and extend from
the point of growth through the middle of the successive layers to
the periphery of the grain. In wheat starch the fissures extend
radially from near the point of growth to near the periphery.

(3) On treating starch grains with weak aqueous solutions of
various aniline dyes, as gentian violet, eosin and safranin, it is
observed that the layers which are less crystalline or colloidal in
character take up the stains (Figs. 3, 4 and 7). The various clefts
and fissures produced in the grains behave toward staining reagents
much like the colloidal layers, and they are probably the tracts or
channels through which liquids are distributed throughout the
grain.

(4) We further find that these two kinds of layers behave differ-

* The reagents used were the following: (i) Chromic acid solution (5 to 15
per cent.); (2) Calcium nitrate solution (5 to 30 per cent.); (3) Potassium hy-
drate solution (one-tenth of i per cent.); (4) Sulphuric acid (10 per cent.); (5)
Silver nitrate solution (2 per cent.); (6) Sodium acetate solution (50 per cent.);
(7) Potassium nitrate solution (saturated); (8) Potassium phosphate solution
(saturated); (9) Hydrochloric acid (5 per cent.); (10) Potassium iodide solu-
tion (I to 10 per cent.); (ii) Tannic acid solution (5 to 15 per cent.); (12)
Saliva; (13) Taka-diastase (saturated solution) ; (14) Chlor-zinc-iodide solution;
(15) Chloral iodine solution and iodine water, equal parts.

1902.] KRAEMER— CONTIXUITV OF PROTOPLASM. 177

ently toward iodine; the one rich in crystalloidal substance
becomes blue with iodine, whereas the other is not affected by this
reagent.

In the studies of the author on the structure of the cell wall, the
following observations tending to show an analogy to the starch
grain have been made :

(i) A similar layering of the cell wall, known as stratification
and striation, is readily observable in the walls of endosperm cells
as well as those cells impregnated more or less with mucilage,
lignin, cutin, suberin and allied substances. In some cases the use
of reagents, as acids and alkalies, may be necessary to bring out
this structure (Fig. 8). While it is not always easy to determine
the nature of the successive layers in the wall, still the structure
seems to correspond in the main to that of the starch grain, the
middle lamella of the cell corresponding to the point of growth.

(2) The same kind of reagents, but in stronger solutions, may
be used to bring out the crystalline or spherite structure in the walls
of thickened parenchyma cells, as endosperm (Plate XXII, Figs. 9
and 13), or lignified cells, as stone cells. In cases where the cell
wall has been metamorphosed into mucilage, simple treatment with
water, as has also been shown to be the case with the starch grain,
is sufficient to bring out this structure.

(3) The differentiation of the layers of the cell wall by the use
of aniline stains,* has not as yet been attended with any marked
degree of success. The use of swelling reagents, as sulphuric
acid, in conjunction with a stain, has, however, produced more or
less interrupted striae resembling the clefts and fissures in the starch

* The methods involving the use of aniline stains in the study of the cell
wall are the same as those used in the study of the continuity of protoplasm, and
embody the three operations of fixing, swelling and staining, between each of
which operations the sections are washed quickly and with large quantities of
water. Fixing is usually accomplished by the use of aqueous iodine solutions
(.5 per cent, of iodine and .5 to i per cent, of potassium iodide); alcohol, osmic
and picric acids may also be employed. The swelling of the specimens is effected
by the use of dilute sulphuric acid (25 to 75 per cent.), iodine being sometimes
added to the sulphuric acid solution ; chlor-zinc-iodide and solutions of the alka-
lies are also employed for this purpose. The stains mostly employed are 5 per
cent, aqueous solutions of gentian violet, eosin or safranin, these being used in
connection with the swelling agents mentioned above. The time required for
each operation is usually from five to ten minutes, but when chloi -zinc-iodide is
used twelve hours may be required for the swelling.

PROG. AMER. PHILOS. 800. XLI. 169. L. PRINTED JULY 28, 1903.

178 KRAEMER — CONTINUITY OF PROTOPLASM. [April 4,

grain. In the case of Nux vomica, solutions of potassium iodide
and iodine produce yellowish-brown striae in fresh sections (Fig.
13), closely resembling in form those produced by aniline stains
(Fig. 14), and which were considered by Tangl as being protoplas-
mic threads, but which are probably due to the precipitation of an
alkaloidal salt in the clefts or fissures in the wall.^

(4) The two kinds of layers behave differently toward chlor-
zinc-iodide ; the one next to the middle lamella and those alter-
nating with it are colored blue, while the others are but slightly
affected.

The observations and comparisons herewith presented lead to
the following interpretations :

(i) The starch grain, as also the cell wall, is made up of alter-
nate lamellae of colloidal and crystalloidal substances.

(2) Physically, the structure of the starch grain and cell wall are
quite similar, although chemically different ; the preponderating
substance in the starch grain being granulose, while in the cell wall
the fundamental substance is cellulose, which may preponderate or
€xist in varying proportions.

(3) The crystalloidal layer in the starch grain, consisting chiefly
of granulose, is colored blue with iodine or chlor-zinc-iodide,
whereas in the cell wall this layer, consisting chiefly of cellulose,
is colored blue only with chlor-zinc-iodide.

(4) The colloidal layers in both the starch grain and cell wall
take up and hold various aniline dyes, the layers being, however,
more clearly defined in the starch grain, particularly potato starch.

(5) In starch grains as in cell walls, there are radial clefts or
colloidal areas which under certain conditions also take up and
hold various aniline stains.

(6) The plastid at the periphery of the starch grain may be
compared to the protoplasm of the plant cell, each contributing to
the growth of successive new layers. In the cell wall the mode of
growth is centripetal, whereas in the starch grain it is centrifugal.

The peculiar bi-convex arrangement of the groups of striae be-
tween contiguous cells in the Nux vomica and vegetable ivory is
rather suggestive of fundamental lines of development corresponding
to chromatin threads, although they may be modifications of the wall

^ This may explain why the iodine method alone has not met with any success
save in the case of fresh sections of Nux vomica.

1902.J KRAEMER — COXTINUITY OF PROTOPLASM. 179

and represent tracts or channels through which liquids are distributed
from cell to cell.

Furthermore, attention should be directed to the fact that the
preparations of both the starch grain and cell wall showing the
colored lamellce and striae, as already described, are permanent only
in Canada balsam and are ephemeral in glycerin or glycerin jelly.

Finally, it may be stated that all authors since the appearance of
Gardiner's work* have fallen into the error of supposing that a
certain aniline dye could be regarded as a differential stain for pro-
toplasm, whereas the fact of the matter is that many colloidal car-
bohydrates, as mucilage and pectin, and oils and other substances
as well, take up these stains. And in this connection we may ask,
If the substance in the cell wall which takes up the stain is proto-
plasm, what is it in the starch grain?

Bibliography.

1 ScHLEiDEN : Grundzuge der wissenschaftlichen Botanik, i. Aufl., 1842-1843.

2 HoFMEiSTER : Die Lehre von der Pflanzenzelle.

3 Sachs: Vorlesungen liber Pflanzenphysiologie, 1882, p. 102.

* Strasburger : Ueber den Bau und das Wachsthnm der Zellhaute, 1882, p. 246.
^ 5 Naegeli : Mechanisch-physiologische Theorie der Abstammungslehre, 1884,
p. 41.
6 Hartig : Botati. Ztg., 1854, p. 51.

' Thuret and Bornet: Etudes phycologiques, Paris, 1878.
8 Fromann : Sitzber. der Jenaischen Gesellschaft fur Medicin und Naturwis-

sensch., 1879, p. 51.
8 Tangl : Vnngsh.Q\m's yahrbilcher filr wissenschaftliche Botanik, 'Qz.yxd, 12,

1880, p. 170.
10 Gardiner: Arb. d. bot. Inst, zu Wurzburg, Bd. Ill, 1884, P- 52.
^^ Kienitz-Gerloff: Bot, Ztg., 1891, p. i.

Schaarschmidt : Botanisches Centralblatt, xviii, 1884, p. 265. (See also
A^ature, xxxi, 1885, p. 290.)

Explanation of Plates.

Plate XXI.

Fig. I. Potato starch grain with point of growth and alternate Iamell::e liaht in
color.

* Gardiner states that « All experiments made with the view of attempting to
detect the presence of protoplasmic filaments in the cell wall when the cell was
normal and intact met with but little success, so that in investigating the subject
of protoplasmic continuity the method of swelling the cell wall and subsequently
staining with a dye which was found to especially stain the protoplasm was
adopted."

180 KRAEMER — CONTINUITY OF PROTOPLASM. [April 4,

Fig. 2. Potato s'arcli grain with point of growth and alternate lamella; dark.
Fig. 3. Pctato starch grain treated with aqueous solution of gentian violet.
Fig. 4. Potato starch grain treated with gentian violet and showing crystalloidal

structure in alternate lamellae.
Fig. 5. Wheat starch grain treated with water at 60O C, or with chromic acid

and other reagents (see footnote *).
Fig. 6. Wheat starch grain treated with water at a temperature of 65O C, or

with the reagents mentioned in footnote *, but for a longer time.
Fig. 7. Wheat starch grain treated with aqueous safranin solution.
Fig. 8. Cells of the endosperm of Date seed (^Phcetiix dactylifera), the one

normal and the other showmg the stratification of the wall after treatment

with chlor-zinc-iodide.

Plate XXII.

Fig. 9. Cell of vegetable ivory [Phytelepkas macrocarpa), showing lamellation
and crystalline structure in the wall after treatment with chlor-zinc-iodide,
clove oil, chromic acid or other reagents.

Fig. 10. Pore of vegetable ivory showing cleft in middle lamella.

Figs. II and 12. Pores of vegetable ivory showing striae between neighboring
cells after treatment with sulphuric acid and gentian violet.

Fig. 13. Cells of endosperm of the seed of Strychnos Nux vomica after treat-
ment with iodine solution.

Fig. 14. Cell of endosperm of seed of Nux vomica treated with sulphuric acid^
and gentian violet.

1902.] MACKENZIE — EQUATIONS OF HEAT PROPAGATION. 181

ON SOME EQUATIONS PERTAINING TO THE PROPA-
GATION OF HEAT IN AN INFINITE MEDIUM.

BY A. STANLEY MACKENZIE.

(Plates XXIII-XXVIII.)
( Read April ^, igo2. )

We may attack a problem in the theory of the conduction of
heat in two ways \ we may make use of a Fourier's series or inte-
gral, or, since the general differential equation is a partial linear
one, we may build up the required solution out of known solutions
for simpler cases. The former way is usually much the more
expeditious if the proper ''trick " can be hit upon, but the method
is a purely artificial one, throwing no light on the process
involved. The student or reader sees at once that this method pro-
duces the required result and that a limited number of very similar
problems might be treated in the same way, but he is apt to feel
instinctively at first that the mathematical tool he has employed is
one of which he has only a superficial knowledge and that will fail
him when he gets out of a certain set of problems ; he wonders what
a Fourier's integral means and why it has a special value in such
problems. The trouble here, as in many other departments of
physics, is that the physical interpretation of mathematical opera-
tions is usually avoided. There can be but one good reason for
this, since all must admit the desirability of such interpretations,
that it is at times exceedingly difficult, if not impossible, to give
the inherent physical meaning of a mathematical operation. Much
more, however, might be done than is done, and there is perhaps
no branch of mathematical physics more suited to the purpose of
introducing to those just beginning such studies the meanings and
the limitations of mathematical operations than heat conduction.
The second method of treating heat conduction problems, by
building up solutions from known solutions for other cases, is full of
suggestiveness, and brings into view the meaning of many of the
mathematical processes employed in any treatment of the conduc-
tion of heat, and the relationships of the equations involved. An
attempt is made in the following pages to point out the necessity
for effort along the lines indicated above, and among other things
to give careful drawings of some of the more important curves of
temperature and current.

182 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. [April 4,

In any heat conduction problem we have ordinarily three sets of
equations, the general differential equation, the initial conditions,
and the surface conditions. For the general purposes of this
paper by taking the medium infinite we can get rid of the surface
conditions without limiting the generality of the methods. Suppose
we wish to study the case of a body of any shape or size maintained
at any temperature in an infinite homogeneous medium of the same
material as the body itself but initially at a uniform low tempera-
ture (which for convenience we take as the zero of temperature),
or of the same body at a given initial temperature put into the
medium and left to cool, we could find their solutions by an
ordinary summation if we knew those for the corresponding prob-
lems in the case of an infinitesimally small particle. We might
begin by assuming as Kelvin does {Math, and Fhys. Papers, Vol.
ii, p. 44), the solution for the case of a quantity of heat, Q, sud-
denly generated at a point r = 0 at time / = 0 ; but it will be
better to see if it can be derived.

We have here to deal with the case of a symmetrical distribution
of temperature about a point. The form of the general differential
equation for this case is

l^^Z=z^ 5F 52^ (I)

k dl r dr dr^

where k = ^r^, J^ being the specific conductivity, C the specific

heat, and D the density of the medium. This equation can be
put in the more symmetrical form

1 ^AZ^ — ^'(^'^'') (2)

This is of exactly the same form as that for the case of the
*' linear flow of heat " of Fourier, that is, of flow in one dimen-
sion only, namely,

J_ 5 F _ d'^V (3)

The distribution of Vr with reference to r for the case of sym-
metry about a point is the same as the distribution of V with
reference to x for the case of symmetry about an infinite plane
perpendicular to the axis of x. This fact will be of assistance in
obtaining and translating results. The ordinary way of treating

1902.] MACKENZIE — EQUATIONS OF HEAT PROPAGATION. 183

any problem of spherical symmetry is to get the simplest kind of a
solution of (1) or (2) and build up from that solution to the
required one. There is of course an infinite number of solutions
of these equations and a great many simple ones, but we can at

once find one by trying Vr =^ e . This gives /5 = ko\ and

ar kd-t —ka-t

hence Vr =^ e e . Changing a to ia we get Vr = e
(cos ar + / sin ar), and so a solution is

Vr=^e cos ar, (4)

where a is any constant. This equation represents a periodic dis-
tribution of Vr along a radius vector dying out with the time ; lor
the case of the infinite plane this would be actually the curve of
distribution of temperature along x. It is seen that the values of
Fin (4) possess maxima and minima; the temperatures are zero at

distances given by ^ = (2/z + 1) -^ at all times. There is a hot

central sphere of radius ^-, surrounded by alternate hot and cold
shells of common thickness — , the maximum numerical tempera-
ture in each falling as we go away from the centre. Calling the
thickness of the shells d, we have a = ^ j so that the constant a is

inversely proportional to the thickness of the shells and deter-
mines it. The central point begins by being, and remains,
infinitely hot ; the hot and cold layers conduct heat to each other
and gradually die down in temperature. At a great distance from
the origin we should have practically the case of a medium made
up of alternate hot and cold infinite plates of the same numerical
temperature and the same thickness left to cool ; and such a prob-
lem could be treated from a consideration of (4).

This case is far from the problem we started out to discuss. We
can, however, get new solutions from the simple one above, and
the common method is now to say that the following is a solution
of (2),

e cos ar da, (5)

0

and then translate this equation as we have just translated (4) ; but

•184 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. [April 4,

instead of doing so we ought rather to be able to say that this opera-
tion means such and such and foretell the distribution of tempera-
ture it will give. This illustrates what was meant above when
saying that we ought if possible to give the physical interpretations
of mathematical processes. What is the meaning of the operation
involved in (5)? Perhaps some light can be had on it from the
following consideration : We are to take a series of distributions
of temperature like that given by (4) and described above, where
the constant a (determining the thickness of the shells) has the

successive values, 0, da, ^da, a, and superpose them on the

medium after first reducing every temperature by multiplying it by
da. We are then to take da indefinitely smaller and smaller, and
finally to make a indefinitely greater and greater. We have thus
the difficulty of a double limit entering, and if we wish to seek the
initial condition it becomes a triple limit. This is sufficient to
prevent any rash prediction in this problem as to the exact nature
of the solution to be obtained ; and this case serves as an excellent
example of the difficulties to be overcome in any such efforts at
physical interpretation. Before the limit is reached the state of
temperatures is given by

p —kt{daY- —Akt{da)"- -I

Vr = d(j\ \ -\- e cos rda + e cos 'irda + etc.

The limiting value of this series, which is equation (5), is not very
evident without considerable study, but on account of the dying-
out factor in each term the series is convergent, and the more
rapidly convergent the greater the value of t, and its value could
be found for any given / and da. Another way of finding this
value at any time and distance required is to take an axis along

which a's are measured and draw the logarithmic curve e and

the curve cos ra, then form the curve whose ordinate at each
point is the product of the ordinates of these two curves at the
point, and the area between this new curve and the axis gives the
numerical value of Vr. Since this area is formed of pieces alter-
nately above and below the axis of a and of decreasing numerical
value, we see that Vr is always of the same sign and that, for any
finite value of ;-, it begins by increasing in value and finally falls
off to zero, and by inference that it is zero at time / = 0 ; but
that at the origin it has initially a value greater than zero. The

1902.] MACKENZIE — EQUATIONS OF HEAT PEOPAGATION. 185

operation (5) therefore promises at least another simple solution
and one much nearer the desired one. Noting that

^'■^—ko.'^t /»^ —koT-t

J-r^—kol-t /»^ —koT-t

e cos ar ^a = 2 I e COS ar da, and that
0

+ »3 +00

J—kaH /» —{kto:-—ira)

e sin ar da = 0, we get \ e da —

— 00 — cc

r'^ ^ ^' f -r ir \1

e

and (5) becomes

Vr = ^ e~^ , (6)

Vkt

where A is an arbitrary constant. This equation says that Vr is
initially indeterminate (evidently infinite, from physical considera-
tions) at the centre and zero elsewhere ; as time goes on the value
of Vr falls off indefinitely at the centre, rises to a maximum at all
other points and then falls off indefinitely also. Now these are
exactly the conditions we want for V itself for the case of an
infinitely hot point cooling in an infinite medium initially of zero
temperature. If we had been studying (3) we would have found
the same equation as (6), with x for r and Ffor Vr, for an infinitely
hot plane cooling in a medium initially zero. The form of the
curves for Fr given by (6) is exhibited on Plates XXIII and XXIV ;
with values of r as abscissae curves A^ to A^ are for values of the

time ^, i , ^^ and ^^ respectively j with values of Ut as abscissae

curves B"" to B'^ are for values of the distance 0, i, \, | and 1
respectively.

We have taken the form (2) of the differential equation in
preference to (1) on account of its symmetry and because we are
solving the case of the infinite plane at the same time; but it
possesses another important advantage. Since either form of the
equation is a linear partial one we can add any number of solutions
for a new solution ; the question arises, therefore, whether F being

a solution -|^and / Vdr are solutions, and what are their physical

186 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. [April 4,
meanings. Without thinking of the special form of the differential

equation, we can find the meaning of -_~ as follows : Let a solution,

F, be/(r,t); then another, F^, is -^/(r,/), where Ar is a small

constant; and another, F^, is — ^/(^O- Superpose on the medium

these two states of temperature, Fi and F^, after first displacing F^
bodily to the positive side of the origin by an amount Ar. When
Ar is indefinitely decreased the limiting state of temperature is that

represented by -^, or — 1— . That is, -^ represents a heating due

to a kind of doublet. We must next find out whether such a state

of temperature as that represented by-^ is a solution of (1). We

d V

saw that -y- was a limiting case, and hence it is not a solution in

the limit (except by some unusual accident) unless it is so just
before the limit is reached. While Ar is still finite, but as small as
we please, the superposed heatings do not satisfy the same differen-
tial equation; for F^ satisfies the equation -j^ ^^^ = j ^^p-^

\ , while F, satisfies the equation -r-^^ — -^ = — ^^^ — - -{-

-, and on account of the variable coefficient these are not

5/-2

d V
the same equation. Hence -j- is not a solution of (1), and is only a

solution of an equation in Fwhen that equation has constant coeffi-
cients, that is, coefficients not containing r. Equation (2) is of that

kind, and hence knowing a solution of it, Fr, we can say that -~

is also a solution. Call this new solution F^r, then F^ is a solution

of (1). Since ^^^= kH- r^-, and since \ — ^ is a solution of

^ ^ dr dt ^ r dr

(1), we have h .- a solution of (1) ; this is what we have just

d V
called F^. Now F satisfies (1), but we have just seen that -^

V
does not, and it can easily be seen that doesnot in general; so we

have the interesting fact that the solution F^ is the sum of two func-
tions of K (itself a solution) neither of which is a solution. We
can at least give a physical interpretation to the method of finding

1902.] MACKEXZIE— EQUATIONS OF HEAT PROPAGATION. 187

a solution of (1) represented by the mathematical operation V--'

where Fr is a solution of (2) and F itself a solution of (1); we
have but to add to the doublet of this V as defined above a heating
at each point ;-, which is F divided by the value of r at the
point.

The meaning of ( F dr, where F is a solution of the differential
equation, is now plain. It simply means finding a new function of
r and /, F^, whose doublet is the solution F. That is, -^ = F, and

or '

/^i = (F dr. This is subject to the same limitations as before,

that the differential equation for Fmust have its coefficients inde-
pendent of r, in order that F^ may be a solution of the equation.

Similarly for equation (2); we have a solution, Fr, to find the
meaning of the new solution, FV, which we get on performing

the integration J v r dr. Since -^^ — ' = Fr, or — — - =z F, we

are but finding the distribution of temperature, F^, whose doublet

added to the heating — gives the distribution of temperature,

F, which we started with.

We thus see that (2) has the great advantage over (1) that when

we find a solution of the former we can differentiate and integrate

it with regard to r for new solutions, but we cannot do so with the

latter.

dV r

The meaning of -j- and of J Fdt as solutions of (1) are of the

same general nature as the similar expressions with r, and are quite

evident ; we now superpose one heating, — /(^/) on another,

— - /ir,/"), after a small interval of time J/, which we make
smaller and smaller indefinitely. We might call this a //>;/<? doublet
and the former a space doublet. Both ^ and J Fdt are solutions

of (1) because the coefficients do not contain t. The same remarks
apply to (2) as regards Fr, with the explanations of the former
paragraph added. Here equation (2) possesses no advantage
over (1).

The meaning of a Fourier's integral may now be given. A
solution of (3) for the flow of heat in one dimension is evidently

188 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. [April 4,

—ka'^t

V= e cos /5(a — x), where a and /5 are arbitrary constants,

—ka^t —ka-t

for it is made up of V = Ae cos ax and F= Be sin ax^

both of which are solutions of (3) as shown above. This equation
denotes a distribution of temperatures which has maxima and
minima values, the latter being at certain fixed points given by the

equation x = a — (2n -\- 1) -^. In general it is very similar to

the distribution represented by (4) already studied. V^ = V (f{a)
is also a solution, where the temperatures are as before except that
they are increased by multiplying every one by ^(a), an arbitrary
constant function of a. Another solution is got, as described
before, by superposing all the heatings formed on reducing the
temperatures in V^ by multiplying each by the very small quantity
da, and giving a all values from — oo to + °^> ^^^ then taking the
limiting case where da tends to zero. Call this new solution V^ ;

then v^ — J ^ cos /5(a — x)<p{a)da. Repeat this last operation

with regard to /5; that is, take the distribution of temperatures
represented by V^ and reduce the numerical value of each by
multiplying by ^/5, then superpose all such heatings formed by
giving /5 every value from 0 to oo, and finally take the limiting case
where d^ tends to zero. Call this new solution V^ ; then

//* — Ka-t

di3 J e cos iS(a — x')(p{fi)da. Still another solution

0 — cc

is got by reducing every temperature in ^ in the ratio of - to 1.

cc 4- cc

if f ~^'^"^

Call this solution F^; then ^ = -J d^ J e cos /5(a — x)<p{a)da;

^ 0 — oo

it has the special importance and peculiarity, as was first shown by
Fourier, that at time zero the distribution of temperature it
represents is the same function of x, (fi^x), that we took originally
of a. Similarly every Fourier integral may be interpreted.

Returning now to equation (6) and the curves drawn for it, we
can find new solutions by addition ; at each point r let us add the
temperature for that point and all other points farther from the
centre, even to infinity, but first reduced in absolute value by
multiplying each by the small quantity dr, which we make ulti-
mately tend to zero. We have but to add on Plate XXIV for any

1902.1 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. ISD

abscissa (time) the ordinates of all possible curves such as B^^ ^%
etc., below any given one, after reducing them as described. For
/ = 0 and r = 0 we would get (co -|- 0 -(- 0 + etc.) dr, which as
^r diminishes indefinitely gives us some finite value; for other
values of r we would get (0 -f- 0 -f- 0 -f etc.) dr, which is zero.
From the way the curves tend to become parallel it is suggested,
and by trial we find, that for r = 0 and any finite value of the
time not zero the sum of all the ordinates would be constant. We
have then the promise of another simple solution, and can foretell
its type somewhat, of the form

Vr

CA^e ^''dr=^B C e d?, (7)

2}/ kt

where B is an arbitrary constant. On studying this equation we
find that Vr at the origin has initially the value B, and maintains
that value ; at all other points it is initially zero and rises asym-
ptotically with time toward the value B. F itself would be' always
infinite at the origin and initially zero elsewhere. For the case
of linear flow equation (7) represents an infinite plane kept at
temperature B in an infinite medium initially zero in temperature.
We can get the solution for an infinitely hot point put into an
infinite medium initially zero and left to cool as follows : At time
zero apply to the medium the state of temperatures represented by
(7) with every temperature increased by multiplying it by the large

quantity—-; after time J/ apply also the state of temperatures

represented by (7) with sign changed and increased numerically as
before ] finally make At tend to zero. We have seen above that
this is equivalent to performing the mathematical operation of
differentiation of (7) with regard to /, that is, taking the time
doublet of Vr. The reason that this solution is the one required is
that the superposition of the two heatings gives Vr a large value
at the origin at first and everywhere else a zero value, and then
instantaneously makes Vr zero at the origin ; that is, at the origin
Fis initially infinite in temperature and then falls off indefinitely,
while all other points begin at zero and rise gradually. These were
the conditions we wanted. Hence we have the solution

190 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. [April 4,

-^[^-/'"*]-S

Vr

2\/ kt

\kt

e ,..(8)

and

E 4^^
V= ^e , (9)

;-

where £ is an arbitrary constant.

Further light can be thrown on this problem by arriving at
equation (8) by other methods. Remembering that equation (6)
gave Fr initially infinite at the centre and zero elsewhere, and
falling in value at the centre and gradually rising to a maximum
elsewhere, we see that by taking the space doublet of this Fr we
get Fr 2it the origin first infinite and then zero; that is, Fat the
origin is at first infinite and then gradually falls off, and is initially
zero elsewhere and rises with time. These are the conditions
required. Hence the solution is

d r A i^i-] Er ^kt

^' ~ ^r \_VJt ' J

{kty

Or we can look at it in this way : We saw that Fr in (6) had
exactly the set of values we want F to have in the problem pro-
posed, and the form of the right-hand member of (6), containing

r-
~ \ki

as it does r in the factor e only, suggests at once that we can

get the desired value of Fby a simple differentiation with regard
to r. This is what we have just done with a good physical reason
for the operation.

Or another method. We saw that equation (G) for the case of
flow in one direction only was that of an infinitely hot plane cool-
ing in an infinite medium initially zero in temperature, and to get
the solution for the similar problem in three dimensions we have
but to multiply that solution by two similar ones with y and z
substituted for x. This gives

'^kt Akt \kt ^ Akt

e e ^ --—5 e (11)

{_ktf {.ktY-

1902.] MACKENZIE — EQUATIONS OF HEAT PROPAGATION. 191
The rate of cooling is given by the equation

Each point of the mass not the centre begins by being zero in

temperature, then rises to a maximum after a time / ^= -^ , and

after this falls off indefinitely toward zero. The forms of the
curves given by (9) are exhibited on Plates XXV and XXVI. With
values of r as abscissae curves I^ to IV' are for values of the time

jTw, -Q^, -j^, and —J respectively ; with values of 4/'/ as abscissae
curves I' to 5^ are for values of the distance 0, i, 1, |- and 2
respectively.

The meaning of the constant £ is determined by finding the
amount of heat supplied initially to the hot point. We have

(2 = r r r CC> Vdx dy dz = ^^^^^ C e ^^\-\ir=-^ CDfJ. . (12)

If we take as our unit of heat that required to raise the niass in a
unit of volume of the substance 1°, the total quantity of heat, (t,
in these units is

ff = 8^-' (13)

We could also get the total heat by taking the integral

C— K ^ i-r'dt. We get from (12) and (13) our equation (11)

0

in the form

Q 4^< ^ ^kt

y — . ^ ^ ^ £ C14)

^CDi^-Kktf. 8(-'^0^

(See Kelvin's Papers, Vol. II, p. 44.)

We cannot build up by summation the solution for the case of a
body of finite dimensions from the above solution for a mathemati-
cal point. We wish to pass to a case which has a physical signifi-
cance, namely, a finitely hot particle left to cool in an infinite

192 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. [April 4,

medium of temperature initially zero. We can get a close
approximation to this problem by putting the same quantity of
heat, ffi into a particle of volume Av which we put into the math-
ematical point, and assuming that the state of temperature produced
in the surrounding medium is the same as that due to the infinitely
hot point and is given accordingly by (14), This equation will
represent the real state the better the longer the time which has
elapsed, in accordance with the fact emphasized by Fourier that
the initial heating is of less and less importance as the time is pro-
longed. The closeness of the approximation for any given time
and distance will be brought out later.

Let the quantity of heat supplied raise the volume Av to the
temperature V^; then Q = CDV.Av, or a r= F,Jv; and (14)
becomes

r-

KAv ^^
V= '—J e (15)

If the volume Av is in the form of a sphere of radius R, (15)
becomes

F,/?3

\kt

V=^^^^-^ e , (16)

6i/7r (i/)i

and it is really for this form of the equation, with R taken as the
unit of length, that the curves referred to on Plates XXV and XXVI
were drawn. They are, as said, approximations only to the true
curves. The latter may be found by the aid of a Fourier's integral.
We know that the solution of (2) subject to the condition V=^f{r)
when / = 0 is

Vr = ^ [J (^ + 21/17 r)/(/' -f Wk^r)

e dy —

•i}/'kt

C» 2—1

J (-r + 2i/I/r)A-^+ V^r)^ \/rJ--(i^)

2\^ kt

Giving f{f) the value V^ from r = 0 to r =^ R, and the value 0
from r =^ R io r =^ ^ y {11) takes the form

1902.] MACKENZIE— EQUATIONS OF HEAT PROPAGATION. 193

r—R

2\/ki

This then is the exact equation for a sphere of any size of initial
temperature Fg put into an infinite medium of the same material as
the sphere of initial temperature zero and left to cool there. The
forms of the curves given by this equation are exhibited on Plates
XXV and XXVI, along with those of the approximate equation
(16). Curves I to IV correspond to V to IV^ and curves 1 to 5
correspond to 1* to 5\

We can get an approximate form from equation (18) by expand-
ing it in": erms of J? ; we find

Tkir- 4^-6

6i/^ W^ L 40 i^ J

(19)

The first term of this is the same as equation (16), found otherwise.
Equation (19) gives us a second approximation, and the second
term within the bracket will enable us to determine the closeness of
(16) as an approximation. In a similar problem, Fourier (Free.
man's translation, p. 380) gives a limit to the time when the
approximation may be used, but he does not give any means of
telling how great the error is in general, and it was for the purpose
of bringing this out distinctly that equation (19) and the curves on
Plates XXV and XXVI were produced. From Plate XXV we see that
the approximate curves are at first steeper and afterward flatter than
the exact curves ; they make the temperatures too high for points
nearer the origin than a certain distance, and too low for points
farther away. Indeed curves I and I^ are very little alike for any value
of/'. As the value of the time for which the curve is drawn is taken
greater and greater the curves approach each other more and more
nearly, even for points less distant than unity (which are inside the
little sphere), for which we might have expected little agreement.
This makes evident the fact to which Fourier calls attention at the
place just cited ; one is very apt to assume that the curves would
approach each other more and more as r is taken greater and
greater, no matter what the value of /; bat just the reverse is true,

PROG. AMER. PHILOS. SOC. XLI. 169. M. PRINTED JULY 28, 1902.

194 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. [April 4,

the curves approach each other more and more for greater and
greater values of the time, no matter what the distance. This is
seen more distinctly from an examination of Plate XXVI. There it
will be seen also that the approximate curves are slower in reaching
their maximum values, as well as that they have different maxima.
For distances less than unity the approximate curves start at go ,
while the exact curves start at F= F^; for the distance unity the

exact curve starts abruptly at-^, while the approximate curve starts

at 0 then gradually rises and has a maximum value less than

-^. For distances greater than unity both curves start at the origin.

From an inspection of the second term of (19) we can foretell
the approximate accuracy of (16). Taking 7? as the unit of length,

if /('/< 15 the error in the value of — ^^ will be everywhere greater

than 1% except in the immediate neighborhood o{r=i/^kf, at which
point the error is practically zero. For instance, for k^ = ^ (curves
IV and IV^) the approximate curve is 33% too high at r = 0, 22 fc
■21 r = 1, correct at about 1.8, and 38% too low at 3. If kt =^ 15,
the error is not more than 1% from r = 0 to r = 13.4. If ^/ ==: 25
the error is not more than 1% from r = 0 to r = ?0. In general,
for any value of kt the error is not more than \% from r = 0 to

r = yfSkt + ^{kt)\ and from r = 0 to r= |/6^ the error

15 *

decreases gradually from -j7% to zero, and after that increases

again. If we want results accurate to .01%, kt must be at least
1500, and in general for any value of ki greater than this the error
is not more than .01% from r =-- 0 to r = y(Skt -J- Yiu ('^0^ ^^^
from r = 0 to r = |/6/C'/ the error decreases gradually from -rrfo
to zero, and after that increases again.

From equation (15) we can build up by summation the equation
for the case of a body of any shape or size initially at Vq cooling in
an infinite medium initially zero. In order to bring out a very
interesting difference between summation and integration we shall
apply equation (15) to the case of an infinite space, one-half of
which is initially at V^ and the other half at zero, the two parts
being separated by an infinite plane surface. We shall first have to
find the solution for a plane lamina. Take the central plane of the
lamina as the plane of ^'2,''and the origin where a perpendicular

1902.] MACKENZIE — EQUATIONS OF HEAT PROPAGATION. 195

from the point P, at which we want to know the temperature,
meets this plane. Call the length of this perpendicular x. Break up
the lamina into concentric rings of radius p about this origin, and
let the distance of every point in one of such rings from the point
P\>^ r and the thickness of the lamina Ax\ then we have

8(

_ -V- + p- f -_

— ^, I e ^r.p. Ax. dp = — 5 — . g /30)

From the symmetry of the problem this is evidently a case of linear

flow, and the solution must satisfy equation (3). Knowing this

solution (we can get it otherwise), the solution for three dimensions

given in (15) can be deduced ; we have but to multiply the value of

/>'
-T7 for the case of one dimension by two similar expressions with

y and z respectively substituted for x.

The corresponding electrical problem is that of an infinite cable
with no lateral loss by leakage touched for an instant to a condenser
of potential V^. If there is lateral leakage equation (20) is still
the solution of the electrical problem; Vis then not the potential,
but the potential can be derived easily from it, as is well known.

If Q or (7, according to the unit of heat used, is the amount of
heat required to raise the mass of a section of the plate of unit
area by V^ degrees, then Q = CDVqJx, or c- = V^Ax, and equa-
tion (20) becomes

Q -Ul ^ --Ikt

V = —-^ ^ e = ^ e (21)

Of course this equation is of only the same grade of approximation
as (15). It will be the more nearly exact the smaller Ax and, since
the product of V^ and Ax measures the heat in a section of unit
area and is to remain constant, the greater V^. In the limit we
should have the solution for an infinitely hot plane. The form of
this solution we have already found ; it is from (G) and the remarks
following it

A

ikt

,-=.e (22)

Calling Q the total heat associated initially with a unit of area of

196 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. [April 4,
the plate, we find (2 = 2 ( CDVdx = lACBx/r.; and this value

0

of A reduces (22) to the form (21). Hence the general form of
equation (21), which is approximate for a plate of actual thickness
Ax, is exact for the infinitely hot plane. We shall revert to this
important fact later.

If we want the exact equation for the plate of thickness Ax we
can get it by the use of a Fourier integral. Making the obvious
changes in (17) to suit it to the case of linear flow, and giving/(:x:)

the value ^o from x =^ -^to^==-^ and the value 0 for all

other values of x, we find

^=77f e dy (23)

2i/ kt

Putting this in an approximate form, we have

2{'Kkt)

V,^x ^ ^^^ p^ , kt - (Ax]

(24)

the first term of which is equation (20). The forms of the curves for
(20) are exhibited on Plates XXIII and XXIV. With values of x as

abscissae curves A'^ to A* are for values of the time -^, -^, -jr and

-^ respectively ; with values of 4/CV as abscissa curves B^ to B'^ are

for values of the distance 0, ^, ^, | and 1 respectively. The second
term of (24) enables us to tell approximately the degree of closeness
of (20) to the exact equation (23). Taking Jx as the unit of

25

length, if kt<.--^ the error will be everywhere greater than 1 %

except in the neighborhood of x --= y^-lkt where it is practically

25
zero. If kt = -^^ the error is not more than 1% from :x: = 0 to

X =z 2.9, being Ifo too high at x = 0, zero at jc = 2, and ifo too
low Sit X ^= 2.9. If /&/ = 25 the error is ^3-% too high at :v = 0,
zero at 7, and 1% too low at 26. This is then a nearer approxima-
tion than the one discussed for the case of a hot particle, as was to

1902.] MACKENZIE — EQUATIONS OF HEAT PROPAGATION. 197

be expected. In general, for any value of kt the error is not more

than \% from ^ = 0 to x = y^Uf -f- ^{kty, and for any value of kt
greater than ^-^ the error is not more than M% from jc = 0 to

2^V -f 23^(^0' y from :\: = 0 to a,- = V 2k/ the error decreases

25
gradually from t^^^ to zero, and after that increases again.

The correspondingly approximate equation for the current or
flow of heat in this case is

J = — A ^= — ^= ^e = -^^ -3 e ....(25)

The forms of these curves are given on Plates XXVII and XXVIII.
With values of a: as abscissae curves C^ and C/, C^ and Q^, and Q^

are for values of the time zrwr, -^ and ^-t- respectively ; with values

of 4^/ as abscissae curves Z>^ and Z?/, D " and Z>i^, and Z>i' are for
value of the distance i, J and 1 respectively.

The exact equation for the flow, found from (23), is

r^rz I— 4:kt Akt — i

/^-ALlF, _, n (26)

2(:7/^/)^L J

the curves for which have not been drawn.

By adding up the effects of an infinite number of such plates we
can get the temperature due to one-half of space initially at a uniform
temperature V^ and the other half at zero temperature. Take the
point F, at which the temperature is desired, in the cold half and
at a distance x from the surface of separation, and take the origin
in that surface at the foot of the perpendicular from F. Let one
of the plates making up the other half of the medium be distant I
from the origin. Then the x of equation (20) becomes x -\- ^,
and Ax becomes J| ; hence the temperature at F due to a series of
such plates extending from c =: 0 to 1^ =: oc , as found by inte-
gration, is

00 TT-, -r, — ft2

V /• ■^^^ V

2V kt

198 MACKENZIE — EQUATIONS OF HEAT PROPAGATION. [April 4,

*['-7^J-''"'*]

(27)

We could arrive at the solution for this case by using Fourier's
integrals, as we did for equation (23), giving /(^) the value F„
from X = — ooto:!t:==0 and the value zero from ;r = 0 to ^ = oo.
We get at once equation (2T) again.

This latter method gives the exact solution for the problem and
yet it gives the same result as the former method, from which one
might expect naturally enough an approximate solution, since we
get it by integrating solutions that were approximate. This is the
point to which attention was called in applying our results to this
case ; we have the integration of approximate solutions an exact
solution. The first explanation offered of this unexpected result is
apt to be that the approximation used is the more exact as the dis-
tance .T -(- ^ is the greater ; but we have seen earlier that just the
contrary is true and that at great distances (20) ceases to be
properly called a solution unless the time is taken very great. The
real explanation is simply that the operations of summation and
integration are not always the same, and this is a case in point.
Nothing is commoner in applying mathematics to physics than to
use mathematical processes with laxity and to test the legitimacy of
the application by the results. It is so uncommon to have a sum-
mation made improperly by integration that we lose sight of the
mathematical fact that the operations are not equivalent. We take
similarly the first two terms of a Taylor's series expansion as a
sufficiently close approximation in almost any piece of analysis,
without questioning whether the function under consideration can
be so expanded and without reference to the value of the terms
disregarded ; we take differential coefficients without asking
whether they can have a meaning, etc. The good excuse offered
is that the chances are overwhelmingly in our favor, and that if we
have made a mistake we shall quickly find it out from the results.
Had we actually made a summation in the above problem we should
have got an approximate result, but by integrating we get the limit
toward which the summation tends as ^^ tends towards zero, and it
happens in this case that this is the exact solution. In finding an
area we take a series of strips of area of y^x and however infinites-
imally small dx is, so long as it is something and not zero, the sum

iy02.] MACKE.VZIE— EQUATIONS OF HEAT PKOPAGATIOX. 199

of such Strips is not the exact area required ; f yt/x is the limit

toward which the sum tends as (/x tends to zero, and we know from
the familiar example of Fourier's series how the value can change
actually in the limit. It happens in the present case that as ^c is
made smaller and smaller, and V^ correspondingly greater and

greater in order to keep cr constant, in the limit — - — ^ is the

exact solution for an infinite plane (see under (21) and (22)). So
in making the integration above, that is, in finding the limit of the
summation, we get necessarily an exact solution because in the limit
each term of the solution is exact. Had we approached the limit
in some other way than in keeping ^ constant we might have got
quite a different result.

The forms of the curves for (27) are shown on Plates XXVII
and XXVIII. Curves £\E^ and £^ are drawn with values of x as

abscissae for values of the time — ^, -jy- and - respectively ; curves

J^\ F"^ and F^ are drawn with values of ikt as abscissae for values
of the distance j, ^, and 1 respectively.

Since the current or flow is got from the temperature by a differ-
entiation with regard to x, and since equation (27) was got from
(20) by an integration with regard to x, it is evident that the
curves for the potential or temperature in (20) are the curves for
currer.t in the present problem.

I=-KZ = ^^.e (28)

dx

%{r:ktY

These curves are given on Plates XXIII and XXIV for points to
the right of the origin ; the form for points to the left is obvious,
since the curves are symmetrical about \.\-\q yz plane.

Physical Laboratory, Bryn Mawr College.
April J, I go 2.

200 SNYDER — A NEW METHOD OF TRANSITING STARS. [April 4,
A NEW METHOD OF TRANSITING STARS.

BV iMONROE B. SNYDER.
{Read April 4, 1902.)

The method of observing transits of stars, here to be described
in a preliminary and general manner, consists in driving the mi-
crometer screw and hence micrometer thread of a transit instru-
ment by means of an electric motor at the uniform speed pertain-
ing to any given declination, at the same time that the observer by
secondary adjustment secures and maintains accurate bisection of
the star, while given positions of the screw and hence thread are
automatically recorded on a chronograph.

It is now more than four years since the writer described the
method to his associates interested in astronomical observation.
In the autumn of 1899 this plan of electrically driving the transit
thread was also mentioned to Professors Wadsworth and Morley
and at some length discussed with the latter. Working drawings
of the special instrument which at present gives concrete expres-
sion to the method were completed in September, 1900. The
"electrical transiter," or more simply *Uransiter," as for brevity
the new device has been named, was mounted on the small me-
ridian circle of the Philadelphia Observatory in February, 190 1,
and there subjected to many tests and improvements since. The
demands on the writer's time have, however, not permitted that
singleness of devotion which the transiter and its interesting
method should receive, and it does not, therefore, seem desirable
any longer to withhold a preliminary communication on the
subject.

The fundamental idea of moving a transit micrometer wire by
means of clockwork synchronously with the star's motion was
proposed in 1865 by Braun.^ But to Repsold is due the persistent
pursuit of the idea that personal equation can be banished from
transit observations by mechanical methods. And although his
practical solutions of the problem have hardly proved adequate,
they have stimulated and permitted serious efforts on the part of
observers.

The first suggestion of Repsold, ^ made in 1888, was to mount the

1 Dr. Carl Braun, Das Fassagen-AIikrofneter^ Leipzig, 1865.

2 F. Repsold, «' Durchgangs-Instrument mit Uhrbewegung," Astron. Nach.y
2828.

1902.] SNYDER — A NEW METHOD OF TRANSITING STARS. 201

base of the transit instrument on a polar axis and within a limited
range drive the instrument to the diurnal motion by means of
clockwork, and in some undescribed manner keep the star bi-
sected so as to determine the meridian passage through electrical
signals automatically made. The plan admirably met the chief
difficulty of the varying rate of motion due to difference of decli-
nation, but was abandoned on account of the great mass to be
moved.

It has to the writer, however, seemed likely that by applying a
powerful electric motor of strictly constant speed, and by using a
second electric motor with regulable speed for driving one element
of a differential gear which engages the shaft driven by the main
motor, or by several other electrical devices not requiring men-
tion, an equatorially mounted transit instrument can be driven to
stellar bisection and readily kept so adjusted.

A second plan, " Neuer Vorschlag zur Vermeidung des person-
liclien Zeit-Fehlers bei Durchgangs-Beobachtungen," was proposed
by RepsoldMn 1889 and tested by Becker'^ in 1891 with moder-
erately favorable result A new form of micrometer, made for the
Madison Observatory by Repsold,^ was described in 1896, and with
the general plan of its construction the writer became acquainted
in the autumn of 1897. This specially designed and rather com-
plicated micrometer requires that s^ar bisection shall be main-
tained by twirling the micrometer shaft alternately v/ith each hand
of the observer. While this twirling is proceeding the ten elec-
trical contacts of a drum mounted on the micrometer screw deter-
mine as many records on the chronograph. This Repsold method,
while not lacking in ingenuity, seemed to the writer to labor under
the following defects : An alternating twirling motion of the mi-
crometer, even when communicated with the greatCbt adroitness,
is not approximately a uniform motion. The observer is attached
to the instrument by both hands, and is incessantly committed to
the most painful attention. Good results could hardly be secured
without the most prolonged and painstaking practice. Through
his special habit of twirling each observer must have a new form of

^F. Repsold, Astron. A^ach., 2940, 1889, September.

2 Prof. E. Becker, " Ueber einige Versuche von Durchgangs Beobachtungen
nach dem neuen Repsold'schen Verfahren," Astron. Nach., 3036, 1891, Marz.

3"Neue ^Mikrometer von A. Repsold u. Sohne," Astron. Nach., 2>Z1h 1896,
Juli.:

202 iSNYDER — A NEW METHOD OF TRANSITING STARS. [April 4,

personal equation. Even the averaging secured by the great
number of electrical contacts does not certainly eliminate the pe-
culiarities of a given habit of twirling. At any rate the great
number of signals to be read from the chronographic sheet consti-
tutes a very serious infliction on time and patience. Finally, the
Repsold method does not, during any given star transit, offer a
ready and direct means of comparison with the ordinary methods
of observation.

The difficulties experienced in acquiring reliable observing
habits with the Repsold transit micrometer are evident from the
reports of Becker,^ Kowalski,* and Flint. ^ The latter is, it seems,
the only American observer who has tried the Repsold device to any
extent, and he says that '^ after considerable practice" he obtained
the same probable error by the method for *' a signal under good
conditions as for a single thread when observing with a fixed
reticule and chronograph."

And yet it is not surprising that among European obser\»ers
engaged in longicude work, the Repsold method should after pro-
longed discipline yield excellent results. Albrecht,^ in an extended
paper on its application to longitude work, points with enthusiasm
to the superior results obtained. He considers the former indiffer-
ent results to be due to lack of practice and insists that the highest
effectiveness, by this method, is attained only after a long season
of active experience. '' Man erlangt das Maximum der Leis-
tungsfahigkeit doch auch bei dieser Method e erst nach langer
Uebungszeit."

These experiences of practiced observers, while pointing to the
value of the plan of micrometer thread motion in eliminating per-
sonal equation and its variations, confirm the anticipations of the
writer as to the inherent defects of the Repsold method. It is
therefore interesting to note that experiments for relieving some of
the imperfections of the method have been going on at the Konis-

1 Loc cit.

2 Ueber das neue selbstregistrirende Mikrometer von Repsold, Petersburg,
1897.

3 Albert S. Flint," The Repsold Micrometer of the Washburn Observatory,"
Astron. Jour., No. 470, 1899, September.

* Prof. Th. Albrecht, « Die Beobachtungsmethode mittelst des Repsold'schen
Registrirmikrometers in ihrer Anwendung auf Langenbestimmungen," Astron.
Nach., 3699, 1 90 1, Marz.

1902.] SXYDER— A NEW METHOD OF TRANcSlTING STARS. 203

berg Observatory, where its Director, H. Struve,^ has successfully-
applied clockwork directly to the Repsold micrometer, and thus
unquestionably improved its usefulness. With this work the writer
became acquainted only after his own plan had been consummated
and the resulting instrument constructed and mounted for use. Dr.
Cohn,^ of the same Observatory, has recently published an extended
investigation which shows marked advances in accuracy over the
usual methods of observing. Struve's method has, however, thus
far involved the unsymmetrical placing of the weight of the appa-
ratus and, while itself possessing serious mechanical limitations,
does not avoid certain peculiarities and limitations of the Repsold
micrometer. The necessity therefore still exists for a method that
shall be flexible in adaptation and use, and not impose unreason-
able conditions on the observer.

The conditions to be attained in an effective method were early
formulated by the writer substantially as follows :

The ordinary micrometer of a transit instrument shall be used,
and its movable wire driven electrically at approximately uniform
speed. The rate of driving shall, as required, vary with the decli-
nation. The direction of motion shall be instantly reversible.
The wire shall be promptly readily started on its course when bi-
section of the star occurs. While in motion the wire shall be
easily regulable for bisection of the star. The automatic chrono-
graphic record shall be made at whole turns or at fractions of a
turn of the screw as desired.

In practically studying the electrical method of determining and
controlling the motion of the thread of a transit micrometer, it has
been found that there are three principal plans of adaptation
available :

I. A small electric motor may be placed on or near the head of
the transit instrument, with its axis parallel to that of the instru-
ment. The varying rate of motion required for change in decli-
nation may then be secured by regulating the field of the motor
and, if necessary, also that of a small dynamo supplying the cur-
rent. The main difficulties in this plan are, the wide range of

1 H, Struve, " Ueber die Verbindung eines Uhrwerks mit dem unperson-
lichen Mikrometer von Repsold," Astron. Nach., 3719, 1901, Marz.

2 Dr. Fritz Cohn, ** Ergebnisse von Eeobachtungen am Repsold'schen Regis-
trirmikrometer bei Anwendung eines Uhrwerks," y/^/rc>«. A'<3r/^, 3766-67, 1901,
November.

20-i SNYDER — A NEW METHOD OF TR.iXSITING STARS. [April 4,

speed regulation required and the interference due to inertia a^
starting.

II. Equatorial speed that is absolutely constant but slightly
regulable may be given the motor, similarly placed, and the differ'
ing rate of motion proper to each declination determined by me-
chanical gearing, consisting principally of two friction disks placed
at right angles to each other, or by some other mechanical equiva-
lent. Both of these plans require special care in the constr^iction
and mounting of the motor, so as to obviate the communication o^
injurious vibration to the transit instrument.

III. It may in some instances be desirable to place the electric
motor on a separate support near the base of the instrument, and
then by means of a light steel shaft entering the axis of the transit
finally communicate the required motion to the micrometer screw.
All the motions and controls peculiar to either of the other plans
may be secured to this form of transiter, excepting that the micro-
meter cannot at all be driven during the time necessary for reversal
of the transit instrument. This limitation would in some instances
be rather annoying, if not destructive of facilities the method
should furnish.

It is also quite feasible to place certain elements of the transiter
on a separate support and communicate the motion to a small slow-
speed alternating current-motor placed on the head of the transit
instrument and connected with the micrometer, and so obviate
practically all the mechanical and electrical difficulties. Experi-
ments in this direction are in progress.

In the attempt made to actualize the electrical method of driving
the transit micrometer contending obstacles and facilities led, for a
first trial, to the selection of the second plan mentioned, namely,
that of gearing from a small motor of fixed speed placed near the
head of the instrument. In reaching this conclusion the writer was
greatly assisted by his friend and former student, Dr. H. G. Geis-
singer, who, immediately the method of the electrical transiter and
the conditions it imposed had been described, became enamored of
the delicate mechanical and electrical problem. Detail drawings of
a transiter of this type were prepared under the writer's direction by
Dr. Geissinger, and he has introduced several ingenious devices
which admirably meet the conditions set. The special aim of the
writer is to construct a form of transiter that may, without serious
modification, be attached to any transit instrument. It should not
involve a special form of micrometer nor in any way vitiate the in-

1002.] SNYDER — A NEW METHOD OF TRANSITING STARS. 205

strument for its usual work however accurate. Excepting in the
unnecessary weight of the parts and general coarseness of the
mechanism, the transiter as now constructed fairly meets all the de-
mands originally set and besides introduces some new conven-
iences. Although it is not the intention at present to give a de-
tailed description of the transiter, it may be allowable briefly to
mention several of its advantages :

Regulation of the bisection of a star is easy and definite. A
record is made but once for each revolution of the micrometer,
and records will be increased in frequency only as special work
demands. A predetermined schedule of recording can be deter-
mined for any given run of the micrometer. Back lash of the
screw on reversal of motion may be completely eliminated by
the adjustment of the electrical contacts. The whole transiter
may be balanced symmetrically on the instrument, and thus
changes in the instrumental constants avoided. Instantaneous re-
versal of the motion of the micrometer permits of many conven-
iences as to method of work. In determinations of time and
longitude the tendency has of late been to reverse the transit in-
strument during the passage of each star, and thus to eliminate a
series of errors and facilitate reductions. The transiter by its
ability to reverse motion instantly, and even automatically, lends
itself readily to this method of work.

From the beginning of 190 1, when it was completed, until the
present the transiter has been the subject of many tests and of
some improvements, and for a year or more it is hoped it may be
destined to progressive change. It is now mounted on the four-
inch meridian circle, for which it is expected a suitable place may
be found at the Suburban Photographic Station of the Observatory,
when this Station shall have been definitely located, but only after
the completion of the present series of experiments with the transi-
ter, and the determination of the latitude and longitude of the City
Station of the Observatory.

Personal equation in all its variations remains a much more
serious factor than many painstaking astronomers, who have not
sufficiently practiced their accuracy even against a simple personal
equation machine, are willing to admit. It is then gratifying to
find that Professor Langley^ has recently been willing to propose

1 Prof. S. P. Langley presided at the meeting, and had at a recent meeting of
the American Astronomical and x\strophysical Society described his new and
very ingenious method of obviating personal equation in any time observation.

206 SNYDER — A NEW METHOD OF TRANSITING STARS. [April 4.

an entirely novel and highly suggestive method for its elimination
in many classes of observation. And it may therefore be .permis-
sible, in this presence, to draw attention to the fact that the
method of the electrical transiter permits for the first time the de-
termination of the absolute personal equation at any and every de-
sired star transit, and on the star itself. While reserving a com-
plete discussion of this subject for a future occasion, it should be
stated that several plans offer themselves to this end in the transiter.
To mention but one : The usual wires are undisturbed, and the
transiter can be adjusted so as to cut itself in and out automati-
cally at certain parts of the run and only there receive the at-
tention of the observer for star-bisection. At other portions of
the run the usual method of chronographic signals, or even of the
eye and ear method may be employed, and so, on reduction to the
middle, be compared with the transiter's automatic signals. Per-
sonal equation may thus be studied with facility on the stars
themselves and its variability traced through a simple observation
or a series of observations, and whatever is sufficiently stable ex-
pressed as a function either of the stellar declination or of stellar
magnitude or even of the physical condition of the observer.

It seems rather likely that finally all such study of the personal
equation, when it shall have clearly demonstrated the unreliable
character of the usual methods of transit observation and the ade-
quate accuracy of the newer method, will be relegated to the
Psychological Laboratory. Certain it is that the banishment of
reaction time from transit observations and the reduction of this
class of errors to those of bisection, either of a star image by a
thread or of a thread interval by a star, means an epoch in ob-
servational astronomy whose actual realization by suitable devices
is a worthy challenge to our best efforts.

With an automatic transiter allowing easy and accurate bi-
sections, a chronograph recording with the utmost accuracy, and a
clock of the best mechanism kept under constant pressure and
temperature, a new field for accurate work in longitude determi-
nation and in the evaluation of stellar position and stellar parallax
would be opened to the activity of the astronomer.

Philadelphia Observatory, March, 1902.

Magellanic Premium.

FOUNDED IN 1 786, BY

JOHN HYACINTH DE MAGELLAN,
OF LONDON.

1902.
The American Philosophical Society,

Held at Philadelphia, for Promoting Useful Knowledge

ANNOUNCES THAT IN

de;cknibe:r, 1902,

IT WILL AWARD ITS

MAGELLANIC GOLD MEDAL

to the author of the best discovery, or most useful invention, relating to Navi-
gation, Astronomy, or Natural Philosophy (mere natural history
only excepted) under the following conditions :

1. The candidate shall, on or before November i, 1902 deliver, free
of postage or other charges, his discovery, invention or improvement,
addressed to the President of the American Philosophical Society, No.
104 South Fifth Street, Philadelphia, U. S. A., and shall distinguish his
performance by some motto, device, or other signature. With his dis-
covery, invention, or improvement, he shall also send a sealed letter
containing the same motto, device, or signature, and subscribed with the
real name and place of residence of the author.

2. Persons of any nation, sect or denomination whatever, shall be
admitted as candidates for this premium.

3. No discovery, invention or improvement shall be entitled to this
premium, which hath been already pubhshed, or for which the author
hath been publicly rewarded elsewhere.

4. The candidate shall communicate his discovery, invention or im-
provement, either in the English, French, German, or Latin language.

5. A full account of the crowned subject shall be published by the
Society, as soon as may be after the adjudication, either in a separate
publication, or in the next succeeding volume of their Transactions, or
in both.

6. The premium shall consist of an oval plate of solid standard gold
of the value of ten guineas, suitably inscribed, with the seal of the Society
annexed to the medal by a ribbon.

All correspondence in relation hereto should be addressed
To THE Secretaries of the

American Philosophical Society,

No. 104 South Fifth Street,

PHILADELPHIA, U. S. A.

TRANSACTIONS

OF THE

American Philosophical Society

Held at Philadelphia,
For Promoting Useful Knowledge.

Vol. XX, New Series. Part /, ^to, pp. 62, with j Plates.
Lately Published.

CONTENTS.

Art. I. — The History of the Pelycosauria, with a Descrip-
tion of the Genus Dimetrodon Cope. By G.
Baur and E. C. Case.

Part II. 4to, pp. 7/^, with 5 Plates.
Just Published.

CONTENTS.

Art. II.— The Chronological Distribution of the Elasmo-
branchs. By O. P. Hay.

Art. III. — Results of Observation with the Zenith Telescope
of the Sayre Observatory from January 19, 1894,
to August 19, 1895. By Charles L. Doolittle.

Art. IV.~A Study of the Chromosomes of the Germ Cells
of Metazoa. By Thomas H. Montgomery, Jr.,
Ph.D.

SUBSCRIPTION— FIVE DOLLARS PER VOLUME.

SEPARATE PARTS ARE NOT SOLD.

Address THE LIBRARIAN OF THE

American Philosophical Society,

No. 104 South Fifth Street,

Philadelphia, U. S, A.

PROCE EDINGS

OF THE

A.MERICAN PHILOSOPHICAL SOCIEIT

HELD AT PniLADELPnU FOR PROMOTING ISEPll KNOWLEDGE.

Vol. XL[,- April, ia02. No. 170

CONTENTS

PAGE

A Cretaceous and Lower Tertiary Section in Soutli Central Montana

(with'^plate). By Earl Douglass 207

Areograpliy. By Perciyal Lowell 225

Systematic Geography, By W. M. Datis 235

Stated Meeting, April 18 259

The First Assembly Account — Philadelphia, 1748. By Thomas

Willing Balch 260

Stated Meeting, May 2 264

Stated Meeting, May 16 * 264

' philadelphia :
The American Philosophical Society,
104 South Fifth Street.
1902.

It is requested that all correspondence be addressed

To THE SeCRETAKIES OF THE

AMEKIC^N PHILOSOPHICAL SOCIETY,
104 South Fifth Street,

Philadelphia, U. S. A.

Members will please communicate to the Secretaries any
inaccuracy in name or address as given on the wrapper of this
number.

It is requested that the receipt of this number of the
Proceedinafs be acknowledged to the Secretaries.

Members who have not as yet sent their photographs to the
Society will confer a favor by so doing ; cabinet size preferred.

OCT

PROCEEDINGS

OF THE

AJ^ERICAN PHILOSOPHICAL SOCIETY

HELD AT PiniADELPflli FOR PROMOTING USEFUL KNOWLEDGE.

Vol. XLI. April, 1902. No. 170.

A CRETACEOUS AND LOWER TERTIARY SECTION
IN SOUTH CENTRAL MONTANA.

(Plate XXIX.)
BY EARL DOUGLASS.
{Read April S, 1902.)

This paper is intended only as a preliminary report of an inter-
esting geological section — an account of what has been done and a
suggestion of what is yet to be accomplished. The points of inter-
est are: (i) The completeness of the Upper Cretaceous which
overlies the older beds, probably Jurassic and Lower Cretaceous,
and underlies the Fort Union, which here contains mammalian
remains, correlating it with the Torrejon of New Mexico; (2) the
excellent exposures of the strata, giving a good opportunity for
study; and (3) the occurrence of interesting fossils, especially verte-
brates, in several different horizons.

The region here referred to lies east of the Crazy Mountains and
south of the Big Snowies, in the basin of the Musselshell River, in
Sweetgrass County. It extends from south of the Musselshell River
southward twelve or fifteen miles, and eastward from a line passing
southward from Harlowton on the Musselshell about the same dis-
tance. This is part of the south limb of a broad anticline, the
general trend of which is south of east. This anticline is dissected
longitudinally by the Musselshell. The lowest strata exposed are
upheaved into a dome-shaped uplift southeast of Harlowton and
four or five miles south of the river, where strata which are appar-
ently of Jurassic age are exposed.

This region is on the western border of the elevated plains coun-
try, and occupies a position intermediate between the plains and

PEOC. AMER. PHILOS. SOC. XLI. 170. N. PRINTED AUG. 29, 1902.

208 DOUGLASS — CRETACEOUS AXD LOWER TERTIARY. [Aprils,

the foothills. The strata are, as a rule, not horizontal, but have
been affected by the disturbances which have elevated the moun-
tains farther to the west or north. In restricted localities the beds
are horizontal and in others nearly vertical, and there are all inter-
mediate grades. The average dip is probably not more than fifteen
or twenty degrees.

The relief beautifully expresses the geological character. Through
the whole section there are alternations of sandstones and shales
and all grades between the two. Sometimes, as in the Fort Benton
and Fort Pierre, the shales predominate and attain a considerable
thickness. Again, as in the Dakota (?), Niobrara, Fox Hills, etc.,
sandstones predominate — at least there is enough indurated sand-
stone to retard erosion and to produce prominent ridges which can
be followed for long distances — fifty miles or more. In all the
formations there is considerable sandstone, and in all there is much
shale j but I have seen but very little limestone in the whole section,
though it sometimes occurs in concretions or in thin layers.

It does not appear that during the whole period of deposition the
sea ever attained any great depth. Probably it was deepest at times
during the Benton epoch, yet even here the great amount of sand
in the shales indicates near-shore deposition. The erosion features
will be given in the descriptions of the different formations.

So far as I am aware this particular region has been described
only by the writer (see Science^ January 3, 1902, p. 31, and Febru-
ary 14, 1902, p. 272). A little to the west is the area mapped in
the Little Belt Folio (No. 56) of the U. S. Geol. Survey, and some
work was done to the eastward on Swimming Woman and Careless
Creeks by W. Lindgren and George H. Eldrege, in connection
with the Northern Transcontinental Survey.^

Of course there is no single section where all the features here
described can be seen, and the depressions or ridges into which the
different strata weather have frequently to be followed for a few
miles to obtain good exposures. Fortunately this is easily done.

The Lake Basin, to which reference will frequently be made, is a
large, depressed area nearly fifty miles long east and west and
twenty-five miles north and south in the widest portion. The
former represents the greatest east and west extension. The east-
ern portion extends northeastward. This portion I have not ex-

^ Tenth Census of the United States, Vol. XV, p. 243.

1902.] DOUGLASS — CKh-TACEOUS AND LOWER TERTIARY. 209

plored. The basin has no outward drainage, but has several small
lakes without outlets, into which small streams empty, when there is
an excess of precipitation. The basin is bounded on the south by
the high rocky bluffs of the Fox Hills, and on the north, at least in
the western portion, by the hard sandstones of the Niobrara and
the Dakota (?). The name Lake Basin seems doubly appropriate,
for it not only contains lakes, but it resembles the bed of some
ancient body of water with bays and inlets, and with capes, pro-
monotories and peninsulas extending into it from the southward.
The scene is spread out like a great panorama ; the southern hills
and northern ridges become hazy in the distance and the farther
border seems a dim ridge on the eastern horizon. At the foot of
the Fox Hills bluffs are the Fort Pierre shales and still farther away
the Fish Creek beds.

As the principal object of this paper is to show something of the
characters of the uppermost Cretaceous and Lower Tertiary forma-
tions in this locality, and to give a little light tending toward the
clearing up of the problem concerning the boundary between the
Mesozoic and the Cenozoic ages in the Rocky Mountains, I will
give only a brief sketch of the formations lower than the Niobrara.

Jurassic, etc.

The supposed Jurassic is exposed in a dome-shaped uplift, so that
the strike of the outcrop is nearly a circle. The beds are sand-
stones and sandy clays. The latter are largely red in color. This
is apparently due to the combustion of coal. There are bones of
large Dinosaufs and of some smaller reptiles, but they have not been
studied. It is possible that this stratum with the sandstones above
may belong to the Lower Cretaceous. There are many hundreds
of feet of hard sandstones and shales between the fossil-bearing
horizon and the Fort Benton. The upper portion probably belongs
to the Dakota formation.

The I'ori Benton Formation.

These beds and their ^contained fossils are much like the cor-
responding ones in other regions. They are principally dark shales
with bands of sandstone in the lower portion, and in one place I
found a half dozen specimens of Frionocyclus Meek in brown con.
cretions in the shales. Higher were Ammonites^ Scaphites, Inoce-

210 DOUGLASS — CKETACEOUS AND LOWER TERTIARY. LApril 3,

ramii small Baculiies and other Mollusca, all of Benton types.
These shales weather into ravines between the sandstones of the
Dakota below and the Niobrara above.

Niobrara.

In the Niobrara gray sandstones predominate, though there are
beds of shale. This differs from the usual character of this forma-
tion in most other regions where it has been observed. It has
usually been described as being composed principally of limestone
and marl, though sometimes containing considerable sand. The
sandstones here are very much like some of those of the Laramie,
and near the middle portion are seams of coal. In two or three
places I looked in vain for any well-preserved plant remains in the
carbonaceous shales and in the sandstones above and below the
coal, and followed ravines cutting through the prominent sandstone
ridges without finding any good fossils. However, about twenty
miles to the southeastward a few plant impressions were found — the
best of which was apparently a Sequoia — in beds which I take to be
Niobrara. Undoubtedly, by careful, continued search, a fair col-
lection could be obtained.

In one place, where Mud Creek cuts through the formation, the
beds approach near to a vertical position. I should not estimate
the thickness to be less than 700 or 800 feet here. It may be
more. The sandstones form a prominent ridge where they are
much inclined. These ridges are sometimes wooded, though the
trees are usually not very large or numerous.

Fisii Creek Beds.

Above the Niobrara are beds which I believe to belong to the
Belly River formation, but until they are certainly correlated with
the latter I give them the above name.

They are best exposed between Fish Creek and Mud Creek, only
a few miles from where the latter empties into the Musselshell
River. Here they are nearly horizontal, while the underlying
Niobrara dips at a considerable angle to the southward. Farther to
the east and west I did not notice any unconformity between the
two formations. In the above-mentioned locality, where they are
horizontal, they weather into *'bad land " forms. The material is
principally rather soft sandy clay, with hard, almost black concre-

1902.] DOUv'^LASS — CRETACEOUS AXD LOWER TERTIARY. 211

tions and hard sandstone layers. In the latter there are, in some
places, plant impressions. The softer layers contain fossil wood,
bivalve moUusks, turtles and bones of Dinosaurs of the genus
Claosaurus. The bones are generally petrified and occur also in
the dark concretions which also contain plant remains. Though
they are, as a" rule, excellently preserved, yet sometimes there is
what seems to be a good portion of a Dinosaur broken into myriads
of little fragments. The beds are probably either fresh or brackish
water.

This formation was observed in several places in this region, and
in all there were bone fragments ; but we found no other equally
good exposures. About twenty- five miles to the southeast, in the
Lake Basin north of Columbus, the formation lying immediately
below the Fort Pierre in one place has a considerable thickness of
sandstone containing petrified logs, but only one or two small
fragments of bone were found. Some of the plants of this forma-
tion are related to Sequoia. The bivalve shells were so fragile as
to crumble with the soft matrix in which they were imbedded.

Lying over these beds is a series of shales and hard laminated
sandstones. Some fossil leaves were seen in the latter. A series of
dark shales, perhaps thirty feet thick, was carefully examined. The
shales were full of carbonaceous plant fragments, and some fairly
good leaves were found in the thin interbedded layers of sand or
sandy concretions. I do not know whether these beds should be
put in this series or in the Fort Pierre. I think it better to consider
them, until they are more thoroughly explored, as belonging to the
Fish Creek series.

Fort Pierre.

Above the beds just described are the Fort Pierre shales. This
represents a well-distinguished horizon, so well marked by litholcgi-
cal characters and by characteristic fossils that its position is
beyond doubt. The description of the Pierre in Colorado, Wyom-
ing, etc., would answer almost equally well for the formation here.
Dark, soft shales predominate. There are occasional thin bands of
sand and many brownish concretions which break into angular
fragments. These sometimes contain marine fossils and sometimes
a network of calcite seams. The best preserved invertebrate fossils
are in these concretions. The shells are those oi Ammonites^ Bacu-
Hies, Scaphifes, Nautili, and small Gasteropods and Cephalopods.

212 DOUGLASS — CRETACEOUS AND LOWER TERTIARY. [Aprils,

Some hard limestone concretions are crowded with these small
molluscs.

What distinguishes the Pierre here from that in other places is
the presence of many vertebrate fossils. Several Mosasaurs have
been found. In the summer of 1900, Mr. Albert Silberling and I
found portions of two individuals, including a skull. In the sum-
mer of 1901, the Princeton Expedition in charge of Dr. M. S. Farr
procured a nearly complete skeleton except the skull.

But the most interesting fossil remains are those of the Dinosaiws.
They have been found to be more numerous here than the Mosa-
saurs. The greater number of them belong to the genus Claosaurus
and apparently to described species. Two portions of skeletons
belong to quadrupedal type, probably to the Cerafopstd(3. A C/ao-
saurtis skull and the greater part of the skeleton was obtained for
the Princeton Museum last summer (190 1). The digging was easy,
but the removal of the bones was slow and tedious, as they had to
be hardened. Nodules had formed around some of them, but
many were in clear shale. The skeleton was just above a layer of
yellowish, partly consolidated sandstone two or three inches in
thickness. There were some thin layers or lenses in the shale, in
which the remains were imbedded. There was also a minute seam
of coal not thicker than cardboard. Cones or ends of twigs of
what appeared to be Sequoia, Ammonites, Scaphites, Baculites and
other molluscs, and shark's teeth were found in the matrix while
removing the skeleton. Only the teeth and a few of the shells
could be preserved, as the fossils in the shale disintegrated on
exposure to the sun and rain. The deeper into the shale excavation
was made, the larger the flakes into which it would break. Quite a
number of other portions of skeletons were found during this and
the previous year. Often the bones are solid, though lying among
the grass roots, where the soil is composed of the disintegrated
shales. Sometimes the nodules surrounding the bones are very hard
and flinty.

The finding of Dinosaur remains in these marine beds was un-
expected, but the sea was evidently shallow. In some places there is
much gypsum in good-sized crystals, or in minute ones scattered
through the shales.

The Pierre beds being soft, have weathered into depressions.
They are usually covered, except in restricted portions, with a good
growth of grass, but are treeless except for a few small willoAvs or

902.] DOUGLASS — CRETACEOUS AND LOWER TERTIARY. 213

cottonwoods that occasionally grow along the streams. They make
grass-clad rolling prairies, with small ravines cutting into the soft
shales.

The transition beds between the Fort Pierre and Fox Hills are
usually obscured by the material washed down from the bluffs of
the latter ; but on the ranch of Mr. B. Forsythe, near the head of
a branch of Big Coulee Creek, they can be nicely seen. The shales
gradually become more sandy, and contain bands of sandstone
until the latter predominates and the shales become shaly sand-
stones or sandy clays. In them I found no trace of fossils.

Fox Hills,

In this formation the hard sandstones form a prominent ridge
adjoining the depression made by the Pierre. It is the next promi-
nent ridge above the Niobrara. I have followed its base for about
thirty-five or forty miles. In only one place was there any confusion
or any difficulty in tracing it, and this was caused by some change
in the geological structure obscuring the Pierre shales. The out-
crop extends southeast and northwest. It forms the southern rim
of the Lake Basin. It furnishes many springs which, uniting their
waters, produce little streams that cut through the rocky ridge and
flow out upon the Pierre flats. In the Fish Creek region they
empty into Fish Creek. In the Lake Basin, if the water does not
soak into the ground, they flow into the land-locked lakes. Where
the streams form little canons and ravines through the Fox Hills
strata, they are fringed with trees and shrubbery. In little valleys
and amphitheatres there are often springs surrounded by groves,
which are very picturesque, and in the heat of summer these places
form a delightful retreat from the almost treeless wastes around.
The trees, which are principally evergreens, cottonwoods, poplars
and willows, follow the streams a little way toward the Pierre flats
and then disappear.

Though these beds usually appear to be sandstone ridges, yet in
places where conditions of weathering are favorable they are seen
to contain much sandy clay, and in places for a short distance
resemble '' bad land " forms.

Fossil leaves and reptilian bone fragments were found in consid-
erable abundance. Dr. Farr brought back some of the fossil leaves,
but they have not yet been determined. Most of the bones are too
fragmentary to be of much use. Some teeth were recognized as

214 DOUGLASS — CKETACEOUS AND LOWER TERTIARY. [Aprils^

belonging to Ciaosaurus. The only fossil plant we were able to
recognize in the field was a species of Salisbu7'ia.

Though this is probably still below the Laramie — at least there
are thousands of feet of what is apparently Laramie above it — yet
this is the highest level in which we found Dinosaur xtmdXns in this
region. This is interesting, as in other regions the Ciaosaurs, with
one exception, have come from beds which have been supposed to
be above the Fox Hills.

It is not certain just where the Fox Hills ends and the Laramie
begins. It is possible that these bones, or at least some of them,
are in the lowest Laramie ; but as the two formations represent
differences in conditions of depositions rather than difference in
age, as distinguished by change or progression of the fauna or flora,
it is not so essential, except as bearing on the more interesting ques-
tion of the extinction of a very remarkable class of animals and
the occupation of their territory by a class that had for millions of
years held a subordinate position.

Above the Pierre, in the Fish Creek region, are alternations of
dark shales and gray sandstones. In places the sandstone is warped,
twisted or made up of imperfectly concentric layers. Above these
are brownish laminated and greenish or brownish unlaminated
sandstones and sandy clays. Provisionally, I place the base of the
Laramie above these latter beds. They contain fossil leaves and
bone fragments.

Laramie,

The lowest beds, which are here taken to be Laramie, are a series
of alternating various-colored shales and gray unlaminated sand-
stones. There are several hundreds of feet of these and no fossils
were found in them. There are in some layers brownish concre-
tions, some of which are large and composed of sandstone. These
beds form a depression, but not so low as that of the Pierre shales.

Over these lies about an equal thickness of similar sandstones and
gray shales. The former are harder and form a bench or ridge.
.There are several thin seams of coaly matter and the shales hold
impressions of ferns and other delicate plants different from what
we observed elsewhere.

Near or at the top of this series there are at least two layers
containing non-marine fossils. In one of the fossils are principally
Gasteropods and in the other bivalves — probably Unio. It

1902.] DOUGLASS — CRETACEOUS AND LOWER TERTIARY. 215

is said that this layer extends for twenty miles up Fish Creek,
but I have not tried to trace it, so do not know whether it is con-
tinuous or not. It is also said that these fossils gave the Mussel-
shell River its name. Here we may be quite sure that we are in the
Laramie, for fresh or brackish water conditions prevail, but it prob-
ably extends between looo and 2000 feet below.

Still higher are shales forming a flat or depression, above which
are conical hills or hog-backs — the remains of dissected ridges cut
through by ravines and by streams which are fed by springs in the
Fort Union sandstone above. These hills or ridges are capped with
brownish, compact, laminated sandstone. No fossils were seen
except fragments of wood in the shale.

Above these sandstones dark shales again predominate. I cannot
tell, at least without more careful study and observation, where the
Laramie terminates and the Fort Union begins. In fact, it looks
as if there were in this section almost continuous deposition from
the Jurassic up. We found here no traces of the volcanic material
of the Livingston formation, which only thirty or forty miles to the
southwest is so well developed. It appears that here deposition
went on quietly and uninterruptedly. There is little doubt that
part of the strata were deposited synchronously with those of the
Livingston. Here, so far as we have discovered, as in other places.
Nature has left no way marks and laid down no boundary line to
distinguish between the great ''Age of Reptiles " and the ''Age of
Mammals." There appears to be no sign of the disturbance that
is supposed to have closed the Mesozoic and brought in a new order
of things; yet only a few miles away there was a region of upheaval
and of intense volcanic activity. The strata in the section under
consideration have been disturbed, but the Tertiary beds are also
involved in the upheaval. Perhaps microscopic or chemical exam-
ination may reveal the presence of fine volcanic material here.

Mr. W. Lindgren made three different measurements of the Lara-
mie to the eastward of this region (see Tejiih Census of the United
States, Vol. XV, p. 744). In none of these does he make the
thickness of the Lower Laramie to be less than 7000 feet. I do not
think that this, as C. A. White ^ thinks probable, includes the Belly
River, or anything lower than Fort Pierre. Lindgren's Upper
Laramie, or Bull Mountain series, is probably Tertiary — apparently

1 « Correlation Paper, Cretaceous," Bull. S4, U. S. Geol. Survey, p. 174.

216 DOUGLASS — CRETACEOUS AND LOWER TERTIARY. [AprU3,

Fort Union. What is supposed to be Laramie in the present sec-
tion is very thick, probably approximating that of Lindgren's
measurements. But here, as everywhere else, the boundaries of the
Laramie are uncertain. Here, however, we have it confined be-
tween certain limits. We have below a characteristic Fort Pierre
fauna and above a characteristic Fort Union flora. Just how much
of that which intervenes is Laramie is not known. I have no
doubt that here deposition was going on at the same time as that of
not only the Livingston, but also of the Arapahoe and Denver beds.
Whether these beds will ultimately be assigned to the Upper Lara-
mie, or included in a separate formation, depends upon the results
of future careful investigation.

Tertiary.

Fort Union.

The dark shales just mentioned continue upward, changing little
in character ; but brown concretions become numerous, then layers
containing shells of bivalve MoUusca, then occasional layers of
sandstone, and above these, often capping the bluffs, heavy gray
sandstones, usually hard, sometimes laminated and sometimes mas-
sive. Above this I cannot speak definitely, but think that the Fort
Union continues much higher. The strata from the top of the bluffs
south of Fish Creek, which make a bench sloping toward Sweet-
grass Creek in the direction of Melville, perhaps belong to higher
members of this formation. The strata are not always continuous
for great distances, but vary locally ; yet a general description can
be given that will apply fairly well to the beds examined. There
are dark gray shales that in many places weather to thin, flaky
particles on the surface. The wind blows away this light material
and leaves bare depressions without vegetation. The sandstones
are usually hard, sometimes massive or imperfectly bedded, and in
some places break into great blocks, which tumble down the steep
sides of the bluffs.

In the Fish Creek region these heavy sandstones, which lie above
the soft shales, form a long line of rugged bluffs extending along
the south side of the creek from the neighborhood of Porcupine
Butte eastward for twenty-five or thirty miles \ then it extends
southeastward, probably forming the divide between the Sweetgrass
on the southwest and the southern branches of Fish Creek and Big

1902.] DOUGLASS— CRETACEOUS AXD LOWER TERTIARY. 217

Coulee Creek ; but I have not examined all of this territory. I ex-
amined hastily the beds on Sweetgrass Creek east and a little north
of Big Timber, where I made a collection of fossil leaves. The
remains of a turtle were also found in the shale.

The portion of the Fort Union described in this paper apparently
represents the upper portion of the Crazy Mountain section, as
given by Weed in the American Geologist oi October, 1896.

Fossil plants, Unios and Gasteropods, are abundant and may
occur in any part of the beds favorable for their preservation. Last
summer (1901) determinable Mammalian remains were found. As
is well known, the exact position of these beds has been a matter of
some doubt and difference of opinion. They have usually been
assigned to the Tertiary, though they have been placed as low as
the Cretaceous and as high as the Miocene.

The bones and teeth of Ma7fi?nals which were found ^ are not
numerous, but are sufficient to show that the beds are of nearly the
same age as the Torrejon of New Mexico. They are :

Mioclcenus acolytus (Cope).

Anisonchus very near to A. sectorius Cope.

Euprotogonia puercensis (Cope) .

Pantolambda caviridis (?).

Pantolatnbda (?), a small species.

Some others are doubtful.

I felt very certain that these beds were Fort Union, but to settle
the matter forever and leave no room for a shadow of doubt, a box
of fossil leaves was sent to Mr. F. H. Knowlton, of the United
States Geological Survey. Mr. Knowlton examined them at once
and sent me a list, which I quote :

Pterospermites cupanioides (Newb.) Knowlton.

Popidus speciosa Ward.

Populus amblyrhyncha Ward.

Ulmus orbicularis ? Ward.
Vitis xantholitJiensis Ward.

Populus dapJmozenoides Ward.

Populus arctica ? Heer.

Platanus aceroides Gopp.

Celastrus sp.

Grewia crenata (Ung.) Heer.

1 Science, February 14, 1902, pp. 272, 273.

218 DOUGLASS — CRETACEOUS AND LOWER TERTIARY. [Aprils,

Viburnu??! asperiun ? Newb. •

Populus cuneaia Newb.

Populus sp.

Plat anus nobilis Newb.

Platanus basilobata Ward.

Viburnum sp.

Paliurus sp.

Grewiopsis viburnifoUa Ward.

Populus .? n. sp.

Mr. Knowlton says : '' The species are all Fort Union beyond a
doubt."

Of a {^\N shells which I enclosed, he writes: ''The shells I
showed to Mr. Stanton, and he says that the two large ones are
Unio Couesi White ; and the other pretty near to Unio Endlichi
White."

The Mammals were found in the shale. The collection of fossil
leaves was made in the sandstone a little higher up, though there
are concretions and layers of sandstone that contain leaves in the
same beds as the Mammalian remains. A portion of the collection
was obtained on Sweetgrass Creek north of east of Big Timber, in
the locality mentioned above.

General Observations.

The problem of greatest interest connected with the study of
this section is that relating to the transition from Mesozoic to
Cenozoic times. Of course, if deposition had been continuous, or
nearly so, and there were no great faunal or floral migrations, there
could be no distinct boundary between the two. There is a great
difference between the Cretaceous as a whole and the Tertiary as a
whole, but where are we to draw the line? If there was a time of
widespread or general upheaval throughout the western portion of
the continent, or of the Rocky Mountain region, this might form
a convenient division. Upheavals and great volcanic activity cer-
tainly occurred in restricted localities, but we cannot at present
prove that such were general or that they did not occur in different
places and at different times. If we could point to any time when
the Di7iosaui's ceased to be and the higher orders of Manwials took
their places, then the matter would be easy ; but heretofore most of
the Cretaceous Dinosaurs, in fact nearly all of them, have been

1902.] ])OUGLASS — CRETACEOUS AND LOWER TERTIARY. 219

supposed to come from the uppermost portion of the Cretaceous —
the Laramie — but the other fossils found in these beds have not
been of a character to settle the doubt concerning the horizon.
There is no direct proof that the Dinosaurs died out before higher
forms of Mammals became numerous. Though they have not yet,
so far as I know, been found in the same beds, yet there seems good
reason for believing that Dinosaurs were contemporaneous with
Puerco Mammals. Were it not for the '^ Ceratops fauna" and the
discovery of a few specimens in the eastern United States and one
in Kansas, we should say that the Dinosaurs died out at the end of
the Jurassic. It would seem that if anything had a chance of being
preserved it would be the large, solid bones of these animals; yet
there are miles of thickness of strata and thousands of square miles
of exposure of Lower Cretaceous, Dakota and Colorado beds, and
nothing, I believe, has been found to tell that these animals still
lived in this great Cordilleran region, except the type of Claosaurus
agilis from the Niobrara of Nebraska. This rock must represent
many millions of years in which Dinosaurs lived, flourished and
progressed. To our view they disappear in their glory, and after
ages appear again in glory but transformed ; again they suddenly
disappear and we see them no more. The morning, midday and
evening of their splendor is lost to us. Until the discovery of the
beds described in this paper almost nothing was known of them in
the Montana formation, at least the beds from which they had been
collected had not been considered as belonging to that age. The
point the writer wishes to make is this : It is extremely unsafe to
say when and where these strange reptiles breathed their last, for
the presence of fossils is certain evidence of the existence of life,
but the lack of them is no evidence of its absence. Dinosaurs may
have continued long in the Eocene, but conditions in the places
where so many Mammalian remains have been found may not have
been favorable for them.

I think we can hardly account for the general absence of Dino-
saur remains in the Kootenai and Upper Cretaceous, below the
Laramie, by the beds being in part marine. Much of the strata is
evidently fresh or brackish water. We should hardly expect to find
them in the Benton and Fort Pierre shales associated with large ma-
rine Mollusca, yet as previously stated we do fitid them in the latter.
This proves that these animals lived near the sea or where they
could float into it. Why don't we get them then in the many

220 DOUGLASS — CRETACEOUS AND LOWER TERTIARY. [Aprils

thousands of feet of sandstone which, if marine, must be near-shore
deposits? It is true that any day we may hear of their being found
in some of these strata, and we may also hear of their being found
in Eocene strata, if they have not been found there already.

As shown by this paper, the presence of Claosauridce, and proba-
bly of Ceratopsldce, is far from showing that the beds in which they
are found are as late as Laramie — I mean as the Laramie as it is
understood. It is true that the Fort Pierre, and in some places the
Fox Hills with it, represents an incursion of the sea, and that con-
ditions of life were not greatly different during the time of the
deposition of the Belly River beds from what they were in the
Laramie.

At present the fossil plants, together with orographic movements
and their results when they occur, are the only things we can use to
distinguish these doubtful formations as the Laramie, Livingston,
Denver, etc. The plants, on account of mixtures of the flora of
different horizons in collecting, have not been available for use until
the material has been carefully separated. As Mr. Knowlton has
been doing this work, his forthcoming monograph on the Flora of
the Laramie and Allied Formations will be looked for with interest.

There is not much doubt that the Livingston in Montana repre-
sents the upper portion of what has been called the Laramie in the
plains region farther to the east. Both have Laramie strata below ;
both are overlaid by Fort Union beds. In Colorado it seems that
the Arapahoe, and probably the Denver, or the greater part of it,
sustains the same relation to Laramie. Mr. Knowlton says : ^' From
these considerations it appears beyond question that the flora of
the Livingston formation finds its nearest relationship with the Den-
ver beds of Colorado," ' If the Livingston and Denver are of the
same age, as has for some time been suspected, then the Denver
must be older than the Fort Union, and therefore older than the
Torrejon. With its apparently Cretaceous Vertebrate fauna, we are
not warranted at present in placing the Denver much higher than
the Livingston. It may be in part contemporaneous with the Fort
Union.

The Puerco should be nearly of the same age, as it lies between
Laramie and Fort Union (Torrejon) strata.

Below is given a table which is intended to show the probable

1 Bull. 105, U. S. Geo I. Survey, p. 63.

1902,1 DOUGLASS — CRETACEOUS AXD LOWER TERTIARY. 221

relations in time of the formations under consideration concerning
which there is doubt :

Table Shotuing Probable Relations of the Laramie and Overlying
Beds in Different Regions.

Cretaceoics. Tertiary,

Laramie of King

In Wyoming ^mmmmmn^mim^mMaimmmm^^^^i^m-

Laramie Fort Union

Plains of Montana •

Laramie Livingston j Fort Union

Crazy Mts., Montana ^_Bai^_iB_ \

Laramie Arapahoe Denver;
Denver Basin .i.^__. ■»«».» ;

Laramie Puerco Torrejon

Puerco River, N. Mex '

The names given are the ones by which the different divisions
have been called. There does not seem to be much doubt that the
Livingston, Denver, Puerco, etc., are contemporaneous with what
in other places has been assigned to the upper portion of the Lara-
mie. Whether all will be included in the Laramie later will de-
pend on the results of further careful investigation. I have indi-
cated the doubtful division between the Cretaceous and Tertiary by
a dotted vertical line passing between the Livingston and Fort
Union and between the Puerco and Torrejon, or approximately so,
not claiming that the time division line between the two sets of
strata would fall exactly in the same place. The horizontal parallel
lines are intended to represent contemporaneity of deposition.
Deposition in the Denver Basin was not continuous and the blank
spaces indicate non-deposition. The broken or dotted lines indi-
cate probable continuity.

Remarks on the Fossil Mammals.

The mammals are represented by about a half dozen species.
Five of these are represented by teeth. Almost any one of these

222 DOUGLASS — CRETACEOUS AXD LOWER TERTIARY. (Aprils,

taken alone would strongly incline one to the belief that the form-
ation containing them is contemporaneous with the Torrejon of
New Mexico. This is made still stronger by nearly every specimen.
There are a radius and ulna which are different from any found in
New Mexico, so that they cannot be assigned to any genus with cer-
tainty, and there is a premolar much like that of Pantolambda, but
indicating an animal much smaller than any species of that genus,
to which, however, I refer it with doubt. The other four are
cogeneric if not cospecific with Torrejon forms.

Miocl<z7iiis acolyius (Cope). (Plate XXIX, Figs. 9 and 10.)

This is represented by a small portion of a mandible with a molar
tooth which is almost unworn. The anterior cusps are connate at
base and much higher than the posterior ones.

Anisonchus Cope. (Plate XXIX, Figs. 3-5.)

This is also represented by a portion of a mandible. There are
two teeth, a fourth premolar and a first molar. They are of nearly
equal length. In size and character the teeth are nearly like A.
sectorius Cope. It may, however, be another species.

Euproiogonia puercensis (Cope). (Plate XXIX, Figs, d-'^.)

Represented by a third premolar and a second molar of the right
side and a third molar of the left. The molars differ somewhat
from the type. Matthew has carefully studied the many specimens
in the American Museum collection and finds a wide range of varia-
tion in the teeth, but no constant characters that will serve to sepa-
rate the various forms which Cope has named. Of the many speci-
mens no two appear to be exactly alike. I have compared the
present specimens with those in the above collection and find that
they do not differ so much from some of the American Museum
specimens, as the latter vary among themselves. What comes near-
est to being a distinguishing character is the smallness of the hypo-
cone as compared with the protocone, but this is at least nearly
paralleled by some of the above-named specimens.

Pantolambda (?) (Plate XXIX, Figs, i, 2, 14.)

There are the greater portions of a radius and ulna, and two
phalanges which are different from anything described from the

1902.] DOUGLASS — CRETACEOUS AND LOWER TERTIARY. 223

Torrejon. They more resemble in some respects the corresponding
bones of Coryphodon.

The ulna is much larger than the radius, is broad antero-poste-
riorly but narrow transversely. The upper portion of the olecranon
is broken off, but a cross section above the glenoid cavity is trian-
gular with the anterior edge thin. The sigmoid cavity is convex
transversely. The outer portion is much less convex longitudinally
than the inner ; it extends lower and its upper portion makes an
oblique emargination on the outer side of the olecranon. There are
two fairly large surfaces for articulation with the radius. The upper
outer surface of the bone has a quite deep longitudinal furrow which
dies out near the middle of the shaft. The inner surface is longi-
tudinally concave from the olecranon to the enlargement near the
distal end of the bone, where there is considerable swelling and
roughening. The distal articular surface is elliptical, slightly con-
cave palmo-dorsally and convex transversely. This surface is very
slightly oblique to the long axis of the bone.

The radius is subcylindrical above. The head is partly broken,
but the surface for articulation with the humerus is shallow and
appears to have been nearly circular. There is a longitudinal
roughening on the ulnar side, to correspond with similar rugosities
on the radial side of the ulna. Below these is a rugosity on the
antero-inner side of the radius and on the opposite side. The bone
has the appearance of being twisted on itself. The form of the
bone suggests freedom of motion of the limb other than a fore-and-
aft movement.

A proximal and medial phalanx apparently do not differ greatly
from those figured in Osborn's paper on '* Evolution of the Ambly-
poda."^

M.

Length of ulna from upper portion of glenoid cavity .1970

Antero-posterior diameter at middle of shaft . .0310

Transverse diameter at middle of shaft . . .0143

Transverse diameter of shaft of ulna at middle . .0195

Pantolambda cavirictis Cope (?).

Fragments of upper jaw, with teeth from which enamel has been
removed. The size is nearly the same as the corresponding teeth

i Bull. Anier. Mus. Nat. Hist., Vol. X, p. 187.
PROC. AMER. PHILOS. 80C. XLI. 170. O. PRINTED SEPT. 23, 1902.

224 DOUGLASS — CRETACEOUS AXD LOWER TERTIARY. [Aprils,

of P. cavirictis, and there is nothing to distinguish it from that
species.

Paniolambda (?) sp. (Plate XXIX, Figs. 11-13.)

An upper premolar, much smaller than P^ of P. bathmodon or P.
cavirictts, but it is possible that it may be a P ^ of nearly as large a
form. It is very doubtful, however, whether it belongs to Panto-
lambda at all. The protocone is more conical, the outer slope on
the median line of the tooth is steeper and the inner less so. The
outer surface near the base of the crown is more concave.

A canine found with the above is not like that of the known
species of Pantolambda, but can hardly be distinguished from that
of modern Carnivores. It probably belongs to some Creodont.

Princeton University, May 24, 1902.

Explanation of Plate XXIX.

Figs, i and 2. L^lna and radius possibly belonging to some species of Panto-
lambda. % natural size.

Figs. 3-5. Anisonchus sectorius (?)

Last lower premolar and first lower molar with portion of mandible. Outer
and inner view of mandible and upper view of teeth. X 2.

Figs, 6-8. Euprotogonia puercensis.

6. Right upper second molar. X 2.

7. Left upper third molar, x 2.

8. Right third upper premolar, x 2.
Figs. 9, 10. AHodcenus acolytus.

A lower molar with portion of a mandible. X 2.
Figs. II-15. Pantolambda {7). Upper premolar. X 2.

13. Canine tooth found with 11.

14. Phalanx found near i and 2. x Yy

J 5. Scale of Lepidosteiis found with mammals, x 2.

1902.] LOWELL — AREOGRAPHY. 225

AREOGRAPHY.

BY PERCIVAL LOWELL.
(^Read April J^, 1902.)

1. Facts familiar to the specialist are often credited with a gen-
eral appreciation they do not possess. Immersed in his own line
of research, the investigator forgets that others are not as intimate
as he with some of the fundamental points of his inquiry, and
omits as truisms what to others are not even known for truths.
Areography is such a subject. Probably no one outside of the pur-
suit is aware how cogent is the conclusion to be derived from an
inspection of the maps that have been made of the planet, as to
the reality and the relation of the markings there depicted. Nor
was it indeed till after I had compared these maps with some par-
ticularity that certain deductions from them forced themselves upon
me. It will perhaps, therefore, not be unproductive of result if I
present at this general meeting a collective view of the maps which
have from time to time been made of Mars and note what they
imply. Maps speak best for themselves, and with very slight intro-
duction can be made to tell their own story better than any amount
of text.

2. Of the maps here brought together, the earlier are taken
from Flammarion's thesaurus La Planete Mars^ Proctor's Dawes'
map from his own book, Schiaparelli's from his memoirs in the
Accademia dei Lincei, and the later ones from my own work. Of
these latter, that for 1896-97 is the result of my own synthesis of
the Flagstaff and Mexican observations of the Lowell Observatory
for those years, while the ones for 1898-99 and 1900-01 I have
but just completed, and they appear here for the first time.

3. All the maps here given marked in their day the point that
areography had then reached. With but two exceptions, that of
Flammarion and Proctor, therefore, they represent original obser-
vations made by the maker of the map himself or under his direc-
tion, and show in procession the evolutionary development of the
subject. Such maps as failed to add to existing knowledge and are
valuable merely as confirmatory documents have not been included.
On the other hand, no map which materially contributed anything
has been left out. Many excellent charts, therefore, have had to
be omitted, not always because they presented nothing new, but

226

LOWELL — AREOGEAPHY.

[April 4,

because contemporaneous ones included practically all their discov-
eries with additions. Of these, Cerulli's maps of 1896-97 and
189S-99 and Flammarion's of 1900-01 deserve special mention.
The omitted maps confirm, not invalidate, the conclusions here
drawn.

4. Twelve maps constitute the series. Of these the ordering
chronologically runs thus :

I.

Map of Beer ar

d Madler

1840

II.

Map of Kaiser

.

. 1864

III.

Map of Dawes,

by Proctor .

1867

IV.

Map resume by

Flammarion .

. 1876

V.

Map of SchiaparelH

1877

VI.

Map of

.

. . 1879

VII.

Map of "

.

1881-82

^III.

Map of

.

. 1883-84

IX.

Map of Lowell

.

1894

X.

Map of "

.

. 1896-97

XI.

Map of "

.

1898-99

XII.

Map of "

.

. I 900-0 I

Mercator's projection is used in all the maps. The zero meridian
is in the same point on the planet in all except Kaiser's, though
that meridian does not always fall on the same part of the plate,
being in I, IX, X, XI and XII on the extreme left, in III, IV, V,
VI, VII and VIII in the centre. Beer and Madler's map is given
by Flammarion on a stereographic projection which, for the sake of
inclusion in the present series from its age and chronological im-
portance, has been changed to Mercator's. All other circumpolar
projections have been omitted.

5. It will be seen from inspection of the maps, and would be
simply corroborated by further additions to the list, that increase
in our knowledge of the surface of Mars falls naturally into four
divisions or stages of development. The first of these is pre-carto-
graphic ; the second extends from 1840 to 1877; the third from 1877
to 1892; the fourth from 1892 to the present day. The first three
of these divisions correspond to those given in Flammarion's La
Planeie Mars. The fourth is since the publication of that bock.

6. Near the end of the several periods are observations which
mark the dawn of the next to come, making as they do adumbra-
tion of phenomena clearly to be revealed in the succeeding stage.
Though not themselves the detection of details which characterize

1902.]

LOWELL — AREOGRAPHY.

227

the period, they make transition to them. Dawes thus made twi-
light to the third period in 1864; W. H. Pickering and the Lick
observers to the fourth in 1892.

7. Distinct phenomena characterize the three periods. Patches
of light and shade make the markings shown on the maps of the
first stage of cartography. Of a piece though these patches are,
their shapes appear well defined. At first one might suppose such
to be due to the handicraft of the draughtsman and to possess no
scientific value. But inspection of the several charts, one after the
other, shows that the shapes are not artistic embodiments of ill-seen
shadings, but are intrinsic traits of the shadings themselves, for
chart after chart reproduces the same turns and twistings.

8. To see this we have but to take up in sequence the maps from
1840 to 1876. No. I of the series shows a cordon of patches

Fig. I.

M

0 » ro JO

fo so. 6O ,

0 eo .

^

-

^ -,

^/a? /S02OOBO2}0iiOtKZ5OBsai»m0lX>JO03/o3X XOMO JS> MO t» Mt m»

--

f...-

■■^..

1*

y

e

^

\

1

1

■r

1

!--??

~

k-

■■^-

,

-ff

%

!.;

/;>

:

%

1

».

E

,f

n

y(

■$} ' • * '. "

^,

■:-.;

S'-

:-•'■

^\

f'jj

&

Su

%

^:

"■

€

^\

n

'!%

ft

i

20

0

i

.

••■'

v^

s

%

.'^r

[€-i

.-?*

^

'««

^^yj

.;^..

,_.,

T'i'

%

w

-i.

^

w

0,

f

'"«!

k

m

^

ri

■■"''

m

d

'■

'v

v,^

w

^^

_,

m

a

w

'1

:

;

m

i^

/

,

X;

] -w

t^

m-

-|

f

=!

- -

"'

:-:-■

'-

If

<

r'

"■'?.

^■

.-'

y

1

r^5

lV>

^

if..

= ;-r:-.-

■aM

;||

A>

r«9 «D ^ gao^oaiaao SSOM> ^SOMOi

MAR5 (640

Map of Beer and Mailer, 1840.

Stretching round the map at about 30° south latitude. Their height
is greatest at 90° of longitude, and from this slopes down through
360° to 20° longitude, whence it gradually rises to the maximum.
At the point of maximum is an oval marked out by broad shading
on the south, by narrow penciling on the north, and holding a
roundish dark spot in its centre. This is the Soils Lacus, the eye
of Mars. To the right of it follows a leech-like patch, the Mare
Sirenum and the Mare Cimmerium seen as one. After this comes

228

LOWELL — AREOGRAPHY.

[April 4,

a large dark area in the shape of a funnel, the Syrtis Major. Then
a ribbon ending in a scroll, the Sabaeus Sinus, the adopted zero
point of Martian longitudes.

Fig. 2.

7 «>59 *<>.*> to m

»

w<

W-;

^

-

^

1 f I-

M U.t. i.

M 1 !•

]

■'^

H

V

■ '4-

.

'

1

'

o 1 1 '

t 1 1

\.v^

lit

,'

h.

S ■ , '1

.^

5?«

■^

^,>

4

^,

^^

^^

^/^

I'f*

^-

r -p.^

'71

fr ^'

' i t

'A^

8

^

^^

«t^

J

>>4

■^

*

"44 llj-i* ^

P^^

\

J

%'

#^

%i

r

1 4

\^1!}.

1 % +-^-

.\

■^'

^t^r^

■^^

""

.-r

"1 ■

^^

V5if

•*

''

,^

,f

^
1
^

■■-, ^^^

X

.

w

»,

~rr^ '

1 1

1

.

\\\ ^1

kWrl.

^

t 1

1

^

^-f^

^

^^

T

i

^^■a. , 4-

1

i

V

-

1 ' ^

■~

1

r"tT^

^M

«>'

■^

j y^

1 1

L^. ,

LJ

L_

u

i,

jS-2

0 I

0

■j.

fO>

swj:

iJJ

OA

o-y

^ 2

wa

^^

»i

WJ^.

iOJi

^i

oaw/s

Map of Kaiser, 1864.

Fig. 3.

Map of Dawes, by Proctor, 1867.

9. In map No. 2, Kaiser's, all these features can be followed,
from the eye with its eyebrow and the curve of its lower lid down
through the chain of seas back to the oculus again. The various
other dark markings on the map can be similarly identified.

1902.]

LOWELL — AREOGRAPHY.

229

lo. A very different set of phenomena stamps the advance made
in the second period. Over the bright portions of the map is now
drawn a network of fine lines. The dark patches remain as before.
These singular lines are what are known as the '^ canals " of Mars.

Fig. 4.

DegrescteSiongitucte

i7T-v<-~^;vy^T /^jp^n:^'^^^ T^TT-Tc-i^

bore a/e3

Map resume, by Flammarion, 1876.

Fig. !^.

Map of Schiaparelli, 1877.

II. The second period was the work of Schiaparelli. Of it are
here given four maps, all that he made on Mercator's projection.

230

LOWELL — AREOG K APH Y ,

[April 4,

After the opposition of 1883-84 he drew only maps with the pole
for^ centre, because of the tilt of the Martian axis which exposed
the northern regions and hid the southern ones.

Fig. 6.

T

y Ir

1

*> '

!

1

^

7. ao

1

FT

"TTT

)

~

■)■■' ■
1

^
^

1^

1

1

52-i

p

,'

ILlJ.li

-*

1

l|^

'

t

iiii

l+f

1
1

.y

■

! J 1 1

'

i

■J.

W

»'

^

1

:ii !;|i,i

i

i

,

^...r-. r-

it

""

"

" . ! "^^ ! 1 1

;.-'■

i

1

11^

*-

■t

/

E.e

..(.

.J.

"

4

1

'f

V

. "^0

• ''' M

t

1

J [

^_..jj--

" 1

.r^

',^

i^

Is

r "^

^
«*^

^

S

C^^

t

^^

L',;^

iif ('■*

Jlli h.___f ._'

..i,..

-H.

.«_

f

*1 1/^

.| /,

.^'

""■i

«i?

^

^^

^f^

t

f^

l;j

^'

1"?

^Jj^* '-

l^

m&-- ' ^

-

U
'

r

-r

r^ ,|.

*

^^^

_"

,'

>.

*1

■^

JF\

V-,J

u

/•

,..,^^'

'

i"

^''

,

^

*

'.

^

^

"

t

/^

-'1

hn

U

<■

""

/

/ '

-"

V

!^

.

■T

»-■•

.

^

* 1

^'•

,

•?

"

h

s

'11

'

;^

r

^r.

Cd.

'.

f

f

1

'

•-J 1

."

tly.

;

'

*

,y

.

"

*■'

'

IST-

»

E,

^s

»

s

<

^^

'^

«H

E

B «

«

E

,

^

'■

*h

>0 I

_

in

b1

» i

*^;

ri

SI

»:»

h

L

J3oJ<

^

—

rs

> '

r~i

n

?7

^

-»

^

i_

_

n

i_

r«

Map of Schiaparelli, 1879.
Fig. 7.

/^ «> /*» 7» (*) /»» MO lAJ rw iw 2*> zf» j<o V 2W i«> i'o *t> .nc .130 jw j

»

,

„ 2C ic 4

^50 « 7^ e^ 50 *<, w /!<> /5^ -.0 /« *o .70 -a, /» J*, i«

I T

w

, ,

,

f

,.

,, : 1

•

'i

M-,

rR

-■1;

■ K

^-l'"

A U |S T

R

Ar3/n

'".

;".'

:r'

1

' 1' 1 -'

\ \

^

1;

>-M

-*

*■•'

't

«.;

v[-

>'

" • 1-

.

^

1.-

: 1: i .

1

^«

oi/tb

"

J

1

<rs

••

-"

-

/

^

.'

.-«

-

..-'

;"'

n,ra

«/)io

"■,4

»«

fO

,.-

■>"

(\rt ^^

^j

"i^

>>,

■i-:'.

'^

'^

:x5

■■

MA

RE

E

.YT

<AE

JM

.-

"

;;:■

'-

„.■

-*i*

4

^,-

,':i.

V/IH

'C^

<

"■^

V

■•{Mt

.';"

-..

"■^t

f^'-.

''

^ ■

'■V,

1

'^"'

' *,

V

=tf'

'-.,

m

!^

..^

'_.,--

-■'

nin

^•ti-

.'■■

^•.

f

'\

/

:,

v-'

f'3

rt

H(^

W^

..,

i-

S

•>,

'^.

;;*

P^

»(/J

■-*

f

\''

,

--^

^

^

^

V>i

A

!?

1

' 1 H

*■«*

^

M,

W.

^

U

^

i'--U

^

m

•**,

'

VA/,

•r--'

*^'

\

■

>

>

tff-.

'\

..-.

^

j

^■

•«(-

S-

'$i

<

^-

:^

■M

-OJI

>-■/''

H

'"■i

^,

■: '\

7

' •i

■>

^-;

'^■-

;.-.,

\

•',

,^

,

"i;

bss

^

/'

^

\ ,

^*

/

>'

■¥

■^

;'

U'

'i

1

.y

'^T

k

''

\

'-■

r-:

^

~

"""

-*

M

k

^N

„(*

>•■

rJ

^v

/ >

\^

t

""'

^:'«

i'.i

1

/

Z

y

;

1

/

/

**^

^v

■-..

/I'

>

N

i^

*-"

74.

C

Af

.i

Ij

^4^'

i

5::?

^

^

,y^

■■:,

■'

^i

„:

■'..

':i

^

L.

— ,

"71

/

-f

^_

-•--■

;7

Ph

'?■'

'

vX

Sa

^

-N^

^
^t

S^

K

*nr

fi'

€

,H>*<

ill

■n^

/

1-*

-«-

-«•

^

r

< '1

5<;?

>*-,

^4

^

»^

r-'"-

,, ■

■■•-•i'

5f*i

*
%

'

•>

m:

^ys

m

.--

«!i

\

■■\

h

1 J

^

'"-

vio-jj

.■ a, >;«. ,» li. »i» «o ^>

«. 3C0 Jio lio X

KJ

i5^

>t«n

trio

ai Ji> -^ ^

' L /o a'c ^ ^ ,)o A

3-^

3-s!

"'^

» /7

» 1 ^^^ ^^

—

Map of Schiaparelli, 1881-1882.
The distinctive features of these maps are the '' canals." The

1902.]

LOWELL — AREOGRAPHY.

231

" canals " are objects as technical in character as they are in name,
being quite unlike any other planetary detail. They are narrow
lines of uniform width, of uniform direction and following usually

Fig. 8.

»€RIOie5

Map of Schiapareili, 1883-18S4.

Fig. 9.

Map of Lowell, 1894,

the arcs of great circles. Tenuity, regularity and intercommuni-
cation are the traits which make them sui generis. Such precision

232

LOWELL — A REOGR APH Y .

[April 4,

is of their essence. But the observations necessary to its apprecia-
tion are not easy. Probably even to-day not above a dozen persons
have seen the canals well enough to make their opinion on the sub-
ject of weight, but all who have done so agree in their dictum.

12. As with the first and second periods, so with the second and
third there was a transition state between the two. What Dawes
had done for the first gap, VV. H. Pickering and the Lick observers
did for the second. In 1892, at Arequipa, Pickering found irregu-
larly narrow markings in the midst of the then called seas, and the
Lick observers detected ^^ streaks " in the same regions. These
played much the same part, though in the case of the Lick ob-
servers much more, to subsequent work that the Dawes' markings
had to Schiaparelli's, so far as ''canal " detection is concerned.

13. For in 1894 Mr. Douglass at Flagstaff found that the irreg-

FiG. 10.

9 20

M «

0 f

t 60 70 fio 9

0 / c f'o n/1

/JO /«» /so /so

no /So /i

»ff f>

''p?

■0 220&.

f f

w

V

io ?

7O2fi0

2^ .1

00 Sib no Mo Mo 350 A

609 to

Jo

M

Si

JC

1

°x

q

'O

1

^ \

J

"

^

-^

-

-

y

y

/

1

r

OfJ

f^

k

HE

L ' A

^

.

\

^

"n,

^^ .

/4

\

' i'

•t^iS."

A-

T

V

:%,

1

\

k

\)

\

.0

/

■44./

-s

f7^

\ /

<^

~^

^

V

LI

iv-

^:X—

1

\

_ ,

^

10

a
10

1 ^

^.-f^]^

>

'1^

d

7

\

^

IX

^

1

P

-

i:

■"1 w

. i/kVK

A

\

1

'\^

X

^1

/

^d

f

f

^

l,f

k f 1

>

^

%\\/

H\

^

k\

"/^

\^

\

^

^

r

\1

\

%

\^

S/\

}

\

f

l\

h

/

y

\

J/

K /

^

^ ^

s>

^

IV

'/■

*?/

f^

In

1

j

\

\^

\

i-^

-1

>^

•-<^

-/

y.

^^

\

7

1

\

/^

A

\/

r

/

1

\

f

4<

i-

M/

^

^

CJ

nv

\)

/ J
/

XV

5

-^

\4

/

—I

^

/

\

V

■4

i.

V

--

—

^/

-ifl

Si

7

^

/'

y^

^

—

J

^

,

'A

4,

^

^^ —

~

^

/

^

/

*

'?

LOi

J. -M

30 «

OBS

0 1

iff

0 6

VAT

Off

0 e

0 3

0 10

0 1

V no

/iO /w

/SV ItO

/TO It
A/

to /i

<Afi

102c

18.

0 2

?6-

02.

7

%oz

ooz.

%0 2

iOi

0»>

zsojt

*7a

OS

\oX

WJi

VJ.

vn

10 I

m

20

Map of Lowell, 1896-97.

ular lines of Pickering and the streaks of the Lick observers were
foreshadowings of something much more peculiar. He found that
a system of lines of the startlingly regular character which gives to
the ''canals" their technical interest, overspread the whole of the
great southern dark areas. Thus the third period marks the detec-
tion of "canals" in the dark regions, and from that a complete
change in the character of the seas, already in part so ably detected
by Pickering. Furthermore, the network of each system showed

1902.]

LOWELL — AREOGRAPHY.

233

itself to be knotted with spots at its intersections. Many of the
spots in the bright regions were detected by Pickering in 1892.
Lastly, the two systems turned out to be connected together, the

Fig. II,

20 JO 'K) so so 70 So ?0 «» "O

Lowell Observatory

MAR5 I89Q-9.

Map of Lowell, 1898-99.

Fig. 12.

10 20 JO «o so 00' 70 80 so 100 110 120
LOWELL O&SERVATORY

• ISO I60 /TO leo 'SO 200 Z/O 220 Z30 2«0 I

MARS 1900-1

Map of Lowell, 1 900- 1 901.

one system running into the other and marked by nick-like points
in the coast line, thus making one united mesh of lines and knots
superposed over the whole surface of the planet.

234 LOWELL — AREOGRAPHY. [April 4,

14. The history of areography may be thus summed up :

1 840-1 8 76. Period of detection of large dark and light markings

on the surface of the planet.
1877-1892. Period of detection of *' canals" in the bright

regions.
1892-1902. Period of detection of '* canals " in the dark areas.

15. Three deductions follow an inspection of the whole series of
maps :

I. The fundamental agreement of the series.
This is evident at once, but can perhaps be made more so by
placing the later maps at a greater distance from the eye, upon
which the fainter markings take on the look they would wear were
the planet less well seen.

16. II. Evidence that the regularity of the '^canals" was not
due to predisposition on Schiaparelli's part, but was forced upon
him by the objects themselves.

Comparing his own maps on the subject, it appears that an evo-
lution took place in his perceptions. His first map, that of 1877,
represents the '^ canals " as straits, more or less irregular, running
up into the land. His next, made in 1879, depicts them narrower,
straighter and decidedly more peculiar. That of 1881-82 shows
them as fully developed geometrical designs, a character they never
afterward lose.

Now, the fact that his representations of the canals grew in regu-
larity as time went on, proves such character to have been no im-
putation on his part. Had he imagined it, he would have depicted
the canals so to start with. As it was, increasing familiarity com-
pelled him to recognize features which he had at first consciously
or unconsciously ignored. We have here, indeed, a record left by
himself of his own conversion to belief in the very qualities that
make the canals so difficult of credence.

17. III. Evidence of an evolution in the detection of the mark-
ings from simple to complex. A steady progression in the matter
of detail can be traced from its beginning to its end. And the
progression is in increasing order of difficulty. The large dark
patches are the easiest of detection, the Schiaparellian *^ canals"
in the bright regions the next so, the '^ canals " in the dark regions
the hardest. This is conclusively shown by the number of times
each class was seen in the many drawings made at Flagstaff. It is
here also evidenced by the way each map, while adding to, also
corroborates its predecessor.

1902.] DAVIS— SYSTEMATIC GEOGRAPHY. 235

SYSTEMATIC GEOGRAPHY.

BY W. M. DAVIS.
{Read April 3, 1902.)

1. Geography lacks System.

2. The Value of Systematic Geography.

3. The Content of Geography.

4. Physiography and Ontography.

5. Comparison of Geography with other Sciences.

6. Subdivisions of Physiography.

7. Classification of Land Forms.

8. Physiographic Classification involves Explanation.

9. Explanation involves Past History.

10. Value of Ideal Geographical Types.

11. Service of Deduction in Geography.

12. Contrasts of Biological and Physiographical Classifications.

13. Examples of Explanatory Description.

14. Distinction of Geography from Geology.

15. Dangers of Explanatory Description.

16. Framework of Physiographic Classification.

17. Complexity of Geography.

18. Relation of Physiography and Ontography.

19. Subdivisions of Ontography.

20. Regional Geography.

21. Conclusion.

I. Geography lacks System. — Geography has not as yet taken so
much advantage from a systematic classification of the facts with
which it is concerned as is the case with the biological sciences.
The botanist or the zoologist is greatly aided in observation and in
description by the effort to refer every organic individual to its
proper place in a comprehensive scheme of classification, whereby
its relationships and its contrasts are most concisely set forth ; and
if he is for a time puzzled by a new species or by a form of uncer-
tain position, he does not for a moment waver in his belief in the
value of the principles of classification, but draws encouragement
from the aid that it has already given him and perseveres until the
systematic position of the new or uncertain species is made clear.
The geographer on the other hand makes no such habitual use of

236

DAVIS — SYSTEMATIC GEOGRAPHY.

[April 3,

systematic methods. The classification that he uses is immature
and imperfect ; many classes of geographical problems are as yet
hardly classified at all. It is with the intention of showing the
need, the possibility and the value of systematic work in geography
that this essay is presented.

If a geographer should come upon such an item as one of the
narrow flood-plain scrolls sketched in Fig. i, he might treat it in
either one of two ways. He might describe it empirically as a
local item of earth form, unrelated to all other items ; or he might
more or less consciously refer it to some appropriate place in a
general scheme of geographical classification, whereby its origin
and relationships would be made manifest. The geographer at

Fig. I. A meandering valley with narrow flood-plain scrolls.

present generally attempts to pursue the second plan, as would be
indicated by the use of such a descriptive phrase as "a. narrow
strip of flood-plain " ; for the term " flood-plain " has a technical
meaning and suggests that the observed example belongs with
other more or less similar examples in a recognized class of geo-
graphical forms. If however we should question different geogra-
phers as to the relation of narrow flood-plain scrolls to flood plains
of other forms, and as the rest of the scheme of classification in
which flood plains form a single group, no approach to agreement
would be found ; for the venerable subject of geography has not
yet established a well-coordinated system of classification for the

1902.] DAVIS — SYSTEMATIC GEOGRAPHY. 237

facts with which it is concerned. The classifications commonly
employed are too often inconsistent, incomplete and immature —
inconsistent in their different parts even as to the larger principles
upon which their subdivisions are based ; incomplete in not
including nearly all the categories of facts which properly belong
under geography ; and immature in making too often only a small
advance over the method and terminology of school days. The
narrow flood-plain scrolls, such as are shown in the figure above,
and such as exist in remarkably perfect development in the valley
of the North Branch of the Susquehanna, would according to the
methods of geographical description and classification usually
current be given no sufficient statement as to their form, no
adequate explanation as to their origin, no appropriate discussion
as to their correlation with adjacent features-, and no systematic
treatment as to their share in constituting the physical environment
of their organic inhabitants. Yet the flood-plain scrolls deserve
due consideration in all these respects from any one who would
clearly portray the geography of their district. Lack of consid-
eration is not due to any serious difficulty that inheres in the
systematic treatment here suggested, but simply to the habitually
unsystematic character of geographical study.

It is the same with the organic items of the broad subject of
geography. A farm or a village, a thicket or an ant-hill, a city on
a bay or a road over a mountain range is too often mentioned as if
it were an isolated and ultimate fact, rather than as if it were a
member of a class, exhibiting the peculiar response of certain kinds
of organisms to their surroundings. Correlation between the
environed organism and the physical environment is coming to be
recognized as an essential part of geographical study, yet corre-
lation is not habitual in the treatment of the organic division of
the subject by those who would wish to be considered geographers ;
and as to classification of the correlations, there has as yet been
made hardly a beginning. Still it can hardly be doubted that all
organic responses are susceptible of a reasonably systematic grouping
in relation to one another, and that every example would be better
seen and appreciated if it were viewed in association with its
fellows.

2. T/ie Value of Systematic Geography.— It may be urged with
much confidence that fuller attention to such items as narrow flood-
plain scrolls, or to any one of the innumerable organic examples

238 DAVIS— SYSTEMATIC GEOGRAPHY. LAprilS,

that might be instanced, would be secured if geographers were
habituated to treat all such items as forming parts of a whole, and
to place every item in its proper place with respect to all others in
a well-arranged system of classification. It is sometimes the case
that the labors of the systematist are decried ; but it is only when
systematization is the master and not the servant of the investigator
that it merits condemnation. The orderly arrangement of the
events in the earth's long history is the goal of all geological study ;
for the facts of physical and structural geology must be dated in
terms of geological chronology if their true relation is to be
appreciated. So with geography : it stands to reason that any
logical scheme for the classification of all the elements that consti-
tute the content of geography would be of practical value in treating
the innumerable items with which the geographer is concerned.
The object of such an arrangement would not be to put facts away,
out of sight, but to expose them in orderly fashion so that they can
be most readily seen, to arrange them so that they shall illuminate
and be illuminated by their neighbors. A result of double value
would thus be gained. Every fact would be seen in logical relation
to its fellows, and its fellows would be seen in logical relation to it.
The attention of the geographer would thus be directed to a
broadened consideration of correlations, instead of being allowed
to limit itself to a narrow view of isolated and unrelated items.
The work of the observer in the field would be greatly aided by
the presence in his mind of an ideally full treatment for every kind
of item that he encounters ; unless indeed he has the good fortune
to come upon an item previously unknown, and in that case the
habit of systematic description already formed would come to his
assistance in the effort to gain a full understanding of the novel
element. There is no other means by which the general principles,
the underlying philosophy of geography, can be so clearly set forth
as by systematic classification.

3. The Content of Geography. — The first step in an attempt at
classification requires an understanding as to the content of geog-
raphy as a whole. Here at the very outset no general agreement can
be expected to-day ; but it is well to note that general agreement
will probably be reached by following the trend of the progress by
which geography has passed through two stages, now to enter upon
a third stage of development. A hundred years ago, geography
was the study of the earth and its inhabitants ; explanation then

1902.] DAVIS — SYSTEMATIC GEOGRAPHY. 239

made a very small part of description, and even a teleological
correlation of the organic and the inorganic divisions of the subject
had not been introduced as a well-defined characteristic of its
methods. With the progress of science during the nineteenth
century, explanation came to constitute a larger and larger share of
the descriptive chapters of geography ; and from the time of Ritter,
geography has been very commonly defined as the study of the
earth in relation to its inhabitants, the relationship being exposed
during the second stage of progress in the light of teleology, of
which abundant traces may be seen to this day.

The third stage of geographical progress is marked by the
introduction of two new principles during the last third of the
nineteenth century. It thus came to be recognized that explanation
must be systematically sought for in every department of the
subject ; for river courses as well as for winds and ocean currents ;
for moraines as well as for sand dunes ; and it is further recognized
that the relationship existing between the earth and its inhabitants
must be explained under the broad principles of evolution. The
earth with its lands and waters was not arranged for the convenience
of its inhabitants ; its inhabitants have had to learn, by more or
less conscious experiment, to live upon the earth as they found it.
As in so many other sciences, the evolutionary philosophy is of
enormous practical import in geography. If the earth has not been
expressly fitted to the convenience of its inhabitants, but if the
inhabitants have had gradually to fit themselves to their slowly
changing surroundings, how essential is it that we should study
those surroundings minutely, with all the intelligence that has been
awakened in the later days of man's history, in order to take the
best advantage of them ; how important is it that we should look
carefully into the real nature of things, so as to avoid an environ-
ment that involves a hopeless struggle against the forces of nature,
and to choose instead an environment in which the inexhaustible
forces of nature will work to our advantage. With the adoption of
the evolutionary pliilosophy, the content of geography can no longer
be defined as the relation of earth and man, but as the relation of
earth and life. The cleared roadway of a colony of pillaging ants
becomes as properly a subject of geographical study as a railroad
that connects centres of human population. Elementary geography
may still deal with the simplest salient facts and place man con-
spicuously in the foreground ; more advanced geography may include

PROC. AMER. PHILOS. SOC. XLI. 170. P. PRINTED SEPT. 23, 1902.

240 DAVIS — SYSTEMATIC GEOGRAPHY. [Aprils,

examples of greater complexity, but always selecting important rather
than trivial matters; but the investigator must study the trivial items
along with the greater ones, and all must be duly scrutinized,
described and classified.

4. Physiography and Ontography. — Let it then be here agreed
that the whole content of geography is the study of the relation of
the earth and its inhabitants. We thus see two prime divisions of
the subject. One includes all the elements of the physical environ-
ment of life \ the other all those responses which life has made to
its environment ; and in accordance with modern methods both of
these divisions should be treated under the explanatory principles
of evolution, inorganic and organic. It is the element of relation-
ship between the physical environment and the environed organism,
between physiography and ontography (to coin a word), that con-
stitutes the essential principle of geography to-day. Mature, fully
developed geography therefore involves the study of physiography
and ontography in their mutual relations. Treated otherwise, the
divisions of the subject lose coherence; they fall apart and are
gathered up by various other sciences. It is only when they are
bound together by the element of relationship that they constitute
a reasonably connected body of study, as well unified a science as
any other. In support of this principle, let us turn aside to note —
as others have done — how largely the principle of relationship is
serviceable in classifying the sciences.

5. Comparison of Geography with other Sciences. — All terrestrial
substances, inorganic and organic, the study of whose relationships
constitutes geography, are also the proper subject of study in relation
to composition by the chemist : rock, water, air and organisms are
all to be analyzed and classified as compounds. Again, all the
activities in the world of geography are the appropriate subject of
study in relation to energy by the physicist. Moreover, as fast as
geography, chemistry, physics and the other sciences advance,
their progress should be duly chronicled by the historian ; for it is
a sad mistake to imagine that the whole content* of history is only
the *' politics of the past." From the discovery of America by
Columbus to the discovery of a narrow flood-plain scroll on the
upper Susquehanna by an early backwoodsman ; from the migra-
tion of races across continents to the settlement of miller by a
waterfall, there is no discontinuity. The historian must regard all
such facts, great and small, as pertinent to his study of the sequence

I

1902,1 DAVIS — SYSTEMATIC GEOGRAPHY. 241

of human events, even though he can make explicit mention of the
greater ones only. The physicist must bring the behavior of a
river in making its flood plains and of a stone in falling from a cliff
under the domain of physical law, although he may not make
mention of every flowing stream and of every falling stone in his
systematic text-book. The chemist must discover all the kinds of
changes in composition caused by the weathering of rocks; he
must learn the composition of everything from the miller to his
flour and his millstone. It is therefore not in terms of the things
studied that a science can be defined, but only in terms of the
relationships involved in the study. The things with which the
geographer is concerned may also concern the physicist, the chemist
and the historian ; but as far as these things enter into the relation
between the earth and its inhabitants, they constitute the content
of geography.

It is particularly in relation to geology that geography has been
needlessly confused. Geology is essentially a historical study; it
is for the earth what history is for man. Geography, on the other
hand, is distinctly not a historical study; what is often called
historical geography might be much better called geographical
history. Geography considers the relationship of existing condi-
tions, inorganic and organic ; and as far as the dimension of time
enters into geographical methods, it is introduced not for the
purpose of studying the sequence of events that lead up to existing
phenomena — that belongs to geology or to history — but for the
purpose of better seeing the existing phenomena themselves, as will
be more fully shown below. Thus understood, geology and geog-
raphy are closely related ; it may be fairly said that geology cul-
minates in geography, and that all geology consists of a sequence
of paleogeographies. Surely, no geologist would dismiss the present
condition of the earth and its inhabitants from consideration as
constituting the last page in the recent chapter of historical geology.
Ocean navigation and cable laying, city growth and railroad build-
ing deserve a place in the geology of the recent period on exactly
the same ground that trilobite tracks and dinosaur prints belong in
the record of the past. Conversely, every geographer should con-
ceive all the geological history of the earth as involving a succession
of geographies, horizontally stratified with respect to a vertical time
line. All the processes of slow erosion, of volcanic eruption, of
rising and falling lands, of organic adaptations, formed elements of

242 DAVIS — SYSTEMATIC GEOGRAPHY. [AprUS,

these successive paleogeographies, just as the slow depression of the
Netherlands, the eruptions of Vesuvius and Pelee, the washing of
neglected fields, and the migrations of Europeans into the open
lands of America constitute elements of the geography of to-day.

The science of geography is therefore, like all other sciences,
concerned with the relationships of things which, when they enter
into relationships of other kinds, belong under other sciences, and
which are known to be pertinent to geography not by their own
qualities but by the relationship in which they are considered. It
is the classification of the elements of a subject thus constituted
that we have to consider.

It is not my purpose however to present here a detailed statement
of a classification, but rather to set forth the nature of a classification
which might, when expanded in a more technical geographical
publication, afford suitable categories for all kinds of geographical
facts. It will suffice therefore to indicate briefly the larger divisions
of the subject, and to pursue only one of these divisions, namely
the lands, into details.

6. Subdivisions of Physiography. — Geography as a whole has
already been shown to consist of two chief divisions, physiography
and ontography. Pliysiography has four chief subdivisions — the
earth as a globe, the atmosphere, the oceans and the lands. Let
us set aside for the present all but the last subdivision. The lands
should first be treated as a whole, and their contrast with the other
exterior parts of the earth considered. A notable contrast, of great
significance in its ontographical relations, is found between the
lands covered by the atmosphere, and the sea floors covered by the
oceans. The latter are monotonously cold, dark and quiet, as well
as remarkably uniform in shape and constitution j while the former
exhibit a variety of forms, such as high and low, smooth and rugged,
flat and steep, and experience a succession of changing conditions,
such as wet and dry, calm and windy, hot and cold. The general
weathering and washing of the lands, whereby their waste gees to
make the gain of the sea floors, results in their being scored by
many branching systems of valleys ; this highly specialized kind of
inequality being as significantly characteristic of the land surface
as is smoothness of the blanketed sea floors. There is nothing
new in all this, but geographers too generally fail to recognize these
general features of the lands as the determining physical environ-

1902.] DAVIS — SYSTEMATIC GEOGRAPHY. 243

ments in response to which many an organic condition has been
called forth.

The lands need subdivision into relatively small areas, for their
forms vary greatly from place to place. It has long been habitual
with geographers to describe and classify these forms empirically ;
but there is to-day a well-defined trend of opinion in favor of
rational, evolutionary or explanatory description and classification,
even though this more modern method has not yet found general
acceptance in practical exploration. An eclectic system of sub-
division, based on the suggestions of various writers, may be briefly
stated as follows ;

7. Classification of Land Forms. — Land forms are classed first as
to kind, according to their rocky structure ; thus one area may be
of horizontal structure; a second may consist of broken and
tilted blocks ; a third may have a domed structure ; a fourth may
be folded ; a fifth may be of volcanic origin, and so on.

Each kind of land form is then to be further classified according
to its stage in the cycle of erosion, to which it is introduced
by initial processes of deformation and (relative) upheaval, and
through which it progresses by the action of weathering and wash-
ing towards an ultimate goal of obliteration in a featureless plain
close to sea level, or in a smooth platform at an undetermined
depth beneath sea level. There is to-day abundant warrant for as-
serting that the sequence of developmental stages through this
destructive cycle of erosion is remarkably systematic, and that very
effective description of land forms may be given by characterizing
them simply as young, mature or old. This is therefore not a
matter of abstract theory, but of practical convenience to the field
geographer.

There is need of distinction between the inert land mass, offered
to erosion by the telluric forces of upheaval, and the physiographic
agencies by which the erosion is accomplished; the chief of the
latter being river systems. There is again need of discriminating
the forms assumed by the slow-moving waste of the land on the
way to the sea, from the inert land mass on the one hand, and
from the more active agencies of erosion on the other hand. With
respect to the active water streams, the land waste is relatively
inert and passive; but with respect to the inert underlying rock
mass, the waste may be treated as part of the superficial river
system. The latter treatment brings forth many interesting homo-

244 DAVIS — SYSTEMATIC GEOGRAPHY. [Aprils,

logics between water streams and waste streams, and from this
arises a simple terminology for waste forms by which the power
that words have of suggesting things is greatly increased.

It is still further necessary to distinguish between the several
kinds of agencies that are chiefly responsible for erosion, as de-
termined by climatic conditions. Thus far, a normal climate has
been assumed, of sufficient rainfall to fill all depressions to over-
flowing and of insufficient snowfall to form glaciers. On one side
of this norm there is the arid climate, where rainfall is small and
vegetation scanty, and where the wind therefore takes a significant
part in the work of shaping the land surface ; here the whole surface
swept by the wind corresponds to the bed of a water stream. On
the other side is the glacial climate, where precipitation is chiefly
in the form of snow and where drainage is chiefly in the form of
glaciers ; here the slender and nimble water streaais of the normal
climate are replaced by clumsy and sluggish ice streams, with the
result of greatly increasing the proportion of drainage channel to
drainage area.

Finally the border of the lands where they dip under the sea is
attacked by waves and currents and appropriately carved ; the cycle
of shore erosion being just as systematic and helpful as the cycle of
rain-and-river erosion. Each kind of land form, as determined by
its rocky structure, exhibits forms peculiar to itself and appropriate
to their stage of littoral erosion. Here, as in the normal and
special cycles of subaerial erosion, such terms as young, mature and
old are highly suggestive because of the systematic correlations of
various elemental forms that they imply.

This system of classification is at present by no means fully- de-
veloped, for it has been directly applied to but a relatively small
part of the lands ; yet it is so efficient where it has been applied
that there is every reason to expect that it will be all the more
efficient when it shall have been more widely applied and more fully
developed. Some of its essential features may now be given fuller
exposition.

8. Physiographic Classification involves Explanation. — Ex-
planation of origin is regarded as essential to a complete descrip-
tion in this evolutionary method of physiographical classification.
Not only must forms of simple and manifest origin, such as sand
dunes and stream gorges, be explained ; but all forms, difficult and
obscure as their origin may be, must if possible be brought under

1902.] DAVIS— SYSTEMATIC GEOGRAPHY. 245

explanatory treatment. Geographers have been slow to accept this
responsibility. True, they have long explained volcanoes by
eruption, because Eruptions have been witnessed ; yet they have
been habitually inattentive to the radial gorges by which extinct
volcanoes are scored. While gorges and water-gaps are still some-
times ascribed to fractures and floods, most geographers of a fair
degree of training explain them more wisely as the result of slow
sawing by the streams that flow through them ; yet most geograph-
ers are still accustomed to adducing a canyon and not a peneplain
in evidence of the magnitude of the work that can be done by rain
and rivers. There is therefore no more wholesome discipline for
the field geographer than to insist on the necessity of explaining
every part of the land form that comes under his observation. His
courage in this respect should be whole-souled rather than half-
hearted ; and whatever difficulties he may encounter, the success
already attained should strengthen his resolution to pursue his task
until complete success is reached.

9. Explanation involves Past History. — It is evident however
that an explanatory method of description involves the considera-
tion of the past history through which land forms have come for-
ward to their present estate ; and thus the subject of physiography
gains a strong savor of geological methods. Some geographers
seem to be disconcerted by this consequence of the explanatory
treatment. They appear to think that description through pro-
cesses of origin involves too serious a trespass on the field of
geology, and they therefore give explanation over to the geologist.
But there is nothing novel in the trespass of one science upon the
methods of another. The chemist is constantly employing physical
methods \ the astronomer is as constantly employing mathematical
and physical methods. Hence no apology is needed if the geog-
rapher employs geological methods whenever they serve his pur-
pose. The real point is that these geological methods serve a
geographical purpose ; the purpose, namely, of aiding the observa-
tion and description of land forms, for which the geographer is
primarily responsible. Any methods that aid this end are ap-
propriate. Much attention as the geographer may give to pro-
cess and time as involved in the sculpture of land forms,
his interest in these geological elements is not aroused simply
from the hope of tracing out the sequence of events that the
past contains, but from the expectation, well warranted by abun-

246 DAVIS — SYSTEMATIC GEOGEAPHY. [Aprils,

dant experience, of being better able to treat existing land forms
by a rational instead of by an empirical method. It is the geolo-
gist who studies the past history of the earth ai an end in itself; it
is his duty to unravel all the tangled skeins of earth history, how-
ever far back they may lead him. The geographer is concerned
with the past not as an end but as a means to an end ; and he cares
only for so much of it as shall serve his present needs.

lo. Value of Ideal Geographical Types. — The addition of expla-
nation to the responsibilities of the geographer brings with it the
need of idealizing actual forms into type forms, for it is chiefly in
terms of type forms that actual forms are in fine described. This is
also a discouragement to geographers of the more conservative
school, who have thought that geography was concerned only with
matters of fact, immediately observable. They must however
come to see that direct observation is entirely insufficient for the
geographer's needs, ^or the simple reason that if he recorded only
what he saw he would be overwhelmed with ungeneralized items.
He must generalize in order to bring the observable items within
the reach of descriptive terms, and as soon as he generalizes, the use
of idealized types is practically unavoidable. Such types have long
been in current use, but they have been too few and too empirically
defined for the best results. They need to be greatly increased in
number, and at the same time they must be correlated with struc-
ture, process and time ; for only by following the path of nature's
progress can we hope to store our minds with types that shall imi-
tate nature's products. It may be fairly urged that the larger the
store of types a geographer possesses, and the more careful and
numerous the comparisons with nature by which the types have
been rectified, the better progress can the geographer make in new
fields of observation.

II. Service of Deduction in Geography. — But the geographer
who adopts the explanatory methods in a whole-souled fashion will
find himself called upon not only to imagine a large series of
type forms ; he must also call into exercise his deductive faculties
and employ them to the fullest, if he would make the best progress
in the newer phases of his subject, however purely inductive he
has imagined it to be. In setting up a store of types, there is need
of deducing one type from another at every step ; and it may be
confidently urged that whoever hesitates to recognize this princi-
ple will fail of his effort to describe through explanation. But as a

1902.] DAVIS— SYSTEMATIC GEOGRAPHY. 247

matter of fact, geography has some time been more deductive than
geographers have supposed it to be ; and the newer phase of the
science is not characterized so much by introducing deduction for
the first time, as by insisting on its whole-souled acceptance as an
essential process in geographical research.

It is only by giving the fullest exercise to the faculties of imagi-
nation and deduction that the cycle of erosion becomes serviceable.
Here the geographer who hesitates is lost. Not only should the
ideal cycle be followed in imagination through all its gradual
changes on a large variety of structures, but the special cycles of
arid and of frigid climate must be similarly treated ; and then each
of these cycles must be broken up by earth-movements into partial
cycles and episodes. It is only in this way that the scheme of the
cycle gains a serviceable elasticity ; and it is highly significant that
among those geographers who find the conception of the cycle un-
fruitful is one who has, with more candid indication of his unex-
ercised imagination than he may have supposed, likened it to a
*' strait jacket."

Those who have not attained some fluency in the verbal transla-
tion.of the various stages of normal and special, simple and inter-
rupted cycles can have little understanding of the practical aid that
is derived from this method of description. The empirical geog-
rapher, unsupplied with a store of carefully imagined and well-
defined type forms, sees only what is before him in the field — if
indeed he sees so much as that. The geographer who calls the
faculties of imagination and deduction to his aid, draws from
his mental store one type after another in the effort of matching the
explained ideal forms with the actually observed forms. Thus com-
paring the partial view of the landscape, as seen by his outer sight,
with the complete view of the type as seen by his inner sight, he
determines, with great saving of time and effort, just where his next
observations should be made in order to decide whether the ideal
type he has provisionally selected fully agrees with the actual land-
scape before him. When the proper type is thus selected, the ob-
served landscape is concisely and effectively named in accordance
with it ; and description is thus greatly abbreviated. It goes without
saying that this relatively advanced stage of investigation is not to
be reached hastily ; that abundant and elaborate description of actual
and of type forms in empirical terms, without a trace of explana-
tion, should be demanded of the tyro who aspires to become an

248 DAVIS— SYSTEMATIC GEOGRAPHY. [Aprils,

expert ; for in no other way can proper training in the use of types
be secured.

12. Contrasts of Biological and Physiographical Classifications. —
It may be worth while to note explicitly that there is little resem-
blance between the basis of the physiographic classification of land
forms, here outlined, and the phylogenetic classification of organic
forms now in vogue. In the latter case resemblance is inherited by
actual derivation from common ancestors ; and if similar forms
arise as a result of similar environment, independent of relationship
by descent, this only serves to emphasize the rule by pointing to
the exception from it. In the former case, resemblance is due to
repetition of physical conditions, and inheritance naturally has no
part to play. Similar structures, acted on by similar processes, at
similar rates for similar times will have similar forms ; but as struc-
tures, processes, rates and times are all variable, it is not to be ex-
pected that identical forms should be developed. All the more
need, therefore, of developing a method of rational generalization,
whereby the essential features of a landscape may be seized upon as
the basis for its description, while the insignificant elements of a
landscape may be set aside. It should further be noted that while
hybridization is of very limited range among organic forms, there
is no limit to it in land forms. All sorts of structures are combined
in all sorts of ways and acted on by all sorts of processes at various
rates for different periods. This is indeed one of the chief causes
of difficulty in physiographical description. Without free crossing
of species, the variety of landscape would be much lessened. Phy-
siography would then be easier and less interesting than it is now.

13. Examples of Explanatory Description. — The flood-plain
scrolls illustrated in Fig. 1 may be instanced as examples that come
very easily under the explanatory description of land forms. It
has been ascertained with a high degree of certainty that a winding
river, revived to renewed downward corrosion by the uplift of its
basin, will increase the radius of curvature of its bends and push
every bend down-valley while it is cutting down to grade with re-
spect to its new baselevel. If the river had a meandering course
when the uplift occurred, the increased width of the meander belt
will be shown by the gentle slope of the spurs that enter each
meander, as well as by the abrupt bluffs by which each meander is
enclosed ; while the down-valley advance of the meander system
will be shown by the extension of the enclosing bluff with decreas-

1902.J DAVIS— SYSTEMATIC GEOGKAPHY. 249

ing height along the up-valley side of each spur, so that the spurs
have an unsymmetrical cross section as shown in the figure. No
flood-plain is developed before grade is reached ; but as soon as
this delicately organized condition is attained, further valley deep-
ening is practically stopped, although the meander belt continues
to widen, and the curves continue to advance slowly down-valley.
As a result, narrow strips of flood plain in scroll-like patterns must
be developed ; a scroll will begin by lapping around the end of a
spur ; it will then follow along the gentle slope on the down-valley
side of the spur and end with reversed curvature shortly after reach-
ing the next enclosing bluff. As time goes on, the spurs are more
consumed and the scrolls are widened. The spurs may be trimmed
into sharp cusps, and later reduced to blunt cusps, and then the
scrolls must have widened into shield-like patterns. As the river
swings more and more freely and opens a valley floor of greater
breadth than the meander belt, the separate flood-plain shields are
joined ; further than this we need not trace them here.

Now it is not conceivable that geographical items as systematic as
these flood-plain scrolls should be treated empirically, after their
origin and their development has once been made out. It suffices
in describing the meandering part of the valley of the North
Branch of the Susquehanna to say that it has reached the stage of
narrow flood-plain scrolls ; for on saying this, the sloping spurs and
the enclosing bluff's at once come to mind as elements of form that
are necessarily correlated with the flood-plain scrolls. The me-
andering valley of the Ranee in Brittany shows a succession of nar-
row scrolls in the most orderly arrangement. The valley of the
lower Seine by Rouen possesses broader scrolls ; nearer the river
mouth, where the tides run strong, the spurs are greatly reduced.
The curving valley of the Evenlode, a diminished headwater of the
Thames system in the Cotteswold hills of England, has sharply
trimmed spurs which prove that the Evenlode was not beheaded
until a somewhat advanced stage of valley development was
reached. The diminished stream now straggles irregularly about
the open valley floor. The valley of the Lot in southwestern
France may be described as having nearly reached the stage of con-
sumed spurs in one cycle, when a moderate elevation introduced a
new cycle in which the stage of wide scrolls is now reached. The
essential features of the valley are thus concisely indicated, al-
though many individual variations from the suggested type are to
be found.

250 DAVIS — SYSTEMATIC GEOGRAPHY. [Aprils,

14. Distinction of Geography fro7n Geology. — If the explanatory
method is adopted as appropriate for the physiographic description
of meandering valleys in the narrow scroll stage, the same method
should be adopted for all other stages of valley carving and for all
other land forms as well. The orderly action of natural processes
through a portion of past time is implied in such a phrase as ^* the
narrow scroll stage," and it is similarly implied in saying that
the AUeghenies of Pennsylvania are of corrugated mountainous
structure, essentially baseleveled in a former cycle ; then broadly
elevated and thus standing long enough for the weaker strata to be
etched out as lowlands, leaving the harder strata to stand up as
even-crested ridges \ and then again moderately elevated long
enough ago for the valley lowlands to have now reached a subma-
ture stage of dissection. The descriptions of the Susquehanna val-
ley and of the Pennsylvania AUeghenies differ in the quantity of
past process and of past time involved ; but such a difference is
only of degree, not of kind. If all the stages of development
through which the Pennsylvania AUeghenies have passed are traced
out for their own sake, as much attention being given to one
stage as to another, then the study is truly geological. If the
changes of the past are introduced only in so far as they illuminate
the present, and with no other object than to secure such illumina-
tion, then the study is geographical. It would be as much a mis-
take to regard such study as geological as it would be to say that a
chemist is studying physics when he uses a balance to weigh a pre-
cipitate, or that he is studying mathematics when he calculates
atomic weights. He is truly enough for the time employing physical
and mathematical methods, but he is studying chemistry. It
would be no more just to regard the explanatory description of
flood plains as belonging under geology because it has to deal with
past time, than to treat it as belonging to the study of physics be-
cause it involves the application of physical principles in the flow of
a stream, in the corrosion of its bed and banks and in the trans-
portation and deposition of detritus ; and surely it would be no more
appropriate to regard such a study of flood plains as a part of
physics than it would be to take away the spectroscopic study of
the stars from astronomy.

15. Dangers of Explanatory Description. — It is sometimes ob-
jected that the explanatory method of description is dangerous, be-
cause the observer who seeks to add explanation to observation

1902.1 DAVIS — SYSTEMATIC GEOGRAPHY. 1 251

may be led to think that he sees things that do not exist. There is
certainly some danger of this kind, but it can be greatly lessened
by good training — without which the explanatory method is indeed
valueless — and in compensation for the little danger that remains,
there is the great increase in the thoroughness and accuracy of ob-
servation that results from bringing forward the various idealized
types to be confronted with the facts in the field. If doubt finally
remains, it may be expressed by the phrase, "as if" : — The Sus-
quehanna valley looks as if it were in the stage of narrow flood-
plain scrolls. The initiated reader is thus concisely put in posses-
sion of the most probable conclusion as well as of the doubt that
accompanies it. As a matter of practical experience, it may be
urged that the gain from attempted explanation far outweighs its
danger ; and in illustration of this conclusion reference may be
made to the curious case of the Connedogwinet, a branch of the
main Susquehanna opposite Harrisburg. The branch has an un-
usually serpentine course, and the tangents between its curves are
of extraordinary length. On visiting it in the spring of 1901, I
expected that it would show normal nanow flood-plain scrolls ; but as
a matter of fact, its scrolls were found to be distinctly abnormal, in-
asmuch as they are nearly all on the down-valley side of the tan-
gents. Truly, this is not a matter of great geographical conse-
quence ; the farmers would cultivate the scrolls, on whichever side
of the tangents they might lie ; but it is certainly of some physio-
graphical interest to note their abnormal position, because it con-
tradicts a generalization that is well supported by the repeated oc-
currence of examples in various parts of the world ; a generalization
that is fully explained by simple processes, perfectly accordant with
the laws of stream flow. No explanation of the abnormal situation
of the Connedogwinet scrolls has yet been suggested ; indeed, as
far as I have read, no mention of them as abnormal features has
ever been made. Their peculiar arrangement seems never to have
been noticed until it was brought out as an exception to the rule of
flood-plain development. This example may therefore be taken to
show that, far from there being serious danger of seeing imaginary
facts by the light of theoretical explanation, a well defined con-
ception of ideal types is a positive aid in correct observation.

16. Framework of Physiographic Classification. — If the explana-
tory method of physiographical description were adopted, it would
result in the construction of a mental framework on which all

252 DAVIS — SYSTEMATIC GEOGRAPHY. [Aprils,

imaginable types would find their appropriate place in a systematic
arrangement. Each of these types might be considered to be the
label on a pigeon-hole ; and actual examples would be placed in
their appropriate pigeon-hole as fast as they were collected. The
compartments designed for common examples would soon be filled ;
while others might long remain empty. Such a plan as this greatly
promotes systematic observation, for the very fact that a certain
pigeon-hole contains no actual form corresponding to its idealized
type urges the observer to search for the missing example in dis-
tricts where its occurrence is most probable. Revision of an
idealized type would naturally be made whenever an example re-
sembling it was found ; for however deductive the method of de-
veloping types may seem when here stated in the abstract, the
actual progress of this sort of study involves repeated oscillations
between induction and deduction, in which each process aids the
other. The types are therefore not to be thought of as fancy
pictures, unreasonably constructed by an ungoverned imagination
and arbitrarily fixed by obstinate deduction. They should be the
very best imitations of nature that the well-trained mind can con-
struct, and they should be held subject to constant revision and
correction as fast as observation is extended.

The conservative geographer will hesitate to construct a frame-
work in which his types shall be more numerous than his examples.
Indeed it sounds at first rather presumptuous to say that the variety
of idealized types can exceed the variety of nature ; but there is no
doubt that it can. The earth is after all not so very large ; and
when all the examples of physiographic items that it contains shall
have been studied out and systematically arranged, it will be easy
enough to construct imaginary types that belong between two
actual examples. Even if all the items that have existed in all
the paleogeographies of the earth's history were brought into
systematic arrangement, it may be doubted whether they would fill
all the pigeon-holes of a well-imagined framework, so easily can the
imagination conceive of a type intermediate with respect to any
two neighboring examples.

It is therefore plainly a profitable exercise for the systematic
geographer to elaborate his systematic framework as far as possible ;
to increase the number of its little compartments, each bearing an
appropriate label ; to arrange all the compartments in as systematic
an order as he can develop ; and to devise every means — verbal.

1902.] DAVIS— SYSTEMATIC GEOGRAPHY. 258

graphic or mechanical — by which the framework shall always be at
his service for practical use. Its value will increase with every
step that is taken towards a vivid realization of its imaginary con-
tents. It may seem cumbersome as long as it is unfamiliar ; but
when it is familiarly known it becomes an indispensable aid in
practical work.

17. Complexity of Geography. — The whole current of thought
changes when the ontographic half of geography is taken up. The
training that is here necessary must be gained largely through
biological study, while the training for the study of the earth as a
globe is associated with astronomy, for the atmosphere with physics,
for the oceans with hydrostatics and hydrodynamics, for the lands
with geology. Whether this diversity of discipline is an advantage
or not need not be answered ; it is certainly a necessity. It is
perhaps true that geography has, by reason of its many-sidedness,
a more complex content than any other science ; but if so it merely
occupies a rank that would be otherwise held by some other subject ;
and certainly there is no impropriety in standing at either end of
the list in this respect. Astronomy ranks well among the sciences,
yet it now calls for mathematical, physical and chemical discipline ;
and if the change of color on the face of Mars follows his seasons
it may be necessary to add a biological discipline as well.

Some have feared that the various parts of geography might fall
asunder from their diversity of content and of discipline. So they
undoubtedly would, but for the bond of relationship that holds
them so strongly together. It may perhaps come to be wise for
the geographer to follow the example of those engaged in other
sciences and limit his attention to one part of his subject. Just as
there are mathematical and physical astronomers, inorganic, organic
and physical chemists, students of ancient, modern and many other
groups of languages, so there may advisedly be physiographers and
ontographers, instead of geographers ; but all this is of secondary
importance. Geography certainly has its inorganic and its organic
side, and both must be understood by any one who would claim to
be a thoroughly trained geographer, versed in the relationships by
which the physiographic and ontographic sides of the subject are
held together. The reason that so few persons can to-day rightly
claim such standing is not so much because there is any inherent
difficulty in the subject on account of its breadth and its complexity.

254 DAVIS — SYSTEMATIC GEOGRAPHY. [Aprils,

as because the subject is not maturely developed ; but this is an
aspect of the question that I shall elsewhere consider.^

1 8. Relaiio7i of Physiography and Ontography. — Unlike physi-
ography, which has been recognized as an essential constituent of
geography for many years past, ontography has to-day hardly gained
an established position. It is best represented in Ratzel's Anthro-
pogeographie, but this subdivision of the science is concerned only
with the human element, and that is manifestly but a part of the
total content of ontography. It is approached in ecology, but
none of the many definitions of that term cover all that is here
intended, for ontography is meant to include all the responses of
organic forms to their physical environment, whether in physio-
logical structure, in individual behavior, or in racial habits.
Whether there is need of this new term, whether it will survive or
not, it serves a present purpose in bringing clearly forward the
organic half of the geographical whole.

The subdivision and classification of ontography has not yet been
well accomplished. Before it can be well done, there must be
much searching; but we may look forward to a time when all
ontographic items shall be arranged on an ontographic framework,
in which every compartment shall have for a label what biologists
might call a type response. I am persuaded from much profitable
experience with the physiographical framework that a corresponding
advantage will come from the construction and familiar use of a
similiar framework for ontography. Still more : the two frame-
works might be brought face to face, and lines might then be drawn
between them, connecting cause with effect, effect with cause. If
then a plane were passed secant to all these lines of relationship, all
the content of geography might be projected along the lines upon
it. If the plane were placed near the physiographic framework,
there would be groups of points, where numerous radiating lines
departing from some dominant physiographic control pass through
the plane on their way to various ontographic effects. If the plane
were passed near the ontographic framework, the grouping of
numerous points of intersection would serve to indicate those
organic forms which respond to many physiographic controls, while
isolated points would indicate forms that respond to few. Accord-

1 National Association for the Scientific Study of Education, Proceedings of
the Minneapolis Meeting, July, 1902.

lW-2.] DAVIS — SYSTEMATIC GEOGRAPHY. 255

ing therefore as the geographical plane is placed nearer to one
framework or to the other, the presentation of the total subject
might be made primarily physiographic and secondarily ontographic,
or the reverse.

19. Subdivisions of Ontography. — It is not an ontographic classi-
fication, but the nature of such a classification that can here be set
forth to best advantage. There should be two chief subdivisions ;
the first includes those responses that were initiated ages ago and
maintained by inheritance till to-day because their controls are
persistent; the second, those of relatively recent origin. Further
subdivision might be made in accordance with the standard classi-
fications of botany and zoology, in which the responses of all kinds
of plants and animals to physiographic controls would be taken up
in their natural order. But in view of the repetition of similar
responses in many different classes of organisms, it will be here
more convenient to follow a physiographic order in the ontographi-
cal classification. Examples of long-inherited responses will be
mentioned first.

As inhabitants of an earth whose mass is very much greater than
that of all its organic population, plants and animals very generally
show a response to the action of gravity in their attitudes as well
as in their structure. As inhabitants of an earth whose opaque
surface is illuminated from without, the distribution of color in
plants and animals is often closely associated with the response to
the downward action of gravity. As occupants of an earth whose
surface is nearly globular, plants and animals have been allowed a
much wider migration than would be possible for the occupants of
a very irregular body, on whose surface gravity would vary greatly.
None of these responses are doubtful as to origin or difficult as to
comprehension ; they ought to be introduced in the elementary
study of the earth as the globe ; and their almost universal omission
from that chapter of geography affords immediate illustration of the
little thoroughness with which the subject is treated. Perhaps
these matters have been omitted because they are regarded as of
less importance than the names of the branches of Siberian rivers ;
but if so, a very singular standard for the measure of importance
has been accepted. Many other long-inherited responses to the
physical features of the earth as a globe might be instanced, but
space is lacking for their presentation.

One of the most universal of all organic habits, that of breathing

PROC. AMER. PHILOS. SOC. XLI. 170. Q. PRINTED SEPT. 24, 1902.

256 DAVIS — SYSTEMATIC GEOGRAPHY. [Aprils,

free oxygen, must be regarded as the long-inherited response to
the presence of oxygen in the free state, whether mixed with other
gases in the atmosphere or dissolved in the waters of the ocean.
Organs of flight, to-day characteristic of many insects and birds,
are extraordinary devices for movement through the air ; this may
seem a valueless truism to some, but it must be explicitly stated if
the ontographic framework is to be thoroughly constructed, and if
conscious attention is to be aroused to it. Vocal organs are
responses to the extreme eclasticity of the air ; human speech may
be reckoned among the responses of modern acquisition under
this class. The adoption of blue as one of the primary colors of the
spectrum is a modern response to the color of the sky ; a physio-
graphic fact that has waited long for its ontographic mate. Pollen
grains, spores and innumerable microscopic organisms of great im-
portance in the economy of the nature, exhibit in their minuteness
a response to the small sustaining power of the winds that bear
them about. If climate were here considered as well as these
simpler physiographic features of the atmosphere, the number of
ontographic responses in this class would be greatly increased.

The greater buoyancy of water than of air has a notable response
in the absence of feet among most of the swimming animals of the
ocean. The flying animals of the atmosphere, on the other hand,
always have legs to sustain them when they alight on the ground
from the little sustaining air. The same contrast between water
and air must account for the much greater size of the floating
inhabitants of the ocean than of the blown-about organisms of the
atmosphere. While the more opaque animals of the sea usually
have a darker dorsal and a lighter ventral surface, many of the
floating animals find relative safety in imitating the transparence of
the waters in which they float. It is the monotony of the cold,
smooth, dark and quiet ocean bottom that has doomed it to be the
home of the less intelligent organisms, while the variety of the
lands has promoted the development of the most remarkable in-
stincts and the highest intelligence among their inhabitants.

The separation of the lands into several large continental
masses has led to the division of mankind into races ; and closely
associated with this division into races go many peculiarities of
government, religion and degree of civilization. All this is most
intimately connected with that phase of ontography commonly
called political geography ; and yet so arbitrary and irrational is

1902.] DAVIS — SYSTEMATIC GEOGRAPHY. 257

the traditional classification of geographical topics that the division
of mankind into races is commonly taught under physical geography.
The races may be fairly enough introduced there as illustrations of
ontographical consequences following from physiographical con-
trols; but to regard them as essentially physiographic topics shows
a regrettable failure to recognize the essential quality of geographi-
cal discipline.

The simple physiographic factor of distance is of great impor-
tance. It involves the separation of the people of a race into
many families, and thus is a determining cause of difference of
language and of many other habits. The unevenness of surface
exhibited in mountain ranges is of small measure in comparison to
the dimensions of the earth, and yet it suffices to make movement
so difficult that the occupants of one valley may have a distinctly
different dialect from those in a neighboring valley. How circum-
scribed would have been the migrations of the earth's inhabitants
if the height of mountain ranges were a large part of the earth's
radius ! The sheet of loose rock waste by which the lands are so
largely covered not only supports the growth of plants, but has
been adopted as a home by many kinds of animals ; and according
as the waste is a coarse talus lying on the steep slopes of a young
mountain side, or a fine, deep soil blanketing a peneplain, its oc-
cupants are of different kinds. Instances of this kind might be ex-
tended without number.

Examples of modern responses to physiographic controls are
best found in those new-fashioned characteristics of mankind that
are seen in sites of settlement, routes of travel, and in the develop-
ment of trades and of commerce.

Settlements in deserts offer particularly striking illustrations of
the dependence of population on water supply. Settlements on
rivers are largely determined by head of tide, by falls, and by
fords. Settlements on coasts are influenced by protection from the
open sea, and by ease of access from sea and land. The routes of
trade and commerce are guided by physiographic factors literally
at every turn. Straight roads are laid out on plains, but winding
valleys are commonly followed in regions of strong relief; tunnels
are driven through mountains ; short-cuts are made through isth-
muses. Here as before, illustrations are endless; yet abundant as
they may be, they have not yet been well classified. At the present
day, ontography is less developed than physiography.

2o8 DAVIS — SYSTEMATIC GEOGRAPHY. [Aprils,

Many examples are individual rather than generic. It was the
shoals remaining where morainic islands once arose that turned the
Mayflower northward from a course that might have led her south
of Cape Cod to New Amsterdam ; it was the greater height of the
mainland where the moraines of Manomet were piled upon it that
led the Pilgrims from their first landing at Provincetown to the quiet
harbor of Plymouth. The varied course of human history affords
innumerable examples of this kind. It would be profitable to
make a long list of them, to classify the items thus gathered, and
to select the best examples of various classes for presentation as
types. A geographer who was well informed regarding such types
would undoubtedly be more observant in his travels than many
travelers are to-day. He would be continually asking questions
and finding answers where he is now silent.

20. Regional Geography. — It is in the prevalently unsuccessful
treatment of regional geography that the undeveloped condition of
systematic geography is made most apparent. It is well recog-
nized in the organic sciences that only after a general understand-
ing of systematic botany or systematic zoology is gained can a
profitable attempt be made to describe the flora or the fauna of a
limited district. The same principle undoubtedly obtains in geog-
raphy; yet nothing is more common in geographical literature than
an attempt to treat the geography of a certain region before any
thorough system of geography has been agreed upon. This error
is in the way of being corrected, but it is still a prevalent error. In
texts on physical geography, for example, it is still common to find
an attempt made to describe the physiographic features of the
several continents before any sufficient understanding has been
gained as to the nature of physiographic features. The year of
study commonly allotted to this subject in the schools is none too
long for a sound systematic course, and by no means long enough
for the addition of a regional course as well. Systematic phy-
siography may be vivified by the introduction of many well-
selected examples from various parts of the world, but there is not
time in a single year to present a substantial account of the con-
tinents or even of a single continent in addition to the systematic
account of the whole subject.

21. Conclusion. — The practical conclusion of all this is that it is
the nature of geography as a whole, rather than the accumulation of
unassorted and uncorrected items, that demands the attention of

1902.] MINUTES. 259

geographers. Careful analysis and arrangement of the content of
the subject is as greatly needed as the exploration of unknown
lands. It must be remembered, however, that the object of analysis
and classification is to render practical aid in the understanding of
geographical items, old and new. There should be no hindrance
placed in the way of the active pathfinders who seek to enter un-
known lands; but there should be every encouragement given to
those who believe that some of the unknown elements of geography
may be discovered without going far from home.

Stated Meeting, April 18, 1902.

President AYistae in the Chair.

Present, 16 members.

Letters accepting membership were read from

Dr. Jolin A. Brashear, Allegheny, Pa.

Dr. Andrew Carnegie, New York.

Prof. William B. Clark, Baltimore.

Dr. Hermann Collitz, Bryn Mawr. Pa.

President Arthur T. Hadley, New Haven.

Prof. George E. Hale, Williams Bay, Wis.

Dr. C. Hart Merriaixi, Washington.

Prof. Theodore W. Richards, Cambridge.

Prof. Felix E. Schelling, Philadelphia.

Prof. Robert Henry Thurston, Ithaca, N. Y.

Prof. Robert S. Woodward, New York.
Mr. Thomas Willing Balch presented, on behalf of his
brother and himself, the MS. account book of the first
" Philadelphia Assembly," 1748, and read a note thereon.

Mr. Alden Sampson read a paper on '' The Ruins of Pal-
myra, with a Brief Consideration of the Ancient Estate of
that City."

260 BALCH — THE FIRST "ASSEMBLY ACCOUNT." [April 18,

THE FIRST '^\SSEMBLY ACCOUNT"
—PHILADELPHIA, 1748.

BY THOMAS WILLING BALCH.
{Read April IS, ld02,)

John Swift, a Manager and the Treasurer of the First Assemblies,
was the eldest child of John Swift and Mary White, his wife, of
London. He was born in 1720. He went to England together
with his younger brother, Joseph Swift, where they visited their
uncle, John White, of Croydon, now a part of London. Return-
ing to America in 1747 he was, as a merchant of Philadelphia, very
successful. He was a member of the Common Council from 1757
to 1776, and Collector of the Port of Philadelphia from 1762 to
1772. During the latter part of his life he lived at ''Croydon
Lodge," in Bucks County, where he died in 1802 and was buried
in Christ Church burying ground, Philadelphia, January 14, 1803.

The Philadelphia Assemblies began in 1748, only five years after
the organization of this Society, They are, I believe, the oldest
dancing organization in the country, their only serious rival, the
Saint-Cecelia Society, of Charleston, dating from several years
later. During the winter of 1748, six Assemblies were given under
the management of four Directors : Lynford Lardner, John Inglis,
John Wallace and John Swift. There is a tradition in the Swift
family that the first meeting at which the Assemblies originated
was held at John Swift's house. There were fifty-nine subscribers
in all, and as an invitation was extended to the families of every
head of a family who subscribed, probably some two hundred
persons were eligible to attend the dances. The subscription was
two pounds sterling. In 1879, ^^' Charles Swift Riche Hildeburn,
a descendant of John Swift, the Manager and the Treasurer, and
Mr. Richard Penn Lardner, a descendant of the first Lynford
Lardner, the Manager, presented to the Historical Society of Penn-
sylvania two documents intimately connected with the First Assem-
blies. Mr. Hildeburn gave the rules to govern the dances, and
Mr. Lardner gave the list of the original subscribers.

A third manuscript relic of those gay festivities is the account
book kept by John Swift. It descended through the lineal descend-
ants of John Swift to that learned and accomplished antiquarian
and bibliophile, the late Mr. Hildeburn, a member of this Society.

1902.] BALCH — THE FIR.-T "ASSEMBLY ACCOUNT." 261

It is a small, thin book, and Mr. Swift used it originally for some
of his own accounts, and for some land transactions for his younger
brother Joseph. On one cover he wrote: ''Account book 1746."
When the Assemblies were instituted and his fellow-managers chose
him the Treasurer, he turned to the other end of this little book and
there kept the "Assembly Account." Owing to age and neglect
— indeed, had it not been for the keen antiquarian eye of Charles
Hildeburn, it would probably long ago have gone to the paper mill
— the Account Book is much worn and somewhat injured. But
now it has been treated by an expert, and every sheet covered with
silk so as to guard it against any future weathering of -time.

An examination of the account shows that the six Assemblies of
1748 were far less costly than the two large balls that are now given
annually at the Academy of Music ; or, for that matter, even the
three balls that were held each season about fifty years ago at
Musical Fund Hall. In one respect, however, those old worthies
were not behind the present generation, for taking all things in
proportion they provided rum liberally. The record kept by Mr
Swift is somewhat injured, so that it is impossible to state exactly
how much he disbursed, but the whole cost of the six dances seems
to have amounted to a little more than ;£i^o. As there were only
fifty-nine subscribers at forty shillings each, which gave a total of
;^ii8, the Managers doubtless, as so often happens nowadays in
all sorts of social and philanthropic undertakings, had to put their
hands in their own pockets. A few extracts from the expenses show
the modest and simple character of the entertainments :

" pd Mary Dicas for China
& Candles as per
Acct ending y^^ March

pd Mr. Inglis for Sundries

1 6th [March] pd for Bisket

pd Musick

pd Diana

And again :

" pd for 2 Gallons Spirit
pd Sharper 5 nights

attendance
pd. Greek for attendance
pd. Mr Inglis for rent

£. s.

8

15

6

—

19

—

—

9

6

I

10

—

—

2

6

£•

s.
15

^.

_

18

9

—

7

6

262 BALCH— THE FIRST "ASSEMBLY ACCOUNT." [April 18,

The Managers of the first Assemblies had to pay a tax, both to
the city and to the county, as may be seen by turning to Mr. Swift's
Account Book ; but as the book is there somewhat torn, it is impos-
sible to know how much.

From the time the Assemblies were first organized to the present,
they have continued with pretty general regularity except when
interrupted by war or other serious drawbacks. They have been
held in various places, and the names on the list of subscribers
have changed very much. Many of the Quaker families — such as

the Kawles, the Norrises, the Logans, the Whartons, etc that

owing to their faith kept aloof at first from such gay and frivolous
pastimes, later joined in with the Shippens, the Willings, the Swifts,
the McCalls, the Hopkinsons, the Lardners, the Francises, the
Bonds, the Lawrences, and others who were among the first list of
subscribers. And some of the old names a/ds have died out.

It is not inappropriate on this occasion to recall to mind a few of
the ladies who took part in those entertainments in the latter half
of the eighteenth century. Lieutenant-Colonel Joseph Shippen, a
graduate of Princeton, who was elected a member of this Society,
January 19, 1768, and who served under Forbes in the capture of
Fort Duquesne, and was a generous patron and benefactor of Benja-
min West, has portrayed for us a charming picture of some of the
belles that reigned supreme at the Assemblies in his day. In the
*' Lines written in an Assembly Room," which he wrote at least as
early as 1774, and very probably in the sixties, he says :

" With just such elegance and ease,
Fair charming Swift appears ;
Thus Willing, whilst she awes, can please ;
Thus Polly Franks endears.

" With either Chew such beauties dwell,
Such charms by each are shared,
No critic's judging eye can tell
Which merits most regard.

" 'Tis far beyond the painter's skill
To set their charms to view ;
As far beyond the poet's quill
To give the praise that's due."

The invitation card to the Assemblies for 1790 for Colonel Ship-

1902.] BALCH — THE FIRST "ASSEMBLY ACCOUNT." 263

pen's daughter Mary — better known as '' Polly " Shippen — has been
preserved, and reads as follows:

" Philadelphia Assembly, 1790.

The Favour of Miss P. Shippen' s
Company is requested for the Season.

J. M. Nesbitt, W. Stewart,

Geo. Meade, Jos. Redman,

John Swanwick, George Harrison,

Another card for the season of 1850, when three Assemblies were
given at Musical Fund Hall, is thus inscribed :

"Assemblies.
" The Honor of

Company is requested for the Season.

" John M. Scott, ^ ^ James H. Blight,

Thomas Cadwalader, 1 B. W. Ingersoll,

Joseph Swift, William T. Twells

Managers. -I

Charles Willing, f ^ ' '\ Alexander Biddle,

Richard Vaux, | | William W. Fisher,

M. G. Evans, ' j [^ Bernard Henry, Jr."

During the Civil War the Assemblies were completely stopped ;
but after the conclusion of that great struggle they were revived in
1866, at the Academy of Music, by Dr. Alexander Wilcocks, who
was a Manager before the war, William Henry Rawle and other
gentlemen.

We Americans, in the rush and stress of every-day life, are too
apt to forget that those things we enjoy to-day are in a measure due
to those who built in an earlier time. It is good to have some
reverence for the experiences of the past as we prepare for the
future. Charles Lamb, in one of his sonnets, tells us:

" 'Tis man's worst deed
To let the things that have been run to waste,
And in the unmeaning present sink the past :
In whose dim glass even now I faintly read
Old buried forms and faces long ago."

While we should not worship what has been so much as to forget

96i MINUTES. [May 16,

to labor in the present and for the future, we should not be entirely
oblivious of those who have gone before. And as unless old papers
and books are properly housed in some collection like that of the
American Philosophical Society, they are pretty sure in the changes
and ups and downs of fortunes of their various succeeding owners
to be eventually destroyed, it is with peculiar pleasure, therefore,
that in behalf of my brother, Mr. Edwin Swift Balch and myself,
I am able to present to the Society the Account Book of the First
Assemblies kept by our great-great-great-uncle, John Swift, and so
insure its future preservation.

Stated Meeting, May 2, 1902,

President Wistar in the Chair.

Present, 7 members.

Mr. Benjamin C. Tilghman, Jr., a newly elected member,
was presented to" the Chair, and took his seat in the Society.
Letters accepting membership were read from

Mr. Grove K. Gilbert, Washington.

Prof. Paul Haupt, Baltimore.

Prof. Albert A. Michelson, Chicago.

Mr. Benjamin C. Tilghman, Jr., Philadelphia.

M. Gaston Darboux, Paris.

Stated Meeting, May 16, 1902.

President Wistar in the Chair.

Present, 19 members.

Prof. Hermann Collitz and Prof. Felix E. Schelling, ne^vl}^
elected members, were presented to the Chair, and took their
seats in the Society.

Letters accepting membership were read from
M. Henri Becquerel, Paris.
Prof. G. Johnstone Stoney, London.

^^^-•^ MINUTES. 265

A letter was read from the Comite de 1' Exposition Inter-
national de I'Art et de la Paix, inviting the Societv to par-
ticipate in the International Exposition, to be held at Lisbon
in May, on the anniversary of the Conference at The Hague.
The decease was anounced of

Mr. Paul Leicester Ford, at New York, on May 8, aged 37.
President Henry Morton at Hobolcen, on May 9, aged 65.
Mr. Stewart Culin read a paper on the Indians of the
Southwestern United States.

Magellanic Premium.

FOUNDED IN 1 786, BY

JOHN HYACINTH de MAGELLAN,

OF London.

1902;

The American Philosophical Society,

Held at Philadelphia, for Promoting Useful Knowledge

ANNOUNCES THAT IN

DKCEMBER, 1902,

IT -WILL AWARD ITS

MAGELLANIC GOLD MEDAL

to the author of the best discovery, or most useful invention, relating to Navi-
gation, Astronomy, or Natural Philosophy (mere natural history
only excepted) under the following conditions :

1. The candidate shall, on or before November i, 1902 deliver, free
of postage or other charges, his discovery, invention or improvement,
addressed to the President of the American Philosophical Society, No.
104 South Fifth Street, Philadelphia, U. S. A., and shall distinguish his
performance by some motto, device, or other signature. With his dis-
covery, invention, or improvement, he shall also send a sealed letter
contaming the same motto, device, or signature, and subscribed with the
real name and place of residence of the author,

2. Persons of any nation, sect or denomination w^hatever, shall be
admitted as candidates for this premium.

3. No discovery, invendon or improvement shall be entitled to this
premium, which hath been already published, or for which the author
hath been publicly rewarded elsewhere.

4. The candidate shall communicate his discovery, invention or im-
provement, either in the English, French, German, or Latin language.

5. A full account of the crowned subject shall be published by the
Society, as soon as may be after the adjudication, either in a separate
publication, or in the next succeeding volume of their Transacdons, or
in both.

6. The premium shall consist of an oval plate of solid standard gold
of the value of ten guineas, suitably inscribed, with the seal of the Society
annexed to the medal by a ribbon.

AH correspondence in relation hereto should be addressed
To THE Secretaries of the

American Philosophical Society,

No. 104 South Fifth Street,

PHILADELPHIA. U. S. A

TRANSACTIONS

OF THE

American Philosophical Society

Held at Philadelphia,
FOR Promoting Useful Knowledge.

/ 'ol. XX, Nezv Series. Pai^t /, ^to, pp. 62, with j Plates.
Lately Published.

CONTENTS.

Art. I. — The 11 i story of the Pel5^cosauria, with a Descrip-
tion of the Genus. Dimetrodon Cope. By G.
Baur and E. C. Case.

Part II. ^lo, pp. 7/^, with 5 Plates.
Just Published.

CONTENTS.

Art. II. — The Chronolog-ical Distribution of the Elasmo-
branchs. By O. P. Hay.

An. III. — Results of Observation with the Zenith Telescope
of the Sayre Observatory from January 19, 1894,
to August 19, 1895. By Charles L. Doolittle.

Art. IV.— A Study of the Chromosomes of the Germ Cells
of Metazoa. By Thomas H. Montgomerv, Jr.,
Ph.D.

SUBSCRIPTION— FIVE DOLLARS PER VOLUME.

SLPAkATB PARTS ARB NOT SOLD.

A.ki;ess THE LIBlfARlAN OF THE

American Philosophical Society,
No. 104 South Fifth Street,

Philadelphia, U. S A.

PROCEEDINGS

OF THE

AxHERICAN PHILOSOPHICAL SOCIETY

UEID AT riilLADElPnU FOE PROMOTING USEFUL KXOffLEDflE.

Vol. XLI. Aphil-December, 1903. Xo. 171

CONTENTS.

PAGE

The Geograpliical Distribution of Freshwater Decapods and its

Bearing upon Ancient Geography. By Dr. A. E. Ortmann. 267

Stated Meeting, October 3 400

A Specialized Cocoon of lelea poli/j}hemus. By A. Radcliffe

Grote 401

Stated Meeting, October 17 40G

Stated Meeting, November 7 40G

Stated Meeting, November 21 406

Stated Meeting, December 5 . , 406

Specific Precipitins and their Medico-Legal Value in Distinguishing

Human and Animal Blood. By Alfred Stengel 407

Stated Meeting December 19 414

philadelphia ;

The American Philosophical Society,

104 South Fifth Street.

1902.

It is ivqiiosted that all oorrcspoiuienco be addressed

To TiiK Skckktariks of the

A]\rERTCAN PHILOSOPHICAL SOCIETY,

104 South Pifth Stkkkt,

Pini.APKLrHiA, U. S. A.

Members will please etnnmunicate to tlie Secretaries any
inaeeiiraev in nnine (->r address as i^iven on the wrapper of t]\is
number.

It is requested that the receipt of this numncr of the
Proeeediniis be aeknowledued to the Secretaries.

Members who have not as yet sent their photographs to the
Society will confer a favor by so doing- ; cabinet size prefeiTed.

iM j3

PROCEEDINGS

OF THE

4MEETCAN PHILOSOPHICAL SOCIETY

HELD AT PHILAIIELPniA FOR PROSOTISG USEFUL KNOTODGE.

Vol. XLI. April-December, 1903. No. 171.

THE GEOGRAPHICAL DISTRIBUTION OF FRESHWATER

DECAPODS AND ITS BEARING UPON

ANCIENT GEOGRAPHY.

BY DR. A. E. ORTMANN.

(/^eai/ Aj>rt7 3, 1903.)

INTRODUCTION.

During the last decennium Zoogeography has developed in a very
peculiar direction, which, in a large part, is directly opposite to
the methods introduced by Wallace. The professed aim of the
latter was the creation of a zoogeographical division of the earth's
surface into regions, realms and the like, the purpose of which was
the subordination of the facts of animal distribution under a fixed
scheme ; and since it was self-evident from the beginning that the
distribution of animals ought to express the physical conditions of
the earth's surface, it was assumed that the proposed zoogeographi-
cal divisions correspond to the chief features of the distribution of
the conditions of life.

Soon, however, it was discovered that it is impossible to give a
division of the earth's surface that could claim general recognition.
It is true that each of the proposed schemes was actually supported
by more or less numerous instances of distribution, and that in
many cases the physical factors influencing and explaining these
divisions were easily understood ; but there was always alongside of
the supposed normal conditions a number of exceptional cases,
where the actual distribution of certain animals or animal groups
was directly the opposite. One of the chief causes of this fact has
already been recognized and carefully studied by Wallace. It is

PROC. AMER. PHILOS. SCO. XLI. 171. R. PRINTED NOV. 19, 1902.

268 OETMANK— DISTRIBUTION" OF DECAPODS [April 3,

the difference of the means of dispersal of the various groups of
animals. On account of these anomalies Wallace constructed his
regions chiefly for Mammals and Birds, excluding all the rest of the
animal kingdom.^

This method, however, can never be satisfactory. It amounts to
nothing but the creation of an arbitrary scheme which may corre-
spond to some of the facts; but if there are any other facts that do
not fit into it — as very often happens — they are simply thrown out
and neglected.

But this is not all. Even the restriction of Wallace's regions to
a single group of animals proved insufficient to cover all cases
within this group. This is true also of all other schemes that have
been proposed by other writers for the same or other smaller
groups. In every single instance there were exceptions to the rule,
and for some time it seemed difficult or even impossible to deal
with these apparent anomalies ; in fact, none of the proposed
divisions into regions can be applied to all cases, even within
smaller groups.

The correct understanding of this fact, that a large number of
animals does not submit to any of the proposed schemes that profess
to comply with the present distribution of the condition of life, was
made possible by the consideration that the actual distribution of
any animal must have originated in the past. Although there are
some animals the history of which does not go very far back, in a
geological sense, there are others which do, and, generally speak-
ing, we may say that the farther back we go in geological history
the more different were the conditions of life from what they are
now, and the present distribution of the respective forms must nec-
essarily appear the more strange and anomalous. Wallace, indeed,
tried to remove this difficulty in a very peculiar way. He simply
propounded his principle of the permanency of the continents,
which means to say that the present distribution of land and water
(and in general of the physical conditions of life) did not change
materially during the earth's history, and that the external features
of the earth's surface have remained practically identical from time
immemorial up to the present. That this principle is without

1 This exclusive restriction to the higher forms of life (Mammals, Birds) is a
principle of Wallace and has been expressly maintained by him as late as in 1894
(see Nature, \o\. xlix, 1894, p. 610).

1902.] AND ANCIENT GEOGRAPHY. 269

proper foundation has now been recognized and the opposite opinion
begins to prevail, that abnormal conditions of distribution are due
to just such changes of the physical conditions during a geological
past, and that cases of this kind may often enable us to draw con-
clusions as to the reconstruction of the old conditions. We may
safely assume that the character of the physical conditions of the
earth's surface has changed continuously and variously in the past
and that we possess among living animals many forms which express
in their present distribution not only the Tertiary state, but which
may also represent Mesozoic or even Palaeozoic conditions. Thus
it is evident that investigation of the present distribution cannot be
used as the starting point for the construction of any scheme. This
has been done, however, not only by Wallace — who entirely disre-
garded the above fact — but also by others, who paid due attention
to it. Indeed Osborn ^ has pronounced it the purpose of Zoogeog-
raphy to unite past and present distribution into one scheme, and
the same idea has led Jacobi - to attempt practically this union.

But if we study the most prominent differences between past and
present we see that they are chiefly found in the different distribu-
tion of land and water, and that frequently in past times land con-
nections existed between parts which are now separated, or vice
versa; and thus it is self-evident that the solution of Osborn's
problem is simply impossible, since there is no way to exoress
separation and connection of the identical parts in one and the
same scheme.^

We consequently arrive at the following three conclusions :

1. Any divisio7i of the earth' s surface into zoogeographical regions
which starts exclusively from the present distribution of animals,
without considering its origin, must be unsatisfactory, since always
only certain cases can be taken in while others refnain outside of this
scheme.

2. Considering the geological development of the distribution of

1 H. F. Osborn, <' The Geological and Faunal Relation of Europe and America
During the Tertiary Period, etc.," in Ann. N. V. Acad. Sci., Vol. xiii, 1900, p.
48, and in Science, April 13, 1900, p. 563.

^ A. Jacobi, « Lage und Form biogeographischer Gebiete " {Zeiisckr. Ges.
fuer Erdkunde, Berlin, Vol. xxxv, I900).

3 This, of course, does not dispose entirely of Osborn's problem. On the con.
trary, it remains "the" problem of Zoogeography, only we have to change its
formal expression and to say that the historical union of past and present distri-
bution is the purpose of zoogeographical study.

270 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

animals, we tnust pronounce it impossible to create any scheme what-
ever that covers all cases.

3. Under these circumstances it is incorrect to regard the creation
of a scheme of animal distribution as an important feature or purpose
of zoogeographical research.

Thus we are justified in saying that zoogeographical study, as
introduced by Wallace, is not directed in the proper channels, and
we are confronted with the question, If the creation of regions of
animal distribution is not a matter of first importance, which is the
vital point in this branch of research ?

This question has been practically answered by many writers. I
name the following: G. Pfeffer, E. von Ihering, H. A. Pilsbry,
R. F. Scharff, C. Hedley, W. Kobelt, H. F. Osborn, A. Jacobi
(besides others), and these we may take as representatives of the
modern tendency in Zoogeography. According to these authors
the chief aim of zoogeographical study consists — as in any other
branch of biology — in the demonstration of its geological develop-
ment. We have to designate this most emphatically, as the final
goal of Zoogeography : the retracing of the present animal distri-
bution to its beginning in the past, and a corollary of this is the
reconstruction of the ancient physical features of the earth's surface,
since these in the first place have guided the development. In the
latter respect the distribution of land and water in past times is all-
important and the easiest to be traced.

Thus Zoogeography becomes a very important aid not only to
physical Geography itself, but also to historic Geology.

The above introductory remarks seem necessary, because the
purpose and methods of the new tendency in Zoogeography have
been frequently misunderstood, and especially because it was not
seen that in this way the fruitless discussions on the limits and
value of the different zoogeographical regions, etc., have been ren-
dered unnecessary. Yet it is a habit among zoogeographers to
create or discuss zoogeographical regions according to Wallace's
ideas, and this is done not only by writers who, like Wallace and
Sclater, are principally opposed to any progress in Zoogeography,
but also by those who are familiar with the new ideas about the geo-
logical development of animal distribution. The old method has
become an integral part of this branch of science to such a degree

1902.1 AND ANCIENT GEOGRAPHY. 271

that any research in this direction is deemed incomplete that is not
finished by the creation or discussion of ** regions."

In opposition to this, we wish to emphasize that we consider it
entirely a matter of indifference whether we accept any regions or
not, since none of the possible schemes can be satisfactory. Only
in a very limited degree and in a modified sense we believe it ad-
visable to divide the earth into regions, and we have proposed such
a division for the marine life districts.^ This scheme, however, is
not intended to represent or to express the actual distribution of
any animals, but it is a scheme of the distribution of the conditions
of existence in the oceans of the present time without consideration
of the past or of any definite group of animals. The only purpose
of these regions is to single out those marine animals which corre-
spond to the normal conditions of life and to separate them from
the abnormal cases ; under '' 7ior77ially distributed,'" consequently, we
mean those animals which shape their distribution according to the
present features of the earth's surface and which belong in their
origin to recent time. All the rest differs and does not fit into
these regions ; but instead of leaving them out of consideration we
know that just these cases are the most interesting, since they
demand closer investigation. In most cases we find that these in-
stances of *' abnormal" distribution are to be traced back into the
geological past in order to be properly understood. This latter
study is the most important branch of Zoogeography, and we see
that the introduction of "regions" in our method is only the
means by which we discover the more interesting and important
cases, but it is not the final aim.

Of course the same method may also be used for land and fresh-
water animals, and it may here be incidentally remarked that the
regions proposed by Wallace are in this respect superior to any
modifications introduced by later authors, since they generally are
well limited and isolated by physical boundaries given on the sur-
face of the earth. But if we are satisfied with the simple statement
of the fact that some animals fit into these regions while others do
not, we do not approach the solution of the question as to how the
actual distribution originated : we are to advance one step further
and investigate those cases which do not submit to the scheme.
The final aim of this investigation is to compare and group together

^ Ortmann, A. E., Grundzuege der marinen Thiergeographie, Jena, 1896.

272 ORTMANN— DISTRIBUTION OF DECAPODS [Aprils,

those abnormal cases which resemble each other. Thus we gain
certain general views as to ancient geography, and we are finally
enabled to trace the distribution of land and water, of climatic con-
ditions and the like in the geological past.

Most prominent among the groups of animals that are available
for these investigations are the Mammals^ and they have actually
been used for just this purpose by various authors (Doederlein,
Zittel, Lydekker, Scharff, Osborn). The palaeontological material
within this group is the most complete of all. But there is one im-
portant drawback : since the history of the Mammals hardly goes
back beyond Tertiary times, at any rate since the palseontological
record of this group is more or less complete only within the Ter-
tiary, we can only draw conclusions from them as to the geographi-
cal conditions of this period, while we have to refrain from an
investigation of those of the Mesozolc times.

This is a very different matter with the land and freshwater Mol-
lusks. According to what we know, it is apparent that many of
these forms can be traced back to Mesozoic times, sometimes even
to Palaeozoic, and, indeed, it is this group of animals that has fur-
nished the material for the studies of von Ihering, Piisbry, Hedley,
Kobelt, and we are to expect that further investigation in this direc-
tion may yield interesting results.

Other groups have also been used. Von Ihering introduced the
study oi Ants, and there may be other promising groups among the
Insects (for instance Spiders). But since the majority of the Insects
possess unusual means of dispersal (power of flight) that are apt to
obscure the original conditions of distribution. Insects in general
are not well adapted to this kind of research. Of other animals the
Earthwofms have been studied in this respect (by Beddard), and
of the Vertebrates, Reptiles, Amphibians, and freshwater Fishes are
very likely to prove good objects, since their history in many cases
goes back to the beginning of the Mesozoic or even to the Palaeo-
zoic time.

In the following treatise I wish to call special attention to certain
groups of Decapod Crustaceans that live in fresh water. In part
these have been discussed previously by other writers as well as by
myself, but it is worth while to go more into detail, since we shall
find them very interesting in this respect.

1902.] AND ANCIENT GEOGRAPHY. 273

The following groups of freshwater Decapods are known :
Family: Atytda.

Palcemonidce (in part).

PotamobiidcB.

ParasiacidcB.

jEgleidce (monotypic).

Potamonidce.

There are, scattered among other families, other forms of fresh-
water Decapods, but the above are the most important groups.
These are found either exclusively in fresh water or possess the
largest number of their members there, and are found only in rare
cases in the sea.

As regards the Atyidce, the present writer has collected the cho-
rological material in a previous paper.^ This is no doubt one of the
oldest groups of freshwater Decapods, and their origin, as is very
likely also according to their morphological characters, is to be
sought for possibly in Jurassic times, although fossil formxS are not
positively known. The chief features of their distribution are
excessively abnormal and even confusing, and therefore the extreme
age of the group is again confirmed. On the other hand, there are
smaller groups within this family, the distribution of which was
apparently formed in later times. Since there is every reason to
believe that our knowledge of the actual distribution of the AtyidcF.
is still more or less defective, we shall refrain from discussing it and
refer only to the latest summary given by the present writer.'

In the family of the PalcemomdcB the genus PalcBmon forms a
group that possesses numerous species which are found chiefly in
fresh water. Their distribution, which has also been previously
investigated by the present writer,* points distinctly to the fact that
this genus is a very recent one, which is at the present time just in
the act of immigrating into fresh water, and that this process is by
no means completed. The different species depend in their dis-

1 Compare Ortmann, A. E., in Bronn's Klassen und Ordnungen des Thier.
reichs. Vol. v, 2, 1899, p. 1^85. We leave out of consideration the families
Coenobitidce and Gecarcinida:, which are more properly land animals. See ibid.,
pp. 1 183 and 1 184.

2 Ortmann, in Proc. Acad. Philadelphia, 1894, p. 397 ff.

3 In Bronn's Klassen tend Ordn., I. c, 190 1, p. 1286 f.

^ In Zool. Jahrb. Syst.,NQ\. v, 1891, pp. 744-748, and in Bronn's Klassen tend
Ordn., I. c, 190I, p. 1291 f.

274 ORTMANN — DISTRIBUTION OF DECAPOI^S [Aprils,

tribution largely on the conditions prevailing in the littoral waters,
and generally they follow the physical regions which we have pro-
posed for the marine littoral district of the present time. To this
there are only a few exceptions, due to special means of dispersal
(crossing over continental divides, for instance). For the investi-
gation of ancient Geography this genus has no value. ^

In the following we shall treat of the remaining four families :
PotamobiidcB^ Parastacidce^ yEgleidce and Fotamonidce.

PART I. CHOROLOGICAL MATERIAL.

A. Chorology of the Families Potamobiid^ and Para-

STACiDiE. (See Fig. i.)

BIBLIOGRAPHY.

(a) General Discussions and Systematic Revisiofis.

Huxley, Th. : The Crayfish, London, 1879.

Faxon, W. : "A Revision of the Astacidne " {Mem. Mus. Harvard, Vol. 10,
1885).

^ Couti^re, H. (" Sur quelques Macrures des eaux douces de Madagascar," in
C. R. Acad. Sci. /'aris, Vol. cxxx, 1900, pp. 1 266-1 268), discussing the Palse-
mons of Madagascar, has advanced some views as to their distribution and con-
cludes by putting the (unanswered) question whether this distribution has
formed under conditions similar to the present ones or not. This question, how-
ever, has been answered in detail by the present writer in the paper quoted above
(1891), with which Couti^re seems to have been unacquainted. This is also evi-
denced by the fact that some of the peculiarities of distribution in this genus,
emphasized by the present writer, are not mentioned by Coutiere — for instance,
the relation of the West African species to those of America. Coutiere holds that
the West African (not South African) ralcemon vollenhoveni Herkl. is most
closely allied to P. brevicarpus Haan from Japan, while I regard the relationship
to the American P. jamaicensis (Hbst.) as more important.

As regards Bithynis hildebrandti Hlgdf (1893) from Madagascar, I believe
it is hardly possible to connect this species genetically with the type species of
this genus from Chile. I think this is a case of convergency. The opinion of
Coutiere, that the theory of a Posttriassic connection of Madagascar with India
and Africa is to be abandoned, has no support whatever. The distribution of
PalcBnioHy which, according to Coutiere himself, does not go back beyond
Miocene times, is absolutely irrelevant to this question, and even the Miocene
age of Palamon seems to be doubtful. The presence of identical species on the
eastern and western sides of the Cordilleras m South America is no evidence for
this, since this distribution is not discontinuous, and the respective species have
apparently crossed this chain of mountains, and are actually found in the moun-
tains high up in the headwaters of the Anazonas river, for instance.

1902.]

AND ANCIENT GEOGRAPHY.

275

276 ORTMANN— DISTRIBUTION OF DECAPODS [Aprils,

Faxon, W. : "Notes on North American Crayfishes, Family Astacidae " (^Pr.

U. S. Mus., Vol. 12, 1890).
Faxon, W. : " Observations on the Astacidae in the U. S. National Museum and

in the Museum of Comparative Zoology, with Descriptions of New Species "

(/>. U. S. Mus., Vol. 20, 1898).
Ortmann, a, E. : «« Ueber Bipolaritaet in der Verbreitung mariner Thiere "

{Zool. Jahrb. Syst., Vol. 9, 1896, p. 588-594).
Ortmann, A. E., in Bronn's Klassen und Ordnungen des Thierreichs, Vol. 5,

Part 2, 1 90 1, pp. 1 288- 1 290.

ib) Special Literature ^ published after Faxon' s Revisions yi88^,

i8gOi ^8g8), or not embodied in them.
Berg, C. : " Datos sobre algunos crustaceos nuevos para la fauna Argentina "

(^Commun. Ahis. Buenos Aires, Vol. i, 1900).
CocKERELL, T. D. A., and Porter, W. : « A New Crayfish from New Mexico "

{Pr. Acad. Philadelphia, 1900).
DOFLEIN, F. : " Weitere Mitteilungen ueber dekapode Crustaceen der k. bayer-

ischen Staatssammlungen " (6". B. Ak. Muenchen, Vol. 30, 1900, p. 132),
Hay, W. P. : *' Description of Two New Species of Crayfish " (^Pr. U. S. Mus.,

Vol. 22, 1899).
Hay, W. p. : «' Synopses of North American Invertebrates — 6. The Astacidae

of North America" (^Ajueric. A^atural., 1899).
Lenz, H. : " Die Crustaceen der Sammlung Plate " (^Zool. Jahrb. Syst., Suppl.

5, 1902, pp. 736,737)-
NoBiLi, C. : " Contribuzioni alia conoscenza della fauna carcinologica della

Papuasia, della Molucche e dell' Australia" (Ann. Mus. Genova, Ser. 2,

Vol. 20, 1899).
Philippi, R. a.: (Descriptions of Three Species of Crayfishes from Chile)

{^Annates Umvers. Chile, Vol. 61, 1882, pp. 624-628, with plate). ^
Philippi, R. A. : " Dos palabras sobre la sinonimiade los Crustaceos, Decapo-

dos, Braquiuros o jaivas de Chile" (Ann. Univers. Chile, 1894).

For the intended publication of the Decapods in the ^'Thier-
reich," edited by the German Zoological Society, the present
writer was obliged to make a complete collection and a critical
review of the systematic literature of these two families. Of course,
the results of these studies are embodied in the following portion of
this article, although it is not possible to refer to this work, the
manuscript of which has just been finished.

I. Family: PoTAMOBiiDiE HuxL'
The family Potamobiidce. is divided into two genera : Potamobius
Sam. and Cambarus Er. The latter is no doubt the more special-

^ Of this rare paper I possess a handwritten copy and sketches of the figures,
through the kindness of Dr. F. Philippi, of Santiago.

2 Those authors (Faxon, Rathbun) who retain for the European crayfish the

1002.] AND ANCIENT GEOGRAPHY. 277

ized one, and its distribution is more sharply limited than that of
Potamobius^ it being found only in the eastern parts of North
America, Mexico and Cuba.

Genus : Caryibai'us Er.

The genus Cambarus contains at present sixty-six well-known
species ; of a sixty-seventh, the group to which it belongs is doubt-
ful {C. clypeatus Hay, Missouri). The species form five groups
within the genus.

Sixteen species belong to the first group, namely :

1. blandingi {Yi2C[\.). 9. versuiiis l^d.g.

2. hayiYz.yi. 10. spicu/ifer (Ltc).

3. fallax Hag. 11. pellucidus (Tell.).

4. clarki 0\x. 12. acheroniis\^o^Vivi}o.

5. troglodytes (Lee). 13. wiegmanni^x.

6. lecontei Hdig. 14. alleniYd^y..

7. angustatus (J-,Qc.). 15. evermanni YdiX.

8. piibescens Fax. 16. penicillatus (Lee).

Eight species belong to the second group :

1. cubensis Er. 5. gallinus Cock, and Port.

2. carinatus Fax. 6. gracilis Bund.

3. ?nexica7ius Er. 7. carolinus Er.

4. simulans Fax. 8. advena (Lee).
To this group possibly belongs clypeatus Hay.

Thirteen species belong to the third group :

1. acuminatus YdiX. 8. tihleri Ysiyi.

2. bartoni (Fabr.). ' 9. setosus Fax.

3. longulus Gir. 10. extraneus Hag.

4. latimanus (Ltc). 11. Jordani YsiX.

5. dubiusYsiX 12. cornutusYd^y..

6. diogenes Gir. 13. ha?nulatus Cope and Pack.

7. argillicola Fax.

generic name of Astactis M. E., claim that Latreille (Consider, gener., etc.,
1810; see Faxon, 1898, p. 662) has made this species, Astacus fluviatilisYz}ox.,
the type of the genus Astactis Fabr. This statement of Latreille, however, is
erroneous, since Astacus of Fabricius is a genus without type, and remained such
until Saniouelle (^The Entomologists'' Useful Compendium, i8i9,p. 95) separated
Astacus and Potamobius (Lobster and Crayfish). See Faxon, 1885 ; Ortmann,
«« Das System der Decapodon Krebse " (^Zool. Jahrb. Syst., Vol. 9, 1896,
p. 430), and Stebbing (in Natural Science, Vol. 12, 1898, p. 239 fif.).

278 ORTMANX — DISTRIBUTION OF DECAPODS f April 3,

Twenty-six species belong to the fourth group :

I.

mississippiensis Fax.

14.

z'/WZ/V Hag.

2.

tmmunis Hag.

15-

72^/i- Fax.

3-

7nedius Fax.

16.

pilosus Hay.

4.

lancifer Hag.

17-

longidigitus Fax.

5-

palmeri Fax.

18.

sloanei Bund.

6.

difficilis Fax.

19.

rusticus Gir.

7-

alabamensis Fax.

20.

?;2^^y^/ Fax.

8.

compressus Fax.

21.

harrisoni Fax.

9-

propinquus Gir.

22.

forceps Fax.

lO.

neglectus Fax.

23-

spinosus Bund.

II.

digueti Bouv.

24.

erichsonianus Fax.

12.

affinis CSay).

25.

putna7ni Fax.

13-

indianensis Hay.

26.

^y/^i- Fax.

Three species belong to the fifth group :

1. montezumcs ^diw^s. 3. shufeidHYdiX.

2. chapalanus Fax.

In discussing the distribution, it is best we take up the single
groups. The species of XhQ first group are restricted chiefly to the
southern parts of the United States and Mexico, and we observe
that all, with two exceptions {blandmgi dindipelluddus), are found in
the region of North America formed by Mexico, Texas, Louisiana,
Mississippi, Alabama, Florida, Georgia and South Carolina.
C. blandingi possesses the widest range ; in the States named it is
wanting only in the farthest southeast, in Florida and Georgia ^ ;
but on the other side it extends beyond those limits along the
Atlantic coast, passing through North Carolina, Maryland and New
Jersey into the neighborhood of New York, and in the Mississippi-
Ohio basin it extends northward through Arkansas, Tennessee,
Missouri, Illinois and Indiana into Ohio and southern and eastern
Iowa. Westward, it has been found as far as Indian Territory.
C. pellucidus is a blind cave species which is restricted to certain
localities in Indiana and Kentucky.

It is apparent that the centre of this group is in the Gulf States
and in the southern Atlantic States, while the region of the eastern
mountains (Allegheny system) is left unoccupied by it, and only
one species advances northward along the Atlantic coast and in the

' It is very likely to be discovered in Georgia.

1902.] AND ANCIENT GEOGRAPHY. 279

Mississippi valley to the neighborhood of the Great Lakes. To this
latter extension of the range also belongs C. pellucidus. South-
ward, this group goes through Texas (here it has been found near
the Mexican boundary line), and is found in the neighborhood of
the city of Mexico (C wiegmanm). Whether this latter locality is
connected with the localities in Texas or not is unknown.

The centre of distribution of the second group is to be found in
the Southwest. We know two species from Mexico, two from New
Mexico, Texas and Kansas. Another species (C gracilis) extends
from these parts northward (in the prairies), and is found in
Kansas, Iowa, Illinois, as far as Wisconsin. In the South we have,
more or less isolated, C. clypeatus in Mississippi, and absolutely
isolated are C. carolinus and advena in South Carolina and Georgia
and C. ctibensis in Cuba.

Within this group we observe a very striking discontinuity ; not
only the Mexican localities are separated from those in the United
States, but also in the Gulf States, the southern Atlantic States and
in Cuba there are representatives of this group, separated from the
rest in the Southwestern and Central States.

Very different is the range of the third group. Here we have
complete continuity, and the centre is evidently in the system of the
Allegheny mountains and in the East. The species are very
numerous in the mountainous parts of Tennessee, Kentucky, North
Carolina, Virginia and West Virginia, Maryland, Pennsylvania, and
in the adjoining parts of Ohio and Indiana. This group is also
well represented in Illinois, and extends, gradually decreasing in
density, westward into Wisconsin, Minnesota, Iowa, Missouri (in the
eastern part only), Arkansas and the Indian Territory. It is very
rare in Texas, Louisiana, Mississippi ; is slightly represented in
Alabama, Georgia and South Carolina, but is wanting in Florida.
In a northeasterly direction, a single species (C bartoni) extends
over New York and New England across the Canadian boundary
into New Brunswick, where it reaches the Restigouche river, a
tributary of the Gulf of St. Lawrence. The same species is found
in the northern affluents of Lake Ontario (Toronto) and the St.
Lawrence river in Quebec (St. John's Lake), where it marks the
northern boundary of the genus. In Michigan this group is repre-
sented in the neighborhood of Lake Huron, but it has not been
found north of the Great Lakes in Canada. The northeastern
extension of the range of this group, on the one hand, is very

280 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

remarkable, while, on the other hand, we have a scarcity of it south
of the Allegheny system and west of the Mississippi. With the
exception of one isolated station of C. argillicola in Texas, this
group is not represented in the Southwest.

The largest number of species is found in the fourth group. In
certain respects it corresponds, in its distribution, to the third,
namely, in its exceeding scarcity in the South and Southwest. It is
wanting in Florida, in the low parts of the Carolinas, of Georgia,
Alabama and Mississippi. It is also wanting in Louisiana, and in
Texas it is found only in the northeastern corner (near the bound-
aries of the Indian Territory and Arkansas). Beginning here, it
extends northward over the Mississippi-Missouri-Ohio basin,
becoming more abundant, the centre being situated, in this region,
in the States of Missouri, Tennessee, Kentucky, Indiana, Illinois,
Iowa and the southern parts of Michigan and Wisconsin. East-
ward this group enters Ohio, Pennsylvania, Virginia, Maryland,
New Jersey and New York, reaching its northeastern limit north of'
Lake Ontario, near Toronto and Montreal. In Wisconsin it extends
to Lake Superior, and one species ( C. virilis) reaches from Minne-
sota, including the northeastern corner of North Dakota, to Lake
Winnipeg and the Saskatchewan river, the most northern locality
known for the genus. Westward, the range of this group includes
Kansas and Nebraska (southern and eastern part only) and the
southeastern corner of Wyoming : this is the most advanced
point for the genus in a northwesterly direction. Entirely isolated
from the range of this group, thus far described, we find a species
(C digueti) in Mexico (Pacific side. State of Jalisco), and another
species (C immunis, known from the prairies of Michigan, Indiana,
Illinois, Wisconsin, Iowa) is said to be present near Orizaba,
Mexico.

Therefore we may say, generally, that the centre of this group is
situated in the central part of the United States, about in that
region where the three large rivers, Missouri, Mississippi and Ohio,
unite. Thence it extends into the eastern and southeastern riioun-
tains, but hardly across them ; northward, it reaches the St.
Lawrence and the Saskatchewan rivers and westward Wyoming.
In a southwesterly direction it hardly reaches Texas, and the Mexican
localities seem to be isolated from the rest.

Of the three species oi\.\\Q fifth group, two are found in Mexico
and one near New Orleans.

1902.] AND ANCIENT GEOGRAPHY. 281

Taking together the distribution of the five groups, we find that
the range of the genus Cambarus extends over the following parts
of North America : In Mexico, the respective species are reported
from the following States : Vera Cruz (near Vera Cruz and
Orizaba), Pueblo, Mexico, Michoacan. This line would represent
the southern boundary of the range.^ Further, the genus has been
found in the States of Jalisco and Sinaloa (Mazatlan) (in the drain-
age of the Pacific Ocean) ; on the central pleateau, in Guanajuato,
San Luis Potosi (Santa Maria) and Coahuila (Parras). This latter
locality forms in a certain degree the connection of the Mexican
part of the range of the genus with that of the United States, since
the Mexican State Coahuila extends northward to the Rio Grande
del Norte, and just across this river, on its left bank, there is, in
Kinney county, Texas, a locality for C. clarki. Thence the range
of the genus is apparently continuous, and reaches eastward to the
sea (Gulf of Mexico and Atlantic Ocean).* Toward the west and
north it is circumscribed by the following line : from Kinney
county, Texas, to New Mexico (including its eastern part), then
receding toward Indian Territory and leaving out Oklahoma,
farther, including Kansas, the southeastern corner of Wyoming
Tpossibly a part of Colorado), the southern and eastern part of
Nebraska, crossing here the Missouri, including Iowa and Minne-
sota and possibly parts of the Dakotas, at any rate the northeastern
corner of North Dakota, crossing over into Canadian territory and
including the region of Lake Winnipeg and Saskatchewan river
(northernmost point). Thence this line recedes in a southeasterly
direction, reaches Lake Superior, and follows the Great Lakes as far
as Lake Erie. At Lake Ontario it advances again northward and
follows at a certain distance the St. Lawrence river, reaching at the
Lake St. John in Quebec the northernmost point in the East.
Then it turns southward, crosses the St. Lawrence and includes, in
New Brunswick, the drainage of the Restigouche and Miramichi
rivers (emptying in the St. Lawrence Gulf) and also the St. John
river (emptying in the Bay of Fundy). Thus the largest part of
New Brunswick seems to belong to the range of this genus, while

^ The genus is said to be represented near Alta Vera Paz, in Guatemala (Faxon,
1885, p. 173). This would advance the range southward beyond the Isthmus of
Tehuantepec. This locality, however, needs confirmation.

2 In Florida, only in the northern half are localities known, southward as far as
Orange, Lake and Hillsboro counties.

282 ORTMANN— DISTRIBUTION OF DECAPODS [Aprils,

Nova Scotia is excluded. Isolated from the continuous Mexican
and United States ranges is the Island of Cuba, where C. cubensis
has been found.

It is hard to say where the centre of the whole genus is situated.
Judging from the number of species represented in the different
parts, it seems to be more in the East than in the West, but for the
rest the genus is pretty evenly distributed in the Southeastern
States, in the region of the Alleghanies and the central basin, and
decreases markedly only in a westerly direction, disappearing
before it reaches the foothills of the Rocky Mountains. In the
Southwest, in Texas and New Mexico, the genus is less abundant,
and in northern Mexico it is found only near Parras, in the State of
Coahuila ; but then again it becomes more abundant in the central
part of Mexico. Whether this apparent scarcity in northern
Mexico and Texas corresponds to the actual conditions, or whether
it is due to defective knowledge of these parts, cannot be decided.
One result, however, is very evident : the genus is preeminently
characteristic of the central and eastern parts of the United States,
there attaining its highest development as regards the number of
species.

Now, what is the origin of this distribution of Cambarus ? Did
this genus originate in these parts, or whence did it come, and
which are its ancestors ?

In order to answer the first question, we learn much by recalling
to our mind the distribution of the single groups as stated above.
We have seen that the centre of the first group is in the Southeast ;
the range of the second group — althougli somewhat discontinuous —
centres in the Southwest. The third group has evidently its centre
in the mountainous regions of the Allegheny system, the fourth
group in the central basin and the fifth in Mexico.

The second and fifth groups are strongly represented in the
Southwest, the first group has distinct relation to these parts, the
fourth group only a few isolated stations, while the third group is
entirely wanting there. ^

1 Faxon (1885, P- •'7^) expresses this in the following way: in the South
(Mexico, Cuba, Gulf States and Atlantic States south of North Carolina) species
of the first, second and fifth groups prevail, while comparatively few species of
the third and fourth groups are present; in the North (Atlantic States north of
South Carolina, Central States and Canada) species of the third and fourth
groups prevail, whilt only a few species of the first and second advance into the
northern provinces.

1902.1

AND ANCIENT GEOGRAPHY. 283

As regards the morphological relations of the five groups, we are
to consider first Faxon's view (1885, p. 19), that the species of the
first group are morphologically the most primitive ones. He draws
this conclusion chiefly from the shape of the male copulatory
organs. If we compare, however, certain species of the second
group {^simulans, viexicanus, cubensis) with those of the first group
in this respect, we see that they chiefly differ from the latter only
in the smaller number of hooks on the pereiopoda of the male (only
on the third pair, not on the third and fourth, as in the first group).
On this account I should prefer to regard the species named as the
most primitive forms of the genus, although, on the other hand, I
agree with Faxon (1885, p. 47) in believing that the other species
of the second group more nearly approach the third group. That
the third and fourth groups, compared with the others, are more
advanced forms is also my opinion. As the most specialized
species I regard those of the third group which have acquired
burrowing habits {dioge?ies, argillicola, dubius). The species of the
fifth group differ from ail the rest in the presence of hooks in the
second and third pereiopods of the male, and thus I think they
represent an early separated side branch. The copulatory organs
of the male in this group resemble in certain respects more those of
the first and second groups than those of the third and fourth, and
the more primitive character of these species is also suggested by
the general shape of the body.

Thus we see that the more primitive forms of the first, second and
fifth groups belong chiefly to the South and point distinctly to a
connection with Mexico, while among the more advanced and
specialized forms of the third and fourth groups this latter connec-
tion is hardly expressed or not at all. Their origin and main dis-
tribution belong to the more northern parts.

This points to an origin of the genus in the Southwest, and we
believe that the genus came from Mexico and immigrated into the
United States in a northeasterly direction.

A few additional distributional facts tend to support this conclu-
sion. It seems that in those groups which possess a large represen-
tation in the Southwest the distribution is rather discontinuous.
This is most evident with the second group. Now discontinuity in
distribution of any animal is very often a sign of the breaking up
of a former continuous range by unfavorable physical conditions.
In the present case it appears that at a certain time the immigra-

PROC. AMER. PHILOS. SOC. XLI. 171. S. PRINTED NOV. 19, 1902.

284 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

tion of Cambarus from Mexico into the United States did not
meet with serious obstacles, but that later in the intermediate
regions (northern Mexico and Texasj more unfavorable conditions
arose which separated the United States more distinctly from
Mexico, and this is possibly due to a more decided development of
the desert character of these parts. Thus the Mexican representa-
tives of the first, second, fourth and fifth groups became more or
less separated from those in the United States, the first and fourth
groups developed more abundantly in the United States, while the
third originated there, possibly out of the second group, which in
these parts did not make any marked progress and was suppressed
and restricted to a few more or less isolated stations, probably on
account of its primitive character. An interesting light is thrown
upon this question by the presence of one species of the second
group (C cubensis) in Cuba. This species is closely related to C.
mexicanus (Pueblo, San Luis Potosi), while it has no closer rela-
tions in the United States, and thus its Mexican origin is most dis-
tinctly indicated. Therefore we may safely say of the second
group that it is a very primitive one and that Mexico, not the
United States, is to be taken as its centre of origin.

The character of discontinuity is more or less noticeable also in
the southwestern part of the range of the first, fourth and fifth
groups. The first possesses an isolated species {wiegmanni) in
Mexico, and the stations of C. blandingi and clarki in Texas are
very scattered. In the fourth group we have an isolated species
{digueti) in Mexico (Jalisco), while C. immmits, a species found
elsewhere in the northern central basin, has been reported from
Orizaba, in Mexico.^ The fifth group has two species in Mexico
and, widely separated from them, a third near New Orleans. If
we compare with this the northern part of the ranges of the first,
third and fourth groups we see everywhere perfect continuity. In
every direction from the centre, except toward the Southwest, the
intensity of distribution decreases gradually. This is especially
true for the first group, the centre of which is in the Southern
States, in the directions northward along the Atlantic coast and
upward in the Mississippi Valley. In the third group, whose centre
is in the Allegheny system, there is a regular decrease in intensity
in all directions, and in the fourth group a very regular decrease is

^ We have to accept this record, however, very cautiously.

190-2.] AND ANCIENT GEOGRAPHY. 285

noticeable from its centre in the middle Mississippi basin toward
the East, North and West.

Thus we are to recognize the fact that the different groups,
chiefly the first, third and fourth, express in their distribution a
regular, continuous advance in a northeasterly direction. Toward
the North and East is continuity, which represents a more recent
stage in distribution, while in the opposite direction, toward South-
west, we observe discontinuity, which characterizes generally a
more ancient stage. In the second group we have a very remark-
able discontinuity, and this group is a comparatively primitive one,
and the fifth group, which is also primitive in some degree, is
chiefly found in the Southwest.

All the foregoing considerations tend to justify our conclusion
that the migration of the genus Cambarus into the United States
started in the Southwest, on the Mexican plateau, and advanced in
a northeasterly direction.

Taking up now the second point to be considered, the question
of the origin and the ancestral forms of the genus Cambarus, we
shall be satisfied — for the present — with the opinion of Faxon
(1885, p. 16), which is also that of the present writer, that this
genus is the most highly specialized within the family Potamobiidce,
a corollary of which is that it must have originated from forms of a
lower type, which probably corresponded to the genus Potamobius ;
in fact, it is easy to imagine that Cambarus is derived directly from
Potamobius by the suppression of the single posterior pleuro-
branchia and the high specialization of the copulatory organs.
However, before entering into a more detailed discussion of the
relation of Cambarus and Potamobius, we shall give a sketch of the
chorology of the latter genus.

Genus Potamobius}

It is advisable here to go more into detail, since, on the one
hand, a synopsis of the more recent publications in this group is
desirable, and since, on the other, the number of species in this
genus is comparatively small and our knowledge of them excellent.
The genus is divided into two subgenera : Potamobius sens, strict.
Ortm. (^Astacus sens, strict. Fax.) and Cajubaroides Fax.

^ The following facts have not been put together since Faxon's review (1885).
I shall use here chiefly the revision of this group which I have prepared for the
" Thierreich."

286 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

Subgenus Potamobius — twelve species :
European group :

\. palHpes (Lereb.)- South and West Europe: Central Spain,
France, England, Ireland, Southwest Germany, Italy south-
ward to Naples, Dalmatia, Greece.

2. torrentium (Schrk.). Central Europe: Switzerland, South Ger-

many, Bohemia.

3. astacus (L.). West Russia (northward to Finland), Austria,
Germany, Denmark, South Sweden and Norway (possibly in-
troduced), France, southward to Northern Italy.

4. leptodactylus{E,sc\\z.). Ponto-Caspian basin : Hungary (Danube,

Theiss), South and Central Russia, northward to the White
Sea ] in Siberia in the region of the Caspian Sea. Further, in
West Siberia in the basin of the rivers Obi and Irtish, intro-
duced, as reported, but possibly indigenous (see Faxon, 1885,

P- 151).

5. pachypus (Rthk.). Estuaries of the Black and Caspian Seas.

6. colchicus (Kessl.). Transcaucasia (upper Rion river).

7. /^^j^/m (Schimk.). Turkestan (Sir Darja).

American group :

8. lenmscidus (Dan.). Washington, Oregon (lower Columbia
river), California (San Francisco).

9. trowbridgei (Stps.). Washington, Oregon (lower Columbia
river).

10. nigrescens (Stps.). California (San Francisco), Washington,
Alaska (Unalaska).

11. klamathensis (Stps.). British Columbia (east of Cascade Moun-
tains), Idaho, Washington, Oregon, Northern California
(mountain rivers).

12. gambeli (Gir.). In the Rocky Mountains: on the Pacific slope

in Utah, Idaho, Wyoming and Yellowstone Park ; on the At-
lantic slope; mouth of Yellowstone river (eastern State line of
Montana).

Subgenus Cambaroides — four species :
i.jchr'enki{Y^t.%'A?). Lower river Amur.

2. daiiricus (Pall.). Upper river Amur. \ '

3. japo7iicus (Haan). North Japan : Yesso.

4. siiiiilis (Koelb.). Korea.

1902.] AND ANCIENT GEOGRAPHY. 287

Generally speaking the range of the genus Potamobius exhibits a
striking discontinuity, which has often been discussed. One group
of species occupies a continuous area in Europe (and Western
Asia); another in East Asia ; a third in Western North America}
It has been said that it is another remarkable fact that the American
species resemble the European more than they do the East Asiatic,
and that the latter more approach Cambarus, which idea is ex-
pressed by their position in a separate subgenus named Ca^nba-
roides. But as regards the gills and the general form of the body,^
Cambaroides belongs without question to Potamobius. The male
copulating organs are as different from those of Cavibarus as they
are from those of the typical species of Potamobius, and the only
character that points decidedly to Cambai'us is the presence of
copulatory hooks on the ischiopodites of certain peraeopods. But
also in this respect Catnb'aroides is rather peculiar, since these hooks
are found on the second and third pair, which case is represented
among Cambarus only in the fifth group (containing only three
species), while all the rest of the numerous species of this genus
possess these hooks either on the third and fourth or only the third
pair.

1 am of the opinion that the resemblance of Ca?tibaroides to
Cambarus does not express very close blood relationship, but is due
to convergency. The development of hooks on the peraeopods of
the male, which serve, as is now known, the purpose of taking hold
of the female in copulation, is easily understood, if we remember
the manner in which copulation is performed, and it is also easily
intelligible that this device has possibly developed independently
in Cambaroides and Canibarus. The shape of the copulating
organs, which shows no doubt in Cambaroides a certain similarity
to the Cambarus type, can be explained in the same way, since it
is quite clear that if they are used in the same manner they may

1 To the latter area belongs an isolated locality of P. nigrescens in Alaska.
According to Hay (1899) this species is found all along the western coast of
North America, from California to Alaska, To my knowledge intermediate locali-
ties between Washington and Alaska have not been published.

2 Faxon (1885, p. 126) calls the shape of the body " subcylindrical," and says
that it resembles that of Cambarus. I cannot concur with him in this opinion;
the form of the carapace in Cambaroides is decidedly rather oval, as in Potamo-
bius, and besides there are variations also in this respect within the genus Ca??i-
barus.

288 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

assume the same or a similar form. If, finally, Faxon says that the
shape of the chelae in Cambaroides resembles those of Cambarus, he
means apparently only the general weak development of them, and
we must bear in mind that many Cambari are more like typical
Fotamobii in this respect.^

Thus the view seems supported that Cambaroides is not so very
closely related to Cambarus , as has been hitherto supposed, and
that the similarities which were emphasized are due only to con-
vergency. If we peruse the comparison of the characters of Camba-
roides, Potainobius and Cambarus given by Faxon (1885, pp. 126,
127), we find that Cambaroides is in some of them more isolated,
and that it resembles in others even more the West American
species of Fotatnobius. (For instance, the lack of a transverse
suture of the telson ; the shape of the second male abdominal ap-
pendage ; the lack of the first abdominal appendage in the female.)

The conclusion drawn from the foregoing is that in certain
respects (telson, second pleopods of male, first pleopods of female)
Cambaroides represents a type that points to the West American
Foiamobii, while the European species are more divergent from it,
and there is nothing that opposes the view that this subgenus-
(which might as well be regarded as a separate genus) forms the
starting point on the one side for the European Foiamobii and on
the other for the American Foiamobii, while subsequently it has
changed itself and become different from both (in the male copula-
tory organs).

The subgenus Cambaroides is restricted to the northeastern parts
of Asia (region of Amur river, Korea, North Japan). The exact
boundaries of its range have nowhere been located positively, and
it is not impossible that in the Siberian and northern Chinese
mountains other representatives of it may exist. For the present
the area from which species of Cambaroides are known is absolutely
separated from the European area of Fotattiobius.

As regards the latter, its centre is apparently in Southern and
Central Russia. From these parts the different species extend into
Western Europe, southward to Central Spain, Middle Italy and
Greece, and in Russia one species passes southward across the
Caucasus Mountains. Eastward a species is found as far as Turke-

1 Some other characters of Cambaroides indicate that this subgenus differs-
from Potamobius as well as from Cambarus^ and these are characters Avhich
approach it to the crayfishes of the southern hemisphere. Compare below.

1902.] AND ANCIENT GEOGRAPHY. 289

Stan, and northward the area reaches the White Sea. East of the
Ural Mountains the genus is said to be lacking, but it is found (the
widely-distributed species P. leptodaciylus) introduced in the river
Obi and its affluents. Some observations, however, have been
made which render it possible that F. leptodaciylus is an original
inhabitant of these parts.

As Huxley (1879) and Faxon (1885, p. 140) believe, thedifferent
forms of Potamobiiis have immigrated into Europe from the East,
and we can distinguish an older immigration on the part of the
group formed by the species P. pallipes and toi-reniium and a more
recent one on the part of P. astaciis and its allies. And even
within the latter group it seems that P. astaciis is older than the
other species and that it is pushed gradually westward by P. lepto-
daciylus, which is spreading in a westerly direction. The writer is
of the same opinion, and we shall see below that this is the only
theory that is admissible, if we consider the origin of Europe as a
continental mass. The occupation of Europe, after it had lost the
character of an archipelago and become part of the Eurasiatic con-
tinent, was possible for these animals only in a west-easterly direc-
tion. This corresponds also to the fact that those forms allied to
the European Potamobii, which are the nearest geographically, are
found to the east of them. They are the forms of Ca77ibaroides in
Eastern Asia, and we can readily imagine that from the area of dis-
tribution of Cambaroides an extension existed formerly in a westerly
direction across Central Asia, which connected with the European
area of Pota7?iobius, and this connection represents the direction of
the migration.

The forms of Potamobiiis which are found in Western North
America possess a continuous area of distribution ^ which is separated
from the rest of the genus. Huxley and Faxon, as has been men-
tioned above, believe that these American species are more closely
related to the European, but I think we have reason to accept a
different view.

My opinion is that a primitive group, which was ancestral to all
three of the living groups, formerly existed in Eastern Asia, which
is to be regarded as the centre of origin of the Potamobudce. This
group sent out a branch in a westerly direction, which finally
reached Europe, and it also sent out a branch in an easterly direc-
tion, which migrated apparently along the northern shores of the

1 Possibly with the exception of the isolated station near Unalaska.

290 ORTMAKN — DISTKIBUTION OF DECAPODS [Aprils

Pacific Ocean and finally immigrated into Northwestern America.
A trace of the direction of this route is preserved in the presence of
Potamobius nigrescens near Unalaska. After the final geographical
separation of the European and American descendants from the
original group in Eastern Asia each of the three groups developed
independently, and the Asiatic group acquired several more advanced
characters (copulatory organs and hooks) which otherwise are
found only in Cambarus, but which do not point to a closer affinity
to the latter genus, but are only due to parallelism.

Further, the West American Potamobii possess a character that is
found also in Cambarus. Faxon mentions that the second pleopods
of the male resemble not only those of Cambaroides, but also those
of Cambarus, while the European species are different in this
respect. This would bring the genus Cambanis into closer relation
to the West American Potamobii, and although this similarity
would hardly be of much value by itself, we have to regard it as
significant, since it agrees well with the distributional facts. The
tracing back of Cambarus to Cambaroides is geographically impos-
sible, and just this latter difficulty has induced the writer to exam-
ine more closely the supposed resemblance of both, and the result
is as has been discussed above. A closar connection of the Euro-
pean species of Potamobius with Cambarus is out of the question,^
and thus only the third group is left, the West American Potamobii.

From the latter group Cambarus is very sharply distinguished
though and no transitional forms are known. Probably this is due
to the fact that the connection of the area of both is far remote
geologically — that is to say, that the migration of Potamobius into
Mexico is very old and that the separation of both genera took
place in very early times, the one becoming restricted to North-
western America (southward to California), the other developing
on the Mexican plateau out of the old Potamobius stock that origi-
nally immigrated thither from the North. Thus the diff'erential
characters of Cambarus became well fixed and no transitions to the
old stock are found any more.

Thus for the family of the Potamobiidoi we may express the fol-

1 Faxon (1885, p. 176) thinks that in former times Cavibai'us and Fotamodius
occupied about the same area, and in order to support this he mentions the sup-
posed existence of a blind Cambarus in the caves of Carniola, Austria. How-
ever, this latter record is entirely erroneous. There exists no Cambartts in the
caves of Carniola (see Haman, Europ(sische Hoehlenfaima^ 1896).

1902.] AND ANCIENT GEOGRAPHY. 291

lowing opinion as to the origin of its distribution, founded exclu-
sively upon systematic and chorological studies.

The oldest home of the PotamobiidcB and their centre of origin is
somewhere in Eastern Asia. This ancestral stock spread chiefly in
two directions : a western extension of the range crossed Central
Asia, finally reaching Europe, while an eastern extension went
across Bering Strait and reached the western parts of North Amer-
ica. The continuity of this wide area, which was once wholly
occupied by the genus Pota??iobms, was interrupted subsequently in
Central Asia and where there is now Bering Sea, and thus three
isolated areas were formed — in Europe, in Eastern Asia and North-
west America. In each one of these parts the genus Potaniobius
continued to develop separately. From the West American stock
of Potamobius finally issued the genus Cambarus, which probably
originated in Mexico and thence invaded the central and eastern
parts of North America. The origin of Ca7fibarus probably lies far
back in time, since it shows no marked special affinities to any of
the three groups of Potamobius, and probably it was separated from
the latter genus before it was divided up into those three groups.

2. Family Parastacidce Huxl.

A systematic revision of this family has not been published hith-
erto. The present writer has tried to collect the necessary data for
a review in the ^' Thierreich," and although it is not possible to
give a complete synopsis, based upon careful criticism of the existing
descriptions as well as upon actual specimens, he has obtained a
fair general idea of the various forms which make up this family.

According to these studies the present state of our knowledge of
the distribution of this group is the following :

I. Genus Cheraps Er. em. Huxl.
Species :

1. quinquecarinatus {Qxx.^. West Australia : Swan river.

2. qtiadricarinatus Mrts. North Australia : Cape York.

3. bicarinatus (Gr.). North and East Australia: Port Essington,
Cape York, Rockhampton, Burnett river, Sydney, Melbourne,
Murray river.

4. preissi Er. Southeast Australia : Victoria.

292 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

Doubtful species : ai/s^raliensis (WL.-E.). Sydney.^

2. Genus Astacopsis Huxl.
Species :

1. franklini (Gr.). N. S. Wales and Tasmania.

2. serratus (Shaw). N. S. Wales : Murray river, Murrumbidgee

river, Richmond river, Brisbane Water and Paramatta river
near Sydney.

The following species represent probably young stages of A.
serratus : paramattensis Bate and sydneyensis Bate, both from
Sydney.

Doubtful species : fasf?tamcus Er. Tasmania.

3. Genus Engceus Er.
Species :

1. fossor Er. Tasmania.

2. cunicularius Er. Tasmania.

4. Genus Paranephrops White.
Species :

1. planifrons White. New Zealand, North Island and northern

part of South Island.

2. zealandicus (White). New Zealand, South Island: Dunedin,
Oamaru (Otago).

3. setosus Hutt. New Zealand, South Island : Canterbury.

This genus possibly is also represented in the Fiji Islands
(Huxley).

A doubtful genus, which perhaps belongs in this neighborhood,
is genus Astaconephrops Nobili.

Species :
I. albertisi^o\y\)\. Southern New Guinea : Katau.

5. Genus Parastacus Huxl.
Species :

I. pilimanus (Mrts.). Southern Brazil : Rio Grande do Sul.
Northern Argentina : Provinces Corientes, Entrerios, Cata-
marca.

1 By Nobili (1899, p. 246) this species is classified with Astacopsis, and is
recorded from the Island of Sorong, west end of New Guinea. It is very doubt-
ful whether this is correct.

1902.] AND ANCIENT GEOGRAPHY. 293

2. brasiliensis (Mrts.). Southern Brazil: Rio Grande do Sul.

3. hassleri Fax. Chili : Talcahuano, Tumbez.

4. defossus Fax. Uruguay. Brazil : Rio Grande do Sul.^

5. saffordiYdiX. Uruguay. Brazil: Rio Grande do Sul.*

6. varicosus Fax. Reported from Colima, Mexico.'

7. mcoleti (V\-\i\.^. Chili: Tumbez.

8. agassizi YdiX. Chili: Talcahuano, Llanquihue (Puerto Montt),

Tumbez. Argentina : Lake Nahuel Huapi.'

Doubtful species: chiletisis (M.-E.), spinifrons (Phil.), biinacu-
latus (Phil.), all three from Chili.

This genus is also found in Sta. Catharina, Southern Brazil,
according to Fr. Mueller.

6. Genus Asiacoides Guer.
Species :

I. niadagascarie7isis (M.-E.). Madagascar.

As regards the detailed limits of the range of the single species
and genera we are very poorly informed, and, further, it is quite
possible that our knowledge of the Australian and South American
crayfishes is very incomplete also on the systematic side, and it is
very likely that there are many unknown species.

It is evident ai, the first glance, however, that the distribution of
the ParastacidcE- is divided into four absolutely isolated areas :
Australia (including Tasmania and possibly New Guinea) ; New
Zealand ; part of South America ; Madagascar. Within each of
these areas are peculiar genera: in Australia, Cheraps, Astacopsis,

^ I have received these two species, defossus and saffordi, from Rio Grande do
Sul through Dr. H. von Ihering.

2 This locality most emphatically needs confirmation. It is very surprising that
this species has never been rediscovered anywhere in Mexico, although large col-
lections of freshwater Crustaceans from these parts have lately reached the
United States Museum.

3 Through Prof. W. B. Scott, of Princeton, I have received from the La Plata
Museum two males and one female of this species from this locality which agree
well with the description, with the exception that in the larger (adult) male the
right (larger) chela is more elongate, with almost parallel margins, and that the
squamif jrm granules of it are more strongly marked. The smaller male and the
female agree perfectly with P. agassizi.

The lake Nahuel Huapi is situated in the Cordilleras, at the southern extremity
of the Argentinian province Neuquen. It drains into the Atlantic through the
river Limay Leofu, which finally forms the Rio Negro. This locality is
directly east of Llanquihue, in Chili, but on the opposite slope of the Cordilleras.

294 ORTiMANN— DISTRIBUTIOX OF DECAPODS [Aprils,

Engceus ; in New Zealand, Paranephrops ; in South America,
Parastacus ; in Madagascar, Astacoides. All these forms are more
or less closely related to each other, only Astacoides from Mada-
gascar is rather isolated morphologically, since its branchial formula
shows peculiar reductions (only one pleurobranchia on the fifth seg-
ment of the thorax, while in all the rest four pleurobranchiae are
present). In this respect Astacoides resembles the Potamobiidce of
the northern hemisphere.

If it should prove to be correct that the genus Astaconephrops of
Nobili, from Southern New Guinea, as its author believes, is most
closely related to the New Zealandian Paranephrops, this, together
with the occurrence oi Paranephrops in the Fiji Islands reported by
Huxley, would indicate a distinct direction of the communication
between New Zealand and the rest of the world. This would have
been over the Fiji Islands in the direction toward New Guinea. As
to the connection of the South American Parastaci with the rest of
the family, we have hardly any systematic or chorological facts
which permit more detailed conclusions. We can only venture to
express the opinion that some kind of a connection between South
America on the one side and Australia or New Zealand on the
other must have once existed.

In order to get an adequate idea as to the geographical relations
of the genus Astacoides we have to recall to our mind a few facts
concerning the morphological relations of the Parastacidcv and the
Potamobiidce (see Ortmann, 1901, p. 1289J. According to Faxon
(1885, p. 126 f.), among the crayfishes of the northern hemisphere
it is only the subgenus Cambaroides which approaches those of the
southern. Not only the characters mentioned above, the absence
of a suture on the telson and the absence of the first pleopods in the
female, are common with the southern forms, but there is also a
peculiarity in the arrangement of Leydig's olfactory organs on the
external flagellum of the antennules which is found in Cambaroides
as well as in the Parastacidce. Moreover, if we consider the fact
that among the Parastacidce it is just the genus Astacoides^ from
Madagascar which shows, in the branchial formula, a similarity to
the Potamobiidce (although in other respects the gills are peculiarly
developed), it is easy to imagine, in trying to construct a connec-
tion between both families — and such a connection must have once
existed — that this was located between the area of Cambaroides
(Northeast Asia) and that o{ Astacoides (Madagascar). This v/ould

1902.] AND AXCIENT GEOGEAPHY. 295

be over India and China, generally over Southern and Eastern
Asia. Under this assumption, that crayfishes formerly existed in
Southeastern Asia, it also becomes clear by which way the rest of
the Parastacidce were geographically connected with the Potamo-
biidcB, namely, by way of the Indian Archipelago, from the conti-
nent of Asia over the Sunda Islands, New Guinea to Australia.

Looking over the various connections between the different
isolated areas of distribution of the different groups of crayfishes,
which have been suggested by the above chorological and system-
atical discussions, we may itemize them in the following way :

1. A connection of East Asia with North America by way of
Bering Sea.

2. A connection of Cuba with Central America (Mexico).

3. A connection of New Zealand with Australia, possibly over the
Fiji Islands and New Guinea.

4. A connection of Australia or New Zealand with South Amer-
ica.

5. A connection of Southeastern Asia with Madagascar and with
Australia.

We need further explanation of the following remarkable facts:

1. The absence of Potajnobiidce in Central Asia.

2. The absence of crayfishes in Southeastern and Eastern Asia.

3. The remarkable geographic restriction and isolation from each
other of the crayfishes of the genera Potamobius and Cambarus in
North America.

4. The remarkable boundaries of the area of Parastacus in South
America.

B. Chorology of the Family ^gleid^^ (See Fig. 2).

Here we shall leave for the present the crayfishes of the families
of the PotamobiidcE and Parastacidce and shall take up the small
group formed by the yEgleidce of Dana. This seems to be a mono-
typic family, consisting only of one genus and one species, ^glea
Icevis (Latr.). The following localities are recorded for it :

Chili : Valparaiso, and between Valparaiso and Santiago ; Lake
Llanquihue, near Puerto Montt.^ Argentina : Provinces Jujuy

1 See Ortmann, 1901, p. 1290.

2Doflein, F. SB. Akad. Muenchen, V. 30, 1900, p. 135.

296

ORTMANN — DISTRIBUTIOX OF DECAPODS

[April 3,

(this is the northernmost point, near the Bolivian boundary), Tucu-
man, San Luis/ Buenos Ayres.^ Uruguay. Southern Brazil : Rio
Grande do Sul and Santa Catharina,

As may be seen, the extremities of
the range on the Atlantic side, Sta.
Catharina and Uruguay, and the
southernmost locality in Chili, near
Puerto Montt, are also mentioned
for the genus Farastacus, and in fact
the distribution of Farastacus and
yEglea are almost identical (see figs.
I and 2), only y^glea seems to ex-
tend a little more to the north
(Jujuy). This similarity is the more
striking, since in both cases the chain
of the Cordilleras, which crosses the
area of distribution from north to
south, has absolutely no effect ; both
genera are found on either side of this mountain range, and in
the case of yEglece Icevis and Farastacus agassizi the identical spe-
cies is found east and west of the Cordilleras. This fact is very
significant, and important conclusions may be derived from it.

Fig. 2, Distribution of ^glea
Icevis (Latr.).

C. Chorology of the Freshwater Crabs of the Family
PoTAMONiD^ (See Figures 3 and 4.)

bibliography.

(^) Revisions, more or less coiiiplete.

Milne-Edwards, A. : «< Revision du genre Thelpbuse " {N'ouv. Arch. Mus,

Paris, V. 5, 1869, pp. 161-191).
Henderson, J. R. : "A Contribution to Indian Carcinology " ( Tr. Linn. Soc,
London, Ser. 2, Zool., V. 5, 1893, P- 3^0 ff.)-
Here a revision of the Indian species.
Ortmann, a. E. : " Carcinologische Studien " {Zool. Jahrb. Syst., V. 10, 1897,
pp. 296-329).
Revision in part, chiefly for the subgenera Potainonatites, Geothelphusa, and
the subfamilies Potamocarciniua; and I'richodactyiince.

iNobili, G. Boll. Mus. Torino, V. ii, No. 265, 1896.

2 I have received Irom the Museum in La Plata specimens that are labeled
Ensenada, Rio de la Plata,

1902.]

AND ANCIENT GEOGRAPHY.

297

298 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils.

Rathbun, M. J. : "A Contribution to a Knowledge of the Freshwater Crabs of
America. The Pseudothelphusinse " (^Proc. U. S. Nat. Mus.,Yo\. 21, 1898,

PP- 507-537)-

Revision of the subfamily PsetidothelphusincB = Potamocarcminoe.
DE Man, J. G. : " Notes sur quelques esp^ces des genres Parathelphusa et
Potamon, recueillies par M. Leonardo Fea pendent son voyage en Birmanie "
{Ann. Miis. Genova,Sex. 2, Vol. 19, 1898, pp. 384-440).

Here a nominal list of the described species of Potamon^ with localities.

{h) More recent sysfe?natic papers, not included in the above revisions.

BORRADAILE, L. A. : *' On a Small Collection of Decapod Crustaceans from

Fresh Waters in North Borneo " {Tr. Zool. Soc. London, I900, pp. 93-95).
DOFLEIN, F. : " Amerikanische Dekapoden der k. bayerischen Staatssamm-

lungen" {SB. Akad. Muenchen,No\. 29, 1899, pp. 187-188).
" Ueber eine neue Suesswasserkrabbe aus Columbien " {Ibid.y Vol. 30,

1900).
" Ostasiatische Dekapoden " {Abh. k. bayerischen Akad. IViss. ,Yol. 21,

1902, pp. 626-628, 662-663).
HiLGENDORF, F. : " Die Land- und Suesswasser-Dekapoden Ostafrikas " (in

Moebius, K. DeniscJi Ostafrika, Vol. 4, 1898).
Lanchester, W. F. : " On Some Malacostracous Crustacea from Malaysia in

the Collection of the Sarawak Museum" (^Ann. Nat. hist., Ser. 7, Vol. 6,

1900, pp. 255-257).
DE Man, J. G. : *' Note sur quelques Thelphusides recueillies par iSL Pavie dans

I'Indo-Chine " {Bttll. Soc. Philom. Paris, Ser. 8, Vol. 10, 1898, pp.

36-52).
. " Description d'une espece nouvelle du genre Potamon Sav. provenant du

pays des Somalis " {Ann. Mits. Geneva, Ser. 2, Vol. 19, 1898).
" Zoological Results of the Dutch Scientific Expedition to Central Borneo.

The Crustaceans. Part 2" {Not. Leyden Mus., Vol. 21, 1899, PP- 67-132).
"Description of a New Freshwater Crustacean from the Soudan" {Pr.

Zool. Soc. London, 1901, pp. 94-104).
NoBiLi, G. : " Viaggio del Dott. A. Borelli nella Republica Argentina e nel

Paraguay. Crostacei Decapodi " {Boll Mus. Torino, Vol. il. No. 222,

1896).
" Di uma nuova varieta della Thelphusa dubia racolta a KazunguU "

{Ibid.,^o. 262, 1896).
"Viaggio del Dr. Enrico Festa nella Republica dell' Ecuador. Decapod^

terrestri^e d'acqua dolce " {Ibid., Vol. 12, No. 275, 1897).
"^Decapodi e Stomatopodi racolti dal Dr. Enrico Festa nel Darien, etc."

{Ibid., No. 280, 1897).
" Supra alcuni Decapodi terrestri e d'acqua dolce" {Ann. Mus. Genova,

Ser. 2, Vol. 19, 1898, pp. 9-14).
" Intorno ad alcuni Crostacei Decapodi del Brasile " {Boll. Mus. Torino,

Vol. 14, No. 355, 1899).
•• Contribuzioni alia conoscenza della fauna carcinologica della Papuasia,

1002.] AND ANCIENT GEOGRAPHY. 299

delle Molucche e dell' Australia" [Ann. Mus. Geneva, Sen 2, Vol. 20,

1899, pp. 261-264).

NoBiLi, G. : " Decapodi e Stomatopodi Indo-Malesi " [Ibid., Ser, 3, Vol. 20,

1900, pp. 499-504).

" Decapodi raccolti dal Dr. Filippo Silvestri nell' America meridionale "

{Boll. Mus. Torino, Vol. 16, No. 402, 1901).
" Viaggio del Dott, Enrico Festa nella Republica dell' Ecuador.

Decapodi e Stomatopodi " {Jbid., Vol. 16, No. 415, 1901),
Rathbun, M. J. : " Descriptions de nouvelles especes de Crabes d'eau douce

appartenant aux collections du Museum d'histoire naturelle de Paris " {Bull.

Mus. Paris, 1897, pp. 58-61).
" Descriptions of Three New Species of Freshwater Crabs of the Genus

Potamon " {Pr. Biol. Soc. Washington, Vol. 12, 1898, pp. 27-30).
"The Decapod Crustaceans of West Africa" {Pr. U. S. Mus., Vol. 22,

1900, pp. 282-285).
" The Brachyura and Macrura of Porto Rico " {Bull. U. S. Fish Conwi.

for 1900, Vol. 2, I901, p. 23).
« Description des nouvelles especes de Parathelphusa appartenant au

Museum de Paris " {Bull. Mus. Paris, 1902, p. 184 ff.).
Weber, M. : "Die Decapoden Crustaceen des Suesswassers von Sued-Afrika "

{Zool. Jahrb. Syst., Vol. 10, 1897, p. 156).

According to Ortmann (1897) the family of Fotamonidce Ortm.
(= Thelphusidte Dan.) is divided into four subfamilies : PotamonincB
Ortm., DeckenimcB Ortm., PotmriocarcinincB Ortm.,^ and Tricho-
dactylmce Ortm. The first two belong to the Old World, the last
two inhabit the New World.*

I. Subfamily : Potamonin^.

The subfamily Potamonince. is in very poor condition, systemati-
cally. Not only our knowledge of the very numerous species is
rather incomplete, but also their arrangement into genera and sub-
genera is by no means satisfactory. Generally, it seems that we can
distinguish two genera: Parathelphusa M.-E. and Potamon Sav.
(= Thelphusa Latr. j, to which possibly a third one is to be added,

^ = Pseudothelphusina Ortmann and Rathbun (1898, p. 508). The division
into genera varies considerably with Ortmann and Rathbun respectively (see
below), and the name of the subfamily depends on the classification accepted.

2 According to Alcock {Journ. Asiat. Soc. Bengal, Vol. 69, 1900, p. 279),
also Gecarcinuc7is (one species in the peninsula of India), which was placed
hitherto with the family Gecarcinidce, belongs to the Thelphusidce (= Pota-
monidce). If this is so, we ought to create, possibly, a separate subfamily for
this genus.

PROG. AMER. PHILOS. SOC. XLI. 171. T. PRINTED NOV. 21, 1902.

300 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

the very incompletely known Erimetopus of Rathbun. The value
of a few other genera, created by various authors, is extremely
doubtful.

Parathelphusa is represented by typical species in the northern
parts of India, in Burma, Siam, Anam, Malacca, Southern China
(Hongkong and Canton), and in the Sunda Islands: Sumatra,
Borneo, Java, extending to Timor and New Guinea. With the
same genus some other forms have been classified which are found
in certain parts of Africa (Congo basin and Nile river; ; but these
have been placed by the present writer in a subgenus {Acanthothel-
phusa) of Fotamon, since they differ in their general shape very
strikingly from the Asiatic species oi Parathelphusa. Unfortunately
these African species are very poorly known ; only of the Nile
species figures have been published (Milne-Edwards and Hilgen-
dorf), and according to these it is impossible to unite this species
and its supposed allies with Parathelphusa}

As regards the genus Potamo?t, it is divided into several sub-
genera, which, however, are not very sharply defined. Aside from
the doubtful subgenus Acanthoihelphusa just mentioned, there are
three of them which are generally recognized : Potamon (sens.
strict.), Potamonautes Macl., and Geothelphusa Stps."^

The centre of the subgenus Potamon is, no doubt, in India and
Farther India. Thence it extends eastward to the greater Sunda
Islands (Sumatra and Java) ; it is found in the Philippine Islands,
but does not advance any farther in this direction. Northward it
enters China, where it is known from the Yang-tse-Kiang (see
Doflein, 1902, p. 662). It does not seem to pass beyond the
Himalaya Mountains to the north, but extends considerably west-
ward (possibly in a single species), going through Persia to the
Transcaspian countries, crossing the Caucasus Mountains and
extending to the Crimea ; from Mesopotamia it extends to Syria and
Asia Minor, where it reaches the Mediterranean countries, and here
it is found in Northern Egypt, Turkey, Greece, Italy, Sicily, and

^ Possibly Platythelphusa A. M.-E. (see Hilgendorf, 1898, p. 21) from Lake
Tanganyika also belongs here.

2 1 disregard, for the present, the subgenus Perithelphusa de Man (1899,
p. 70), which contains apparently rather primitive forms of Geothelphusa, and,
on account of its exclusive occurrence in Borneo, may be left united with
Geothelphusa. As to Platythelphusa, see the last note. As to Hydrothelphusa
A. M.-E., see below.

1902.] AND ANCIENT GEOGRAPHY. 301

farther in Algiers as far as Oran.^ It is a remarkable fact that this
subgenus is entirely absent from Africa proper, i.e., the part of it
that lies to the south of the Sahara Desert.

The subgenus Potai?ionautes, on the contrary, has its chief centre
of distribution in tropical Africa. It has been found, beginning at
Liberia, all along the western coast as far as Mossamedes. It is
found in the interior, in the region of the upper Zambesi (Kazun-
gula), extends over Transvaal to the Cape Colony, and northward
all along the eastern coast (Natal, Mozambique) to German East
Africa. Also in the eastern part of the interior it is represented,
for instance, in the headwaters of the Nile (Victoria Nyanza) and in
the Somali country. From the upper Nile it extends down the
Nile valley as far as Bahr-el-Gebel in the Egyptian Soudan. It is
also found on the Island of Socotra and in Madagascar, although
the species of the latter island do not seem to belong to the typical
form of this subgenus.^

^A. Milne-Edwards reports a species that is identical with an Indian (P. le-
schenaudi (M.-E.)) from i^Iauritius : this locality, however, lacks confirmation.
As regards the Madagassian species of Potamon, their systematic position is
doubtful, and they possibly do not belong to this subgenus. Compare next note.

2 Three species of Potavion are known from Madagascar. P. goiidoti (M.-E.)
(see A. Milne- Edwards, 1869, p. 172, PI. 8, Fig. 4) is a pecuhar form, but its
postfrontal crest distinctly points to Potamonautes, A. Milne-Edwards compares
it with P. obestim A. M.-E. from Zanzibar, and indeed it seems to be closely
related to it. The latter species is also an abnormal type of Potamonautes, and
forms with several others a group that is peculiar to East Africa; but there is
no reason to separate this group from Potamonautes, and thus we may safely
regard P. goudoti as a Potamonautes. The second species is P. madagas-
cariense (A. M.-E.) (^Ann. Sci. Nat. Zool., Sen 5, Vol. 15, 1872). As to this
form, the diagnosis of which is very brief, and which has not been figured, its
author says that it is a true Ihelphusa {i.e., subgenus Potamon'^, but this seems
hardly correct according to the descripion of the postfrontal crest, which is said
to be simply interrupted in the middle, while the median parts of it are not
advanced beyond the rest. This would better agree with Potamonautes. The
third species is regarded by A. Milne-Edwards {Ibid., 1872) as the type of a
separate genus, HydrothelpJmsa {^H. agilis A. M.-E.). This genus is said to be
characterized by the flat carapace, which is scarcely dilated and almost quad-
rangular, and by the horizontal front. The postfrontal crest is distinct and
interrupted. Since no figure is given, it is hard to form an opinion as to the
relation of this form to others, but it seems to be very peculiar.

Thus it seems that the Madagassian species of Potamon shoiv, in some
respects, a distinct relation to East Africa and the subgenus Potamonautes,
while in others they appear quite peculiar. (This is opposed to the opinioii
expressed by myself in 1901, p. 1290, footnote.)

802 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

The main range of Potamonautes in Africa seems to be almost
continuous, but absolutely isolated from it is a secondary centre in
South Asia. Here this subgenus is represented in India, and
thence it extends to Farther India, and reappears on some of the
islands : Pulo Condore on the coast of Cochin China, in the Philip-
pine Islands, Celebes and New Guinea. These latter localities are
distinctly discontinuous.

The third subgenus, Geothelphusa, undoubtedly has its centre in
the extreme East, and it is most characteristic for the Malaysian
Islands. On the Asiatic continent it seems to be absent ; but it is
found abundantly in Sumatra, Java, Borneo, and extends eastward
over Aru Island and New Guinea to North Australia, where it is
found on the Cape York Peninsula, and in Queensland as far as
Port Mackay.^ Northward this subgenus ranges over the Philippine
and Loo-Choo Islands to Japan, where it reaches its northernmost
station in the neighborhood of Tokyo.

On the continent of Asia typical species of tliis subgenus have
not been found ; indeed a few small species from India have been
described which might be united with this subgenus, but this is by
no means sure.

But this identical subgenus, Geothelphiisa, is apparently found in
another locality isolated from the rest of the range : this is
P. berardi (Aud.) from Egypt (Nile river). This species, however,
is also morphologically isolated from the rest ; and further, this
subgenus is recorded by Rathbun from Liberia (/*. w^^r^/^/j- Rthb.,
1898), and some species from East and Central Africa, related to
P. obesum, mentioned above, resemble, in the reduction of the
postfrontal crest, the subgenus Geofhelphusa,'^ while on the other
hand they are undoubtedly relited to the subgenus Poiatnonautes.
It is quite possible also that P. berardi from Egypt (Kairo south-
ward to Mount Elgon) belongs to this East African group. In my
opinion, all these species do not properly belong to Geothelphusa,
and we have to deal here again with a case of convergency : the

1 According to de Man, an Australian species i^P. transversum (Mrts.)) is also
found in the Fiji Islands ; but this lacks confirmation.

2 These are P. obesiim (A. M.-E ), Zanzibar; P. emini Hlgdf., P. new-
manni Hlgdf., P. pilosum Hlgdt. (Hilgendorf, 1898), all three from East
Africa and the region of the Great Lakes. Possibly P. socotrense Hilgendorf
(1883, Zeitschr. d. Natiirw.^ Ser. 4, Vol. 2) == P. granomm Koelbel i^SB.
Akad. Wien, Vol. 90, 1885) belongs here.

1902.] AND ANCIENT GEOGRAPHY. 303

tendency to reduce the postfrontal crest has developed in the East
African forms independently from the typical GeothelphuscB^ and
the East African (possibly also the Liberian) species form a pecu-
liar branch of Potamonautes.

The genus Erimetopus of Rathbun is found so far only in the
Congo basin.

Considering the distribution of the subfamily Potatnonince
in general, we see that it is continuous over the whole of
tropical Africa, then it extends through the Nile valley into the
Mediterranean regions and connects with the Asiatic range, which
goes from Syria over Mesopotamia, Persia to India, China and the
Malaysian archipelago, over which it finally reaches Northern Austra-
lia and Japan. This whole range is practically continuous, only
the larger continental islands (disregarding the smaller ones), Mada-
gascar and the Sunda Islands, the Philippines, New Guinea and
Japan, constituting breaks in the continuity.

V/ithin this large area, however, we are able to distinguish two
main divisions : an African, characterized by the prevalence of the
subgenus Pota?fwnautes, the complete lack of the subgenus Potamoii
(and possibly of Geothelphusa), and an Asiatic-Australian division,
characterized by the prevalence of the subgenus Potawon, the pres-
ence of Geothelphusa (in its eastern part), and the scarcity of
Poia77ionautes. Both divisions are practically connected by the
Nile valley ; this connection, however, does not seem to represent
the original condition, but suggests a secondary one, since different
types are here associated which are not at all related to each other.
Species of Potamonautes, to which subgenus, according to our opin-
ion, P, berardi also belongs, migrating northward from the Soudan,
have met here in Lower Egypt a species of the subgenus Potamon
{P. Jluviatile), which had migrated westward from India. Both
subgenera entered the Nile valley from different directions and
accidentally became occupants of the same territory, but the Nile
valley is not the route of migration by which African species
migrated into Asia or vice versa.

Aside from this narrow connection, the fauna of freshwater crabs
of tropical Africa is very sharply characterized and isolated from
Asia,^ and the fact is worth special mention that North Africa

1 The peculiarity of the African fauna is emphasized by the doubtful forms of
ParathelpJmsa (or Acanthothelphusa), and by Erimetopics.

30i ORTMANN— DISTRIBUTION OF DECAPODS [Aprils,

(Lower Egypt and Algiers) points, like the whole of the Mediter-
ranean region, to India, from which locality the species present there,
P. fluviatile (Latr.), has apparently migrated in an east-westerly
direction over Persia, Mesopotamia and Syria. P. fluviatile has
been actually recorded from western India; at any rate the most
closely allied species to this one are found in India and China.

Other remarkable facts in the distribution of this subfamily may
be summed up thus :

1. The Asiatic as well as the African part of the range is occu-
pied by the subgenus Potamonatites. It is impossible to say which
was the original home of Potamonautes, but this much is evident,
that it must have been present in both parts at a comparatively
early time, it being probably older than Potamon sens, strict. In
Africa Potatnonautes attained its highest development, being the
prevailing type there and showing great variety.

2. Madagascar, while belonging distinctly to Africa in its fauna,
possesses some rather peculiar types.

3. The subgenus Potamon originated in Asia, apparently at a time
when there was no connection any more with tropical Africa or
Madagascar. The immigration oi Potamon into the Mediterranean
countries, across Persia, etc., is probably a comparatively recent
one, since the route of immigration is easily traced and occupied
by one single species.

4. The Malaysian and Philippine Islands, Japan and North Austra-
lia possess in Geothelphusa a very peculiar group. This distribution
of GeotJielphusa does not correspond to that of ParathelpJiiisa^
Potamonautes and Potamon sens, strict., which are also found in the
Malaysian Islands. Potamo?iautes and Parathelphusa are similar in
this respect, possessing on the Sunda Islands only scattered stations
(as far as New Guinea), which by their discontinuity express an
ancient condition. Potamon points directly to an Indian origin,
extending only to Sumatra, Java and the Philippines, but going not
any farther to the east.

5. The position of Parathelphusa is hard to understand. If it
is really absent in Africa, as we believe, its distribution in Asia is
rather eastern than western, being chiefly found in Farther India.
Its extension over the Sunda Islands to New Guinea points to old
conditions. Since the morphological relations of Parathelphusa to
the rest of the subfamily are not well understood, it is better to
exclude it from our further consideration.

1902.] AND ANCIENT GEOGRAPHY. 305

Supposing that this subfamily must have had once a more or less
continuous distribution, we are to draw from this the following
conclusions as to the geographic conditions of the past :

1 . Africa and India ?nusi have been connected once. This connec-
tioHj however J was not by way of North Africa, Arabia and Persia,
and is possibly identical with that from Africa over Madagascar to
India, discussed above (see No. 5, p. 295).

2. Madagascar must once have been a part of Africa.

3. TJie Indo-Malaysian Islands, including the Philippine Islands,
Loo- Choo Islands and Japan, must have been once connected not only
between themselves, but also with New Gui?iea and North Australia
(as indicated by Geothelphusa). On the other hand, the distribution
of the typical forms of Potamon indicates that some of these islands
(Sumatra, Java, Philippines) were once connected with the continent
of Asia. Then, again, by Poianionautes (and Parathelphusa) the

former continuity of the whole region fro^n India to New Guinea is
indicated (see p. 295). It is evident that here repeated and important
changes of the mutual connections have taken place at different
periods of the past.

The history of the subfamily of Potayjionince would then be this :
Its centre lies in an Afro-Indian continental mass, which was divided
subsequently into two parts, tropical Africa and India. From India
the subfamily extended at a very early period over the Sunda
Islands, Philippine Islands, which consequently must have formed
a part of the continent, and this continental connection extended
as far as New Guinea and Australia, but not without repeated inter-
ruptions and changes. In the region of unstability and change lies
the home of the subgenus Geothelphusa, which was able at a certain
time to go as far north as Japan. A separate branch of the sub-
genus Potafnon was sent out from India westward, which finally
reached the Mediterranean countries, where it met in the lower
Nile valley a branch of the African subgenus Pota?nonautes which
came down the Nile from the south.

2. Subfamily: Deckeniince.
The second subfamily of the Old World, the Deckeniince, contains
only one genus, Deckenia Hlgdf. (see Ortmann, 1897, p. 3i4)> of
which three species have been described :
D. imitatrix Hlgdf. Interior of British East Africa : Taro (Hil-

306 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils.

gendorf, 1898, p 23) and Somali country (de Man, 1898,

p. 270).
D. mitis Hlgdf. (1898, p. 24). German East Africa and British

East Africa (Mombas).
D. alluaudi A. M.-E. and Bouv. (^ cristata Rthb.). Seychelle

Islands.

The Deckeniina: are, as is expressed by their morphological char-
acters (Ortmann, 1897, p. 297), a highly specialized group of the
family which may be connected without hesitation with the genus
Fofamon, and possibly with the African branch of it. This sub-
family is a group localized in East Africa, and the presence of one
of the species in the Seychelles indicates a former connection of
these islands with East Africa. It is quite probable that this con-
nection is an additional proof for that old Afro-Indian landbridge
discussed above, which included Madagascar (see No. 5, p. 295,
and No. i, p. 305).

3. Subfamily : PotamocarcinincB.

The subfamily Potamncarcinince (= Pseudothelphusince) is re-
stricted to America and is wanting in the Old World. The syste-
matic arrangement of it is a matter of discussion, since the two
revisions published by Rathbun and Ortmann do not agree as to the
principles of division.

Regarding the subfamily as a whole, its range comprises the
following parts : West Indies — Greater Antilles : Cuba (including
the Isle of Pines), Hayti, Porto Rico (including Santa Cruz) ; Les-
ser Antilles: Guadeloupe, Dominica, Martinique, Sta. Lucia. On
the continent its range begins in Mexico ; the northern boundary
is marked by a line beginning in Tepic Territory, running through
the States Jalisco and Guanajuato to Vera Cruz. Thence the range
covers the southern parts of Mexico, Guatemala, Nicaragua, Costa
Rica and Colombia, and extends eastward over Venezuela (includ-
ing Trinidad) and Guyana. In a southerly direction it passes from
Colombia into Ecuador, Peru and to Northern Bolivia. In the lat-
ter region it is found in the Cordilleras and the tributaries of the
upper Amazonas river. An isolated locality is Para, on the southern
side of the mouth of the Amazonas river {Pseudothelphiisa agassizi
Rthb.).

In order to get an idea of the distribution of the different genera

1902.] AND ANCIENT GEOGRAPHY. 307

of this subfamily, it is necessary to discuss the systematics of it.
Ortmann distinguishes four genera: Potamocarcinus, Epilobocera^
Hypolobocera and Kingsleya, while Rathbun accepts the following :
Epilobocera, Foia?nocarcinus, Pseudothelphusa and Rathbunia.
Generally, Ortmann 's Pota7nocarcinus corresponds to the genera
Poiamocarcinus and Pseudothelphusa of Rathbun, and the close
affinity of these two is also admitted by Rathbun, so that their
union (under Poiamocarcinus') is well supported. But in this case,
we are to exclude from Potamocarcinus the species sinuatifrons
Kgsl. (and Ortm., nee A. M.-E.) == haytensis Rthb., which be-
longs to Epilobocera. If we add this latter species to Ortmann' s
Epilobocera, this genus corresponds exactly to Epilobocera Rath-
bun. Hypolobocera of Ortmann is classed by Rathbun with Pseudo-
thelphusa {Potamocarcinus of Ortmann), and rightly so, as we now
believe. Kingsleya Ortmann is put by Rathbun with Potamocar-
cinus (sens, strict.); this, however^ does not seem to be justified,
since then the very peculiar shape of the orbita is neglected. While
in all other forms of the subfamily the lower orbital margin pos-
sesses on the inner end a suborbital lobe, which may unite with the
front, in Kingsleya the lower orbital margin itself joins the front,
while the suborbital lobe is hidden. This character, connected
with the extremely reduced condition of the exopodite of the third
maxilliped, which also does not find its like in the whole subfamily,
fully warrant, in our opinion, the creation of a separate genus. The
genus Rathbunia of Nobili is founded upon a single species, and
its chief character is taken from the shape of the meropodite of the
third maxilliped, which is narrower than usual at the proximal end.
In all other respects this genus agrees absolutely with Pseudothel-
phusa (resp. Potamocarcinus of Ortmann), and a generic separation
does not seem to be necessary.

As a compromise between both generic divisions I should like to
suggest the following:

Genus: Epilobocera Stps. (corresponding fully to Epilobocera
Rathbun).

Genus: Potamocarcinus M.-E. (= Potamocarcinus Ortm. (ex-
cluding sinuatifrons Ortm. = haytensis Rthb.) -f Hypolobo-
cera Ortm.).
I. Subgen. Potamocarcinus M.-E. (genus, according to
Rathbun, excluding the species latifrojis Rand.).

308 ORTMANX — DISTRIBUTION OF DECAPODS [AprU 3,

2. Subgen. Pseiidothelphusa Sauss. (= genus Pseudothel-
phusa Rthb.).

3. Subgen. Rathbunia Nobili (= genus Nobili and Rathb.).
Genus Kingsleya Ortm.

It is entirely a matter of taste whether one prefers to regard
Potamocarcinus , Pseudothelplmsa and Rathbunia as genera or sub-
genera. This much, however, is evident, that they are much more
closely allied to each other morphologically than to either Epilobo-
cera or Kingsleya. Judging from the third maxillipeds (which
furnish a good criterion in this respect), Epilobocera should be re-
garded as the most primitive form, Pota7nocarcinus (in the largest
sense) would be typical and Kingsleya the most specialized.

This division into three genera corresponds well to the geographi-
cal distribution of the different forms (see Rathbun, 1898, pp.

532-537)-

Epilobocera contains six species which are restricted to the Greater
Antilles: Cuba, Isle of Pines, Hayti, Porto Rico and Santa Cruz
Island.

Potamocarcinus (in the widest sense) contains 47 species*, which
cover the whole continental range of the subfamily from Mexico to
Bolivia and Para, the Lesser Antilles and of the Greater Antilles,
Cuba and Hayti. The subgenus Pseudothelphusa has the same
range, while ot the two species of Potamocarcinus (sens, strict.)
one is found in Guyana, the other in Costa Rica. Rathbunia is
known only from Darien. Kingsleya is so far known only from
Guyana.

The range of the subfamily on the continent seems to be perfectly
continuous ; only P. agassizi from near Para appears to be more or
less isolated. The most closely allied forms to this one {reflexi-
frons Ortm. and denticulatus M.-E.) are found in the region of
the upper Amazonas and in Guyana respectively, so that this
locality (Para) is possibly connected with Guyana. There is, how-
ever, the other possibility, that along the course of the Amazonas
river a connection exists between its lower part (Para) and its
upper (upper Amazonas). A very important fact is that Para is

^ Forty-two species mentioned by Rathbun, one described subsequently by
Doflein (1900, P, principessce, Colombia), one described by Nobili (1901, P.
caputii, Ecuador) ; these forty- four belong to Pseudothelphusa. Two species
belong to Potamocarcinus and one to Rathunia.

1902.] AND ANCIEXr GEOGRAPHY. 309

the only locality known for this subfamily to the south of the Ama-
zonas river, at least in Brazil. Generally, we may call this river
the southern boundary of the range of the subfamily, although in
the Cordilleras of Peru and Bolivia Poiamocarcinince are found more
to the south.

The localities of this subfamily in the West Indian islands are
now separated from the main range on the continent. Here we
can distinguish two groups : the Greater Antilles possess as a charac-
teristic type the genus Epi/oboce?'a, which is found nowhere else.
At the same time we have in Cuba three species of Pseudothelphusay
of which one {americana) is also found in Hayti. This same spe-
cies, F. americana Sauss., is found largely distributed in Mexico
(States of Guanajuato, Morelos, Puebla, Guerrero, Oaxaca), and,
further, another Cuban species {terrestris Rthb.) has also been
reported from Mexico (Jalisco and Tepic), while the third species
{affijiis Rthb.) is restricted to Cuba.^

The second group within the West Indies is formed by the islands
of Gaudeloupe, Dominica, Martinique, St. Lucia, where one species
{^P. dentata (M.-E.)) is found. According to Rathbun (1898, p.
524), the most closely allied forms to this 2.xq P. garjtiani Rthb.
from Trinidad and Venezuela, and P.fossor Rthb. from Venezuela.

The above chorological and systematic facts justify the following
conclusions :

1 . The distribution of the PotamocarcinincE in Central and South
America is remarkable, in so far as it does not go southward beyotid
the Amazonas river.

2. The West Indian islands must have been once connected with
Central and South America. The freshwater crabs of the Greater
Antilles poi7it to a connection with Mexico, as well as to a connection
between themselves, after they were separated from the mainland
{Epilobocera). The freshwater crabs of the Lesser Antilles point
to a connection with Tri?iidad a fid Venezuela.

Connection of the Poiamocarcinince and Potamonince.

As is accepted by all authors, the affinity of the PotajJiocarcinince
of the New World with the Potamonince of the Old World is
beyond question, and this affinity is expressed by their position as

1 This locality, given for a specimen from the old collection of Guerin in
Philadelphia, needs confirmation.

310 ORTMAXN — DISTRIBUTION OF DECAPODS [Aprils,.

two subfamilies within the same family, Potamonidce, which has
never been disputed.^ Consequently the idea suggests itself that
both subfamilies have a common origin, or have descended the one
from the other. Transitional forms between them are not known ;
this, however, is not astonishing if we consider their geographic
isolation.

The present writer has called attention to the presence in Cen-
tral Africa of a group of Potamon, v*^hich he has designated as the
subgenus Acanthothelphusa. These species have been united by
others with Parathelphiisa, which classification we do not consider
to be correct. Although these species are very poorly known, it
seems impossible to unite the type-species of Acantliothelphusa
(from the Nile) with Parathelphusa^ and it would be well to
examine the other species more closely with a view to their possible
relation to the American Potamocarcinince.

Whether this prove to be so or not, this much is unquestionable,
that the West African PotamonincB are geographically most closely
approached by the South American PotaiJiocarcinitue^ and thus a
former connection of the respective parts ^ West Africa and npi'thern
South America, is suggested (see Ortmann, 1901, p. 1291).

4. Subfamily: Trichodactylince. (See Fig. 4.)

Finally, we are to consider the subfamily Trichodactylince, Ortm.,
which is divided, according to Ortmann (1897, p. 298), into two
genera, Trichodactylus Latr. and Orthostoma Rand., which latter
generic name, however, is to be abandoned as preoccupied. Its
place is to be taken by Sylviocarcinus or Dilocarcinus M.-E,, 1853.
But even Trichodactylus and Dilocarcinus (in its largest sense in-
cluding Sylviocarcinus, according to Ortmann, and being identical
with Orthostoma^ are not always sharply defined, and, further, the

^ According to Ortmann {Zool. Jahrb. Syst„Vo\. vii, 1893, p. 430), the Thel-
phusidcB {Pot a VI 071 idee) are possibly derived from MenippidcB — i.e., primitive
Xanthida (in Alcock's sense). They are primitive Cyclometopa, which, how-
ever, in certain characteristics, probably connected with their liabits^ are more
highly and abnormally developed, and exhibit (due to convergency ?) similarities
to the Catometopa.

Alcock {Jour. Asiat. Soc. Bengal, V. Iviii, Part 2, No. i, 1899, p. 3) is inclined
to regard the Thelphusidce as descendants of the Oziince or Eriphiince (higher
Xanthidce'), and takes them for very highly specialized Cyclometopa.

Both views agree in that the family Xanthiidce is supposed to be the ancestral
stock of these freshwater crabs.

1902.

AND ANCIENT GEOGRAPHY.

311

?

Fig. 4. Distribution of the Crabs
of the subfamily Trichodactylince.

distinction of species seems to be very arbitrary within these
genera. Up to the present, about five or six species of TricJio-
dactylus and about fourteen species of Dilocarcinics have been de-
scribed. In the following we shall discuss them all together.

The subfamily covers an area
that comprises the larger south-
ern half of Brazil (Bahia, Rio de
Janeiro, Goyaz, Minas Geraes,
S. Paulo, Sta. Catharina, Rio
Grande do Sul). It is found in
Paraguay, and in the Argentin-
ian provinces : Missiones, Chaco
and near La Plata (Ensenada).^
Further, species of Trichodacty-
lince are very abundant in the
Cordilleras, in the region of the
headwaters of the Amazonas
river, namely, in Bolivia (pro-
vince Beni, Yocuma river, be-
longing to the upper Madeira), in
Peru (rivers Ucayali, Huallaga), and in the Maranon at Nauta and
Loreto (Ecuador). Since there are also representatives known from
the lower Amazonas (Island Marajo), all these localities named seem
to form a continuous area, which extends from the Amazonas
river southward to La Plata, and from the Atlantic Ocean westward
over Brazil, Paraguay and Argentina to Bolivia and Peru, where it
reaches the eastern slopes of the Cordilleras. Apparently isolated
from this area, several species are found in Guyana, and, finally, one
species {Trichodactylus quinquedentatus Rthb.) is known from the
upper parts of the river Magdalena in Colombia, and from the
Escondido river in Nicaragua.

It is quite possible that the isolated stations in Guyana, Colombia
and Nicaragua will be connected by subsequent discoveries
(Colombia is very near to the localities of the upper Amazonas),
and then we would have for this subfamily a continuous range,
which comprises the whole of South America southward to La
Plata, and from the Atlantic Ocean to the eastern slopes of the
Cordilleras, and which extends in Central America as far as

1 I have received from the La Plata Museum specimens of Dilocarcinus
panoplus (iSIrts.) from Ensenada.

312 ORTMANN — DISTKIBUTIOX OF DECAPODS [Aprils,

Nicaragua. It is to be remarked that none of the localities is
situated in the drainage of the Pacific Ocean, but all are in that of
the Atlantic.^

This distribution does not offer any remarkable facts. The
TrichodactylincR seem to belong to the tropical parts of the
Atlantic slope of South America, and their centre is somewhere in
Brazil; from Brazil they extend in every direction until, in the
east the Atlantic Ocean, in the west the Cordilleras, in the
south the climate of Argentina form barriers. To the north the
most advanced station is in Nicaragua ; here no natural boundary
(climatic or topographic) is marked.

Further speculations as to the distribution of this subfamily do
not seem to be very promising until we are better acquainted with
the chorological facts. The whole appearance presented by the
distribution is a recent one ; probably it is continuous and, in
most directions, limited by natural boundaries. In this respect it
is strikingly distinguished from the other groups of the family
Fotamonidce discussed above,

I have the impression that the TrichodactylincB are not so closely
connected, systematically, with the other subfamilies of the Fota-
monidce as was believed hitherto. In fact, transitional forms to any
of the other subfamilies are not known, and the Trichodactyli7ice
are morphologically isolated and sharply defined. Moreover, the
whole *' habitus" of these crabs is so entirely different from that
of the PctamocarcinincE that it is worth while to revise the syste-
matic relations of these groups. As I venture to imagine, it will
be found, possibly, that the Tj-ichodactylincE form a group that is
much more sharply isolated, systematically, and that has little to
do with the family Potainonidce. This much is evident : according
to its morphologic isolation, we ought to expect that the Tricho-
dactylijice are a comparatively ancient group ; but this is contra-
dicted by their distribution, which possesses a remarkably recent
character.

These are the reasons why we shall exclude the Trichodaciylince
from our further discussions.

^ This is contrary to what we have in the PotamocarcinincE, which are found
also on the Pacific slope in Ecuador, and especially in Central America and
Mexico.

1902.] AND ANCIENT GEOGRAPrf:Y. 318

PART II. RECONSTRUCTION OF ANCIENT GEOGRAPHIC CON-
DITIONS.

BIBLIOGRAPHY.^

(a) Papers Written from a Zoogeographical Standpoint.

Hedley, C. " Considerations on the Surviving Refugees in Austral Lands of

Ancient Antarctic Life" (^Proc. R. Soc. N. S. Wales, 1895).

«' A Zoogeographical Scheme for the Mid-Pacific" [Proc. Linn. Soc. iV. S.

IVales, 1899, pp. 391-417).
Huxley, T. 77/.? Crayfish, London, 1879.
Ihering, H. VON.2 " On the Ancient Relations between New Zealand and South

America" {Trans. New Zealand Instil., Vol. xxv, 1891, pp. 431-445).

" Die Ameisen von Rio Grande do Sul " {Berlin. Enlomol. Zeitschr., Vol.

xxxix, 1894, pp. 321-446).
Jacobi, A. " Lage und Form biogeographischer Gebiete " {Zeitschr. Gesellsch. f.

Erdkunde Berlin, Vol. xxxv, 1900, pp. 147-238),
Kobelt, W. " Studien zur Zoogeographie — Die MoUusken der palsearktischen

Region," Wiesbaden, 1897).
Ortmann, 4- E. "Ueber Bipolaritaet in der Verbreitung mariner Thiere "

{Zool, Jahrb. Syst., Vol. ix, 1896, pp. 588-594).

" Raeumliche Verbreitung," in Bronn's Klassen und Ordniingen des Thier-

reiches. Vol. v. Crust., 1901, pp. 1285- 1293.

" The Theories of the Origin of the Antarctic Faunas and Floras " {Avier.

Natural., Vol. xxxv, 1901, pp. 139-142).

«' Tertiary Invertebrates," in Rep. Princeton Exped. Patagonia, Vol. iv,

Part 2, 1902, pp. 310-324.
OsBORN, H. F. " The Geological and Faunal Relations of Europe and America

during the Tertiary Period and the Theory of the Successive Invasions of an

African Fauna" {Science, April 18, 1900, pp. 561-574; see also^;z;/. AL Y.

Acad. Sci., Vol. xiii, 1900, pp. 1-72).
PiLSBRY, H. A. "Distribution of Helices in Time and Space," in Manual of

Conchology, Ser. 2, Vol. ix, 1894, pp. xxxviii ff.
SCHARFF, R. F. " Etude sur les Mammiferes de la region holarctique " {Mem.

Soc. Zoo I. France, 1895),

"On the Origin of the European Fauna" {Proc. R. Irish Acad., Ser. 3,

Vol. iv, 1897).

(^) Papers of a Geological Character.

Dana, J. D. '^ Manual of Geology i' Fourth Edition, 1895.

Hill, R. T, « The Cretaceous Formations of Mexico and their Relations to

1 Only the more important papers are given in the following list. Others,
quoted only incidentally, shall find their place in footnotes.

2 I quote only the following two papers of von Ihering, although he has pub-
lished several more on these and kindred subjects. But these two contain the
essence of his theories.

3] 4 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

North American Geographic Development" {Amer. Jotirn. Sci., Vol. xlv,

1893)-
Hill, R. T. " The Geological History of the Isthmus of Panama and Portions

of Costa Rica" (Bu//. Mus. Harvard, Yo\. xxviii, 1898).
KoKEN, E, Die Vorwelt t*nd ihre Entwicklungsgesc/iichtey l?>g-^.
KossMAT, F. " Die Bedeutung der suedindischen Kreideformation fuer die
Beurteilung der geographischen Verhaeltnisse waehrend der spaeteren
Kreidezeit" {Jahrb. k. k. Geol. Reichanst., Vol. xliv, 1895).
Meddlicott, H. B,, and Blandford, W. T. A Manual of the Geology of

India, Vol. i, 1879,
Neumayr, M. Erdgeschichte, 1890.
SUESS, E. Das Antlitz der Erde, Vol. i, 1883- 1885 5 Vol. ii, 1888.

"Beitraege zur Stratigraphie Central- Asiens " {Denksckr. Akad. Wiss.

Wien,Vo\. Ixi, 1894).

In the following we shall endeavor to answer the questions :
What con7tections are suggested by the distribution of the freshwater
Decapods^ and Is there any other evidence, in the first place, of a
geological character to support them ? The solution of these ques-
tions will furnish us the key for the reconstruction of the old geo-
graphic conditions.

If we recall the connections suggested by the distribution of the
freshwater Crustaceans, we can collect them in the following list :

1. Connection of nortlieast Asia with 7iorthwest America across
Bering Sea (see pp. 290, "291, 295).

2. Connection of east Asia with Australia (see pp. 295, 305).

3. Connection of south Asia with Madagascar a?id Africa (see
pp. 295, 305, 306).

4. Connection of New Zealand with Australia (see p. 295).

5. Connection of Australia {resp. New Zealand^ with South
Ai7ierica (see p, 295).

6. Connectioti of the West Lidies with Ce?itral, resp. South
America (see pp. 295, 309).

7. Connection of South America witli Africa (see p. 310).
Other important questions arose out of the distributional facts,

which may be classified under the following heads :

8. General relations of North, Central and South America (see
pp. 295, 309).

9. Relations of Africa to the rest of the world (see pp. 303, 304).

10. Relations of Europe to Asia {and Africa) (see pp. 291,

295. 304).
We shall take up these different items in the order here indi-

1902] AND ANCIENT GEOGRAPHY. 315

cated. But before we do so, we have to say a few words by way
of explanation and introduction, characterizing the value of the
study of the freshwater Decapods for these purposes.

In all the following discussions, the fundamental supposition has
been made that freshwater crayfishes, as well as freshwater crabs,
do not possess any exceptional means of dispersal ; that is to say,
that they are restricted to fresh water and cannot exist in salt water ;
that they cannot leave the water for any continued period, and con-
sequently cannot migrate over land to any extent ; and, finally,
that they do not possess in any stage of their life, and especially not
in the egg or larval stage, any means or devices which permit their
passive transport. We may specify these three points in the follow-
ing way :

1. The restriction to fresh water is not absolute. There are a few
exceptions, namely :

Potamobius pachypus (Rthk.) is found in the Black and Caspian Seas

in brackish and salt water.
Potamobius trowbridgei (Stps.) has once been collected in salt water

at Monterey, California (Faxon, 1898, p. ddd).
Cambarus uhleri Fax. is characteristic for the marshes of the coast

of Maryland, and lives in fresh, brackish, and salt water.
Cambarus montezumcz Sauss. has been found, in one case, in a salt

lake in Mexico (Lake Tezcoco, near City of Mexico; see

Faxon, 1885, p. 123).
Potamon fluviatile var. ibericum (Bisb.) is found in fresh water and

salt water of the Caspian Sea (see Ortmann, 1897, p. 302;.

On account of the small number of these cases, we have to re-
gard them as exceptional, and they are, no doubt, secondary
adaptations. In fact, none of these species is a true saltwater
form, they being always more or less euryhalin, and frequenting
also brackish or freshwater. Thus we may say, generally, that
both, crayfishes and crabs, discussed here, are true freshwater ani-
mals, and preeminently so, and that a migration across oceans or
parts of oceans is practically excluded.

2. Being animals breathing by gills, crayfishes and freshwater
crabs cannot leave the water. This rule is without exception with
the Potamobiidce and Parastacidce ; they may leave the water for a
short time, but a prolonged stay outside of it is always fatal. There
are only a few species in North and South America, and in Aus-

PROC. AMER. PHILOS. SOC. XLI. 171, U. PRINTED NOV. 22, 1903.

316 OKTMA^'N — DISTRIBUTION OF DECAPODS [Aprils,

tralia, burrowing in mud, which leave the water habitually; but
they always have to return to the water to moisten their gills, and
their burrows end in water. The forms most adapted to a subter-
restrial life are probably the two species of Engceiis in Tasmania. In
general, yi?r the crayjishes, tracts of land without water {^deserts) are
absolute barriers.

The PotamobiidcE lead a rather amphibic life and leave the water,
in many cases, habitually. Yet they always depend on the presence
of water and cannot go far out of easy reach of it. Some of the
species (^Potamon fluviatile in Persia, etc.) live in steppes, where
there is a scarcity of water, but here they always are found near
some kind of water supply. In general, they also cannot exist in
deserts.

3. As in all other Decapods, also in crayfishes and freshwater
crabs the eggs are carried and hatched under the abdomen of the
female. There is, as far as we know, no free metamorphosis of the
young (known in PotamobiuSj Camba^^tis, Potamon), and the young
hatch in a stage similar to the parents. Thus there seems to be no
means which effect, under normal conditions, an increased facility
of dispersal in an active or passive way among the young ones.
There may be, occasionally, a passive transport by other animals
(water fowl), but such cases can only be exceptions and have never
been observed. The whole character of the distribution of the
different species is against the assumption of exceptional means of
dispersal.

I. CONNECTION OF NORTHEAST ASIA WITH NORTHWEST AMERICA BY
WAY OF BERING SEA.

A connection of northeast Asia with northwest America is pos-
tulated, as we have seen above, by the presence of Potamobiidoi in
the region of the Amur river, Korea, and north Japan on the
one side, and in western North America on the other ; the direc-
tion of this connection is indicated by the presence of Poiamobius
nigrescens (Stps.) in Unalaska.

This connection is mentioned by Jacobi (1900) under his
"regions of dispersal " (" Ausbreitungsgebiete "), and is called
by the name of *' Berings-Strassen-Ausbreitungsgebiet." This is
well known among zoogeographers. In fact, for an explanation of
the very peculiar conditions of distribution of many animals of the
northern hemisphere, a former connection of the northern land

1902] AND ANCIENT GEOGRAPHY. 317

masses of the Old and New Worlds is absolutely necessary, and the
similarity of the land faunas of both parts, which is not explained
by the present conditions, is so strong that these regions (northern
Eurasia and North America) have been united by certain authors
into one zoogeographical region, the Holarctic. As to the location
of this connection, two ways are possible : either from Siberia to
Alaska, or from Labrador over Greenland to Scandinavia. The
latter connection, which has been discussed, from a geological
standpoint, chiefly by Suess and Neumayr (for older times, Meso-
zoic and Tertiary), and, from a zoogeographical view by Scharff
(for the Pleistocene), may be disregarded for our present purpose ;
there is no indication for its existence among the crayfishes. But
the latter support strongly, as has been said, the other connection
over Bering Strait.

Viewed from the tectonic side, this connection is quite possible.
The old rocks of northeast Asia are continued into northeastern
Siberia (east of the rivers Lena and Aldan) to the river Kolyma, ^
and farther, toward the Arctic Ocean and Bering Sea, and similar
rocks are found in Alaska ; and further, the chain of the Aleutian
Islands is, according to Suess, another proof for the tectonic unity
of the lands east and west of Bering Sea.

As regards the time of existence of this land bridge, we have to
assume it during almost the whole of the Tertiary period. Osborn
(1900) takes its existence for granted and demonstrates (p. 568)
that during Eocene, Miocene and upward to the Pliocene, a regu-
lar exchange of the faunas of Eurasia and North America took
place. In the older Pleistocene (p. 571) this connection still
existed, but was interrupted in the middle Pleistocene (p. 572).

If we put the question, whether and how far this land bridge goes
back in Pretertiary times, we have to consult first Neumayr's opin-
ion as to the distribution of the Jurassic oceans and continents
(1890, map, p. 336). It is true, in Siberia, deposits of Lower
Jurassic age are not known, and possibly Siberia was land during
this time ^ There are found here, however, deposits belonging to

iSee Tscherski, Sap. Akad. St. Petersb., Vol. 73, Append. 5, 1893 (Russian) ;
Review in N. Jahrb. Mineral., etc., 1896, Vol. 2, p. 318.

2 Land and freshwater deposits of Jurassic age are largely distributed in Sibe-
ria as coal-bearing strata. Compare the geological investigations connected
with the great Siberian railroad, by Obrutchew, Gerassimow, Gedroiz, Jawor-
owksy. Reviews in AL Jahrb. Mineral., 1899, Vol., 2, p. 111-116.

318 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

the Upper Jurassic (Neumayr, 1890, p. 329), which reach the
Pacific Ocean. Beds of the same age are known on the Aleutian
Islands and in Alaska. These deposits exhibit a peculiar chaiacter,
which has been called the boreal or arctic type, and in this respect
the Jurassic beds of the western coast of North America are very
important, since they agree with the boreal type. Neumayr con-
cluded from this that the Upper Jurassic Polar Sea sent an exten-
sion southward along the western coast of North America into the
North Pacific, and its fauna also extends in this direction; by this
extension of the Polar Sea, east Asia was separated from North
America. Consequently there was no land bridge.^

These conditions of Upper Jurassic times continued, according
to Neumayr, into the Lower Cretaceous ; the Wolga stage, with its
characteristic Aucella-beds, belongs in part to the Lower Cretace-
ous, and the Polar basin was also in the beginning of the Creta-
ceous in open communication with the northern Pacific. This is
represented in Koken's map (1893, pi. i), although Asia and
North America approach each other considerably. This same
map, however, expresses, for the Upper Cretaceous, a separation of
the Polar Sea from the Pacific, and this land connection between
Asia and North America is preserved in Koken's map for the older
Tertiary (/. r., pi. 2). The evidence for this disconnection of the
oceans in the Upper Cretaceous time is given by Neumayr (1890,
p. 389-391) ; palseontologically, we can trace a continuous Upper
Cretaceous ocean, including the northern Pacific from California
to Japan, which was connected with south India. This province
differs strikingly from the American-European (Atlantic) province ;
the Polar Sea was much reduced in size, and, to all appearances,
Siberia was largely dry land and was connected with North
America.

Thus there is some evidence of the existence of a land connec-
tion between Siberia and Alaska, beginning at about the middle of
the Cretaceous period, and continuing up to the end of the Ter-
tiary. Whether this connection was continuous in time, or inter-
rupted at certain periods, is hard to decide ; at all events, it was
of such a character that an easy and free communication was pos-
sible between the respective parts, and this is expressed very dis-
tinctly in the faunas of the northern land masses, although the

1 This Jurassic ocean forms apparently the continuation of the old Triassic
basin, comprising the Pacific and Arctic Oceans (see Neumayr, p. 266).

1902.] AND ANCIENT GEOGRAPHY. 3l9

geological evidence is very slender on account of our defective
knowledge of the Beringian countries.

For our present purposes, this has the following meaning. The
Potamobiidcc of eastern Asia, the remnants of which are known as
the subgenus Ccmibaroides, had easy access to northwestern America
by way of the Beringian connection, from the beginning of the Upper
Cretaceous to the end of the Tertiary. Since Cavibaroides is to be
taken, as we have seen above (p. 288), for the more primitive group,
the migration must have been in an easterly direction. It cannot
have taken place in recent times, since this way is now rendered
impossible, and just this recent (or Pleistocene) interruption (prob-
ably connected with a change of climate) has separated the Asiatic
and American range of Potamobius. Before the middle of the
Cretaceous it was also impossible for the crayfishes to pass along
this line, since then this connection was not yet formed, and thus
we obtain a very important lower limit for this process in the dis-
persal of the PotainobiidcB about the middle of the Cretaceous period.
Consequently, the Potamobiidce may go back, in their geological
history, at least as far as this time. We shall see later that we are
able to define also an upper limit for the time of immigration into
North America.

2. CONNECTION OF EASTERN ASIA AND AUSTRALIA.

Another geographic postulate of the distribution of the Potamo-
biidce. and Parastacidce is the connection of east Asia, the region
of Cambaroides, with Australia, the main region of the family
Parastacidce. This same connection, from Farther India and south
China over the east Asiatic islands to north Australia, is suggested
by the distribution of the subfamily of the PotamonincE.

Other zoogeographical facts point the same way. Pilsbry (1894,
p. xlv) says that eastern Asia and China, southward to Australia,
constitutes a great division in Helix distribution, and many other
writers have emphasized the close affinity of the fauna of Australia
to that of southeastern Asia, although this is true only for certain
groups of animals. The opposite opinion, which generally pre-
vails, that Australia is sharply isolated from the rest of the world
in its faunal relations, is founded chiefly on the highest forms of life,
the Mammals. Other groups of animals, which permit us to draw
conclusions in this respect, indicate clearly that a large part of the

320 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

Australian fauna is derived from Asia (see von Ihering, 1894, p.
406, and Hedley, 1899).

This connection between east Asia and Australia (Sino-Austra-
lian) is not well expressed in Jacobi's scheme. The apparent
reason for this is that Jacobi considered chiefly those groups of
animals (Mammals, Birds) which do not bear upon this question.
Nevertheless, some of his *' regions of dispersal " come under this
head, namely, the ninth, tenth, eleventh and twelfth (Papuan, Far-
ther Indian, Philippine, southern Japanese ; see Jacobi, 1900,
pp. 208-210), and discussing the Papuan, he directly mentions the
Oriental origin of certain elements of it, thus indicating its 'rela-
tion to southeastern Asia.

Studying the tectonic configuration of the repective parts, we are
to remember that Australia belongs to the old. Palaeozoic Gond-
wana land of Suess (1888, p. 317 ff.), which also comprised Africa
and India. But we cannot refer to this old connection of Austra-
lia with India, since India in turn was not then united with the rest
of Asia, and since this connection was destroyed in very early
times, possibly in Palaeozoic. For a tectonic connection of Aus-
tralia and eastern Asia (excluding India) we have only evidence to
the contrary.

On the other side, the eastern parts of present Asia, especially
China, northeastern Siberia, and Farther India, form a more or less
complete tectonic unit. Suess (1888, pp. 206-242) has shown that
this whole region consists largely of old archaic and palaeozoic
rocks, which form, in northern China, an old continental mass, in
the south a series of folded mountain ranges (p. 287), which con-
tinue into the mountains of Tonkin and Anam as far as the mouth
of the river Mekong. In this whole region no Mesozoic deposits
(with the exception of Rhaetic beds in Tonkin) are known. Ac-
cording to Koken,^ a Triassic ocean extended from the region of
the Himalaya mountains and Central Asia to the shores of the
present Pacific, covering a large part of China. The latter may
have been land before Rhaetic times; but at present we have only
evidence that it was surely land in the Jurassic period.^

1 Neues Jahrb. f. Mineral.^ etc., 1900, Vol. I, p. 196.

^ See Loczy, L, von, PVissenschaftliche Ergebnisse der Reise des Grafen
Bela Szeche7iyi in Oslasien,Wo\. 2, 1899; the Central-Chinese sea (south of the
Kuen-Lun mountains) disappeared at the end of Triassic and in Jurassic times.

1902.] AND ANCIENT GEOQEAPHY. 321

Thus it is clear that we may assume the existence of this conti-
nent, the Sinic, from the Jurassic upward.

Further, according to Suess, the chains of islands accompanying
this old continent on its southern and eastern sides are tectonically
connected with the latter. One of them is formed by the moun-
tain ranges which form the Japanese and Philippine Islands, con-
sisting of old rocks, and, in the south, we can trace a similar chain
(Suess, 1885, PP- 579-5^^)' which begins with the Burmese ranges,
and extends over Malakka, Sumatra, Java eastward, possibly as far
as New Guinea.

Thus, nothing in the tectonic configuration is opposed to the
theory that at least a large part of the Indo-Malaysian islands belongs
to the continent. But this does not give us any proof for an actual
former connection of these islands with the Sinic continent. This
can only be decided by geological investigation of the respective
parts. Unfortunately, our knowledge in this respect is very
scanty.

Neumayr (1890) constructs in his map, mentioned above, an old
Jurassic continent, the Siiio-Austi-alian, which, with reference to
eastern Asia, is well supported, and the Australian part of which
is also established by the fact that large parts of Australia possess
a very old age (Gondwana land). The connection of both goes
over the present Indo-Malaysian Archipelago, and, according to
the map, this region was largely land during Jurassic times.
Further (Neumayr, 1890, p. 419), Australia became separated from
Asia and the rest of the world before the end of the Mesozoic time,
that is to say, probably in the Cretaceous. This same idea is ex-
pressed by Koken (1893) ^^ ^^s map of the distribution of land in
the latter period. Here we see that Asia and Australia were dis-
connected during the Lower as well as the Upper Cretaceous, but
Australia comprises parts of New Guinea and the Sunda Islands as
far as Java, Borneo and the Philippine Islands. In the older Ter-
tiary, Koken includes Farther India into Asia, but then follows an
archipelago and Australia remains isolated.

This wide connection, drawn by Neumayr between Asia and
Australia during the Jurassic period, does not seem to be well sup-
ported, since marine Jurassic deposits have been discovered in the
region of the Malaysian islands.^ On the other hand, it is settled

1 In Borneo, according to Krause and Vogel {Sainml. geol. Reichsmus. Ley-
den, Vol. 5, 1897). The so-called "old slates " of Borneo are said to belong to

322 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

that a number of these islands possess very old, possibly Archaic
rocks, which are overlaid directly by Tertiary beds, thus giving
evidence of an intervening extended land period, during which
no sedimentation took place. This has been demonstrated for the
Philippine Islands, where an Eocene, Miocene and Pliocene series
follows on top of old crystalline schists.^ Similar conditions are
said to prevail in Java (Martin). This, however, seems to be
doubtful, since Verbeck and Fennema,^ although they do not posi-
tively deny the possibility of the existence of Archaic rocks, pro-
nounce the schists of Java Cretaceous, upon which, unconformably,
Eocene and younger Tertiary beds are deposited. Archaic rocks
are found in the Island of Amboina, where they are overlaid by
Tertiary and Quarternary coral limestones. Between both there
are, locally, older sediments of undetermined age.^

Aside from the supposed Cretaceous schists in Java, we know of
beds of this period in Borneo, and, according to Kossmat (1895,
p. 469 f.), only such that belong to the Upper Cretaceous, corre-
sponding to the Ariyalur group (Senonian) of India. This fact is
the more important, since, as Kossmat points out, it demonstrates
that the Upper Cretaceous of southern India can be traced over
Assam and Borneo to Japan and the Island of Sachalin (and thence
to the western coast of North America). This indicates a contin-
uity of the oceans in this direction, and consequently Australia and
Asia must have bee?! disconnected in the Upper Cretaceous.

From the foregoing, the conclusion may be drawn, that the
geology of the Indo-Malaysian Archipelago is 100 scantily known to
form an adequate idea of the former connection of Australia and
Asia. This much, however, is settled, that large parts of this
archipelago were once land, and the single islands were in many
cases connected with one another. Verbeck (/.r.) has shown that
of the Island of Java, in Miocene time (that is to say, very late), only
the western part existed as a unit, and that it was continued east-
ward by a series of small islands. At the end of the Tertiary these

the Lias (Martin, ibid.^ Vol. 5, 1898; see also MolengraafF, G. A. F., Geolo-
gische Verkennigstochten i7i Central Borneo^ 1900).

1 Martin, ibid.^ Vol. 5, 1896.

2 Geologiske beschryving von Java en Madera, 1897. See also Verbeck, in
Petermanns geograph. Mitteil.^ 1898,

3 Martin, K., Reiseii in den Molukken, in Ambon, aen Uliassern, Seran und
Buru. Geolog., Teil i, Leyden, 1897.

1902.] AND ANCIENT GEOGRAPHY. 323

islands became connected, and tlie whole was united with the conti-
nent of Asia; subsequently, a new (Quarternary) subsidence took
place. According to Weber,^ Celebes was connected in early times
(beginning of the Tertiary ?) with eastern Asia, but was separated
later and dissolved into smaller islands, and assumed its present
form at the end of the Tertiary. If changes of this character took
place during the comparatively short Tertiary period, we are to
expect, in Pretertiary times, much more varied conditions, and it is
by no means impossible that the different islands, of which certain
parts (for instance central Borneo) were never submerged after the
beginning of the Mesozoic era, were variously and repeatedly con-
nected with each other and the Asiatic mainland.^ Such changing
conditions existed probably during the whole of the Mesozoic time,
and it seems, on account of the scarcity of Jurassic deposits, that
during the Jurassic period land-conditions prevailed, although the
land may not have had the extent assumed by Neumayr, It may
have been similar during the Cretaceous period, but it seems that
the land bridge began to dissolve ; at least, in the Upper Cretace-
ous, we have positive indications that the connection between Asia
and Australia was interrupted. This bridge probably was never
again completely restored ; the single parts of it, however, were
not stationary in 1 ertiary times, and communicated with each other
in various directions. These changing conditions are noticeable as
far as New Guinea, and, as regards the latter island, we know
through Haddon, Sollas and Cole,' that it is closely connected,
tectonically, with Queensland. The archaic and palaeozoic rocks
of the '' Australian Cordilleras" continue across the islands of
Torres Straits into the southern part of New Guinea, which belongs
undoubtedly to Queensland, and was separated from it .at a very
recent period. On the other side, the larger Sunda Islands
(Sumatra, Java, Borneo) must have also been united with the
Asiatic mainland in very recent time, as is positively shown by
their fauna of higher land animals.

1 Weber, M., Zool. Ergebn. Reise Niederlaend. Ost-Indien, Vol. 2, 1892 ;
Vol. 3, 1894.

2 According to Molengraaff (^Geologische Verkenningstochten in Central
Borneo, iQco), Borneo was submerged in Precretaceous times, but part of it was
land in the Middle Cretaceous. At the end of the Cretaceous a subsidence took
place, then again an elevation. The different parts of Borneo were subject in
various degrees to these changes, which continued through the Tertiary.

3 Trans. R, Irish Acad., Vol. 30, 1894.

324 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

If the first sharp separation of Australia and Asia belongs to the
Upper Cretaceous, it is consequent, for the Parastacidce and
Potamobiidce, that their area of distribution, which before the
beginning of the Upper Cretaceous extended over the Sino-Austra-
lian continent, was cut in two ; of course, the ancestral forms occu-
pying this old continent could not possibly have been divided into
these two families, and their differentiation was directly connected
with this separation of the geographic range. After that, there
was a chance for either family to develop, since there was no longer
communication between the Asiatic and the Australian stock. This
forces us to the conclusion that the ancestors of these two families
7nust have existed before the beginning of the Upper Cretaceous time,
and that during the Upper Cretaceous the division into Potamobiidce.
and Parastacidce. took place. It is impossible to place the origin of
these families at a later period, since, as we shall see below, any
crayfishes of late Mesozoic or early Tertiary age, in any part of the
world, belong either to the one or the other family. Although
there was at least a partial connection of Asia and Australia in Ter-
tiary times, the two families never came into contact again : with
the cause of this remarkable fact we shall become acquainted
below.

With reference to the Pota77ionince, their distribution over the
Indo-Malaysian Archipelago is only partly explained by the
assumption of a former continuous land bridge. The distribution
of the freshwater crabs is by no means simple, and does not extend
uniform.ly from eastern Asia to Australia, but there are numerous
complications and peculiarities. In the first line, we have to
emphasize the fact that only a single group, which is apparently
highly specialized, the subgenus Geothelphusa, reaches the conti-
nent of Australia, and that this group (in its typical forms) is
restricted to the Indo-Malaysian islands, and is wanting on the
Asiatic continent. This is the more remarkable, since this group is
most abundant just on the large islands of Sumatra, Java, Borneo,
and extends northward over the Philippine and Loo Choo Islands
to Japan. On the other hand, we have seen that the typical species
of the genus Potamon (subgenus Potamon~), which are found in both
India and China, reappear in very closely allied forms in Java,
Sumatra and the Philippines, but do not pass farther to the East.
Then again, the subgenus Potamonautes possesses scattered stations

1902.] AND ANCIENT GEOGRAPHY. 325

(probably strongly discontinuous) as far as New Guinea, and the
same is the case in the genus Parathelphusa.

For the present, these very strange conditions defy explanation,
and especially the eastern boundary of Potamon (sens, strict.) and
the western boundary of Geothelphusa are puzzling. But this much
we may say, that the distribution of the Potamonince over the Indo-
Malaysian Archipelago is apparently due to the varying relations of
the different islands between themselves and to the continents dur-
ing the Tertiary period, and that it furnishes additional proof for
the complexity of the changes that took place during this time in
this region.

Another fact is to be especially mentioned. Among the Pota-
monifice we do not have such a sharp separation of Australian and
Asiatic types as we have found among the crayfishes : on the con-
trary, the species of Geothelphusa, found in northern and eastern
Australia, are all closely related to those found in New Guinea and
on the other islands. Also the different forms of Pofa?non (sens,
strict.), found in Java and Sumatra, are very closely allied to con-
tinental species. All this points to the conclusion that the separa-
tion of the respective parts from each other, which brought about
the present conditions, must be of comparatively recent date, and
that at a time not very far remote from the present the distri-
bution of land and water in this archipelago must have been con-
siderably different from what it is now. Thus it seems that the
causes of the distribution of the PoiamonincR in the Indo-Malaysian
Archipelago are to be sought for in later times, presumably in the
Tertiary, and that during this period, and possibly up to a very
recent time, conditions prevailed here which — although they may
not have amounted to a continuous land bridge — constituted a cer-
tain unstable connection between Asia and Australia. Probably
there was a maze of larger and smaller islands, channels, straits and
the like, which was not permanent in its parts, and changed
repeatedly.^

Our final result on this question would be the following: South-
eastern Asia was" connected with Australia in the Jurassic, and
probably also at the beginning of the Cretaceous period. In the

1 According to von Ihering (1894, p. 406), Australia was connected with Asia
during the Eocene and Oligocene. Hedley (1899) connects New Guinea with
Australia in the later Tertiary ; but a similar connection existed also in the
Eocene, and through the latter Oriental elements were brought to New Guinea.

326 ORTMANN — DISTRIBUTION" OF DECAPODS [Aprils,

Upper Cretaceous, a sharp separation between both continents was
formed, which continued possibly up to the Eocene. Then the
connection was, at least partially, reestablished, but it was of a very
changing character, which is expressed by the great complexity in
animal distribution. These changing and unstable conditions pre-
vailed all through the Tertiary, and up to the present time, and it
is hard to trace them under the present imperfect state of our
knowledge of the geology of the respective parts.

The Upper Cretaceous separation of Asia and Australia is
expressed in the distribution of the Potamobiidce and Farastacidce :
the formerly continuous area of their ancestors, which comprised in
the Lower Cretaceous the Sino-Australian continent, was divided,
about the middle of the Cretaceous, in a northern (East Asia-
Potamobiidce) and a southern {kM'itx2X\2.-Parastacidce) part. The
varying conditions of the Tertiary are expressed in the distribution
of the PotamonmcB ; the detail§, however, cannot be made out, and
further study of the freshwater crabs of these regions, as well as a
more thorough study of the geology of these parts, is very desirable.

3. CONNECTION OF AFRICA AND INDIA.

The occurrence of crayfishes (genus Astacoides) in Madagascar
has led us, as we have seen above (p. 295), to the assumption that
there once existed a connection of this island with southern Asia
(respectively with the Sino-Australian continent). The same con-
nection is suggested by the distribution of the Potamoni7ice., of
which the subgenus Potamonautes is found in Africa as well as
India. The Madagassian forms of the PotamonincB (see above, p. 301)
indicate a relation of this island to Africa, while a closer connec-
tion with India is not so striking. A genetic connection of the
ranges of this subfamily in Africa and India by way of the Nile
valley and Syria is improbable, although, geographically, this con-
nection actually exists ; this, however, is apparently due to second-
ary migrations, different branches of the subfamily, coming from
India and Central Africa respectively, meeting in lower Egypt.

Thus we have to regard Madagascar as a stepping-stone between
Africa and India, and, with reference to the Poiamonince, its rela-
tion to Africa is closer than that to India.

This supposed connection is well known in zoogeography under
the name of the Leniurian continent. Jacobi (1900, p. 169 ff.)
quite recently has doubted this Lemuria-hypothesis, although he

1902.J AND ANCIENT GEOGRAPHY. 327

introduces among his regions of dispersal, as a seventh, an Indo-
African, which occupies this geographic position. He believes,
however, that it is not correct to explain certain similarities of the
faunas of India and Madagascar by a land-bridge, but prefers to
accept the existence of a chain of islands, which permitted, in later
Tertiary times, a migration of animals possessing the power of
flight (Birds, Bats) in this direction. On the other hand, he grants
a connection of Madagascar with Africa upward to the Miocene.

Jacobi's assumption of a series of islands instead of a continental
connection from Madagascar to India seems to be well founded
only for this particular time, the younger Tertiary. But the simi-
larity of both faunas has apparently been underestimated by him,
even if he takes into consideration only Mammals and Birds, and
there are no doubt numerous relations between both parts among
other animals not possessing the power of flight. This fact has
been urged by Pilsbry (1894, p. xlv) for the Helices, and he says that
Madagascar is much more closely allied to Ceylon and Australia
than to South Africa.^ The present cases offered by the genus
Astacoides and within the family of the PoiamonidcB are also very
important for this question, since the idea of a migration of these
forms over a chain of islands and across parts of the ocean is
entirely out of question. Thus it seems that we have to assume a
continental connection — if not during the later Tertiary — in earlier
times.

The parts under discussion belong to the old Gondwana-land,
which, according to Suess, existed in Palaeozoic times, and was par-
tially destroyed in the same period through the disconnection of
Australia from it. Africa, however, remained intact, and formed
an ancient table-land, to which was added as a peninsula the
Lemurian bridge, which extended from Madagascar to India, and
traces of which are preserved up to the Eocene (Suess, 1885,
p. 538). This same peninsula is accepted by Neumayr for the
Jurassic period, and is represented in his map ; it is separated from
the main part of Africa by a great gulf extending southward, the
Ethiopian Mediten-anean Sea, includes the present peninsula of
India, and is not connected with the Sino-Australian continent, the
Indian Gulf and the Strait of Bengal forming its northeastern
shores. According to Neumayr (1890, p. 390), this Indo-Mada-
gassian peninsula existed up to the end of the Cretaceous, and even
^ In part, this may be due to old-Mesozoic, and even Palaeozoic geography.

328 OKTMANX — DISTRIBUTIOX OF DECAPODS [April

to the beginning of the Tertiary, but was destroyed in the older
Tertiary {I.e., p. 397).

The same view is expressed by Koken : for the Lower and Upper
Cretaceous he gives to this peninsula about the same shape it had in
the Jurassic (Neumayr), and in the older Tertiary he draws —
instead of this continuous land-bridge — a chain of islands.

There are not many cases where we possess such ample evidence
for the former existence of a land mass that has now disappeared,
at least as regards such a remote epoch. The chief arguments for
this land-bridge are taken from the character of the marine deposits
found at the supposed southeast and northwest sides of this penin-
sula, and they are especially convincing for the Cretaceous period.
The South-Indian Cretaceous, as it is found typically in the neigh-
borhood of Pondichery, is known similarly developed in Madagas-
car and Natal, and belongs to the ocean to the east and south of
this peninsula, while contemporaneous deposits of the western
Indian Ocean (in East Africa) and in northwestern India are
strongly contrasted to it, and are related to the Mediterranen type.
We even may obtain further information as to the shape of this
peninsula. According to Newton and Boule,^ the Jurassic beds of
the western coast of Madagascar belong to the Ethiopian Mediter-
ranean Sea (possessing the Mediterranean type), while the Creta-
ceous beds (Cenomanian-Senonian) of the same parts exhibit the
South-Indian type. This indicates that the Ethiopian Mediterra-
nean Sea extended, during the Jurassic period, farther south than
during the Cretaceous. The respective maps of Neumayr and
Koken agree well with this : according to Neumayr, the southern
extremity of Madagascar was united with Africa, while, according
to Koken, the connection was situated at its northern end. This
latter bridge continued to exist apparently during part of the
Tertiary time. We have seen above that the connection of East
Africa and India continued up to the very beginning of the
Tertiary, and was destroyed soon after. This destruction, hovvever,
affected only the parts between Madagascar and India, while Mada-
gascar itself remained connected with Africa : according to Jacobi,
up to the beginning of the Miocene. Lydekker^ is of the opinion

^ See review by Boehm in Neties Jahrb. f. Mineral.^ etc., 1897, Vol, i,
p. 489.

2 A Geographical History of Mammals, Cambridge, 1 896.

iy02.] AND ANCIENT GEOGRAPHY. 32^

that Madagascar became separated from Africa in the Oligocene or
Miocene ; at the same time he connects Madagascar with India, and
believes that this connection was not severed before the beginning
of the Pliocene, In opposition to this we maintain that the
connection of Madagascar with India was interrupted before that
with Africa.

As the only remnants of this old bridge, the Seychelles have been
preserved. They consist, according to Bauer/ chiefly of granitic
rocks, which are accompanied by dikes and sheets of volcanic
origin. Only traces of sedimentary rocks are found, and these
point to a very old age. While we thus may safely take the Sey-
chelles for a remnant of this old bridge — and this is confirmed by
the presence of the East- African genus Deckenia — the other islands
of the Indian Ocean (Chagos group, etc.), are coral-formation.
They may rest upon the highest peaks of the submerged Lemuria,
but the latter itself has disappeared here. Consequently the fauna
of these islands — at least as regards freshwater Decapods — does not
contain any forms indicating this old bridge, since they must have
all been drowned.

The northeastern extremity of the Indo-Madagassian peninsula
was formed by the present peninsula of India. According to Neu-
mayr and Koken, this latter was separated, from the Jurassic to the
older Tertiary, from the rest of Asia, that is to say from the Sino-
Australian continent, by the Strait of Bengal, and, during the
older Tertiary, India was, according to Koken (/. c, p. 452), an
island (also disconnected from Madagascar). It seems, however,
that this separation of India from the rest of Asia was not so per.
manent as is believed by these authors. It is true, as regards its
tectonic configuration, India has nothing in common with Asia,
but it seems that there was a connection, at least at certain periods.

That the ^' Central Mediterranean Sea " of Neumayr extended
during the Jurassic period across northern India to the Bengal
Strait, separating India and Asia, seems to be correct, since no
evidence to the contrary has been brought forth, and the latest in-
vestigations have shown that Jurassic deposits are widely distrib-
uted not only in the western but also in the central Himalayas.'
But during a part of the Cretaceous, this strait does not seem to

1 Neues Jahrb. f. Mineral., etc., 1 898, Vol. 2.

2 See Griesbach, Rec. Geol. Surv. India, 26, 1893, and Diener, Verh. k. k.
geolog. Reichsanst., 1895.

330 ORTMANX — DISTRIBUTION OF DECAPODS [Aprils,

have existed. Already Meddlicott and Blanford (1879, p. Ix)
have doubted that the plain of the Ganges river was covered by
the Cretaceous ocean, and, although these authors generally disbe-
lieve the existence of such a strait during Jurassic, Cretaceous and
Tertiary times, Diener (/. c, 1895) has demonstrated that there
exists, in the central Himalaya mountains, an almost complete
series of sediments from the Cambrian to the Eocene, among
which Triassic and Jurassic beds are well represented, while Creta-
ceous beds apparently are missing and Eocene again is known.
This is very much in favor of a connection of India with Asia
during the Cretaceous. A very positive opinion on this question
is expressed by Kossmat (1895, P- 463)- He says that the Middle
and Upper Cretaceous ocean of southern and eastern India was
not connected over northern India with Europe.

Therefore, it seems to be well to assume only for the Jurassic
period and for the Lower Cretaceous a separation of India and the
Sinic continent ; that is to say, during these times Lemuria (Mada-
gascar-India) was a peninsula connected with Africa. In the Mid-
dle and Upper Cretaceous, this peninsula became united with the
Sinic continent, forming a land-bridge between the latter and
Africa. This connection, however, was apparently interrupted
again in Eocene times. According to Neumayr (/. c.^ p. 481),
the Eocene deposits of the Central Mediterranean Sea (Nummulite-
beds) are continued across the whole of northern India to the
Gulf of Bengal (and farther to Java, Borneo and the Philippine
Islands), and indicate thus a continuous ocean, which isolated
India from the rest of Asia. Since, at about the same time
(Eocene), the destruction of the Lemurian bridge took place,
India became an island, as is first pointed out by Koken. In Post-
Eocene times, this strait separating India and Asia disappeared,
and we have, in northern India generally, at about this time (cer-
tainly from the Miocene upward), a regression of the ocean (see
Meddlicott and Blanford, 1879, p. liii). The island of India was
definitively joined to Asia and never again separated.

After the destruction of the connection of India with Madagas-
car, in the beginning of the Tertiary, of the southwestern parts of
Lemuria only Madagascar remained, which was still connected, as
a peninsula, with East Africa. Then this connection was also
severed, but not before the Oligocene or the beginning of the Mio-
cene. Thus the main outlines of the present distribution of land

1902.] AND ANCIENT GEOGRAPHY . 331

and water were established at about the beginning of the Miocene.
After the destruction of ihe Lemurian bridge in the Eocene, its
northeastern portion, India, became part of Asia, while its south-
western portion, Madagascar, which at first remained a peninsula
of Africa, became an island.

The application of these geographical results to the distribution
of the freshwater Decapods is the following: First, we have to
emphasize that before the middle of the Cretaceous it was impos-
sible for the genus Astacoides to reach Madagascar. Since the
separation of the Asiatic and Australian group of the crayfishes
took place in about the Upper Cretaceous and since the morpho-
logical differentiation of the Potimobiidce and FarasfacidcB wdiS cot\-
nected with this separation, and further, since Astacoides must
have immigrated into Madagascar from the Asiatic part of the old
Sino- Australian continent, this latter process must have gone on
shortly before the completion of this separation, that is to say,
about the middle of the Cretaceous. This assumption is supported
by the morphological characters of Astacoides^ which are, in a
certain degree, intermediate between the present two families and
favor the view of an early separation from the original stock.

Thus there is nolhing that prevents us to assume an immigration
of Astacoides from southeastern Asia into Madagascar in the mid-
dle of the Cretaceous period. At a later time this does not seem
to have happened, since, in this case, we should have different
morphological characters in Astacoides, At an earlier time this
immigration was impossible, since then India was not connected
with the Sino Australian continent. After the Eocene this migra-
tion was absolutely impossible, since then the land connection
between India and Madagascar had disappeared.

Although we may thus fix the time of immigration of Asta-
coides rather exactly, there arise other questions. We want prin-
cipally an explanation of the absence of similar forms in Africa
itself, and for the absence of such in India and generally in
southeastern Asia.

Regarding the Potamonince, their presence in Madagascar, and
the close relation of the Madagassian forms to East-African, is
easily explained by the former connection of Madagascar with
Africa. The freshwater crabs of Madagascar thus indicate geo-
graphical conditions which are older than Miocene. The presence

PROG. AMBR. PHILOS. SOC. XLI. 171. V. PRINTED DEC. 20, 1902.

332 ORTMANN — DISTRIBUTION OF DECAPODS [AprU3,

of Deckenia on the Seychelle Islands connects also this group more
closely with Africa than with India. Possibly this connection is
identical with that over Madagascar, although Deckenia has not
been found on the latter island.

The presence of Poiamonince in India, corresponding to the
African type (subgenus Potamonautts), indicates the full develop-
ment of the Lemurian peninsula, that is to say, conditions prevail-
ing in the oldest Tertiary, if not earlier. Potainonince, represented
by forms which resembled the subgenus Potamonautes, must have
existed at least in the beginning of the Eocene, and their distri-
bution extended over Africa and the Lemurian peninsula, includ-
ing India. During the Eocene this range was separated into two
parts, an African (to which Madagascar belonged) and an Indian,
and, beginning in the Miocene, the PoiamonincB had a chance to
expand over southern and eastern Asia (Farther India and Chma*).
At the same time they availed themselves of the various and chang-
ing connections within the region of the Indo Malavsian archipe-
lago, occup>ing the latter and reaching Australia. The opening
of this region of dispersal offered to this group a new opportunity
for a rich development, and the origin of the subgenera Potamon
and Geothelphusa was probably the outcome of it.

We cannot leave this chapter without saying a few words on the
Arabian region of dispersal of Jacobi. This extends from north-
eastern Africa across Arabia to India. Jacobi mentions the simi-
larity of the Siwalik-fauna of India with the Ethiopian. Ihis,
consequently, refers to a very recent period, the later Tertiary.
Before this time, in the older Tertiary and in the Mesozoic, this
connection is out of question. The Potamo7iirice, which, as we
have seen, existed in the older Tertiary, show no tiace of this con-
nection across Arabia, and, as we shall see below, our knowledge
of the ancient geography of these parts is a very fair one. Arabia
itself formed originally a part of Africa, and the Red Sea did
not exist at all in the earlier Tertiary, it being quite recent (see
below). Toward the north, northeast and east Arabia was circum-

1 There was, possibly, an earlier cliance to reach the Sinic continent, in Upper
Cretaceous times, and I am inclined to t>elieve that the discontinuous localities
of I otamonautes (.md /'araihelphusa) in the Indo Malaysian archipelago point
to an immigration of these foims that precedes in time that of Potamon sens,
strict.

1902.] AND ANCIENT GEOGRAPHY. 333

scribed by sea — the Central Mediterranean Sea and the Ethiopian
Gulf of Neumayr. A connection with India in this direction,
and a migration of PotamonincB from India to Africa (or vice
versa) by this route was then impossible.

Further, I should lil<:e to point out that we have to be careful
about this Indo Madagassian bridge. A case which has occurred
to me, and which might lead to misinterpretation, is furnished by
the distribution of the Reptile-family Chamaleontidce. According
to Gadow, ^ this family is found in Africa, Madagascar and India,
a distribution which is quite analogous to that of the Potamoni-
nce, and might induce us, at the first glance, to trace it back to
this old Indo-Madagassian connection. A closer study, however,
reveals the fact that the Chameleon of India has nothing to do with
the Madagassian species, but is related to the form widely dis-
tributed in North Africa, Syria and Asia Minor. Here the con-
nection apparently goes from North Africa over Syria and Arabia
to India, and this distribution belongs to a much later period when
Lemuria no longer existed.

4. CONNECTION OF NEW ZEALAND WITH AUSTRALIA.

We have seen that a genus of the family Parastacidce, Parane-
phrops, is found in New Zealand, and this fact points to a former
connection of these islands with Australia. We further are to
pay attention to some adaitional facts, which, although they do
not seem to be sufficiently established to be accepted witiiout com-
ment, are apt to throw some light upon this connection.

First, according to Huxley ( 7> Zool. Soc., 1878, p. 771),
Paranephrops is said to be found in the Fiji Islands. This
locality is supported by two specimens in the British Museum,
which are in a very bad condition ; moreover, there is no report
as to the authenticity of the locality, and the genus has never again
been reported from these islands.

Further, Nobili (1899) describes from southern New Guinea a
genus and species, Astaconephrops alhertisi^ which is said to be
closely allied to Paranephrops. It is impossible, however, to con-
trol the systematic position of this form, since only external char-
acters are given, and the most important one, the branchial

1 Gadow, H., " Amphibia and Reptilia,' in The Cambridge Natural His-
tory, Vol. 8, 1 90 1, map, p. 568.

1902.] AND ANCIENT GEOGRAPHY. 335

about a possible tectonic connection of these parts. Neuraayr,
however (1890), draws in his map of the Jurassic continents, men-
tioned repeatedly above, a peninsula, which is connected with his
Sino-Australian continent, and which corresponds closely to Hed-
ley's idea of Melanesia. This peninsula is missing in Koken's map
(1893) of the Cretaceous continents, and even New Zealand is not
given as land there. But Koken does not seem to have paid much
attention to these parts of the earth's surface in Cretaceous times,
since it seems quite sure that at least parts of New Zealand were
land then. In the 01<ler Tertiary, New Zealand and New Cale-
donia were islands, according to Koken, while Australia extended
far to the east, including Lord Howe Island.

Although, in general, the geological evidence for the connection
of New Zealand with Australia is very scarce, we certainly have to
assume it according to the characters of the fauna and flora of New
Zealand, and the material at hand points distinctly to the fact that
this connection was interrupted at a comparatively early period.
Thus there is nothing that is opposed to the view of von Ihering,
that the final isolation of New Zealand took place not later
than the beginning of the Eocene, and there is no objection to the
demonstration on the part of Hedley that this connection with
Australia was by way of New Caledonia and New Guinea. Our
present case, the distribu ion of Paranephrops in New Zealand, fits
well into this theory: t'lis genus reached New Zealand in Pre-
Tertiary times, probably in the Upper Cretaceous, and very likely
by the way indicated by Hedley; since the Eocene it has become
isolated on this island group.

5. CONNECTION OF SOUTH AMERICA WITH AUSTRALIA (rESP. NEW

ZEALAND).

The genus Parasiacus in the temperate and subtropical parts of
South America points to a connection of this continent with those
parts in which allied forms are found, namely, with Australia and
New Zealand. Numerous instances of a similar distribution,
which suggest a relation of the same parts, are known, not only
among land and freshwater animals, but also among the marine
littoral fauna. This remarkable fact has been noticed at a very
early time, and has suggested various theories, which have been
reviewed and classified by the present writer (Ortmann, igoi).
The views of the majority of the later authors now agree more or

336 ORTMANN— DISTRIBUTION OF DECAPODS [Aprils,

less in that this connection is placed across the Antarctic conti-
nent, and this idea is chiefly supported by Hedley (1895, 1899),
von Ihering (1891, 1894), Osborn (1900), Pilsbry (1894), and
Ortmann (1901, 1902).

While Pilsbry only generally expresses the opinion that the sup-
position of an old Antarctic continent connecting the respective
parts of the present southern continents would furnish the condi-
tions necessary for the explanation of the zoogeography of the
land-mollusks, and while Osborn only tries to give an approximate
idea of the mutual relations of these land-masses by pointing out
that a subsidence of the ocean level of a certain amount would
connect these parts, von Ihering (1894, p. 438) gives a more
detailed theory of this connection. He unites not only South
America over Antarctica with Australia, but continues this (Meso-
zoic) land mass beyond the Indo-Malaysian islands to east Asia,
thus including the Smo Australian continent di^5cussed above. He
calls this vast continent by the name of Archinotis.

As to the details of the special connection of Australia and South
America, Hedley's opinion is the most important ; according to
him (1895, p. 6), during Mesozoic and older Tertiary times a
stretch of land extended from Tasmania over the South Pole to
Terra del Fuego ; the shore line of this land ( \ntarctica) formed a
wide gulf between Tasmania and Cape Horn, and approached the
Pole. This land-bridge, however, was not very solid, but was sub-
ject to various changes resulting in a repeated breaking up and
becoming reunited of the different parts. As regards New Zealand,
he believes that during the Tertiary time it was not directly con-
nect d with Antarctica. In another paper, however (1899, p. 399),
Hedley also assumes a connection of New Zealand with Antarc-
tica, but this was of an older date than that from Australia over
Tasmania to Antarctica, and consequently is to be placed in the
Mesozoic time.

That Australia was once connected with Antarctica, especially
with what is now called Wilkes' and Victoria Land, can be imag-
ined as possible on tectonic grounds. Australia itself consists,
according to Suess (1888, p 188 ff. ), in its eastern part of a very
old range of mountains, running in a north-southerly direction;
its larger western part is an old Archaic and Palaeozoic plateau
(part of Gondwana Land). Both parts are fractured and cut off
toward the south, and the southern parts have disappeared ; a line

1902] AND ANCIENT GEOGRAPHY. 337

of faults at the southern margin of the Australian Plateau indicates
that Australia undoubtedly extended once farther southward, in the
direction toward Antarctica. Whether it was really united with
the latter cannot be said positively, chiefly because the geological
structure of Wilkes' Land is entirely unknown.

The time of the subsidence of the southern parts of the old
Australian continent can be determined according to the condi-
tions known to exist on the shores of the great Australian Bight.
Here, on the foot of the broken edge of the Australian Plateau, a
series of Tertiary deposits is found, the age of which is not yet
positively ascertained, but which seem to belong to both the older
and younger Tertiary. The fact that no older (Mesozoic) beds
are found in this region seems to indicate that such were not de-
posited, and that means to say that up to the end of the Mesozoic
time the southern part of the Australian Plateau had not subsided,
and that this process took place at the very beginning of the
Tertiary,

Thus we have reason to believe that the connection of this part
of Australia (the western plateau) with Antarctica existed up to the
end of the Mesozoic time.

The Tertiary deposits of the south coast of Australia are lacking
from Tasmania along the eastern coast of Australia; here is a frac-
ture toward the south and east, the age of which cannot be deter-
mined at present. Hedley believes that there was here a connec-
tion with Antarctica that persisted up into the Tertiary (over
Tasmania), but he gives no geological evidence for it. It is
entirely unknown whether the East Australian Cordilleras find a
continuation in Antarctica. So Hedley's assumption may or may
not be correct.

Another tectonic line in these regions has been pointed out by
Gregory.^ He also emphasizes the former southward extension of
the Australian Plateau ; but besides, there seems to be, according
to him, a very important tectonic line marked by the volcanoes of
New Zealand and. Victoria Land, and this, possibly, fmds its con-
tinuation in the volcanoes of the region of Graham Land, and
passes thence over Terra del Fuego to the South American Cordil-
leras. Of course, this is no evidence at all that this line from
New Zealand over Antarctica to South America has ever been a
continuous mountain range actually connecting these parts, but the

1 Nature^ Vol. Ixiii, 1901, pp. 6io-6ir, with map, p. 611.

338 ORTMANJSr— DISTKIBUTION OF DECAPODS [AprU S,

existence of such a line would in a large degree facilitate the
imagination of such a connection, and would force us — if we have
other evidence pointing to a former connection of these parts — to
construct this old land-bridge nowhere else but along the direction
of this line. That is to say, the connection of Australia and New
Zealand with South America, which is probable on account of cer-
tain facts in the distribution of life, ivas acj-oss the Pole, and not
in lower latitudes in the southern part of the Pacific Ocean, as
accepted by some authors.

The tectonic connection of Graham Land with Terra del Fuego,
indicated by Gregory, is much emphasized by Fricker.^ Accord-
ing to him, it is formed by the arc of islands running from Terra
del Fuego over South Georgia and the South Sandwich Islands to
Graham Land. This line, however, again indicates only the
general direction of this possible connection, but does not give any
hints as to its actual existence, nor to the possible time of it.

We know that a large part of South America (the Brazilian
Plateau, see below) is a very old continental mass, which ex-
tended southward into northern Argentina, but not into Patagonia.
What is now the chain of the Cordilleras was certainly ocean dur-
ing Mesozoic times, since here we find Jurassic and Cretaceous
deposits largely developed, and the latter have been traced far to
the south and over almost all of Patagonia; the Tertiary beds of
southern Patagonia rest, wherever this has been observed, upon
Cretaceous deposits.'^ The Patagonian Cretaceous, in its upper divi-
sions, consists of rocks formed apparently under continental condi-
tions (littoral, freshwater, or eolian), and these latter (Guaranitic
beds) were subject, after their deposition, to erosion, indicating a
land period at the close and after the Cretaceous. Thus, there
seems to have been an upheaval, beginning at the end of Mesozoic
time and continuing into the Tertiary ; during the Eocene these
regions probably were land to a large extent.'

West of the Mesozoic beds known in the tract of the Cordilleras
there are, in the so-called Coast Cordilleras of Chili, rocks of
another character; they are apparently metamorphic, but their age
is disputed. According to Steinmann,* they are Mesozoic ; snd

1 The Antarctic Regions, London, 1900, p. 140 fF,

2 See Hatcher, J. B., in Amer. Jour. Sci., Vol. ix, 1900, p. 95 ff., and
Ortmann, /^ep. Princeton Exped, Patagonia, Vol. iv. Part 2, 1902, p. 285.

3 See Ortmann, /. c, p. 317.

* Neiies Jahrb. f. Mineral., etc., Beil., Bd. 10, 1895, P- ^-

1902.] AND ANCIENT GEOGRAPHY. 339

according to WoUf,' they are at least older than Jurassic. This
coast range is continued southward across the Straits of Magellan,
and forms the southwest and south coast of Terra del Fuego,
where similar rocks are found, and here it curves more and more
in a west-easterly direction.

However, the old and even Mesozoic age of the rocks composing
this chain is not generally accepted, and also the identity of the
Fuegian rocks with those of Chili has been doubted ; Norden-
skjold,* for instance, takes the metamorphic roi ks of the outer
(western and southern) side of Terra del Fuego for Cretaceous.

Thus we see that there is considerable uncertainty about the con-
figuration and geology of southern South America in the Mesozoic
era; but this much seems to be settled, that the Chilean coast
range existed as early as the Cretaceous period," and that the Cor-
dilleras in Terra del Fuego were not formed later than in the Cre-
tacoeus. It is just this latter chain that continues over Staten
Island, South Georgia, etc., and finally connects with Graham
Land ; and if there was connection at any time, it was by this
way and in the Cretaceous.

Further, there is no doubt that at the end of the Cretaceous
period large tracts of Patagonia became dry land, and the maximum
of land extension falls probably in Eocene times.

Consequently we have to put the chief connection of the southern
parts of South America with Antarctica at the end of the Creta-
ceous and in the Eocene. But we are to emphasize here that thus
far we have been able only to connect the Chilean coa-t range
with Antarctica. According to von Ihering, this connection also
comprised old Archiplata (the Brazilian mass) and existed during

I Wollf, F. von, in Zeitschr. deutsch. Geolog. Geselhch., Vol. 51, 1899.

' Geological Map of the Magellan Territones (Svenska Exped. till Magellans-
laend, Vol. i, No. 3, 1899).

'And possibly earlier. Burckhardt, C. ("Traces geologiques d'un ancien
continent Pacifique," in Rev. Mus. de la Plaia, Vol. 10, 1900, p 177 ff ), has
brought forth some evidence for th- assumption that in Chili, west of the present
Cordilleras, which were sea during the Upper Jurassic, there existed a continent,
the eastern shore of which was formed by the Chilean coast range. Thtre is no
means, however, of deciding how far this Jurassic continent extende I to the
west. The Jurassic age of this range, together with the corres|)onding rocks of
Terra del Fuego, etc., is quite likely if Gregory's theory of the tectonic connec-
tion with New Zealand is correct; also, the mountains of New Zealand aie said
to possess Jurassic age (see above, p. 334)-

S-kO ORTMANN — DISTRIBUTION OF DECAPODS [AprUJ,

the whole Mesozoic era, but this seems to be doubtful ; at any rate,
a connection of the southern and western land in South America
(Chilean coast range) with the rest of South America is improbable
during a large part of the Cretaceous time, since marine deposits
belonging to this period are found in the present Cordilleras, indi-
cating separation by sea. This sea apparently was a strait running
in a north-southerly direction, and coinciding approximately with
the present direction and location of the Cordilleras. This strait
became dry at the beginning of the Tertiary, in the Eocene, since
Tertiary deposits are not found here, and thus a connection of the
main mass of South America, the Brazilian Plateau, was formed in
an east-westerly direction with the Chilean coast range. This
completed the connection of South America with Antarctica in
Eocene times, and, in our opinion, is very important and serves to
explain the numerous zoogeographical peculiarities of South
America.

To sum it up, we are justified to draw the following conclusions:
There is nothing that opposes, on tectonic or geological grounds,
the assumption of a connection of Australia with Antarctica, as far
as the evidence at hand goes. This connection t/e longs pre e?ninently
to the Mesozoic time, and was interrupted, at least for a large part, at
the end of this era, definitively at the beginning of the Tertiary.
We have no positive evidence for a permanent Mesozoic connection
of South America and Antarctica (but such may have existed) ;
but a connection of these parts is very probable at the end of the
Mesozoic time, and especially during the Eocene between the Antarctic
lands and the old Brazilian mass ; the southernmost parts of South
America (southern Patagonia and Chili) were connected in the
Cretaceous with Antarctica, forming part of it, but were still
separated from Archiplata. In the Eocene they were also con-
nected with the latter. This latter union was brought about by
the upheaval of the Cordilleras, which began toward the close of
the Cretaceous and continued almost all through the Tertiary.

We are not going to follow up this idea any further, although we
believe that it will prove to be very important with regard to the
origin of the South American fauna. For our present purpose, the
explanation of the presence of the genus Faiastacus in South
America, we arrive at the conclusion that the family Farasfacidcs,
which existed, as we have seen, during the Upper Cretaceous
period in Australia, had a chance, during this same time, to spread

1902.] AND ANCIENT GEOGRAPHY. 341

into Antarctica, and consequently into the southernmost parts of
America (Chili). Thence it extended, in the beginning of the
Tertiary, into Northern Argentina and Southern Brazil (Archiplata).
This west-easterly direction of migration from Chili to Brazil is in
a certain degree expressed in the present distribution of the genus
Parastacus, and the distribution of the genus ^glea seems to have
been formed under similar conditions, although its Antarctic
origin does not seem probable. The present southern boundary of
Parasiacus is possibly due to the present climatic conditions, it
having died out in the south of Chili and Patagonia on account of
the unfavorable climate of these parts.

The fact that the genus does not extend northward into the
truly tropical parts of Brazil needs further explanation. We shall
return to this later.

The presence of the genus Parasiacus on both slopes of the Cor-
dilleras (even the identical species is found in one case on both
sides, and in this respect the genus j^^^lea agrees with Parasiacus)
points to a time when the Cordilleras had not yet attained their
present elevation. As v. Ihering has shown, for many groups of
animals this chain forms a very sharp barrier, and it does not seem
probable that these freshwater Crustaceans are able to cross these
high snow and ice-covered mountains. In the case of Patastacus
agassizi 2i shihmg oi the continental divide (by the capturing of
the headwaters of a stream belonging to the drainage of the oppo-
site side) cannot explain its presence on both slopes, since in this
region the original divide seems to be intact (the waters of Lake
Nahuel Huapi drain to the Atlantic Ocean). Thus also this fact is
in favor of an early origin of the distribution, since the elevation
of the Cordilleras, although beginning at the end of the Creta-
ceous, did not attain its maximum till about the Miocene.^

6. CONNECTION OF THE WEST INDIES WITH CENTRAL AND SOUTH

AMERICA.

We have seen above that Cambarus cubensis of Cuba finds its
most closely allied species in C. mexica?i'/s of Mexico. Similar
conditions prevail among the species of Pseudothelphusa from the
Greater Antilles, two species {P. ajnericana and terreslris) being
also found in Mexico. On the other hand, the six species of the

* According 1o Hatcher, the Miocene Patagonian and Santa Cruzian beds are
largely disturbed in the region of the Cordilleras in Southern Patagonia.

342 OKTMANN— DISTRIBUTION OF DECAPODS [Aprils,

genus Epilobocera (from Cuba to Porto Rico and Sta. Cruz) are
restricted to these islands and do not possess any closely allied
forms on the mainland, although there is no doubt that they are
distantly related to the Central and South American freshwater
crabs and must have been derived from these parts. Thus it seems
that we are to distinguish two groups among the freshwater Crusta-
ceans of the Greater Antilles, pointing to two migrations from the
mainland of Central America — an older one, represented by
Epilobocera, the higher age of which is supported by the fact that
this genus possesses in some respects the most primitive characters
among the whole subfamily, and a younger migration, represented
by the identical species of Pseudothelphusa. It is doubtful to
which of these groups Cambanis c/bensis belongs, since it is differ-
ent from but closely allied to a Mexican species.

Entirely different in its relitions is the Pseudothelphusa (/*.
denfata) of the Lesser Antilles. This one points beyond doubt to
South America (Trinidad and Venezuela) and bears no relation at
all to the Greater Antilles, not to speak of Mexico.

The geology and tectonics of the West Indies and Central
America are only poorly known, but lately some very important
contributions have been published. Nevertheless, the conditions
prevailing here are far from being clear, and the opinions of differ-
ent authors vary frequently. This much seems to be sure, that the
history of this section of the earth is a very varied and complex one.

In the first line we are to consider the fact that the general feat-
ures of the main mountain ranges of Central America and the West
Indies possess much in common and differ sharply from both North
and South America. Especially the west-easterly strike of the old
ranges of Central America (Guatemala, Honduras), as well as of
the Greater Antilles (Cuba, Hayti, Porto Rico, Jamaica), is very
remarkable (see Suess, 1885, p. 698 ff.), and indicates a former
tectonic unit. According to Hill (1898), old rocks exhibiting the
same west-easterly strike are found largely distributed also in Nica-
ragua and Costa Rica, and, further, in the region of the Isthmus of
Panama, especially to the east of Colon, in the Cordilleras of San
Bias. Furthermore, the whole northern shore of Venezuela, from
Puerto Cabello to the northeastern end of the island of Trinidad,
consists of old granitic ranges with the same strike. All these obser-
vations, although apparently incomplete and scattered, most likely

1902.] AND ANCIENT GEOGRAPHY. 343

indicate that this whole region, i.e., Central America, the Greater
Antilles and the northern coast of South America, possesses an
**old basement of granitic rocks of earlier age than the oldest
determinable sedimentary rocks" (Hill, 1898, p. 241), which, in
its west-easterly strike, differs entirely from the present mountain
ranges of North and South America, and it is possible that these
parts once formed a unit, a solid continental mass. As to tiie age
of this continent we may form an idea if we consider that (Hill,
/. c, p. 243) Jurassic beds are absent in this region, so that during
this time at least this continent was in existence. The same seems
to be true for a large part of the Cretaceous time. Cretaceous
deposits are wanting in Central America from Costa Rica eastward
in the Isthmian region an^ on the West Indian islands.' On the
other hand, Cretaceous beds are found in Guatemala to the north
of the old granitic mountains. They are also extensively developed
to the south and southeast of the old granites in Colombia and
Venezuela (see below), and thus it seems that this old continent
was washed to the north as well as to the south by Cretaceous seas.
But at about this time (toward the end of the Cretaceous) tids
old Mesozoic Antillean continent must have been destroyed, prob-
ably by the formation of the Caribbean Sea. We do not know
much about the exact time when this happened, but we know that
the subsidence forming this basin within this old mass was accom-
panied by a faulting along the margins of the subsiding area. This
fault is cle.irly seen at the coast ot Venezuela, where, according to
Suess (1885, p. 687 f.), the old coast range breaks off to the north.
The fact that within the whole region of the Caribbean Sea no
Cretaceous deposits are positively known makes it very probable
that the formation of this depression falls at the end of the Meso-
zoic age.^

1 There are, however, Pre-Tertiary sediments, belonging possibly to the upper-
most Cretaceous, in some of ihe Greater Antilles (see Hill, Amer. Journ. Sct.j
Vol. 48, 1894, p. 197) ; but this again demonstrates that sedimeniauon in these
parts did not begin till the very end of the Mesozoic time,

2 There are Cietaceous deposits of Lower Senonian age in western Venezuela
which po>sess the Mediterranean type (see Gerhardt, m A^. Jahrb. Mineral. ,
etc., Beil., Bd. 11, 1897, p. 87). This possibly is the first indication ot the exist-
ence of the Caribbean Sea. But we must not forget that the Lower Cretaceous
of Colombia and Peru also exhibits Mediterranean character, which is due, no
doubt, to the Orinoco connection. It is remarkable that the relation of these
Lower Cretaceous beds to Texas is not very evident, they probably being sepa-

344 ORTMANN — LISTRIBUTIOX OF DECAPODS [Aprils.

In subsequent times, at the beginning of the Tertiary, the Carib-
bean Sea must have existed, since Tertiary deposits are largely
developed in this region, not only on the Antilles but also on the
Isthmus of Panama. It seems that in the beginning of the Tertiary
the old Antillean continent was divided into two main sections —
the Greater Antilles with Honduras and Guatemala to the north,
and the coast range of Venezuela to the south. The remnants of
this continent in the Greater Antilles and Central America re-
mained first in a large part land, bat apparently they were subject
to various changes during the Tertiary period and subsided and
were elevated repeatedly.

We have seen that the geographical distribution of certain fresh-
water Decapods demands in the first line a connection of the
Greater Antilles with Mexico, and according to the foregoing con-
siderations this connection can have been situated only in the
direction over Honduras and Guatemala. We have further seen
that a Mesozoic connection of these parts is very likely, and that
the connection of Venezuela with Central America existed almost
up to the end of the Cretaceous. As we shall see below, we have
reason to believe that the freshwater crabs reached Venezuela in the
second half of the Cretaceous, and consequently it was also possible
for them to extend during this time into Central America (and
Mexico). If the latter parts were then or later connected with the
Greater Antilles, this would account for the presence of the most
primitive genus of the subfamily, Epilobocera, in these islands. On
the other hand, Potam)biid(Z vvere probably present at the end of
the Cretaceous times in western North America. These parts were
connected with Central America in this period, Mexico being dry
land, and thus there was also a chance for the Potamobiidce (repre-
sented here by Cambarus) to reach finally the Greater Antilles.
Therefore we reach the conclusion that the first immigration of
freshwater Decapods into the Greater Antilles, represented by
Epilobocera, belongs to the end of the Cretaceous or the beginning
of the Tertiary, and that Cambarus cubensis possibly also belongs
to it \ but since this form is a true Cambarus^ although a primitive
one, I should prefer to put its immigration rather in the Tertiary
than in the Cretaceous.

rated from Texas by ihe Aniillean continent, while the Upper Cretaceous of
Western Venezuela shows close affinity to Texas, the Antillean continent having
disappeared.

1902.] AND ANCIENT GEOGRAPHY. 345

The same zoogeographical question has been investigated by
Simpson* with reference to the land and freshwater Mollusks. He
points out that among this group in the Greater Antilles we find
quite a number of species which are identical with species from
Central America and Mexico (list p. 488, /. c.^, and, besides, there
are in both parts numerous and more or less closely allied forms.
Simpson does not distinguish very sharply these two categories,
identical and allied forms, but they correspond very likely to the
same two groups among our Decapods.

Now Simpson draws the following conclusions : Sometime during
the Eocene the Greater Antilles were elevated and connected with
each other and with Central America by way of Jamaica (and pos-
sibly across the Yucatan channel). Then a period of subsidence
followed, culminating in the Miocene and submerging the Antilles
with the exception of their highest parts, which ended the connec-
tion with Central America. In Postmiocene times the Greater
Antilles were elevated again and attained their present shape.

For the Lesser Antilles the matter was entirely different. These
islands did not exist at all in Eocene times or were submerged sub-
sequently, since their Mollusk-fauna, with the exception of a few
forms which may have reached them by drift, shows no affinities to
that of the Greater Antilles. After the formation of this island
chain, during the course of the Tertiary,* it was populated chiefly
from South America, and, as Simpson believes, by drift. The
South American (Venezuelan) origin of the fauna of the Lesser
Antilles is also confirmed by our material. Piitamocarcinus denta-
tus points directly to Trinidad and Venezuela and not to the
Greater Antilles. I should doubt, however, that this species has
reached these islands by drift, and I am inclined to assume a con-
tinental connection of these parts, which may have been of short
duration, during the later Tertiary. I am loatli to believe that it is
possible for these freshwater crabs to be transported across salt
water, and the fact that one species is found on the islands of
Guadeloupe, Dominica, Martinique, St. Lucia, another in Trini-

1 Simpson, C. T. : « Distribution of the Land and Freshwater Mollusks of the
West Indian Region, and their Evidence with regard to Pa-.t Changes of Land
and Sea" {Pr. U. S. Nat. Mus., VoJ. 17, 1895).

2 That these islands were formed during the Tertiary is also the opinion of
Hill. See Report by Robert T. Hill on the volcanic d sturbances ia the West
Indies in Ike Mation. Geograph. Magaz., Vol. 13, 1902, pp. 229, 240, 265.

846 ORTMANJSr — DISTRIBUTION OF DECAPODS [Aprils,

dad and a third in Venezuela is entirely opposed to the drift
theory, since under the latter we ought to expect only 07ie species
in this whole region.

Simpson's theory of the origin of the West Indian faunas is sup-
ported exclusively by zoogeographical evidence, and, as we have
seen, it agrees admirably with the facts presented by the Decapod
Crustaceans. But the various changes undergone by the West
Indian islands have been investigated also from a geological
and physiographical standpoint. I shall disregard the views of
Spencer* on the Antillean continent, which are certainly exag-
gerated, since he makes this whole region land during the Pliocene,
even including the floor of the Mexican Gulf and the Caribbean
Sea. According to him, the Pliocene land would have been ele-
vated above the present level to the amount of one and one-half to
two and one half miles, and this would result in a wide connection
of both North and South America with the Antillean land. But
this is simply impossil)le. If such a land connection had existed
in Pliocene times, it should have left not only unmistakable traces
in the present fauna of the Antilles, but the Antillean fauna ought
to be practically identical with that of the southern parts of North
America and the northern parts of South America; but this is by
no means the case. Nevertheless, one of the items in Spencer's
theories is important for our purposes. This is the assumption of a
Pliocene elevation of these parts, succeeded by the opposite move-
ment at the end of the Pliocene and in the Pleistocene.

On the other hand, HilP assumes for Cuba a subsidence at the
beginning of the Tertiary. Tins is followed, in the older Pleisto-
cene, by a rapid elevation, continuing more or less continuously up
to the present time. This late Tertiary and recent elevation influ-
enced also the neighboring parts of the Gulf of Mexico and the
Caribbean Sea, and Hill concedes that it was possible that Cuba
extended then as far as Yucatan, thus connecting with Central
America.

Tiie views of Hill and Simpson agree only in part as to the gen-
eral movements of these regions. Simpson assumes an Eocene
elevation and land connection, while Hill's elevation is Pleisto-
cene. But it is quite possible that both are correct. We have

' Spencer, J. W. : " Reconstruction of the Antillean Continent " {Bull. Geo-
log. Soc. America^ Vol. 6, 1895).

' Bull. Mus. Harvard, Vol. 16, 1895.

1902.] AND ANCIENT GEOGRAriiY. b47

seen that our material points to a double connection of Cuba and
Central America, an older and a younger one, and it is very likely
that the one is identical with Simpson's and the other with Hill's.
Between them there is a period of subsidence, the maximum of
which belongs probably to the Miocene. This agrees with both
Hill's^ and Simpson's views. The upheaval assumed by Hill for
the end of the Tertiary and the corresponding connection with the
mainland has been indicated previously by Neumayr (1890, p.
541), and the same theory is proposed by Spencer. And, further,
Simpson also advocates a Postmiocene elevation, which, however,
did not result in a connection with Central America.'*

According to the foregoing, the history of the development of
the Central American and West Indian region, as supported by the
freshwater Decapods, is the following :

Central America, the West Indies and the northern margin of South
America formed in the Mesozoic period {certai^ily during Jurassic and
Cretaceous) a continental mass (^Antillean continent), which was
bounded by sea to the north and south. This ccntinent broke up at
the end of the Cretaceous, the chief factor in its destruction beifig the
formation of the Caribbean Sea. The northern remnant of this coti-
tinent, consisti?tg of the Greater Antilles and paj'ts of present Central
America, probably remained a unit up to the Eocene. But at the end of
the Eocene and during Oligocene and Miocene the connection between
the Greater Antilles and the mainland was severed. But it was re-
established toward the end of the Tertiary {^Pleistocene) and again
destroyed in the recent time.^

1 The subsidence of Cuba at the beginning of the Tertiary, mentioned by Hill
(I. c, 1895), refers to the beginning of the Cuban Tertiary — that is to say, to
deposits including Eocene and Miocene. See Hill, in Ajner. Joitrn. Sci., Vol.
4b, 1894, p. 201.

2 T. Wayland Vaughan [Science, January 24, 1902, p. 148) doubts the Pleisto-
cene connection of Cuba with the mainland, since the recorded finds of Pleisto-
cene Mammals in Cuba are open to discussion, and possibly did not come from
this island. But the cases of identical species among the MoUusks, mentioned
by Simpson, and the identical species of freshwater crabs discussed here are
beyond doubt, and the tendency of the evidence furnished by them is in the same
direction as that of the Mammals. We do not believe, however, in a connection
of Cuba with North America, but with Central America. (Simpson accepts an
Eocene connection with the istand of Florida, by way of the Bahamas, which
ended in the Miocene.)

'^ This only partly agrees with what we know about the history of Jamaica.

PROC. AMER. PHILOS. SOC. XLI. 171. W. PRINTED DEC. 26, 1902.

348 ORTMANN — DISTRIBUTION" OF DECAPOIS [Aprils,

It seems that part of the freshwater Decapods (the identical spe-
cies) found their way from Central America to the Greater Antilles
during the Pleistocene connection, while the genus Epilcbocera
reached the same parts in much older times, in the beginning of
the Eocene or even at the end of the Cretaceous. How all these
forms were able to get into Central America we shall discuss
below.

To which of the two immigrations Cainbarus cubensis belongs
remains doubtful. I am inclined to classify it with the older
(Eocene) immigration.

The freshwater crab of the Windward Islands, Potauiocarcinus
dentatus, confirms the view of Simpson that these islands and their
fauna have little to do with the Greater Antilles, but rather that
they are related to South America. But, while Simpson believes
that the (late Tertiary) population of the Lesser Antilles was ac-
complished by drift, I believe that a land connection is indicated.

7. CONNECTION OF SOUTH AMERICA AND AFRICA.

The presence of freshwater crabs belonging to the family of the
Potamonidce. in the Old World (subfamilies Foiamonince and Deck-
eniince), as well as in the tropical parts of the New World (subfamily
Potamocarcinince), has led us above (p. 310) to the assumption that
there was once a land connection between South America and the
West Indies on the one side and Africa on the other. Similar
zoogeographical facts have been emphasized chiefly by von Ihering
(1891, p. 438, and 1894, p. 406), and, according to him, ''all
affinities of the freshwater fauna of northern South America direct
us to Africa." He believes (we shall discuss this later) that the

Hill (^Bull. Mus. Harvard, Vol. 34, 1899) says that at the end of the Cretaceous
and the beginning of the Eocene there was an extensive continental period, but
that there was a subsidence at the end of- the Eocene and in the Oligocene, and
then again an uplift at the end of the Oligocene and in the Miocene. The latter
is jus-t the opposite movement from what is known for Cuba. It is quite likely
that a different fate is to be assumed for the different islands, and it seems that
Spencer's idea of contemporaneous subsidence or e'evation of the whole region
between North and South America is entirely wrong; the orogenetic movements
and the changes of level connected with them were, after the first great subsi-
dence of the Caribbean basin, more or less local and affected only limited parts,
so that at the same time we may have had opposite movements in different sec-
tions of this region.

iiXVj.] AND ANCIENT GEOGRAPHY. 349

northern parts of South America (Archiguiana) once formed, during
Mesozoic times, a part separated from the rest of South America,
which, however, continued eastward across the Atlantic Ocean
connecting with Africa. Fernando Noronha and St. Helena are
remnants of this land-bridge, which he calls by the name of Arch-
helenis. This connection was destroyed, according to von Ihering,
in the Eocene, or, at any rate, not later than in the Oligocene.

To the numerous instances quoted by von Ihering in support of
his theory the distribution of the family of the FotamonidcB adds
another one, and the fact that two different subfamilies are found
in the Old and the New Worlds, and that the affinities of the Ameri
can forms with those of Africa and Asia are somewhat obscure,
indicates that the connection of both is to be regarded as an old
one and that it has been severed long ago. Therefore its existence
in Mesozoic times and destruction in the beginning of the Terti-
ary, as maintained by von Ihering, has much in its favor.

Taking up the geological side of this question, we first hai/'e the
broad Jurassic connection between Aftica and South America
assumed by Neumayr (1890). According to this author, and also
according to Suess (1888, p. 677 ff.), the whole of the southern
Atlantic Ocean ^did not exist neither during the Jurassic nor
during the older Cretaceous (Naumayr, /. c, p. 376), since no
traces of deposits belonging to these periods are found in West
Africa or on the eastern shores of South America It was not until
the beginning of the Upper Cretaceous that sea washed the eastern
parts of Brazil (/. c, p. 389). But the connection of both conti-
nents persisted even then, although in a limited degree, and dis-
appeared entirely as late as after the beginning of the Tertiary (/. c,
?• 397)- Its last remnant (/. c.^ p. 493) was formed by a chain of
islands which extended in the Oligocene from tropical Africa to
South America and the West Indies.

This view, however, is not accepted by Koken. In his map
(1893, P^- ^) t^^ Cretaceous continents of South America and
Africa are absolutely separated in the earlier as well as in the later
part of this period, and the Atlantic coast lines of both generally
agree with the present ones. In the older Tertiary Koken (pi. 2)
draws an island chain (Brazilo-Ethiopian islands) from the West
Indies to Africa.

As far as it refers to the Cretaceous period, Koken seems to be
mistaken. Although formerly it was supposed that Lower Creta-

350 ORTMaNN — DISTRIBUTION OF DECAPODS [Aprils,

ceous deposits are found in West Africa, it was soon recognized ^
that the respective beds are younger, and are certainly not older,
than the Middle Cretaceous (in Cameroon); and especially Kossmat
(1895) ^^^5 demonstrated that the Cretaceous beds of West Africa
(Angola, Elobi Islands, etc.) belong to the Middle and possibly
the Upper Cretaceous (Cenomanian and Lower Senonian), and
that they unmistakably possess a South Indian character, being
connected probably around the Cape of Good Hope with the Indian
Ocean. According to Kossmat, also the Brazilian Upper Creta-
ceous deposits in Sergipe, Pernambuco, etc.,'"^ are of the South Indian
type. Farther north, on the coasts of Morocco and Algiers,
typical Mediterranean Cretaceous beds are present. The upper-
most Cretaceous beds of Angola, however, are said to exhibit
traces of the influence of the Mediterranean province (Kossmat, p.
465).

According to these facts we are to form the following idea as to
the destruction of the old Brazilo-Ethiopian continent: It existed
in its full development during the Jurassic and in the beginning of
the Cretaceous time, being the western remnant of the old Paleo-
zoic Gondwana Land, and probably it had the extension assigned to
it by Neumayr — that is to say, it connected Africa with the north-
ern as well as with the southern parts (Brazil) of South America.
In the middle of the Cretaceous time the southern Atlantic Ocean
was formed and the sea extended from the south (connected around
the Cape of Good Hope with the Indian Ocean) toward the
equator. About the same time, or rather a little later (in the
Upper Cretaceous), a branch of the new South Atlantic extended
into what is now the valley of the Amazonas river, separating the
southern part of the Brazilian mass from the northern (Guiana)
(compare below). But Guiana remained connected with Af?'ica

I See Koenen, A. von, in Abh. Ges. IViss. Goettingen,S&x. 2, Vol. i, 1897,
1898.

» Described by White {Arch. Mus. Rio Janeiro, Vol. 7, 1888). Although
some of these beds (marine beds in Sergipe and Parahyba) are without any
doubt Upper Cretaceous, Branner {Canadian Meeting Americ. Jnstit. Min.
Engin., 1900, p. 17 f., and Jhtll. Geol. Soc. America, Yo\. 13, I902) has lately
demonstrated that other marine sediments in Sergipe, Alagoas, Pernambuco
Parahyba, Rio Grande do Norte and Para belong to the Eocene Tertiary (I902 •
pp. 47, 64, 85, 91, 96), and also that the freshwater deposits of the Bahia basin are
probably Eocene (1900, p. 18).

1902.J AND ANCIENT GEOGRAPHY. 851

and this restricted land-bridge going across tiie middle part of
the Atlantic existed probably during the rest of the Cretaceous
time and was not destroyed until the beginning of the Tertiary, a
chain of islands remaining as late as the Oligocene.

This means, with respect to our freshwater crabs, that their age
goes back at least to the Upper Cretaceous. During this period the
last remnant of the continental connection between Africa and
Guiana still existed, and the absence of Potamonidce in South
America south of the Amazonas valley further substantiates this
assumption, that these crabs did not reach South America prior to
the Upper Cretaceous, when the main part of Brazil also took part
in this old continental connection. Aside from this fact, we have
the consideration that it is not very likely that the age of the fresh-
water crabs goes far back in Cretaceous times. Although we have
no definite information as to the latter point, we may say, from a
morphological standpoint, that the Potamonidce represent a pecu-
liarly specialized side branch of primitive Cycloinetopa. Cyclovie-
topa existed in the beginning of the Cretaceous, but were rare.
Thus an Upper Cretaceous age of the Potamonidce is admissible.

The subsequent fate of the PotamonidcE in South and Central
America, after they immigrated (or originated) in these parts in the
later Cretaceous, will be discussed in the next chapter.

5. THE MUTUAL RELATIONS OF NORTH, CENTRAL AND SOUTH

AMERICA.

Aside from the peculiarities in the distribution of the freshwater
Decapods of America, discussed above, there are several other
features which need explanation. They are the following (see pp.
295, 296, 309):

- I. The remarkable restriction of the genus Potamobius to the
western parts of North America, while Canibarus is found in the
east and south (Mexico).

2. The southern limit of the range of Ca/nbarus.

3. The distribution of the Poiainocarciiiince over the West Indies,
Central America and the northern parts of South America; their
presence in the mountains of Ecuador and Peru and their absence
in Brazil south of the Amazonas.

4. The peculiar shape of the areas of Parastacus and ^glea,
which are almost identical, and extend, in the subtropical and tern-

S54»

So4 U«ii V

laTrx. tW lc£:^rts^ idn M to

'%gi

fW U^UftA i>cf«i| rnrnM the

l*^

1902.J AND ANCIENT GEOGRAPHY. 353

exist, and was not brought about until the mutual relations had gone
through various and entirely different stages.

a. North America.

If we want to get an idea of the configuration of North America
during Mesozoic times, we have to consult in the first line Neu-
mayr's well-known map (1890). According to this, in the Jurassic,
the northern and eastern parts of North America formed a conti-
nental mass, which extended well to the west (Utah peninsula),
while the northwest was covered by the sea that separated America
from Northeastern Asia. At the same time this continent (Nearc-
tic) was bounded by sea to the south, Mexico and the West Indies
being submerged.^ This representation, however, needs correction,
chiefly as regards the West Indies, as we have seen above.

Differing but little from the view taken by Neumayr is that of
Koken (1893, P^- ') '^^^'^ respect to the Lower Cretaceous period;
but here the land extends considerably to the northwest and
includes parts of Mexico, a conception v/hich is also to be modi-
fied, as we shall presently see.

The general history of North America during the Cretaceous
period is best represented by Dana (1895, PP- ^^3' ^74' ^Si).
According to him. Western North America was largely land during
the Lower Cretaceous and continuous with the rest. In the Upper
Cretaceous, however, chiefly in its earlier part, a central depression
became evident, which extended from the south (Gulf of Mexico)
northward and possibly reached the Arctic Ocean, dividing the
continent into an eastern and a western half. The western half, as
we have seen above (p. 318), became connected across Bering Sea
with Asia at about this time,^ At the end of the Cretaceous
(Laramie) and in the beginning of the Tertiary an extended eleva-
tion began, which culminated in the formation of the Rocky Moun-
tains, and by this process the interior Cretaceous sea became land
again, which resulted in the reconnection of Western and Eastern
North America. But, although there was a geographical union,
Eastern and Western North America remained separated bionomi-

1 Compare, also, Logan, W. N., in Journ. of Geology, Vol. 8, I900, but here
the Jurassic ocean of the Northwest is considerably reduced in size and repre-
sented only by a shallow bay.

^ Temporarily the Cretaceous sea of the interior was connected in British Co-
lumbia with the Pacific (see Kossmat, 1895, p. 474)-

354 ORTMANN — DISTRIBUTIOX OF DECAPODS [April 3,

cally, the Upper Cretaceous sea barrier being replaced by a barrier
formed by the Rocky Mountains.*

Looking now toward Mexico and its continuation southward, we
shall refer in the first place to the papers of Hill (1893 ^^^ 1898).
The history of Mexico in Pre-Cretaceous times is very obscure.
Possibly it was covered by sea, as is also assumed by Neumayr, in
the Jurassic, at least in part (Hill, 1898). But it seems to be well
established that in the Lower Cretaceous (Hill, 1893) almost all of
Mexico was submerged from the Atlantic to the Pacific side. This
Lower Cretaceous sea was limited on the north by the southern
coast of the North American continent, which extended from the
old Appalachian region across the present Indian Territory and
New Mexico to the Mexican province of Sonora.-

In the middle of the Cretaceous period (at the end of the
Comanche series, Gault) a large part of Mexico became land,
forming a southern continuation of the western part of North
America, which was separated in the Upper Cretaceous from the
eastern, and which therefore extended from British Columbia' to
the Isthmus of Tehuantepec. This strip of land formed during this
period a very important barrier, separating the marine faunas of
the Pacific and Atlantic Oceans. While both faunas were more or
less connected during the Lower Cretaceous across Mexico, they
became separated later and never again communicated in this
region.

The Isthmus of Tehuantepec consists, according to Spencer,* of
the identical Lower Cretaceous deposits found in Mexico, and, .
further, according to Sapper," Cretaceous rocks are found in the

^ This barrier was probably emphasized by the development of desert condi-
tions in and at the foot of this mountain range. Compare Scott, W. B,, An In-
troduction to Geology, 1897, p. 500: " Probably the upheavals at the end of the
Bridge r and at the end of the Eocene had made the climate much drier by cut-
ting off the moisture-laden winds."

- In 1898 (pp. 243 and 259) Hill qualifies his views, and says that it is doubt-
ful whether the whole country (Mexico) was entirely submerged at anyone time
during this period, ffe thinks it was a mere shifting of the barrier between the
Atlantic and Pacific. Compare, also, Stanton, T. W., in Joinn. of Geology , Vol,
3, 1895, p. 861. .

3 And these parts must have been connected, as we have seen above, with
Northeastern Asia.

* Bull. Geolog. Sac. America, Vol. 9, 1897.

5 /^oll. Instit. Geol., Mexico, Vol. 3, 1896.

1902.] AXJ) ANX'IENT GEOGKAPllY. 3o5

Mexican State of Chiapas, which adjoins Guatemala. As regards
Guatemala, we know that here old rocks appear which belong to
the system of the Antillean continent (see above, p. 342). Thus we
have reason to assume that, while Mexico was covered by the Lower
Cretaceous seas which separated North and Central America, this
whole region became land at about the middle of the Cretaceous,
thus effecting a connection of Western North America with Central
America (Guatemala) or with the old Antillean continent. This
seems to be also the view of Hill, and he likewise believes that this
connection was never subsequently interrupted.^

The result of the foregoing discussion is that during the Jurassic,
and especially during the Lower Cretaceous, North America formed
a unit, which was separated from Asia and which was also circum-
scribed by a shore line in the south, being disconnected from
Central America. In the middle part of the Cretaceous Mexico
was elevated, and this new-formed land connected the western part
of North America with the AndUean continent. At about the
same time a connection of Western North America with Northeast-
ern Asia was established (by way of Bering Sea), and the Mexican
Gulf extended northward, separating Western from Eastern North
America.

Thus ue have, in the Upper Cretaceous, a strip of land extending
from Northeastern Asia over Bering Sea and over the western side
of North A?nerica to Mexico and the Antillean continent. Eastern
Nofth A?n erica was separated from this strip.

In the beginning of the Tertiaj-y Eastern P7orth America became
reunited to this western section.

At the end of the Tertiary the Beringian co/inection 7vith Asia was
interrupted {sQQ above, p. 317).

This would lead us for our Crustaceans to the following conclu-
sions: We have seen above (p. 319) that at any time, beginning in
the Upper Cretaceous, Fota??tobius may have invaded the western
parts of North America. This is again supported by the preceding

1 See Hill, 1893, p. 323. Spencer (1897) assumes that there was a reestab-
lishment of the connection of the Atlantic and Pacific Oceans across the Isthmus
of Tehuantepec in late Tertiary times. The evidence for it, however, is entirely
insufficient. The gravels found on the passes of the isthmus are of no value,
since their marine character has not been demonstrated. Compare, also, Hill,
1S98, p. 262, footnote.

356 ORTM ANN— DISTRIBUTION OF DECAPODS [Aprils,

considerations, in so far as it is confirmed that Potanwbius cannot
have been present in North America during the Lower Cretaceous,
otherwise the remarkable restriction to the west would be inexplic-
able. But the genus must have immigrated during the Upper Cre-
taceous, since fossil remains of PoiaiuobiidcE^ are known from the
Eocene of North America. This latter fact, therefore, narrows
down the time of immigration to more definite limits (those of the
Upper Cretaceous), and at the same time explains its restriction to
the west. During this period the western parts of the country
were separated from the eastern by sea. At the same time there
was a possibility for the crayfishes to reach Mexico, and it is easily
understood that Potamobius then sent a branch southward, which
subsequently developed on the Mexican plateau into Cambarus.
After this, in the beginning of the Tertiary, Cambariis had a chance
to migrate by way of Texas into Eastern North America, where it
reaches its culmination in the present time.

The morphological differentiation of Cainbarus from Potamobius
probably took place in the beginning of the Tertiary, after the
ranges of these genera had become separated. This separation is
apparently due to a climatic change in the region between Mexico
and central California, where desert conditions developed. This
desert climate is not so pronounced on the eastern side of the con-
tinent, near the Atlantic coast in Texas, and, consequently, the area
of Cambarus is not here interrupted between Mexico and the
United States. A subsequent connection of the ranges of Potamo-
bius and Cambarus in the interior ol North America Hn the region
of the Rocky Mountains and the plains adjacent to their eastern
slope) was impossible on account of the topographic and climatic
barrier existing there in Tertiary times, which has been mentioned
above (p. 354). The Rocky Mountains themselves and the arid
regions are not favorable for the freshwater crayfishes. Thus the
areas of both genera remained separated, and only in one case a
species {P. gambeli) has crossed the continental divide in the
region of the Yellowstone National Park. This, however, is very
likely due to the capturing of streams that originally belonged to
the Pacific slope by the Yellowstone river.

b. Central America.

The tectonic unity of the old Archaic and Paleozoic rocks known

1 Cambarus primcEOHS of Packard, from the Eocene of Western Wyoming,
which is, however, according to Faxon (1885, p. 155), rather a Potamobius.

1902.] AND ANCIENT GEOGRAPIIY. 357

from Guatemala to Venezuela is also emphasized by Hill (1898, p.
239 ff.), and he also thinks that, during Mesozoic times, a continu-
ous continental mass may have existed here, which reached as far
as Trinidad. We have seen above that the destruction of this con-
tinent was probably due, in the first place, to the formation of the
Caribbean depression, at the end of the Cretaceous. This also
agrees with Hill's view (1898, p. 260 f.) that during the whole of
the Cretaceous, or at least during the larger part of it, the Atlantic
and the Pacific Oceans were separated in the region of Central
America — that is to say, that there was a land connection between
the northern parts of Central America and northern South America.
But, according to Hill, this connection is not identical with the
present isthmian region, but was situated chiefly to the west of it.

We have nothing to say against a western extension of this
Cretaceous land (which probably extended as far as the Galapagos
Islands), but we believe that the isthmian region and the present
Caribbean Sea also were land during this time ; the main point is,
that there was a connection between Guatemala, Honduras, Nica-
ragua, and the Greater Antilles on the one side, and northern
Venezuela on the other.

These conditions changed considerably during the Tertiary.
First, the Caribbean Sea was formed, and possibly it extended
farther to the west and southwest than it does now. At least, parts
of the present land-bridge, the Isthmus of Panama, were covered
entirely by sea in the earlier Tertiary, and that this sea reached
from the present Caribbean Sea across to the Pacific is beyond doubt.
In the first line, the part through which the Panama canal is to be
built is composed entirely of deposits that are not older than
Eocene and Oligocene (Hill, 1898, p. 236), and this well agrees
with the investigations of Douville/ and Bertrand and Zurcher :* the
Old Tertiary sea (Eocene and Oligocene) must have here extended
entirely across the isthmus, from the Atlantic to the Pacific.

The same seems to be true for the Nicaragua canal. According
to Hayes,^ there are no rocks along this route that are older than
Tertiary, and the Tertiary deposits probably belong to the Eocene
and Oligocene. The remarkable discovery has been made that

1 C. R. Soc. geolog. France, 1898.

2 Bertrand, M. et Zurcher, O. Etude geologique sur V hthme de Panama, 1S99.

3 Hayes, C. W., " Physiography and Geology of Region Adjacent to the Nic-
aragua Canal Route " {Bull. Geol. Soc. America, Vol, 10, 1899).

358 ORTMANX — DISTRIBUTION OF DECAPODS [Aprils,

sediments on the Pacific side contain the same fossils as the corre-
sponding ones on the Caribbean side, which is an important addi-
tion to Hill's observations.

Between these depressions of the isthmian region, filled out by
older Tertiary deposits, there are Archaic rocks at various places ;
we know of such not only from Guatemala and Honduras, but also
from northern Nicaragua (Hayes), Costa Rica (Hill, Hayes), and
even farther east, beyond the Panama canal, in the Cordilleras of
San Bias. Thus it seems that the present isthmus, from Nicaragua
to Colombia, consisted during the older Tertiary of a series of
islands separated by ocean straits.

According to the unanimous opinion of Hill, Hayes, Bertrand
and Zucher and others, these straits (Nicaragua and Panama)
became dry in the Middle Tertiary, i.e., in the Miocene, and, con-
sequently, the cofinection of North and South America was then
established.

Although this Eocene and Oligocene communication of the
oceans is admitted by Hill, he is inclined to minimize its import-
ance. Moreover, he assumes (1898, p. 263) that to the southward
and westward, toward the Pacific, a large land mass must have
existed, from which the material of the marine deposits of the
isthmus was derived, and, further, he believes that this land mass
chiefly extended in a north-southerly direction, probably connect-
ing North and South America. I think we do not need this land,
and even if we accept its existence,^ it hardly formed, in the earlier
Tertiary, a connection of the Americas. Be that as it may, the
insular elevations of the isthmus, and the masses of old rocks to the
north of it, in Nicaragua, Guatemala, Honduras, and in the
supposed connecting land with Jamaica and Cuba (see above, p. 347),
were in our opinion sufficient to furnish material for those Old Ter-
tiary sediments in the isthmian region."' On the other hand, any

1 Since we need, as I naost emphatically believe, a connection with Galapagos
Islands; this subject,, however, is outside of the present question.

2 Hill himself (1898, p. 263) discusses the idea that the land to the north of the
isthmus may have furnished the material, but dismisses it, since here "we are
confronted by great depths." Now, in my opinion, great depths are no funda-
mental objection, and just in this case the character of the sea bottom in the
region between Honduras, Jamaica, Cuba and Hayti indicates that important
disturbances have occurred here, and, in the first place, the deep submarine rift
valley, known as " Bartlett deep," may be of a very recent age.

1902.] AND ANCIENT GEOGRAPHY. 35U

land mass to the south and west of the isthmus cannot have formed an
Old Tertiary barrier completely separating both oceans, since we
need an interoceanic communication during this time, as we shall
presently see.

Our opinion is, that during the Cretaceous there luas a connectio?i
between northern Central America and northern South America, the
Antillean continent still being tnore or less intact. At the beginning
of the Tertiary y however, and after the formation of the Caribbean
Sea, an oceanic connection existed between the Atlantic and Pacific
in the isthmian region, and this communication existed up to the
Miocene, separating North and South America. But afterward,
beginning in the Miocene, the isthmus was elevated, reconnecting the
separated chief remnants of the Antillean continent, and at the same
time JVorth and South America. The Atlantic and Pacific Oceans
were separated, and never again communicated, either here or else-
where}

We here arrive at a result which differs considerably from von
Ihering's ideas as to the relations of North and South America:
von Ihering believes (1894, p. 405) that both continents were sepa-
rated by Cretaceous sea, and that Central America was entirely
submerged at this time ; the origin of the Isthmian land-bridge is
also placed by von Ihering in the Miocene.

For our Crustaceans, we are to draw from this the following con-
clusions :

The presence of PotamocarcinincB in the present continental parts
of the old Antillean continent, in Nicaragua, Guatemala (to which
we must add the southern parts of Mexico), the isthmian region and
Venezuela, is due to the Cretaceous connection of these parts ; the
presence of the genus Epilobocera in the Greater Antilles is due to
the former connection of these islands with the mainland, and
belongs to the same land period, or to the continuation of it in the
earlier part of the Tertiary. After the separation of the Greater

1 This idea well agrees with the character of the present Atlantic and Pacific
marine liuoral faunas in the Central American region. These faunistic facts are
often incorrectly represented and understood, and Hill's argument against the
importance of the interoceanic communication in older Tertiary times is based
upon such a misunderstanding. I have studied this question chiefly with refer-
ence to the marine Decapod Crustaceans, and shall give below a correct repre-
sentation of the actual conditions of the faunal relations of both oceans. See
Appendix.

360 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

Antilles from the mainland, there was left on these islands an iso-
lated stock of primitive freshwater crabs, now known under the
name oi Epilobocera. On the mainland, these primitive forms dis-
appeared, or changed into what is now known as the genus Potamo-
carcinus, and although in the beginning of the Tertiary the conti-
nental range of this genus was much cut up, chiefly in the region of
the isthmus, the different parts were reunited in the Miocene, form-
ing a unit that extended from northern Central America to Trinidad
and Guiana. This explains the uniform distribution of Potamocar-
cinus over this region. In the later Tertiary we had a second union
of the Greater Antilles with northern Central America, which
explains the immigration of identical species of Potamo car cinus
from Mexico into Cuba and Hayti. The Lesser Antilles were
probably connected in the later Tertiary with Venezuela, and a
species of freshwater crabs reached them by this way.

c. Relation of Venezuela to the rest of South America. The
Orinoco Valley.

The northern coast range of Venezuela belongs, as has been
stated, to Central America. To the south, on the slope toward the
Orinoco, it is fringed by extensively developed Cretaceous deposits,
which are also known from Trinidad in a similar position. These
deposits are said to belong to the Lower Cretaceous (Suess, 1885,
p. 688), and to extend vvestward far into Colombia. To the south
of this zone, in Venezuela, there are (Suess, ibid.^ younger Tertiary
marine beds, which, in part, enter this region through a depression
extending southward from the Bay of Barcelona on the northern
coast of Venezuela.

This would indicate that during the Lower Cretaceous, the old
Antillean continent was bounded on the south by sea (see above, p.
343), which separated it from the old granitic masses of Guiana.
The apparent lack of Upper Cretaceous deposits, with the excep-
tion of a small region of western Venezuela, points to the assumption
that at the end of the Mesozoic time (Upper Cretaceous) both
regions were connected. Then again, in the later Tertiary, they
were separated, at least in part, by sea that entered into the
Orinoco valley (Suess, 1888, p. 161).

The Lower Cretaceous sea not only separated Venezuela and
Guiana, but apparently continued westward, into Colombia,
Ecuador and Peru. Indeed, there are in the western chain of the

1^2.] AND ANCIENT GEOGRAPIIV. f^6l

Cordilleras in Peru and Bolivia many exposures of old (Archaic and
Paleozoic) rocks; but the fact that in this region (from Colombia
to Bolivia) Lower Cretaceous of the Mediterranean type has been
discovered right in the Cordilleras/ renders it possible that those
older rocks were originally covered by Mesozoic deposits, which
were removed subsequently by erosion ; and this is also the view of
Suess (1885, p. 684, and 1888, p. 68^), since he takes it for
granted that the Cretaceous beds of the Upper Amazonas (and
Orinoco) valley once continued across the whole continent to the
Pacific Ocean.

Thus there would result, in Lower Cretaceous times, a complete
separation of Central from South America by a sea, which extended
from the region of the mouth of the Orinoco westward to Ecuador
and Peru, cofinectifig the Atlantic and Pacific Oceans : In the Upper
Cretaceous, however^ Guiana was united with Venezuela.

The Fotamocarcinince, which, as we have seen above (p. 351),
arrived in Guiana in the Upper Cretaceous (by way of the connec-
tion with Africa), found at this same time a land connection with
the northern parts of Venezuela, and generally with the Antillean
continent, and this explains their general distribution over Central
America and the West Indian region, as set forth above (p. 308),
and the origin of this distribution consequently falls in the Upper
Cretaceous.

d. South America.

The separation and isolation of South America from Central,
resp. North America during Mesozoic times as well as in the begin-
ning of the Tertiary, forms the fundamental idea of von Ihering's
Archiplata and Archhelenis theory (1891). But, according to him,
the line of separation was situated in the present Amazonas valley,
and existed during the Jurassic, Cretaceous and the Eocene; in the
Oligocene the elevation of the Cordilleras began, and Archiplata
(the southern part) was united with Archiguiana (Guiana and
Venezuela), and it was not until the beginning of the seond half of
the Tertiary (Miocene) that these latter parts became united with
North America by the formation of the Isthmus of Panama.

We can accept this view only in part, since the very important

1 Hyatt, A,, in Proc. Boston Soc. Nat. Hist., Vol. 17, 1875, p. 3^5 ^-'^ Stei-
mann, in N. Jahrb. Mineral., etc., 1881. 2 p. 130 ff"., 1882, i p. 166 flf., and
Gerhardt, ibid., Beil., Bd. ii, 1897.

362 ORTMANX — DISTRIBUTION OF DECAPODS [Aprils,

interoceanic connection through the Orinoco valley, discussed
above, is not taken account of, and since, as we shall see presently,
the relations between Guiana and Brazil and between Guiana and
Venezuela are much more complex than v. Ihering assumes.

Considering the tectonic configuration of South America, we are
to mention, in the first place, that the whole eastern part is formed
by the so-called Brazilian mass (Suess, 1885, p. 655 ff.) : this is an
old Archaic-Paleozoic plateau, which was possibly connected,
from very early times up to the Lower Cretaceous (see p. 350), with
Africa. Part of this mass is formed by the mountains of Guiana
(Suess, 1885, p. 658), and the present valley of the Amazonas is a
symmetrical syncline within the old plateau, in the centre of which
are Carboniferous beds and, on top of the latter, Upper Cretaceous
deposits. Thus the Amazonas valley was apparently land during
most of the Mesozoic time, and Guiana was connected with Brazil ;
but in the Upper Cretaceous it was a wide sea, the northern and
southern shores of which were formed by Paleozoic rocks (Suess,
1885, p. 660). This sea extended from the Atlantic westward into
the region of the Upper Maranon, in the Cordilleras, and probably
connected with the Pacific Ocean (Suess, 1888, p. 683), which is
very likely, since the western shore of the old Brazilian mass hardly
extended to the eastern foothills of the Cordilleras (in a certain
region, the present river Madeira marks the western boundary), and
since it is quite sure that the Cordilleras were sea during the
Jurassic as well as the larger part of the Cretaceous. This results
in an Upper Cretaceous interoceanic connection between the
southern Atlantic and the Pacific, which was situated about where
the Amazonas valley now is. This Upper Cretaceous strait agrees
with the sea that separated von Iheri?ig's Archiplata and Archi-
guiana, but it is well to emphasize the fact that it is restricted, as a
separating strait^ to the Upper Cretaceous period : during previous
times, especially the Jurassic and Lower Cretaceous, it did not
exist at all, and later it was changed into a bay, as we shall see
below. The interoceanic connection during the earlier Cretaceous
was not situated here, but went by way of the Orinoco valley (see
above, p. 360). The directions of both straits converge to the
westward, and it is possible that they actually met, if they coexisted
at any time : but generally, we are to maintain that the separation
of Central and South America during the Loiver Cretaceous was
effected by the Orinoco Strait^ and that at this time Guiana was

1902.] AND ANX'IENT GEOGRAPHY. 363

united luiih Brazil, to which it belongs tectonically, while, i?i tlie
Upper Cretaceous, Giiia?ia was ufiited with Central America, and
was separated from Brazil by the transgression of the Atlantic
Ocean in the Amazonas valley.

This latter strait thus formed a continuation of the South Atlantic
Ocean, which came into existence, as we have seen above (p. 350),
at about the middle of the Cretaceous.

The Upper Cretaceous conditions were generally preserved in this
region during the beginning of the Tertiary, and the Eocene and
Oligocene sea extended, in the Amazonas transgression, far to the
west fbrackish Oligocene deposits are known near Pebas, Peru). But
during this time (older Tertiary), the elevation of the Cordilleras
must have become evident ^ in the western parts of this interoceanic
connection, since older Tertiary deposits are wanting in this region.
Thus those parts which now comprise Colombia, Ecuador, Peru
and Bolivia became land, and the Amazonas Strait was shut off from
the Pacific Ocean, being transformed into a deep bay, which occu-
pied the Amazonas valley as far as the foothills of this new eleva-
tion (Cordilleras). Therefore, this interoceanic connection was
interrupted in the beginning of the Tertiary, the main part of South
America, or the old " Archiplata" of von Ihering, becoming united
with northern South America ("Venezuela and Guiana). But we
have seen above (p. 344) that at the same time (earlier Tertiary or
uppermost Cretaceous) another interoceanic connection had formed
in the isthmian region, and this replaced the Amazonas connection
of the Upper Cretaceous era.

The old connection of the Brazilian mass with Africa continued
in part as we have seen (p. 350) during the Upper Cretaceous, for
its northern portion, Guiana. That is to say, an intermigration
of the faunas of Guiana and Africa was yet possible in the Upper
Cretaceous. The fact that during this time (and in the beginning
of the Tertiary) a strait or bay extended along the region of the
Amazonas river as far as the Pacific Ocean (or as far as the
Cordilleras), furnishes the explanation for the zoogeographical fact
that animals immigrating from Africa into Guiana during the
Upper Cretaceous could reach Central America and the West
Indies, but not those parts of Brazil which are to the south of this
old Amazonas Strait : this seems to apply to our Potamonidce, and

^ The first signs of an elevation belong to the Upper Cretaceous.

PROC. AMER. PHILOS. SOC. XLI. 171. X. PRINTED DEC. 29, 1902.

364 ORTMANX — DISTRIBUTION OF DECAPODS [Aprils,

explains their general absence south of the Amazonas. The exten-
sion of the range of these freshwater crabs into Colombia, Ecuador
and Peru was not obstructed during the older Tertiary, since
during this time these parts became land and were connected with
Venezuela.

Regarding the extension of the old Brazilian mass (Archiplata)
to the south, we know that the old Archaic, Paleozoic and Old-
Mesozoic rocks continue in southern Bolivia and northern Argen-
tina, into the eastern Cordilleras (Suess, 1885, p. 661) ; in Argen-
tina, these rocks prevail in the northern parts : they are also found
in the Pampean Sierras,^ but do not seem to extend southward beyond
the province of Buenos Ayres (Suess, 1885, p. 664). To the south
of these parts the whole of Patagonia was apparently covered by the
Cretaceous sea (Suess, 1888, p. 68;^, and above, p. 338J. The
Brazilian continent was also surrounded in the west by Jurassic and
Cretaceous sea, as is demonstrated by the presence of the respective
deposits in the region of the Chilian-Argentinian Cordilleras (see
p. 338). As we have seen above {ibid.), it is very probable that
during the Jurassic and a larger part of the Cretaceous era, the
Brazilian mass was separated by this sea, which occupied present
Patagonia and the site of the Cordilleras, from another continental
mass lying to the west, southwest and south of it, which was formed
by the present Chilian coast range and its southern continuation,
which belonged, at least during the Cretaceous, to the Antarctic
continent. At the end of the Cretaceous a land period began in
these regions which culminated in the Eocene, and which effected a
connection of the old Antarctica with Archiplata, chiefly in the
region of the Chilian-Argentinian Cordifleras. This connection
made possible the immigration of Parastacus into the southern parts
of Archiplata (Argentina, Uruguay, southern Brazil), and it has
remained up to the present time, although parts of Patagonia were
again submerged during the course of the Tertiary.

The results obtained in the foregoing concerning the history of
the American continent may be summed up as follows.

I. America originally consists of three parts : North America (its
nucleus being in the East), the Antillean continent (comprising the
West Indies, Central America and the northern coast of Venezuela)
and the old Brazilian fnass (Archiplata). Also a fourth part enters

* Valentin, J., Bosqiiejo geologico de la Argentina, 1898.

V.m.] AND ANCIENT GEOGRAPHY. 365

the present boundaries of South America, which is formed by the
Chilian- Fuegian coast range, once z. part of Antarctica.

2. North Afnerica was separated during the Lower Cretaceous
from Central America. During the Upper Cretaceous it was divided
into an eastern and a western portion ; the western was definitively
connected at this time with Central America. In the beginning of
the Tertiary iht eastern portion was reunited with the western, and
thus the whole of North America, from the Arctic Ocean to the
Gulf of Mexico and the Caribbean Sea, became a unit.

3. Central America existed as a continental mass up to the end of
the Cretaceous. Being originally separated from North America, it
became united with it in the Upper Cretaceous. By the formation
of the Caribbean Sea it was broken up and consisted, in the begin-
ning of the Tertiary^ of two main parts: a northern, belonging to
North America, and a southern, which became united with South
America, then undergoing the process of construction. Both parts
were separated by the Old Tertiary interoceanic connections at
Panama and Nicaragua.

The southern part of Central America was originally (Lower
Cretaceous) bounded on the south by sea, which occupied the
region from the Orinoco valley westward. In the U^per Cretaceous
Guiana was connected with Venezuela, and thus Central America
was connected also with Africa. To the south of these parts was
the Upper Cretaceous interoceanic connection of the Amazonas
valley. In the beginning of the Tertiary, what was left of Central
America in the south (Venezuela and Guiana) was united with the
Brazilian mass by the beginning of the upheaval of the Cordilleras,
by which parts of Colombia, Ecuador and Peru became land.

In the middle of the Tertiary (Miocene) the interoceanic connec-
tion in the isthmian region became land, and thus North America
and the northern remnants of Central America were united with the
southern remnants of Central America and South America.

4. South America consisted in the beginning (Jurassic and Lower
Cretaceous) of the Brazilian mass (Archiplata), which included
Guiana, and a smaller part which is perhaps of Cretaceous age, rep-
resented now by the Chilian coast range. Archiplata was con-
nected with Africa up to the middle of the Cretaceous. In the Upper
Cretaceous, Guiana was separated from Brazil by the interoceanic
connection of the Amazonas valley and Archiplata became an
island. At the end of the Cretaceous, and chiefly during the

366 ORTMANN — DISTRIBUTION OF DECAPODS [April:',,

Eocene, Archiplata became united with the Chilian coast range by
the elevation of the Cordilleras, and it was thus connected with
Antarctica. And, further, in the beginning of the Tertiary, Archi-
plata connected, by way of Peru and Ecuador, with Central
America. This resulted in the final formation of South America
(in its rough outlines) which, however, was still in communication
with Antarctica. Finally, in the middle of the Tertiary, South
America was united with North America (in the isthmian region)
and was severed from Antarctica, and this represents the chief
features of the present conditions.

We have seen that during the geological development of the
Americas interoceanic connections, which were directed east-westerly,
and united the waters of the Atlantic and Pacific Oceans, have
repeatedly played a part. These connections being extremely
important for marine zoogeography, have often been referred to by
various authors, but have generally been misunderstood, the value
of a determination of the exact time of their existence being
neglected. So it will be worth while here to put them together by
themselves.

Interoceanic coiiuections of the Atlantic and Pacific Oceans.

1. In the Loiver Cretaceous there were two connections : a. across
Mexico, and b. through the Orinoco valley. Both probably united
the marine fauna of the Mediterranean province with that of the
(Indo-) Pacific.

2. In the Upper Cretaceous we have the connection through the
Amazonas valley. This united the South Atlantic fauna, which, in
this period, formed part of the Indo-Pacific, with the identical fauna
of the eastern Pacific.

3. In the Older Tertiary there existed the Pananiic connection,
which united the fauna of the Atlantic, the chief element of which
is Mediterranean, with that of the Indo-Pacific.

4. In the Later Tertiary no interoceanic connection existed, the
Atlantic and Pacific faunas being sharply separated. These condi-
tions continued up to the present time.

It is impossible to say at present whether there were any transi-
tions between these different stages. A coexistence and union of
the connections i and 2, at about the beginning of the Upper Cre-
taceous, is possible in the region of the Upper Orinoco and Upper

1902.] AND ANCIENT GEOGRAPHY. 367

Amazonas. But we have do evidence for this, the Geology of the
respective countries being too incompletely known.

9. THE RELATIONS OF AFRICA TO THE REST OF THE WORLD.

We have seen (p. 303) that for the two main divisions of the
range of the Potamonifice in the Old World Egypt and the Nile
valley form an actual connection ; but examining this more closely
we find that this subfamily cannot have migrated along this route
from Africa to India (or vice versa), but entered Egypt from two
opposite directions, from the south C Central Africa) and the north
(resp. northeast) over Persia, Mesopotamia and Syria.

The causes why this way was not open in former times have been
briefly mentioned above (p. 333), and we shall here try to investi-
gate the relations of Africa and Asia with respect chiefly to this
northern connection. For this purpose we are to discuss also the
northern boundaries of Africa with reference to Europe. This is
the more important, since we have to consider the alleged fact that
fossil forms of the Pota7noiiincB have been found in Miocene fresh-
water deposits of Oeningen (Switzerland), Sigmaringen (Southern
Germany) and Northern Italy. ^

Very important for a study of these questions is the former exist-
ence of a Central Mediterranean Sea, as Neumayr calls it (1890, pp.
332, 333, and map p 336), or the Tethys of Suess (1894). This
ancient sea goes back to Paleozoic times and covered in Mesozoic
times the whole of Middle and Southern Europe, the present Medi-
terranean Sea, Northern Africa and extended eastward over Asia
Minor, Syria, the Caucasus Mountains and Mesopotamia as faj as
Northern India. In the east a large bay extended southward along
the East African coast, which separated the Indo-Madagassian
peninsula (Lemuria) from Africa. In a westerly direction the
Tethys was broadly connected with the Atlantic Ocean, leaving
only the island of Spain (Meseta) uncovered.

In these general outlines the Tethys existed in Jurassic as well as
in Cretaceous times, thus completely circumscribing the African
continent toward the north and northeast. Europe did not then
exist at all as a continental mass and Africa was separated from the
Sinic continent by an eastern continuation of the Tethys, the

^ Thelphtisa speciosa Mey. and Th. qttenstedti Zitt., see Zittel, Handbuch d.
Palaontol., Vol. 2, 1885, p. 714. These forms have a remote resemblance to the
subgenus Potamouautes, if they belong here at all.

868 ORTMANX — DISTRIBUTION OF DECAPODS [Aprils,

Strait of Bengal} The only connection of Africa during these
times was with South America, the old Archiplata (Jurassic and
Lower Cretaceous) and the old Archiguiana (Archhelenis, Upper
Cretaceous). On the southern margin of the Tethys, as sketched
above, there is a zone in the desert region of North Africa and
A^rabia, where Jurassic deposits are wanting and Cretaceous directly
overlies Paleozoic beds. This indicates a farther extension of
Africa northward in Jurassic times and a transgression of the sea
southward in the Cretaceous (Neumayr, 1890, p. 386). The
deposits of the Cretaceous sea can be traced very distinctly in a
broad belt from Syria over Arabia, Persia, Afghanistan and Beluch-
istan to Northern India.

Also in the Older Tertiary (Neumayr, p. 480) the Central Medi-
terranean Sea reaches from the Atlantic Ocean to India, and it was
not until after the end of the Oligocene that its unity was de-
stroyed. In the beginning of the Miocene Western Asia became
largely land, and thus a broad connection was established from
Asia to Africa (India to Arabia), and at the same time from Asia
to Europe, which was then forming (Neumayr, 1890, p. 501 f.).

In detail the processes in the northeastern part of Africa were the
following : Arabia during Mesozoic and the greater part of Tertiary
times was broadly connected with Africa. The Red Sea did not
exist, according to the unanimous opinion of all writers (Neumayr,
Suess, Gregory, Blankenhorn and others). The origin of the Red
Sea falls late in Tertiary times, after the connection of Africa with
India was long established, and thus, in the second half of the
Tertiary, a regular exchange of the faunas of Africa and India
could take place, for which we possess ample evidence.

The Red Sea is a rift valley, which is tectonically connected
with the valley of the Jordan river in Palestine.'^ The most de-
tailed investigations on this question have been published by
Blankenhorn.' According to this author, the Mediterranean Sea
(the western part of the old Tethys) in Miocene times sent a wide
bay to the southeast, which extended as far as the southern end of
the Gulf of Suez, which, of course, dicl not then exist, and the Nile

1 Which, however, was temporarily interrupted during the Upper Cretaceous.
See above, p. 330.

■■' See Gregory, J. W., in Proc. Zool. Soc. London^ 1894, p. 165.
^ Blankenhorn, M., in Centralbl. f. Mineral., etc., 1900, p. 209 ff.

1502.] AXl) ANCIENT GEOGRAPHY. 369

valley.' The latter and the Red Sea originated in the Pliocene.
Into the Nile valley entered the Pliocene Mediterranean Sea. It
then changed into a series of inland lakes, and finally, in the
middle Diluvial time, it became a river valley. The depression of
the Red Sea was occupied first (Pliocene) by inland lakes, and
finally, toward the end of the P'iocene, by the Indian Ocean,
which entered it from the south. ^ The present separation of Africa
and Arabia (Asia), which is nearly complete, belongs, therefore, to
a very recent date. In the later Tertiary Southern Asia and Africa
were not distinguished zoogeographically, while in older times
(Pre-Miocene; there was a complete separation of Africa (including
Arabia) from the Sinic continent, and only during the second half
of the Cretaceous was there a limited connection by way of Mada-
gascar and the Indian peninsula.^

The old isolation of Africa was* ended not only in these eastern
parts during the Tertiary, but also in the northwest changes
occurred which extended Africa and brought it into contact with
Western Europe.

The Cretaceous sea covering Northwestern Africa was no doubt
considerably reduced in the Tertiary. Indeed, there are Tertiary
deposits in this region, and according to Suess (1888, p. 155), the
Middle Tertiary sea probably also covered the Western Sahara.
But about this time apparently a land connection was formed to the
north toward the old Spanish Meseta. According to Bergeron,*
the Algerian Sahara possesses deposits from the Senonian to the Plio-
cene, but these are bounded in the west by a Cretaceous mountain

^ In the region of the Nil 3 valley there was a river, but this was not the Nile,
but came from the west out of the Libyan desert.

2 An actual ccmnection of the Red Sea and the Mediterranean Sea is very
doubtful, but was possibly established for a short time in the beginning of
Diluvial times, when the Mediterranean Sea became cold. The improbability
of a connection of both seas is especially emphasized by Jousseaume {Ann. Set.
Nat., Ser. 7, Vol. 12, 1891). According to hija, the Red Sea is Quarternary
(Diluvial).

3 The oceanic connection from the Gulf of Aden across the Sahara desert to
the Atlantic (Senegambia), advarced by Jousseaume (/. <r ), has no geological
support. It is founded upon an alleged similarity of the Mollusk faunas of both
parts, which, however, needs closer inves igation and might possibly find its
explanation in the configuration of the Pre-Miocene Tetkys, which reached from
the Persian Gulf to the Mediterranean Sea and Atlantic Ocean.

* In Mem. Soc. Itigen. civ. France, 1897.

370 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

range running north-south. In Algeria we have, according to Lap-
parent/ deposits of Cretaceous and Eocene age, but only traces of
Oligocene, and thus we can place the upheaval of these parts at the
end of the Eocene, and probably at this time the connection with
Southwestern Europe began to develop. The mountain range
along the northern coast of Algiers, as far as the Strait of Gibraltar,
consists of rocks which (see Suess, 1883, P- 293 ff.) are also found
in the so-called Betic Cordilleras {ibid., p. 298) in Southern Spain,
and the tectonic unity of these ranges of Algiers and Spain is
especially emphasized by Suess, as well as their tectonic connection
with the Apennines and the Alps. The origin of all these moun-
tain chains was near the. end of the first half of the Tertiary, about
the Oligocene time.

But this connection of the northwestern parts of Africa and of
Southern Spain with the rest of Africa did not constitute a com-
plete union with Europe. We know that the central and northern
region of Spain, the Iberian Meseta, is an old land, but that to the
south and north of it, on the one side along the valley of the
Guadalquivir river in Spain, on the other in the region of the
Garonne river in France, connections of the Mediterranean Sea with
the Atlantic Ocean existed. According to Suess (1885, P* 3^1 ; see
also Neumayr, 1890, p. 516), in the valley of the Guadalquivir
there are Tertiary deposits, reaching from the Atlantic to the Medi-
terranean Sea, which belong to the first and second Mediterranean
stage — that is to say, to the Miocene — while deposits of the third
Mediterranean stage (Pliocene) have not been found. Conse-
quently this strait (Betic Strait) became dry at the end of the
Miocene, and by this process the northern part of Spain was united
with the southern and with Algiers and Africa.

The disappearance of this strait was the last step which resulted
in a definitive connection of Africa with Europe, since the strait in
the region of the Garonne river, in Southern France, became prob-
ably land a little earlier, namely, at the end of the Oligocene (see
Suess, 1885, P- 3^2 ff., and Neumayr, 1890, p. 516).

But it is to be borne in mind that possibly the conditions were
not so simple as has been represented above. According to Lap-
parent (/. c, pp. 1 291 and 1313), the Betic Strait (detroit betic,
also called Andalushn connection) was dry during the Oligocene,

^ Traite de Geologie, Vol. 2, 1893, p. 1291.

1902.] ' ANJ) ANCIENT GEOGRAPHV. 871

while sea again occupied it in the Miocene, and thus we would
have a chance for the African fauna to reach Northern Spain as
early as the Oligocene. This connection, however, scarcely
amounted to a complete communication of Africa and Europe,
since at that time the Oligocene strait to the north of the Iberian
Meseta was still in existence, forming a barrier to the further
advance of the African fauna. Thus, even under this assumption,
a final connection of Africa and Europe was not established until
the end of the Miocene, after the second obliteration of the Betic
Strait. Subsequently the connection of both continents was again
interrupted by the formation of the Strait of Gibraltar; but this
belongs to very recent times. ^

Another tectonic line goes from Northwestern Africa to Sicily
and Italy, and is marked by the eastern continuation of the same
mountain range that curves in the west from Africa into Southern
Spain. This system belongs to Post- Oligocene times, and as a
land-bridge apparently underwent repeated changes. Moreover, it
is doubtful whether it existed at any time as a complete and solid
bridge, but it is represented as such by Scharff (1895, maps pp. 465
and 470) for the Pliocene time,'^ while Neumayr (1890, Vol. i, p.
330), for the Lower Pliocene, gives only a series of islands.

From the above discussion we are to draw the conclusion that —
aside from a connection with' the Sinic continent during the Upper

1 Kobelt, W. {Sttidien zur Zoogeographie — " 2. Die Fauna der mediterranen
Subregion," 189S), arrives at a diffeient conclusion. According to him, the
Mediterranean Sea was separated from the Atlantic Ocean in the Older Tertiary
by a connection ot Central Spain with the Atlas mountains (the Sierra Nevada
or Betic Cordilleras did not then exist). The Miocene Tethys reached from India
to Spain and Central France, but did not communicate with the Atlantic, the
connection along the valley of the Guadalquivir being of Pliocene age.

This result is contrary, however, as we have seen, to what is known of the
Geology of these parts. Just the opposite is the case. In the Older Tertiary the
Tethys and the Atlantic were broadly connected, and in the Miocene they still
communicated through the Andalusian or Betic Strait, as is positively shown by
the presence of Miocene beds there. But in Pliocene times this strait was dry
land.

2 According to Scharff (in 1897, pp. 461 and 466), this bridge belongs to the
Upper Pliocene and Gla:ial times. We shall become acquainted below with the
evidence for its exis^tence as an actually connecting bridge.

372 ORTMANN — DISTRIBUTION OF DECAPODS ' [April 3,

Cretaceous — Africa, for a very long time, was isolated from the rest
of the Old World. After it had become disconnected from South
America, at the beginning of the Tertiary, it was absolutely
isolated, but soon during the course of the Tertiary it became
united with Asia and the new continent of Europe. The most impor-
tant stages in this process were that of the elevation of Western Asia,
in the Miocene, and the elevation of the northwestern parts of
Africa and southwestern parts of Europe at about the same time.

This has the following bearing upon the origin of the distribution
of our freshwater Crustaceans: The African types of the subfamily
Potamonince (chiefly the subgenus Potainonautes) could not reach
Europe before Miocene times, and, on the other hand, an immigra-
tion of the Asiatic types (subgenus Fotanion) into Africa (and
Europe) was also impossible before the Miocene and after the de-
struction of the Madagassian land-bridge in the earlier Eocene.

Whether the alleged fossil species of Potamon from the Miocene
of Europe indicate this Miocene connection of Asia, Africa and
Europe remains doubtful. The lack of African types in the Medi-
terranean region of the present time, as well as the general absence
of a northerly and easterly advance of them (except in the Nile
valley, where the immigration no doubt belongs to a very recent
period), is opposed to the above assumption, and it is quite possible
that these fossil forms do not belong to relations of the Potamonina.
It seems that the desert zone of the Sahara already existed in
Miocene times, at least that it began to develop at the same time
that Western Asia became land, since just this process furnished
the conditions for the origin of an arid climate in North Africa and
West Asia. On the other hand, we see that a species of the sub-
genus Potamon^ belonging to the Indian fauna, advanced in a west-
erly direction across the new land areas formed in Miocene time,
and that it reached by this route Northern Africa (Egypt and
Algiers). But the distribution of this species {Potamo?i fluviatile)
needs further explanation, since it is also found in certain parts of
Europe, and we shall discuss this question in the next chapter.

lO. RELATIONS OF EUROPE TO ASIA.

In discussing the distribution of Potaiiion fluviatile in Southern
Europe, just referred to, we are also to consider the presence of the
genus Potaifwbius in Europe, the area of which is separated from
that of the rest of the genus (in Northeast Asia and Northwest

H*02.] AND AN'CIENT GEOGRAPHY. 37S

America). The essential point in tliis respect is the investigation
of the geological relation of Europe to Asia.

Europe did not exist as a continent — i.e.., as a continuous mass —
from the beginning of the earth's history up to about the middle of
the Tertiary. Indeed, there was a number of larger and smaller
islands in the old Tethys, but they never were connected so as to
assume continental shape. To the north, however, we had the
large Scandinavian mass, which probably was connected over
Greenland with North America, but we shall disregard this possible
junction, since our present material, the Decapod Crustaceans, do
not furnish additional facts which bear upon it.

The Tethys, as we have seen (p. 367), covered the whole region
of the present Mediterranean Sea and extended over Western Asia,
reaching, in the older Eocene, not only as far as the Indian Ocean,
but in an easterly and northerly direction as far as the eastern side
of the Kuen-Lun mountains and the Gobi desert.^ Subsequently,
in the Miocene, the western parts of Asia (from Asia Minor and
Syria to India) became land (Neumayr, p. 501), and the Tethys
was cut into a western (the present Mediterranean Sea) and an
eastern section (forming part of the Indian Ocean). But during
this time, and even afterward, the northern and northeastern parts
of the old Tethys persisted. The Miocene Mediterranean Sea
(Neumayr, 1890, p. 516) sent a strait from the basin of the Rhone
river (France) through Switzerland into Austria, which there
widened out into the Pannonian basin and in the Upper Miocene
became a huge inland sea, the Sarmatian, which was cut off from
its former western connection with the Mediterranean and reached
from Austria over Southern Russia into the region of the Caspian
and Aral Seas (Neumayr, p. 523). To the south of this sea the
present Balkan Peninsula, the ^gean Sea and Asia Minor were
largely land, but in Eastern Asia Minor a continuation of the
Mediterranean Sea approached almost to the Black Sea. The
region of the Caucasus mountains was probably sea up to Miocene

1 In the Kuen-Lun mountains there are, according to Bogdanowitsch
[Geolog. Untersuch. im oestlichen Turkestan, 1892, Russian. Review in
Neues Jahrb. f. Mineral., eic, 1895, Vol. 2, p. no), Archaic and Paleozoic
rocks and traces of Jurassic deposits (coal-bearing strata), and then again marine
Cretaceous beds. Thus it seems that these mountains were land since beginning
of the Mesoziic times and formed part of the Sinic continent. During the Cre-
taceous there was a temporary transgression of the sea.

374 ORTMANN — DISTRIBUTION OF DECAPODS [Aprils,

times, as is shown by a continuous series of sediments lying upon
Azoic rocks. Here ^ has been found Jurassic, Cretaceous, Eocene,
Oligocene and Miocene. The latter deposits (Miocene) belong
to the Sarmatian inland sea. Beginning with the Pontic stage, the
sea recedes on the southern side of the Caucasus (freshwater
deposits), while on the northern side marine deposits, belonging
to the Ponto-Caspian Sea, continue. The latter disappear after the
Glacial period.

The map, given by Neumayr, of the Eastern Mediterranean
countries during the Lower Pliocene (1890, Vol. i, p. 330, see
also Vol. 2, p. 526) exhibits a much more extended development
of land than at the present time. Especially striking is the direct
connection of Asia Minor with the Balkan Peninsula and Central
Europe. The corresponding rrap for the Later Pliocene, given by
Scharff (1895, p. 465, and 1897, p. 461), indicates an additional
land connection from Dalmatia over Southern Italy and Sicily to
Algiers (see also 1895, P- 47°» ^^^ 1897, p. 461), which is repre-
sented in Neumayr's map only by a series of islands." Thus we
obtain a continuous land connection from Asia Minor to North-
western Africa, belonging to the Pliocene age.

In the Pleistocene (Glacial) time, according to ScharfF (1897,
map p. 466), this connection is still present.

In the northern parts of Europe we have no land connection in
an easterly direction during the Cretaceous time. According to
Koken (1893), however, North Asia was connected with Scandi-
navia in the Upper Cretaceous, forming part of a huge circumpolar
Arctic continent ; but the evidence for its existence seems to be
very doubtful. For the Older Tertiary, Koken again indicates a
separation of Northern Europe from Asia. In subsequent times, up
to the Later Pliocene, the Sarmatian Sea covered the whole of
Northeast Europe (Scharff, 1897, map p. 461), thus perpetuating
the separation from Asia. During Glacial times this separation
was maintained by the ice sheet covering Northern Russia and by
the existence of the Aralo Caspian basin, and it was not until
Interglacial times that a communication of Asia and Central

' See Fournier, in Ann. Fac. Sci. Marseille, Vol. 7, 1896.

■•* The large extension of the Mediterranean Sea to the south of Algiers over
the Sahara desert in a westerly direction, as shown by Scliarft's map (p. 470), is
probably erroneous.

1902.] AND AN'CIENT GEOGRAPHV. ^^^75

Europe was established north of the Aralo Caspian basin over
Southern Russia (Scharff, 1897, map p. 466).

The gradual origin of Europe, beginning with the formation of
the chief mountain chains in the Oligocene, its connection first of
the southern and central parts with Western Asia across the Balkan
Peninsula (Miocene and Pliocene) and with Northern Africa over
Spain (end of the Miocene), and subsequently the connection of
the- central and northern parts with Siberia (over Russia), by which
processes Europe assumed the shape of a continent (part of Asia),
have been largely used by previous authors for the explanation of
the zoogeographical conditions of Europe.

Osborn (1900, p. 569) mentions a repeated immigration of
Mammals into Europe and indicates the Upper Eocene, the
Miocene and the Pliocene times as most important in this respect.
But we m.ust always bear in mind that during the Older Tertiary
Europe was not a unit at all — in fact did not exist as a zoogeo-
graphical section. The Old Tertiary Mammalian faunas of Europe
(chiefly in the Northwest, in. France) probably belong to the British-
Scandinavian mass, which was connected, as has been mentioned
incidentally, with North America.^ Then, in the Miocene, we
have in Europe, which assumes a more consistent shape, a fauna of
new character, the origin of which is to be sought in the East and
Southeast (Asia), and possibly during this time the first African
types reached Europe, either by the roundabout way over Western
Asia or more directly over Algiers and Spain.

Kobelt (1897) also assumes an isolation of Europe at the begin-
ning of the Tertiary, and discusses the immigration of an Indo-
Chinese fauna from the East in Pliocene times, while the Nile
valley formed a route by which freshwater animals immigrated
from the South (Africa).^

The most detailed investigations on this question have been pub-

1 As to this connection, which is not treated here, I refer the reader to Neu-
mayr (1890, Vol. 2, pp. 497 and 504) and to Scott {An hitroduction to
Geology, 1897, P- S^S)-

2 Contrary to this, Pilsbry C1894) is inclined to assume, for the Helices, a Cre-
taceous immigration from Southeastern Asia into Europe and Africa. But
according to the present state of our knowledge, as set forth above, the history ot
the d'-velopraent of Africa and Europe, as well as of Asia, does not warrant ihis
assumption. There was no possibility, on geological grounds, for the old Sinio
fauna to reach Europe and Northern Africa before ihe Miocene.

376 ORTMANX— DISTRIBCTIOX OF DECAPODS [Aprils,

lished by Scharff (1897). He distinguishes — aside from an Arctic
migration — two main routes of immigration into Europe during
the Later Tertiary period : i, a southern one during Miocene and
Pliocene, which was directed from Western Asia over Asia Minor,
the Balkan Peninsula, Italy, Sicily, Algiers and Spain (and which
apparently sent a branch from the Balkan Peninsula into Central
Europe), and 2, a Siberian migration from Western Siberia through
Southern Russia to Central Europe, which belongs to the Pleisto*
cene (see map, /. c, p. 466) and was impossible before this time
(in Miocene and Pliocene), the Sarmatian and Ponto-Caspian Sea
forming barriers.

Comparing our freshwater Decapods with the above, we see at
the first glance that the present distribution of the European fresh-
water crab, Potamon fluviatile, unmistakably agrees with that land
connection which began in the Miocene and culminated in the
Pliocene and which extends from Asia Minor over the Balkan
Peninsula to Italy, Sicily and Algiers. Even the minor features of
it are traceable. Potamon fluviatile is found everywhere in West-
ern Asia, in the Caucasus region and in the Crimea, but is missing
in the rest of Southern Russia. This corresponds to the fact that
the Crimea was connected in Pliocene times with the Caucasus and
was not in communication with the rest of Russia (see Scharff, 1897,
map p. 461). Potamon fluviatile is found in Asia Minor, Syria, on
the island of Cyprus and in Egypt. All these parts were then con-
nected. Along the tract of the land-bridge, from Asia Minor to
Italy and Algiers, this crab has been everywhere found. ^ This
relation of the supposed Pliocene land exteision with the distribu-
tion of Potamon fluviatile is so close that there is no objection
whatever to the assumption that the immigration of this species
falls in the Upper Pliocene, when this land connection was fully de-
veloped, and not in the Lower Pliocene, when there was only a
series of islands (see p. 374).

Turning now to the crayfishes of Europe, we see that the centre
of the range of the group of Potamobius astacus (the Russian cray-
fishes) is just in that region which, during Miocene, Pliocene and

^ That this species extended, in former times, farther to the north from the
Balkan Peninsula is shown by the discovery of it in lossil state in diluvial cal-
careous tufa near Siiito, Com. Komarom, Hungary (see Loerenthey, E., Nat.
Hefte Ungar. Nat. AJus., 1898. Review in Neiics jfahrb, f. Mineral., etc.,
1900, Vol. 2, p. 473).

1902.] AND ANCIENT GEOGRAPHY. 377

the beginning of the Pleistocene, was covered by the Sarmatian and
Ponto-Caspian Sea. This group consequently can only have
reached these parts at a later period, namely, in Interglacial or
Postglacial times, and its immigration no doubt corresponds to the
Siberian of Scharff (/. c, pp. 448 and 466). In regard to the other
group formed by the two species, P. pallipes and to?'rentium, which
occupy the South and West of Europe, we have to call attention to
the important fact that this group is found not only in Southern and
Western Europe, but also in England. Now we know that England
was connected with the continent in Preglacial and even during
Glacial times, and that this connection existed up to the beginning
of the Siberian migration (Interglacial). It was interrupted later —
according to Osborn (1900, p. 572), at about this time (Middle
Pleistocene) and possibly even later. ^

The fact that Potamobius astacus is found in France, but not in
England, while P. pallipes passes over into the latter country,
points to a difference in time of the immigration of either species.
P. pallipes arrived in these parts before the end of the Glacial time,
P. astacus at the end of it or even later. The latter consequently
without doubt belongs to the Siberian migration, but rather to the
later part of it. P. pallipes may belong to the earlier Interglacial
part of the Siberian migration and have come from East and Cen-
tral Europe; but it is also possible that it belongs to Scharff's
southern migration and came from Asia Minor over the Balkan
Peninsula. It is true, forms of the pallipes group have not been
found in Asia Minor nor in Algiers, but it is not impossible that
such may be discovered in these parts, or that they once existed
there and have now disappeared. We shall see below why this
latter assumption is admissible. The crayfishes in Asia Minor,
Southern Italy and iVlgiers may have been exterminated by the
freshwater crabs subsequently occupying these parts. Until this
question is finally settled it is impossible to decide whether the
gxoxr^oi P. pallipes has reached its present area by the southern
route (Miocene-Pliocene) or by the route of the Siberian migration
(end of the Pleistocene) ; but however that may be it arrived in
Europe before the group of P. astacus.

The connection of the European crayfishes with the Sinic conti-

^ Suess (1888, p. 528) thinks it possible that this happened in historic time or
shortly before the beginning of it.

378 ORTMANX — DISTKIBUTIOX OF DECAPODS [Aprils,

nent, where presumably their original home was located, is not yet
established. It must necessarily have gone over Central Asia. Cray-
fishes of the European type are found eastward as far as Turkestan.
It is doubtful whether crayfishes are absolutely lacking in the region
between Turkestan and the Amur river. None are reported, but
these parts are very poorly known. For the present I cannot imag-
ine any reason for their disappearance in this region, in which they
must have once existed, and therefore it is well to suspend judgment
until these parts have been properly investigated.^

SUMMARY OF RESULTS OF PART II.
A. History of the Continents.

a. Lower Cretaceous. (See Fig. 5, p. 379.)

I. During the Lower Cretaceous there existed a Sino- Australian
continent, comprising eastern Asia, the Indo- Malaysian Archipelago
and Australia, and which was continued to Antarctica. We may
retain for this continent the name Sino-Australian, although it
is larger than that drawn by Neumayr for the Jurassic time. We
are justified in doing this, since probably the Jurassic Sino-Austra-
lia also included Antarctica.

II. Besides, we have 2. Nearctic coniifient : this consists of the
larger part of present North Ai7terica, and extended probably across
Greenland to the Scandinavian mass of North Europe. This con-
tinent also corresponds closely to Neumayr's Nearctica of Jurassic
times, and consequently we have retained this name for it.

III. A third continent was formed by Central America, and we
shall call this by the name Antillia. Its remnants are now found
in Central America, the West Indian Islands and northern South
America (excluding Guiana). This continent is not given by
Neumayr for the Jurassic, but probably existed then.

IV. A fourth continent was formed by the western portion of old

' A theory lately propounded by C. Y. Wright (see Science^ Vol. 16, Aug. 15,
1902, p. 262 f.) would go far toward an explanation of the causes leading to the
destruction of the Central Asiatic crayfishes if properly supported. Wright be-
lieves that Northern and Central Asia was largely coverad by water in recent
geological time, but the evidence introduced lor this is, in my opinion, entirely
inappropriate. Of the five points mentioned by Wright two (Nos. 3 and 4)
have no bearing at all upon this theory, and the value of the other three, espe-
cially of the fifth, is highly questionable.

1902.]

AN"D AXCIENT GEOGRAPHY.

87^

PROC. AMEK. PHILOS. SOC. XLT. 171. Y. FEINTED DEC. 31, 1902.

380 ORTMANN — DISTRIBUTION OF DECAPODS [AprD3,

Gondwana Land. It comprises the Brazilian mass (including
Guiana) and tropical Africa with the Lemuriatt Peninsula (Mada-
gascar-India). This continent corresponds, generally, to Neu-
mayr's Jurassic Brazi/o-Elhiopian continent, but comprises a
smaller part of South America (also, for Jurassic times, the section
of South America that entered it, according to Neumayr, is too
large). It agrees to a certain degree with what v. Ihering has
called Archhelenis, although it is larger, and it may be permitted
in this sense to modify the conception of Archhelenis.

Thus in Lower Cretaceous times we have the following four con-
tinental masses : Sino-Australia, 'Nearctica, Antillia, Archhelenis,
which were mutually isolated. Besides, there were smaller islands,
chiefly in the region of present Europe.

b. Upper Cretaceous. (See Fig. 6, p. 381.)

The following changes took place :

Sino Australia was divided into a Sinic (Asiatic) and zn Australian
part, the latter comprising Australia and Antarctica.

The Sinic section of Si no-Australia became united, across Bering
Sea, with the western part of Nearctica.

The western part of Nearctica was separated from the eastern.

The western part of Nearctica became united with Antillia.

Guiana became united with Antillia and separated from the
Brazilian mass.

Brazil became disconnected from Africa.

The Lemurian bridge was connected with the Sinic continent.

The result of this is :

I. An irregular ri?ig of land around the earth, which, generally,
encircles it in the direction of the equator, but curving far to the
north in the region of the Pacific Ocean. This ring, beginning at
the Sinic land, goes across Bering Sea to western North America,
thence to Antillia^ Guiana, Africa and the Lemurian land bridge,
which latter completes it by its union with the Sinic land. We
m.ay call this ring-shaped continent Mesozonia.

Aside from Mesozonia we have, separated from it, the following
continental masses :

IL Upper Cretaceous Nearctica. Smaller than the Lower
Cretaceous continent of the same name, since its western part is
cut off and enters Mesozonia.

19()2.]

AND ANCIENT GEOGRAPHY.

oyi

382 OKTMAXN- — DISTRIBUTION OF DECAPODS [Aprils,

III. Archiplata of von Ihering, comprising Brazil, south of the
Amazonas, and northern Argentina.

IV. Archinotis, comprising Australia and Antarctica. As will
be noticed, the term Archinotis (as used by von Ihering and others)
does not exactly correspond to the meaning given to it here, but we
think it convenient to define this term in this way, applying it only
to the truly notal regions ; aside from Australia and Antarctica^ a
part of South America belongs to Archinotis, namely the western
(Chilian coast range).

Note — The existence of this ring-shaped continent Mesozonia in
Upper Cretaceous times is extremely important for marine zoogeo-
graphy. The distinction of two types of marine faunas, the Medi-
terranean and the Pacific, is well known among geologists, and
this continent furnishes an expla