!|5

•COfi dcJ

O ~ O

ES^SMITHSONIAN'^INSTITUTION NOlifUliSNI NVINOSHiiWS S3iyVHan LI B R AR I ES“SMITHS0NIAN‘^1NSTITUTI0N
Z f~ ^ ^ 2 ^ „ 2 r- 2 r~

CB ,- /{ir^vSx g p 1 /0^ t •« "

>

~ ro

2 |T1 ^ m

CO — CO — £/) ^ (/5

Ni NViNOSHilWS sBiavaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiins S3iavaan

2 » CO 2 CO

O

I;; I’' ■ ■ i --'y I i ' 5 '■^S^ |-'

ES^SMITHSONIAN INSTITUTION N0liniliSNI_NVIN0SHilWs‘^S3 I a Va 8 11 LI B RAR I ES^SMITHSONIAN^INSTITUTION

CO “ CC7 2 \ ^ 5 ^

I Itz^ -«aA>r>l ^ loS^ — I » ^Ib^x.^'SSSSx ^ /Cfi^

i\ q:

*<<

Q 'vJ^OSUiiV^ ^ ^ Q X^OSK.^ ~ o

NrNVINOSHimS S3iavaan‘^L!BRARIES SMITHSONIAN INSTITUTION NOIiniliSNl“*NVINOSHill/yS^S3 I a Va a 11
r- 2 f“ z <“ 2 r-

m m x2v_DXi/ " m

CO \ ^ CO ± CO ± — CO _

ES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHillNS S3iavaail LIBRARIES SMITHSONIAN INSTITUTlO^
z -V 5 ^ ^ 5 <••■ ^ z V

•«. . ■ ^ x<av!soa^X S . <

2

£\ O

o) ^

/ h“

;ni NviNOSHiit^s^^sa I ava a n libraries Smithsonian institution NoiiniiiSNi NvmosHiitvs'^sa i avaa r

CO 2 CO CO — (/>

ES SMITHSONIAN INSTITUTION NOliDiliSNI NVINOSHilWS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION

2 <2 ^ 1. ^ ^ ^ *”

O jirJSiiSi, _ /TolljoX O xSvriryX „ O “ y<''Ealv?;'x O

XV*

ro '' >^' ^ rn X^osk^ ^ m x

CO !Z CO ' ' _ CO _ CO

NI NVINOSHillMS S3iaVaail LIBRARIES SMITHSONIAN INSTITUTION NOlifUliSNI NVINOSHilWS S3iaVaai1

2 CO 2 , CO Z » CO 2

— ~ ^ E

2 /^jSWl^is H .5^ -H 2 H 2 /WisSWiiaX H

CO

■- -.

ES^SMITHSONIAN lNSTITUTI0N‘^N0liniliSNI_NVIN0SHilWs‘^S3 I a Va a 11 LI B R AR I Es'^SMITHSONIAN INSTITUTION

CO — CO 2 \ ^ 5 CO ~

>o'^n *T>v t • ( ✓'fTcrirTrSw . _ 1 1 1 >■'''^^^IT7T»^v i * t

A

<
q;

0 ™ O — '^Xt!iAS»^ Q X^^osjA^ ~ O

<»J ^ _j ^ 2 «ol 2

;ni NviNOSHiit^s S3iavaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws ssiavaai

1 CO ° / . ^ I

m xc CO m X^cac>' ^ rn xiVDVx' ^ m

CO A 2 CO ± CO ± CO

ES SMITHSONIAN INSTITUTION NOliOiliSNI NVINOSHillNS S3iavaatl LIBRARIES SMITHSONIAN INSTITUTlOf

CO Z ■ CO 2 CO 2 .,■. CO Z ,vv

? V.- 2 .< E X ^ —

tn

O xii'v Q ~ o 'vi^v_£v^ _ v//~ O ''5!;<j2S

iliSNl""NVIN0SHiIIAlS^S3 iavyan~^LIBRARI ES^SMITHS0N1AN"^INSTITUTI0N NOIiniliSNl“'NVINOSHillAIS S3 I y vy

r- , z r- z <“ 2 ^

7rT:t>^ .V O X5?fir(})x ^ y<-ia O „ xfo<3>oX m x>x 2

^RIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3iavyan L I B R A R I E S SMITHSONIAN INSTITU

I
CO

O

iliSNI NVIN0SHillMs‘^S3 I ava a n^LI B RAR I ES^^SMITHSONIAN INSTITUTION NOliniliSNI _ NVINOSHillNS^ S 3 1 a V }

CO = CO — CO z \ 2 ^

UJ x55iij<ovX uj caJ m cu

ARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHillNS S3iavaail LIBRARIES SMITHSONIAN INSTITU
z 2 ^ z r- z ^ ^

12^ 2 .m 2 ;;; o cd

m w pi ^

_ CO ± CO \ ^

UliSNI NVINOSHilWS S3iaVaail libraries SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3 1 a VI

7 CO Z C/> ^ CO ^

< s: < .vv^o S 2 .<

ARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHlIiAIS $ 3 I a V a a ll_ ® ^ ^ ^ ^^'^''^^SONIAN _ I NSTITL

H

Ql C

Q -s^/y pc>^ _ Xj'UixS^ Q 'X — X^OxixS^X ^ _ v- O

U!lSNl"^NVIN0SHllWS^S3 I avaa n”'LI BRARI ES^SMITHSONIAN“'lNSTITUTION^NOIinillSNI NVINOSHIIWS S3 I av)

r~ , z i— z ^ L'^

o XisirnTX ~ O Z X^SaT??x 2 m ^ 2

I

» O

UliSNI _NVIN0SH1I/Js‘^ S3 I ava a ll^LI BRAR I E S^SMITHSONIAN INSTITUTION NOliniliSNI _ NVINOSHlIiAIS^^ S 3 I a V

o

z

H

^ c

_ /*'

O — X^WiX Q o

ARIES SMITHSONIAN INSTITUTION NOIinillSNl“^NVINOSHlllAIS ^S 3 I a V a a 11 ^L I B R A R I E S ^SMITHSONIAN^I NSTITL

- ro 00 °

PI ^ m ^ ro V°\_dc>^ ^ m

CO „ co^. ? ^ ^

niiiSNi NVINOSHlIiAIS S3iavaaii libraries Smithsonian institution NoiiniiiSNi nvinoshiiiais S3iav

Z c/5 Z CO Z . . CO Z CO

V.

iK ' ' r

■[f,' 1^. I

O-i/'.''

•'^. t. ' 4t

■■v<l

'"*''■ X' '
• ,™''>; . ■

"■' r'jK

'.!?v ' '

'>v^

4

■A'<

V

-jf*;'

'A

relationships OF SOME INSECTIVORES AND
RODENTS FROM THE MIOCENE OF
NORTH AMERICA AND EUROPE

BURKART ENGESSER

NUMBER 14 PITTSBURGH^ 1979

BULLETIN

of CARNEGIE MUSEUM OF NATURAL HISTORY

RELATIONSHIPS OF SOME INSECTIVORES AND
RODENTS FROM THE MIOCENE OF
NORTH AMERICA AND EUROPE

BURKART ENGESSER

Resident Museum Specialist, Section of Vertebrate Fossils
{permanent address: Naturhistorisches Museum,
Basel, Switzerland)

NUMBER 14

PITTSBURGH, 1979

BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY
Number 14, pages 1-68, figures 1-12, table 1, plates 1-20

Issued 14 May 1979

Price: $5.00 a copy

Craig C. Black, Director

Editorial Staff: Hugh H. Genoways, Editor; Duane A. Schlitter,
Associate Editor; Stephen L. Williams, Associate Editor;
Barbara Farkas, Technical Assistant.

(c) 1979 by the Trustees of Carnegie Institute, all rights reserved.

CARNEGIE MUSEUM OF NATURAL HISTORY, 4400 FORBES AVENUE
PITTSBURGH, PENNSYLVANIA 15213

CONTENTS

Abstract 5

Zusammenfassung 5

Resume 5

Introduction 6

Acknowledgments 7

Relationships of Insectivores 7

La n tlia noth eriu m 7

Plesiosore.xIMeterix 9

Subfamily Heterosoricinae Viret and Zapfe 1951 12

Limnoeciis 19

Relationships of Rodents 20

'"Sciuropterus" 20

Petauristodon, new genus 23

Family Eomyidae Deperet and Douxami, 1902 23

Pseudotheridomys 24

Leptodontomys 25

Eomyops, new genus 27

Cotimus, Leidymys, Eiimyaiion. and Eucricetodon 28

Democricetodon and Copemys 31

Plesiosminthus 35

Discussion 39

Migration 39

Parallel Evolution 41

Conclusion 42

Literature Cited 42

Plates 46

<

1

ABSTRACT

Ten genera of Miocene small mammals from Europe are com-
pared with similar forms, regarded as congeneric by many au-
thors, from North America. These are Lanthanotherium, Ple-
siosore.x. "Trimylus" IHeterosorex, Limnoecus, "Sciiiropterus,"
Pseudotheridomys , Leptudontomys, "Cotimus," Dcmocricelo-
don/Copemys, and Plesiosminthus. On the basis of morpholog-
ical differences, the European eomyids hitherto assigned to Lep-
todontomys are here separated from this genus and united in the
new genus Eomyops. For the North American rodents so far
described as "Sciiiropterus" the new genus Petauristodon is
introduced. The newly discovered upper dentition oi "Cotimus"
ulicae shows that Schaub’s "Cricetodon heveticus" and related
forms from Europe, assigned to the American genus "Cotimus"
i=Leidymys) by several authors, have nothing to do with this

genus. The European forms are merged within another genus for
which the name Eumyarion Thaler, 1966, should he used. The
European heterosoricine genera Quercysore.x, Heterosore.x, and
Dinosore.x are differentiated morphologically from the American
genera Domnina and Pseudotrimylus. Some small soricids from
Europe assigned by many authors to the American subfamily
Limnoecinae are compared with representatives of the genus
Limnoecus. These comparisons led to the conclusion that at
present there is no reason to include any European form in this
American subfamily.

The discussion attempts to show that in addition to migration
other possibilities, such as parallelism and persistence of prim-
itive characters over long periods, must be considered to explain
the close similarities of some forms from both continents.

ZUSAMMENFASSUNG

Zehn miozane Kleinsauger-Formen Europas werden mit ahn-
lichen Formen aus Nordamerika verglichen, mit welchen zusam-
men sie bisher in die gleiche Gattung gestelit wurden: Lantha-
notherium. Plesiosore.x, "Trimylus" /Heterosore.x. Limnoecus.
"Sciiiropterus, " Pseudotheriodomys, Leptodontomys, "Coti-
mus." Democricetodon/Copemys, Plesiosminthus. Aufgrund
morphologischer Unterschiede werden die bisher Leptodonto-
mys zugeordneten Eomyiden Europas von diesem Genus abge-
trennt und im neu geschaffenen Genus Eomyops vereinigt. Fiir
die bisher als "Sciiiropterus" beschriebenen Formen Norda-
merikas wird das neue Genus Petauristodon aufgestellt. Funde
der bisher unbekannten Maxillarbezahnung von "Cotimus" al-
icae haben gezeigt. dass Schaubs "Cricetodon helveticiis" und
verwandte Formen Europas, welche von manchen Autoren dem
amerikanischen Genus "Cotimus" (=Leidymys) zugeordnet
wurden. nichts mit diesem Genus zu tun haben. Sie sind im

Genus Eumyarion Thaler 1966 zu vereinigen. Die europaischen
Heterosoricinae-Genera Quercysore.x, Heterosore.x und Dino-
sore.x werden gegen die amerikanischen Genera Domnina und
Pseudotrimylus morphologisch abgegrenzt. Die von manchen
Autoren der amerikanischen Unterfamilie Limnoecinae zugeord-
neten kleinen Soriciden Europas werden mit Vertretern der Gat-
tung Limnoecus verglichen. Dabei zeigt sich, dass gegen wartig
keine Veranlassung besteht. irgendwelche europaischen Formen
in diese amerikanische Unterfamilie zue stellen.

Im allgemeinen Teil wird zu zeigen versucht, dass als Erkla-
rung fiir das Zustandekommen der erstaunlichen Aehniichkeiten
von Formen beider Kontinente, neben der Migration auch noch
mit andern Mbglichkeiten wie Parallelentwicklungen und Bei-
behaltung primitiver Merkmale fiber lange Zeitraume zu rechnen
ist.

RESUME

Dix micromammiferes miocenes d" Europe sont compares a
des formes semblables de FAmerique du Nord qui figuraient
jusqu'ici dans le meme genre: Lanthanotherium, Plesiosore.x,
"Trimylus" /Heterosore.x, Limnoecus, "Sciiiropterus," Pseu-
dotheridomys, Leptodontomys, "Cotimus," Dernocricetodon/
Copemys, Plesiosminthus. En raison de differences morpholo-
giques, les eomides d’Europe sont separes du genre Leptodon-
tomys auquel ils avaient ete attribues et sont reunis dans le genre
Eomyops nouvellement cree. Pur les formes americaines de-
crites auparavant, comme "Sciiiropterus ." un nouveau genre,
Petauristodon. est etabli. La decouverte de dents de maxillaires
de "Cotimus" alicae inconnues jusqu’a present a montre que
le "Cricetodon helveticiis" de Schaub et d’autres formes appar-
entees d’Europe, attributes par plusieurs auteurs au genre amer-
icain "Cotimus" {=Leidymys), n’avaient rien a voir avec ce gen-

re. Ces formes doivent etre reunies dans le genre Eumyarion
Thaler 1966. Les genres europeens d'heterosoricines, Quercy-
sore.x. Heterosore.x et Dinosore.x sont distingues morphologique-
ment des genres americains Domnina et Pseudotrimylus. Les
petits soricides d’Europe, attribues par plusieurs auteurs a la
sous-famille des limnoecines sont compares a des representants
du genre Limnoecus. Ces comparaisons ont demontre qu’il n’y
a actuellement pas de raison de placer Tune ou I’autre des formes
europeennes dans cette sous-famille.

Dans la partie generale, on tente de demontrer que I’etonnante
ressemblance des formes des deux continents peut etre expli-
quee outre la migration, par d’autres possibilites telles que
revolution parallele et la persistance des caracteres primitifs
pendant de tres longues periodes.

5

6

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

INTRODUCTION

This paper is an experiment. A great number of
questions are asked, but to only a few can an an-
swer be given.

During the last few years several species of Eu-
ropean mammals have frequently been united in the
same genus as North American species, almost al-
ways on the basis of similarities in tooth morphol-
ogy. The presence of such similar forms on both
continents was usually explained by intercontinen-
tal faunal exchange or, as most authors state, by
"migration."

In this study I tried to approach the problems
from a different point of view from that of most
authors having worked in this field. In many cases
I intentionally took the part of the advocatus dia-
holi. My main object with this was to re-open sev-
eral issues so that some problems will be discussed
again. Many problems of relationships between
Tertiary mammalian faunas from North America
and Europe are regarded as solved and are not dis-
cussed anymore. In my opinion it is premature to
give a solution to such complicated problems on the
basis of the scanty available comparative material.
For the solution we need more complete and abun-
dant specimens. In addition, in the course of this
study I began to doubt whether tooth morphology
alone can be trusted for the clarification of phylo-
genetic relations, especially for such higher system-
atic categories as genus, subfamily, or family.

In this paper I try to show that in addition to
faunal exchange other possibilities must still be con-
sidered, including parallelism and the persistence of
primitive forms on both continents over long pe-
riods. In most cases no material was available other
than that from which my colleagues had drawn their
conclusions, so I tried to apply other morphologic
criteria. I also examined earlier faunas of both con-
tinents to determine whether there are forms from
which the latter forms, showing similarities to those
of the other continent, could be derived (without
supposing a lateral communication). Thus, starting
with upper Miocene genera being reported to occur
on both continents, 1 also studied small mammals
from the Lower Miocene, the Oligocene, and, in
some cases, the Eocene.

For this study 1 selected ten genera that appeared
to me to be especially rewarding and with which I
had the possibility of comparing American and Eu-
ropean original specimens. I endeavored to com-
pare as many different groups as possible, because

conclusions can follow from a broad spectrum of
forms that cannot be drawn from one group. Be-
cause of the abundance of forms, and because the
magnitude of the subject required me to follow the
different groups over long periods — sometimes
through the whole Tertiary — I confined myself to
the rodents and insectivores.

Assuming that most readers will not read through
the whole paper, but rather will confine themselves
to chapters interesting to them, the different chap-
ters are self-contained. This involved, of course,
some repetition in the different chapters.

Because the time correlations of North American
and European mammal faunas still are very much
a matter of discussion, I abstained from giving a
correlation table and correlated only on the level of
epochs. The term "Miocene" is used in the mod-
ern sense, that is, the period between ca. 23 and 5
million years ago; likewise the Pliocene, which cov-
ers the period between 5 and 1.8 million years ago.
For North American stratigraphy, I used the cur-
rent land mammal ages such as Arikareean, Hem-
ingfordian, and so forth, but for the European Neo-
gene I largely refrained from using terms like
Helvetian, Tortonian, Sarmatian, and others. These
are mostly defined on the basis of marine forma-
tions and therefore in continental stratigraphy they
can mean very different things, according to the
country and school of the different authors. So for
the European Neogene I used the mammal units
NMU 1-16 (NMU = Neogene Mammal Unit), in-
troduced and worked out by Mein (1976). For col-
leagues not acquainted with this new mammal zo-
nation, I sometimes added the misleading stage
names in quotation marks.

Abbreviations of the institutions owning the specimens de-
scribed and figured in this paper are as follows: AMNH, Amer-
ican Museum of Natural History. New York; BSPM, Bayerische
Staatssammiung fiir Palaontologie und historische Geologic,
Miinchen; CM, Carnegie Museum of Natural History, Pitts-
burgh; FSL, Faculte des Sciences de Lyon; KU, University of
Kansas Museum of Natural History, Lawrence; MO, Museum
Olten (Solothurn, Switzerland); MHNP, Museum national
d'Histoire naturelle, Paris; NHW, Naturhistorisches Museum
Wien; NMBS, Naturhistorisches Museum Basel; NMNH, Na-
tional Museum of Natural History, Smithsonian Institution,
Washington, D.C.; SDSM, South Dakota School of Mines, Mu-
seum of Geology, Rapid City; UCMP, University of California
Museum of Paleontology. Berkeley; UNSM, University of Ne-
braska State Museum, Lincoln; UO. University of Oregon, Mu-
seum of Natural History, Eugene.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

7

ACKNOWLEDGMENTS

Without the generous help of the Carnegie Museum of Natural
History, which offered me a nine month stay as a Resident Mu-
seum Specialist. I would not have been able to accomplish this
study. Among the staff of this museum I am especially indebted
to Dr. Mary Dawson. Dr. M. G. Netting. Dr. J. Swauger. and
Mr. W. Woodside. In addition. Dr. Dawson's discussions and
criticism have been most helpful in the realization of this work;
her help was very much appreciated in correcting the English
text. 1 also wish to thank Dr. L. Krishtalka and Dr. J. H. Hutch-
ison for going over parts of the manuscript. 1 am also very grate-
ful to Dr. C. C. Black for having the paper published here.

Further, 1 am indebted to the Stipendienkommission des Kun-
tons Basel-Stadt for having provided an additional grant for my
stay in the United States and to the Janggen-Pohn-Foundation
(St. Gallen) which enabled me to study European comparative
materials in the collections of Fyon and Munich after my return
to Europe.

Dr. Johannes Hurzeler followed this study from its beginning
with great interest and was always at hand with valuable infor-
mation and suggestions. 1 also wish to thank Dr. F. Wieden-

mayer and Mr. J. B. Saunders for having helped me with the
English translation and Miss H. Pouget and Miss E. Hill for
having typed the manuscript. I am grateful to my wife Wies for
her help and her patience during the time of this work and es-
pecially for having made excellent casts of specimens during our
stay in America.

To the following 1 am indebted for the loan of comparative
materials, providing literature and casts, for discussions, infor-
mation, and criticism, as well as for other kinds of help: Prof.
Dr. R. Dehm, Munich; Dr. R. J. Emry, Washington, D.C.; Dr,
V. Fahibusch, Munich; Dr. O. Fejfar, Prague; Dr. E. Heizmann,
Stuttgart; Dr. M. Hugueney, Lyon; Dr. D. E. Savage, Dr. J. H.
Hutchison, and B. Waters. Berkeley; J. Yarmer. Pittsburgh; Dr.
L. Kittleman, Eugene; Dr. E. Lindsay, Tucson; Dr. L. D. Mar-
tin. Lawrence; H. McGinnis, Pittsburgh; Drs. M. C. McKenna,
J. H. Wahlert, and M. F. Skinner. New York; P. Mein. Lyon;
D. L. Rasmussen, Denver; Dr. C. Ray, Washington, D.C.; Dr.
P. Robinson, Boulder; Drs. R. Zangerl, W. D. Turnbull, and W.
Segall, Chicago; Prof. Dr. H. Tobien, Mainz; Dr. R. W. Wilson,
Lawrence; Prof. Dr. H. Zapfe, Wien.

RELATIONSHIPS OF INSECTIVORES

La nthanotherwm

(Earn. Erinaceidae, Bonaparte, 1838, Subfamily
Echinosoricinae Cabrera, 1925)

History of Investigations

The genus Lanthanotherium was introduced in
1888 by Filhol for Lartet’s Erinacens sansaniensis
from Sansan. Outside of L. sansaniense four more
species have now been described from Europe — L.
rohnstnm Viret, 1940, from La Grive; L. longi-
rostre Thenius, 1949, from Leoben; L. sanmigneli
Villalta and Crusafont, 1944, from Viladecaballs;
and L. piveteaui Crusafont, Villalta, and Truyols,
1955, from Can Cerda. A second species from San-
san, described in 1972 by Baudelot as L. tobieni.
is here considered as a synonym of L. sansan-
iense.' Butler (1969) reported Lanthanotherium
from the Miocene of Africa (Songhor, eastern Af-
rica). I know of no fossils of Lanthanotherium from
Asia, where all Recent representatives of the Echi-
nosoricinae live. Webb’s reference (1961:1085) to
a passage in Viret ( 1940:53) in which a questionable

' Baudelot (1972) introduced the species L. tohicni chiefly on the basis of a fourth
premolar in two lower jaw fragments {L. sansaniense usually has three lower pre-
molars). It is generally known that the number of premolars in Echinosoricinae varies
considerably within a population, often even in the two halves of one jaw. Van Valen
(1967:272). for example, described three skulls of the Recent species Hylornys suil-
lus — one specimen had three P on both sides; another, four P on both sides; and a
third specimen, three P on the right upper jaw and four on the left. According to Van
Valen. the same variability can be observed in the Recent Neotetracus sinensis and
I found the same variability also in lower jaws of Echinosoricinae in our collection.

Lanthanotherium specimen from China is said to
be mentioned, apparently is due to a mistake, be-
cause in this passage Viret discusses only the Re-
cent Neotetracus sinensis Trouessart, 1909, and the
Oligocene Tetracus from Ronzon (France). Since
1961 several fossil insect! vores from North America
were incorporated into the genus Lanthanotherium.
Webb (1961) assigned a jaw fragment with three
molars from the Bopesta Formation (Barstovian) of
California to this genus. Subsequently. James
(1963) described two new species of Lanthanothe-
rium— L. sawini and L. dehmi. Since then remains
of these animals were also reported from the Red
Basin in Oregon (Shotwell, 1968) and from the Bar-
stow Formation in California (Lindsay, 1972).

Among the numerous currently described species
of Lanthanotherium the dentition is known to some
extent from only three — L. sawini. of which even
a well preserved skull fragment is known; L. san-
saniense; and L. longirostre. All the other forms
are hitherto documented only by fragments of man-
dibles. Therefore, in Lanthanotherium, as in many
genera treated in this paper, all comparisons are
limited to the dentition.

As in most insectivore groups, in the Echinoso-
ricinae the molars behave very conservatively dur-
ing evolution, whereas the anterior part of the den-
tition undergoes various differentiations. This
becomes obvious on comparing the dentitions of

8

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

Neiirogymnuriis and Lanthanotheriiim, for exam-
ple. Therefore, at the generic level more systematic
value must be attached to differences in molars than
to those of the anterior dentition. The same obser-
vation can be made comparing the European
species of Lanthanotheniim — the anterior dentition
shows very important differences, the molar den-
tition hardly any.

Comparisons of American and European Forms

Because, as James (1963) pointed out, the best
documented American species, L. sawini, shows a
highly specialized anterior dentition, the following
comparisons are restricted to molar morphology.
Thus, I tried to distinguish the features in common
to forms on one continent and compared them with
the characteristics of the forms on the other conti-
nent. Determination of the typical features of the
American species of Lanthanotherium was greatly
aided by examination of excellent material from an
unpublished fauna (I was requested not to mention
this material, specifically, but as a check on my
conclusions it was very helpful). Viret (1940) and
Butler (1948) showed that the third upper molar in
the Erinaceidae is taxonomically significant. I found
M^/M3to differ between the species on the two con-
tinents— these differences I interpret as general
ones.

M^ of American Lanthanotherium is more re-
duced compared with those of European species.
The parastyle and anterior cingulum are less devel-
oped and, sometimes, lacking, a feature never ob-
served in European forms (see Plate 1). The ante-
rior crest of the protocone extends directly to the
paracone, rather than to the anterolabial corner of
the tooth (the parastyle) as in European species of
Lanthanotherium. On M® of the American species
the cusps are less distinct and fused to a ring-shaped
wall that is open only on the lingual side. The cusp
on the posterior corner of the tooth“, especially well
developed in L. sansaniense, (Plate la) is less well
differentiated in American forms. The posterior
crest of the protocone in European species extends
to the metacone, leaving a distinct lingual shelf. On
M^ of American specimens this posterior protocone
crest ends shortly beyond the protocone or extends
to the cusp on the posterior corner of the tooth.

^ Butler (1948) and James ( 1963) designate this cusp as a hypocone. This designation
seems somewhat daring to me. because this cusp is lacking in earlier Echinosoricinae.
for example, some Galericini or Neurogymnurus. and therefore, in my opinion, has
to be regarded as a new acquisition. For that reason it would hardly be homologous
with a real hypocone.

The shelf between the posterior branch of the pro-
tocone and the lingual tooth margin is lacking or
only weakly developed.

On M^ of American species of Lanthanotherium
the anterior cingulum is less well developed than
that of European forms. The paracone and meta-
cone are connected by a longitudinal crest, which
is weakly expressed on corresponding teeth from
Europe (Plate la-c). In addition, the valley in the
posterior half of the tooth between the metacone
and hypocone is larger in American forms.

M' of American forms is relatively short and its
metaconule is crescent-shaped in occlusal view,
whereas it is more nearly round in European
species. Otherwise M' of American and European
forms are the most similar of all the upper molars.

Even fewer differenees can be observed between
the lower molars of American and European
species of Lanthanotherium (Plate 2). This is not
too surprising since the lower molars of different
echinosoricines resemble each other very closely
(for example, those of Lanthanotherium, Neuro-
gymnurus, and Hylomys).

As already mentioned the species on each con-
tinent vary considerably as far as the anterior den-
tition of the lower and upper jaw is concerned.
Therefore the anterior dentition seems to me to be
very important for the distinction of different
species, but not for the definition of higher taxo-
nomic units. However, the robust anterior denti-
tions exhibited in L. san ini and especially L. dehmi
(see James, 1963) are not known in European
forms.

Einally, all described species of Lanthanotherium
from North America are distinctly smaller than
most European forms of the same age. The only
exception among European species is that from Can
Llobateres, which, although one of the latest forms,
has the smallest dimensions (see Plates Ic and 2d).^

Discussion

The comparisons of American and European
species of Lanthanotherium demonstrate that in
addition to the great similarity of the dentitions
there are also differences. Even if these differences
are not very apparent, their significance for system-
atics should, in my opinion, not be underestimated.

^ Whether the form from Can Llobateres is identifiable as Lanihunotherium san-
migueli (Vilialta and Crusafont. 1944) cannot be decided on the basis of the descrip-
tions and figures in the original publication. However, on the basis of size the form
from Can Llobateres corresponds to L. sanmigideli from Viladecaballs, which is of
about the same age.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

9

If, in addition, one takes into consideration how
little various genera of the Echinosoricinae often
differ dentally one can question the incorporation
of the Lanthanotlu'iium-Wke forms from North
America into the genus Lanthcinotherium. Com-
paring, for example, the dentition of American Lan-
thanotherium with those of certain Recent echino-
soricines such as Echinosorex, N e at et ranis, and
Hylomys (see Plates Id and 2e), one notices that
the correspondence with the Recent forms is the
same as with Lanthanotheriiim from Europe. With
the same justification the American Lanthanothe-
riiim could therefore be assigned to Hylomys or
Neotetracus. Eor this reason, it seems appropriate
to separate generically the American and European
species of Lanthanotheriiim. However, I refrain
from this action until better skull remains of the
European forms are known. New conclusions about
the phylogenetic relations can also be expected fol-
lowing study of the unpublished Lanthanotheriiim
material in American collections. Einally, the pos-
sibility exists that if more complete remains of
Ocajila from the Wounded Knee area (Macdonald,
1963) become known, generic identity of the latter
genus with the American forms hitherto assigned to
Lanthanotheriiim may result, so that all American
forms could be united into the genus Ocajila.

Even if the dental differences of American and
European species of Lanthanotheriiim seem to jus-
tify a generic separation, close relationship between
the forms from both continents cannot be contest-
ed, and with good reason are united with the Recent
"moon rats” in the tribe Echinosoricini (Butler,
1948; Van Valen, 1967). How closely these forms
are related cannot be judged for the moment, be-
cause of the general conservative nature of the den-
titions in echinosoricines. A pertinent question here
is when the Echinosoricinae (sensii Butler, 1948)
appeared in the New World. Either this group had
already reached America from Europe in the Upper
Eocene, or it originated there. An Ocajila-Vike eri-
naceid is known from the Uintan Tepee Trail Eor-
mation of Wyoming (I am indebted to Prof. Robert
Wilson for sending me a photocopy of a letter from
Dr. Peter Robinson, in which this form is men-
tioned). This genus, originally described from the
Sharps Formation (lower Arikareean), is generally
regarded as a member of the Echinosoricinae and
is therefore the earliest known representative of this
subfamily in the New World. Unfortunately, I have
not seen specimens of the Eocene Erinaceid, and
only part of the lower dentition of the Wounded

Knee species is known (the jaw fragment of Mac-
donald’s figure 4 [1970] is obviously not of Ocajila
but of a heterocoricine; see also Hutchison,
1972: 13). The lower molars of Ocajila and Lantha-
notheriiim are remarkably alike (see Plate 2c), ex-
cept that the latter have a somewhat longer trigonid.
Based on these, admittedly poor, remains there is
no feature in the dentition of Ocajila which would
exclude this genus from the ancestry of Lanthano-
theriiim. In Europe the earliest known echinosori-
cines are the Middle Oligocene genera Neiirogym-
niiriis and Tetraciis, either of which may have given
rise to Lanthanotheriiim.

The first appearance of Lanthanotheriiim in Eu-
rope is in the Lower Miocene of Vieux Collonges
(NMU 4), but in America it is not until the Upper
Miocene (Barstovian). On both continents Lantha-
notheriiim is recorded until the Upper Miocene
(Turolian in Europe, Clarendonian in North Amer-
ica).

If we attribute the dental resemblances between
European and North American species assigned to
Lanthanotheriiim to intercontinental migration, the
Bering Strait is the only possible migration route
after the end of the Eocene. There, as with other
groups (for example, Sciiiropteriis) treated in this
paper, a problem arises — ail Recent Echinosorici-
nae are dwellers of tropical or subtropical forests.
From this we can conclude, if conclusions based on
analogy with living forms are valid, that the fossil
forms ecologically did not behave much differently.
For dwellers of tropical forests the far north Bering
Strait region should not be underestimated as a cli-
matic barrier (see Simpson, 1947:65 1 : “there is no
evidence from the mammals that any truly tropical
or subtropical animal ever migrated between Eur-
asia and North America”). Other ecological bar-
riers likely also existed on this long route. In ad-
dition it has to be remembered that no Tertiary
Echinosoricinae have yet been found in Asia.

Really we are still groping in the dark with the
reconstruction of the history of Lanthanotheriiim.
The only statement on which we are on firm ground
is that of the morphological differences of the den-
tition, and from these the conclusion that there
should be generic separation for the American and
European forms seems to be justified.

Plesiosorex/Meterix
(Family Plesiosoricidae Winge, 1917)

The discussion on Plesiosorex is confined to
points bearing on the relationship between North

10

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

American and European forms. A monographic
treatment of Plesiosorex is in preparation. Also not
discussed is the systematic affiliation of this genus,
which is still not clear. Wilson (1960) thoroughly
compared Plesiosorex coloradensis with European
species, and therefore 1 confine myself to some few
interesting, comparative points.

Historical Introduction

De Blainville first described in his “Osteogra-
phie” (1838) a representative of this genus under
the name Erinaceiis soricinoides. The type was
from the Stampian locality of Chaufours (Limagne,
Erance). Pomel (1854) classified the species from
Chaufours with the myogales and introduced the
generic name Plesiosorex. Dietrich (1929) seems to
have been the first author to notice the similarity
between the European Plesiosorex and the Ameri-
can genus Meterix described by Hall ( 1929). Strom-
er (1940) incorporated the species Myogale ger-
manica (Seemann, 1938) in the genus Meterix Hall.
In the same year Viret (1940) discussed the phylo-
genetic relationships of Plesiosorex and Meterix,
and took the view that probably at the end of the
“ Vindobonian” or the beginning of the “Pontian"
representatives of the genus Plesiosorex “emigrat-
ed” to America. In his monograph on the Quarry
A fauna, Wilson (1960) described an insectivore as
Plesiosorex coloradensis, stating that the Quarry A
species showed more similarities with European
Plesiosorex than with the American Meterix.

Comparisons of American and European Eorms

Comparisons between European and American
plesiosoricids are confined to the dentition, because
no other remains of these animals are now known
from Europe. Species of Plesiosorex from North
America and Europe are surprisingly similar and
exhibit all the dental features typical of the genus —
the paracone and metacone of the upper molars are
shifted medially compared with those of other in-
sectivores. and the occurrence of cusps labial to the
paracone and metacone. Extensive comparisons of
European and North American material failed to
reveal fundamental differences in the dentition.
Nevertheless, some differences do occur on MVM,,
the most diagnostic teeth of Plesiosorex apart from

U.

M' of all American Plesiosorex known to me
shows a relatively strong and well-defined hypo-
cone, whereas this cusp is distinctly lower and less
clearly separated from the protocone on M’ of Eu-

ropean species (see Eig. Ic, f, i, and 1), such as P.
schaffneri where the hypocone is essentially a shelf
on the posterior side of the protocone (Eig. Ic and
Plate 5c).

On M‘ of American species of Plesiosorex the
cusp labial to the paracone is higher than the latter,
whereas in European forms the paracone is always
higher (Eig. lb, e, h, and k). In the North American
species the cusp labial to the metacone on M‘ is
more developed than in European ones, where it
occurs as a crest rather than as a distinct cusp (Eig.
la, d, g, and j).

Species of Plesiosorex from both continents do
not differ fundamentally in remaining known parts
of the dentition. However, the anterior parts of the
dentition are too poorly known for extensive com-
parisons.

The mandible in American and European Plesio-
soricidae are very similar too (see Plate 3).

It is interesting that both the primitive and ad-
vanced species of Plesiosorex from North America
and Europe are very similar to each other. P. co-
loradensis, the earliest and morphologically the
most primitive North American form, surprisingly
resembles certain primitive species from Europe,
for example that from the Rumikon. Both have a
relatively short M' and a posterior trigon leg (Tri-
gonumkante) that runs via the metaconule to the
metastyle (Plate 5a and d). More evolved forms
from both continents, such as Upper Miocene and
Pliocene species of Meterix and Plesiosorex sp.
from Grosslappen have a more elongated M', with
a posterior trigon leg that ends at the posterior mar-
gin of the crown without reaching the metastyle (see
Plate 5b and e). Similar observations can be made
in the lower dentition — more primitive species on
both sides of the Atlantic have an M, with a short
trigonid compared with the talonid (see Plate 4b and
d). In more evolved forms an elongation of the tri-
gonid angle occurs (Plate 4a and c). Also, the num-
ber of antemolar teeth apparently becomes more
reduced in evolved forms of North America and
Europe.

Einally, compared with the more primitive
species, the mandible of advanced species from
both continents bears a posteriorly shifted mental
foramen — the anterior foramen of P. coloradensis
is below P2; that of Meterix is below P3 (plate 3a-
b). Similarly, among European species of Plesio-
sorex the earlier forms, such as that from St- Andre
(Viret, 1946) and Chaveroche (Plate 3d) bear the
anterior foramen below the space between P, and

u

Fig. I. — Different views of LM' of North American and European Plesiosoricidae. Upper row: lingual view; middle row: front view; lower row; labial
view, a-c) Plesiosorex schaffneh Engesser, a and c; NMBS. Al. 145 (invers), b; NMBS, Al. 538, Anwil (Switzerland), Middle Miocene; d-f) Plesiosorex
sp., BSPM 1926 E 81 (invers), Grosslappen (Germany), Middle Miocene; g-i) Plesiosorex coloradensis Wilson, KU 9990 (invers). Quarry A, Martin
Canyon, Colorado, Hemingfordian; j-1) Melerix sp., a and c; OU 22329 (invers), b: OU 22326 (invers), Ove, Quartz Basin, Oregon, Barstovian. All
figures I2x.

12

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

P2, but in more advanced forms, such as that from
Aumeister, it occurs below the space between P3
and P4. Also, the posterior mental foramen of Eu-
ropean forms seems to have moved posteriorly dur-
ing the course of evolution — in the species from St-
Andre and Chaveroche it occurs below P4 (or below
the interspace between P4 and M,), whereas, in
more evolved forms (for example, from Viehhausen
[Seeman, 1938], Riimikon, and Aumeister) it is be-
low the protoconid of M,. This posterior migration
of the posterior foramen in European forms hap-
pened apparently between the Upper Oligocene and
the Lower Miocene. In America this trend seems
to have been terminated in the Lower Miocene, be-
cause, like the geologically younger Meterix, P. co-
loradensis has this foramen below M, (Plate 3a-b).

As is probable for soricids, the movement of the
mental foramen in plesiosoricids can be correlated
to an enlargement of the large lower incisor (U),
which unfortunately is known in only a few species.

The intercontinental similarity among primitive
and evolved plesiosoricids cannot, in my opinion,
be explained only by faunal exchange, because
these would have to have involved frequent “inter-
migrations.” An analogous case involves Democri-
cetodon and Copemys, where the advanced species
(on either continent) are more similar to each other
than to the primitive ones. It is very difficult to find
an explanation for these phenomena. In my opinion
it seems worth considering whether evolutionary
trends could be involved, which, proceeding from
a common origin, can be conserved over long pe-
riods.

Of all described species of Plesiosorex, P. color-
adensis from Quarry A is by far the best known.
Wilson ( 1960) concluded that compared with P. so-
ricinoides from St-Andre and P. cf. soricinoides
from Chaveroche, P. coloradensis shows more pro-
gressive features, but compared with P. styriaciis
(Steiermark) and P. germanicus (Viehhausen) was
more primitive.

Unpublished Plesiosorex material from Europe
are similar to P. coloradensis and, considering its
age, imply the progressive nature of the latter. The
strongly reduced labial cingulum of the lower mo-
lars is an especially progressive feature in the den-
tition of P. coloradensis. P. coloradensis most
closely resembles the Riimikon Plesiosorex (canton
of Zurich, Switzerland, see Plate 5a). On the basis

* Thenius (1949) recommended synonymy of P. germanicus and P. styriacus. in
which he was followed by Wilson (1960). Later I (1972:59) advocated for maintenance
of both species.

of the cricetids, the Riimikon fauna (NMU 5) has
to be regarded as somewhat older than that from
Sansan (NMU 6), but according to Wilson’s strati-
graphic placement of Quarry A, P. coloradensis
would be distinctly older than the Riimikon species.

How can the great similarities between North
American and European Plesiosorex be explained?
As mentioned above, Viret (1940) suggested “mi-
gration” from Europe to North America immedi-
ately before the “Pontian.” Wilson (1960) also pre-
ferred the possibility of a “migration,” but he fixed
the time for it somewhat earlier — Lower Miocene —
because he regarded P. coloradensis as a possible
ancestor of Meterix.^

At present, P. coloradensis is the oldest known
plesiosoricid in North America; no suitable ances-
tors are known from the Oligocene or Eocene of
North America. Nevertheless, such Paleogene
forms as Geolahis (=Metacodon) show similarities
with Plesiosorex. As stated by Wilson (1960) most
of these similarities have to be considered as com-
mon primitive features. In Europe Plesiosorex is
known from earlier levels, that is, from the Stam-
pian. This and the fact the American and European
forms are very similar, at least in those parts that
we can compare, implies a faunal exchange.

On the other hand, Plesiosorex is dentally prim-
itive, having retained these features over very long
periods. That the evolved, latest forms on both con-
tinents are surprisingly similar implies the possibil-
ity of parallelism. Thus, it is possible that Plesio-
sorex reached North America before the time of its
earliest known occurrence, and then developed for
a long time in parallel to its European relatives.

Subfamily Heterosoricinae Viret and
Zapfe, 1951

(Eamily Soricidae Gray, 1821)

History of Investigations

Since my study of the Miocene mammal fauna
from Anwil (1972:73) reported on the history of the
European Heterosoricinae, this discussion is limit-
ed to the opinions of different authors regarding the
relationships between North American and Euro-
pean forms. Since Cope’s (1873) description of the

' If a close relationship of American and European plesiosoricids will be confirmed
in the future, the generic name Meterix probably will be declared a synonym of
Plesiosorex, because for the present it is difficult to differentiate the two genera
morphologically. We hope that we soon will know more about these relationships
because some North American collections contain extensive unpublished material,
possibly permitting determination of the phylogenetic lineages from P. coloradensis
to the latest forms of the Upper Pliocene.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

13

genus Domnina, heterocoricines have been re-
covered in North America, but, for a long time,
were not considered to have a close relationship
with European forms. Patterson and McGrew
(1937) were the first to point out the similarity of
Domnina to the European Heterosorex, but they
did not alter their systematics. Mawby ( 1960) noted
the great resemblance of Domnina" compressa
(Galbreath, 1953) to Heterosorex'^ Gaillard and as-
signed “D.” compressa and an undescribed species
from Oregon, =Pseudotrimylns mawby i (Repen-
ning) 1967, to Heterosorex. Wilson (1960) referred
two species from Quarry A to Heterosorex — H. ro-
peri and Heterosorex sp. — and compared them ex-
haustively with European species. Wilson claimed
that "Heterosorex" roperi could not have devel-
oped independently from European species since
the Oligocene. Since 1960, the names Trimylus and
Heterosorex have also been used for American
forms (Hutchison, 1966, 1972; Doben-Florin, 1966;
Repenning, 1967). In his revision of the family So-
ricidae. Repenning (1967) was of the opinion that
the European and North American Heterosoricinae
likely represent two separate evolutionary lines, but
he still included species from Europe and North
America in the genus Trimylus. Gureev (1971) in-
troduced a new genus, Pseudotrimylus , for Heter-
osorex roperi from Quarry A. Crochet (1968) had
referred three heterosoricines from the European
Oligocene to Domina — "Amphisorex" primaevus
Eilhol, "Sorex" herrlingensis Palmowski and
Wachendorf, and the heterosoricine from Ricken-
bach (Switzerland). In 1975, I introduced the genus
Quercysorex for the former two, and placed the lat-
ter in the genus Dinosorex.

The known diversity of the subfamily Heteroso-
ricinae has greatly increased in the last 15 years.
Although new North American and European
species have been described, their phylogenetic af-
finities still are not clearly understood. Several au-
thors oversimplified this problem by considering all
later forms to have descended from earlier ones.
Their nice phylogenetic trees hardly show real phy-
logenetic relationships. Typical examples are the
cases of Saturninia and Dinosorex huerzeleri
{ = Heterosorex aff. neumayrianus) from Ricken-
bach, which, being the oldest known forms, are
used and abused as ancestors for all sorts of groups.
Among the present abundance of known heterosor-

^ As far as the problematic nature of the names Heterosorex and Trimylus Is con-
cerned, see Engesser (1975).

icines there are hardly two species of different age
that can be derived with certainty from one other.
The conclusions obtained by observing the trends
of one feature always seem to be opposed by those
from other features. Their systematics are also
complicated by the fact that many of the taxa are
incompletely known, for example from one frag-
mentary mandible — and by and large differ very lit-
tle. The latter can probably be explained in the fol-
lowing manner: the Heterosoricinae (analogous to
the rodents) specialized very early — probably al-
ready in the Eocene — and apparently very success-
fully too; this specialization considerably limited
their spectrum of evolutionary possibilities, where-
as in return, as in rodents (Wood, 1947), increased
their chances for parallel evolution. In the Miocene
of Europe and North America there are a great
number of different evolutionary lines, often very
hard to distinguish. The diversity of taxa in the
Oligocene of both continents also argues in favor of
the existence of different lineages. Therefore, the
phylogeny of this group is at present unclear.

The relationships between North American and
European heterosoricines are also unresolved and
these taxa are compared below.

Comparisons of Oligocene
Heterosoricines of America and Europe

The Oligocene soricid faunas of North America
are dominated by the genus Domnina Cope, 1873,
of which five species have been described. After
the Middle Oligocene (Orellan) other forms, dis-
tinctly different from Domnina, occur. These are
here provisionally designated as Pseudotrimylus
Gureev, 1971 (see below for detailed discussion of
the nomenclature). In the European Oligocene two
genera of heterosoricines are distinguishable —
Quercysorex Engesser, 1975, and Dinosorex En-
gesser, 1972. These will be discussed later.

Differences between Domnina and the Oligocene
Pseudotrimylus.^ — The Oligocene Pseudotrimylus
(the only known specimens are those of P. com-
pressus from Cedar Creek and Pine Ridge [see Re-
penning, 1967:Fig. 5| and that of Pseudotrimylus
sp. from Lawson Ranch; see Fig. 2b and Plate 7d)
differs from Domnina in its stouter mandible; its
lowest point is immediately behind M3 and its height
increases distinctly anteriorly. The mandibular
height of Domnina is about the same from front to
back, except that below Mi a gentle inlet can be

’ For use of the name Pseudotrimylus see p. 15.

14

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

a

b

c

Fig. 2, — a) Dinosorex huerzeleri (Engesser), L mandible with I, M.-Mj, (type) MO, Rb. 101, Rickenbach (Canton Solothurn, Swit-
zerland), Late Oligocene; b) Pseudothmylus sp., L mandible with M,-M2, CM 10922, Lawson Ranch, Goshen Co., Wyoming, Lower
Orellan?: c) Domnina sp., L mandible with Mj-Ma, CM 10403, Pipestone Springs, Jefferson Co., Montana, Chadronian. All figures
8x.

observed (Fig. 2c). The last antemolar of Domnina
is largest, whereas in Pseudotriniylus the first is
largest. The M1-M2 of Domnina are of equal size,
and have slender, pointed cusps with the protoco-
nid highest (Plate 7e). \n Pseudotrimylus the lower
molars decrease in size posteriorly and are more
squat than in Domnina, with the protoconid only
slightly higher than the other cusps. The hypocristid
in all species of Domnina extends behind the en-
toconid (Modus B, Engesser, 1972:76), whereas in
Pseudotrimylus it mostly joins the entoconid (Mo-
dus A). Finally, the teeth of Domnina are intensely
pigmented, whereas those of Pseudotrimylus are
either not at all (for example, Pseudotrimylus from
Lawson Ranch) or only slightly pigmented (P. com-
pressus). (For more differences, see Repenning,
1967:10.)

Domnina and Quercysorex. — Crochet (1968),
noting the great similarity between Domnina on the
one hand and '"Amphisorex" primaevus Filhol
from the Quercy and '^Sorex" herrlingensis from
Herrlingen on the other, placed the two European

forms and the Rickenbach soricid in the American
genus Domnina. However, later comparisons (En-
gesser, 1975) of the three European taxa with dif-
ferent species of Domnina illustrated that they dif-
fered in some important features, and that their
similarities were due to common retention of prim-
itive features. Consequently, the genus Quercyso-
rex was erected (Engesser, 1975) for the shrews
from the Quercy and Herrlingen. The Rickenbach
form (Plate 7c and Fig. 2a), which in my opinion
does not resemble Domnina, was referred to a new
species — D. huerzeleri — in the genus Dinosorex.

Pseudotrimylus compressus and Dinosorex huer-
zeleri.— This Oligocene species of Dinosorex
shows more similarities with the Oligocene Pseu-
dotrimylus of North America (for example, P. com-
pressus)— the massive horizontal ramus of the man-
dible and the inflated, low crowned lower molars
(see Fig. 2a and Plate 7c). Nevertheless, there are
also important differences — Dinosorex is much
more primitive with five antemolars and a hypo-
cristid extending behind the entoconid (Modus B);

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

15

Table 1. — Comparisons of American and European Heterosoricinae (except for such primitive forms as Domnina and Quercysorexj.

Characters

Typical for North American forms

Typical for European forms

Position of the mental
foramen

below M,; exception: P. roperi

below M2 or below the space between M,
and M2

Pigmentation

tooth apices unpigmented;
exceptions: P. compressus
P. mawhyi

distinct pigmentation of tooth apices

Mesostyle of the upper
molars

undivided; exceptions: Heterosorex sp.
(Quarry A), Paradomnina

divided

Hypocone of M‘ and

very well developed

much weaker

Posterior labial cusp of P'*

very well developed (P. mawbyil)

very weak or lacking

Course of the posterior
crest of the hypoconid
of M, and M2
(hypocristid)

directly to the entoconid (Modus A);
exceptions: P. dakotensis, possibly
Paradomnina

behing the entoconid, often separated from
this cusp by a small valley (Modus B);
exceptions: Dinosorex zapfei from
Neudorf and Vermes

Proportions of M, and M2

M2 comparatively small; exception:
Paradomnina

M2 comparatively large; exception:
Dinosorex. Pachygnathus

P. compressus is derived in that only three ante-
molars occur and the hypocristid joins the entoco-
nid directly (Modus A). In addition, the mental fo-
ramen in Dinosorex is situated below the trigonid
of M2, whereas in P. compressus it is below Mj.

The Miocene Heterosoricinae

In a revision of the European Heterosoricinae
(Engesser, 1975), I proposed suppression of the
poorly defined genus Trimylus Roger, 1885. Het-
erosorex Gaillard, 1915, on the other hand, is very
well defined. However, the type species of the lat-
ter— H. delphinensis from La Grive — shows many
morphological peculiarities (for example, the hardly
divided fossa masseterica on the mandible), shared
only by H. neumayrianus among known heterosor-
icines. Therefore, the genus Heterosorex should be
limited to the species H. delphinensis and H. neu-
mayrianus (for the new diagnosis of Heterosorex,
see Engesser, 1975). Ail other European species
hitherto included in Heterosorex or Trimylus were
referred to the genus Dinosorex, for which I gave
an emended diagnosis.

North American Miocene Heterosoricinae seem
to be more diverse than those from the Miocene of
Europe. Such extremely specialized North Ameri-
can forms as Ingentisorex and Pseudotrimyius
mawbyi, and such primitive ones as Paradomnina
have no corresponding forms in Europe. As noted
by different authors, there is no difficulty in mor-
phologically distinguishing most of the latter from
European taxa. However, among the New World
shrews one is out of place — P. roperi from Quarry

A shows similarities in some features to the Euro-
pean heterosoricines, especially Dinosorex zapfei
from Neudorf (see Engesser, 1975).

Apart from Ingentisorex and Paradomnina all
other North American Miocene species without ex-
ception were classified as Trimylus or Heterosorex.
Because Trimylus is invalid and because Hetero-
sorex has been redefined to include only the two
European species H. delphinensis and H. neumay-
rianus, to which genus do these North American
Miocene shrews belong? Gureev (1971) introduced
Pseudotrimyius for "'Heterosorex" roperi from
Quarry A. However, as already mentioned, the
Quarry A form is out of place among American sor-
icids. Also, the taxa hitherto placed in the genera
Heterosorex and Trimylus are very heterogeneous.
Nevertheless, I abstain from introducing a new ge-
nus at this time and provisionally use the generic
name Pseudotrimyius for all North American forms
except Domnina, Paradomnia, and Ingentisorex.

Comparison of Miocene Heterosoricinae of
Europe and North America

Considering the diversity of heterosoricines on
both continents, it is remarkable that no general
feature distinguishes among forms from America
and Europe.

Position of the mental foramen. — With one ex-
ception (P. roperi), the mental foramen in all known
North American Heterocoricinae is situated below
Mj. In all European taxa, except for Quercysorex,
it occurs more posteriorly, either below M2 or be-
low the interspace between M, and M2 (Table 1 and

16

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

Plate 8). The position of this foramen is one of the
most important systematic features because, as dif-
ferent investigations showed, it is remarkably stable
within populations and species. According to Steh-
lin ( 1940), a more posteriorly placed foramen in sor-
icids is more evolved than a more anteriorly placed
one, because it indicates the presence of a large
incisor. This observation, certainly correct in vaster
connections, does not apply to the subfamily
Heterosoricinae. For example, the mental foramen
in species with relatively small lower incisors, such
as those from Rickenbach, Wintershof-West, and
Vieux Collonges, is not placed more anteriorly than
in those showing a large incisor (species from An-
wil, Sansan, Neudorf).

Pigmentation. — The use of pigmentation of fossil
soricid teeth as a diagnostic feature is still contro-
versial. In certain cases it cannot be stated with
certainty whether fossil teeth showing no pigmen-
tation were formerly pigmented or not. Teeth of
known European Heterosoricinae, except Quercy-
sorex, are pigmented. American heterosoricines, on
the other hand, show no pigmentation except for
Domnina which has heavily pigmented cusp apices,
and P. mawbyi and P. compressus, which possess
slightly pigmented teeth (see Table 1).

Mesostyle of the upper molars. — In all European
species known to me, the mesostyle of M‘, and in
a smaller measure also of M“, is more or less divid-
ed. In American heterosoricines such a division
generally does not occur (see Plate 6). In Wilson’s
(1960) "^Heterosorex sp.,” the mesostyle is more
nearly divided than in any European shrew. This
shrew still is so aberrant in this feature and in oth-
ers— its upper molars also show a well-developed
mesocone — that it is of no account to our investi-
gations. According to Hutchison (1966) the meso-
style on the upper molars of Paradomnina is also
slightly divided.

Direction of the posterior crest of the hypoconid
(hypocristid) relative to the entoconid. — In Europe
and North America both orientations of the hypo-
cristid branch occur (Modus A and B, Engesser,
1975). Among European heterosoricines (Quercy-
sorex, most species of Dinosorex, Heterosorex) the
indirect course (Modus B) is far more common,
whereas in American species the indirect course is
more common, excluding Domnina, Paradomnina,
and Ingentisorex from these comparisons (see Ta-
ble 1). As Repenning (1967) noted, both modi rep-
resent different stages in an evolutionary trend,
with Modus B the more primitive. Modus A oc-

curred in America much earlier (Middle Oligocene
[Orellan, in P. compressus]), than in Europe (Mid-
dle Miocene [NMU 6 with Dinosorex zapfei from
Neudorf a.d. March]). With the exception of the
aberrant Ingentisorex and Paradomnina, I do not
know any Miocene heterosoricine from North
America with Modus B. In Europe, on the other
hand, heterosoricines with Modus B occur through-
out the Miocene.

Proportions of Mi compared with those of M.^. —
Another distinctive feature appears to be the pro-
portions of the M, and M2 — in American het-
erosoricines except Paradomnina and Domnina,
M.2 is smaller than M,; in European taxa there is
generally less difference in size between M, and M2
(Plate 7). This distinction is obvious in very early
forms on both continents (for example, between P.
compressus and Dinosorex huerzeleri from Rick-
enbach [Plate 7c and dj).

Comparisons of Pseudotrimylus roperi
with European Forms

(see also Wilson, 1960; Doben-Florin, 1964)

As mentioned above, of all known North Amer-
ican heterosoricines P. roperi from Quarry A is the
most similar to European forms. This fact led sev-
eral authors (Doben-Florin, 1964:70; Thenius,
1969:135) to consider P. roperi either as an immi-
grant from Europe or the ancestor of many Miocene
European heterosoricines (for example, D. sansan-
iensis). P. roperi cannot be the ancestor of D. san-
saniensis despite many features in common. On Mj
and M2 of P. roperi, the hypocristid runs directly
to the entoconid (Modus A, Plate 7a), a more
evolved condition than in D. sansaniensis , which
still shows an indirect course of the hypocristid on
M, and M2 (Modus B, Plate 7b). Thus, in all prob-
ability, the two species belong to two different evo-
lutionary lines.

As far as P. roperi being a European element
within the American Heterosoricinae is concerned,
we have little evidence. P. roperi has a number of
features in common with American forms, in which
it differs from European ones — (1) the tooth apices
of P. roperi do not seem to be pigmented*, whereas,
except for Quercysorex, all European heterosori-
cines known to me have pigmented teeth; (2) the
mesostyle on the upper molars of P. roperi is un-

“ Wilson's assumption that the teeth of P. roperi were originally pigmented ( 1960:28)
is based, as of this writing, on analogy to the pigmentation of D. sansaniensis.

17

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

Fig, 3. — Left mandibular condyle of a) Domnina thompsoni Simpson (after McDowell 1958); b) Quercysorex primaevus (Filhol); c)
Dinosorex sansuniensis (Lartet); d) Dinosorex zapfei Engesser; e) Pseudotrimylus roperi (Wilson). All figures ca. lOx.

divided^, whereas mesostyle division occurs in all
European heterosoricines for which the upper den-
tition is known (see Plate 6); (3) the hypocone of
M' and is much more developed in P. roperi
than in any European form (Plate 6b); (4) the pos-
terolingual cusp on P^ is strong in the American
species, but much weaker or lacking in European
species. With the stocky shape of the teeth and M,
much larger than M2 compared to European forms,
P. roperi resembles American Pseudotrimylus. In
addition, there seems to be no morphological im-
pediment to deriving P. roperi from certain Oligo-
cene North American heterosoricines, such as P.
compressus', similarly, in my opinion, the Miocene
European heterosoricines can be easily derived
from Oligocene representatives of the group, such
as Dinosorex huerzerleri of Rickenbach (Stampian).
The authors not considering P. roperi as a Euro-

^ It is not certain that the mesostyle on the upper molars of P. roperi was not
divided, because as far as I know there is no known unworn upper molar of this
animal. However, in European heterosoricine upper molars the division of the me-
soslyle can often also be observed in relatively worn teeth, especially in labial view.
So. the conclusion of an undivided mesostyle in P. roperi seems justified.

pean immigrant were probably influenced in their
opinion by the position of the mental foramen.

However, the posterior migration of the mental
foramen can occur on forms of different continents
as a parallel development.

Pandlel Developments and Analogous
Trends in North American and
European Heterosoricinae

Some new morphological acquisitions observed
on both sides of the Atlantic cannot be explained
by a faunal exchange between North America and
Europe after the Eocene, unless one assumes con-
tinuous traffic in both directions across the Bering
Strait. In my opinion, these "migrations” must be
regarded as exceptions (see Discussion). An ex-
ample of these, probably parallel, developments, is
the gradual division of the mandibular condyle,
which occurred on both continents. The American
Heterosoricinae apparently developed this division
earlier than the European shrews. Pseudotrimylus
dakotensis from the middle Arikareean already
shows a distinctly divided condyle (Repenning,

18

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

X

o

c

(U

O)

c

ro

c

c

E-

o

XI

E

Cl

CO

S B)

^ ro

tact

i|‘

O TO

QO

^■c

• p CD"

§o)

Q

S

•ro"

X

E

O)

CO

c

■ C ■

E

o

Q

c

o

•CO

X

E

o

o

Q

o ^

0

X

2

CO

>s

o

X

D

0

w

X

cn

CD

D

>

>^>^

0)

O

s- $

(D

X c:

E

0

c

0

0)

CO

cn

X

>

o

o

CO

C

>

0<

:lr

O D •
0Q

W

B

a;

o ^

"O -C

- U

C3 P

.£ o

o e

^ u

P ,o

E

<

o

X

<v

E

o

OD

e

W

^ w

c o>

‘E ^

GO c/3
cd g

o y

iH

uu

yjj ^ —

o 2 i-*

^ B- £

a, o .

z

<

z

<

>

Q

X

O

X

z

<

X

ijj

z

z

z

<

z

o

Q

O

X

X

<

X

1 —
CO
X
<
CD

z

LJJ

I

<

X

X

<

z

h-

I

X

X

X

X

o

Q

<

I

O

GO

<50

3

dj

(U

X, ?

a o ~a

H c u

I ^

.SP P

E

(u

D,
P 0>
Q. u.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

19

1967:Fig. 6), whereas that of the "Burdigalian”
(NMU 3) Heterosorex neunmynanus from Winter-
shof-West, besides Quercysorex the first European
form, of which the condyle is known, is still undi-
vided. The Middle Miocene (NMU 6) Dinosorex
from Neudorf shows a more distinct separation of
the articular faces, which are still coherent (Fig. 3).
Probably in European forms the division of the con-
dyle was independently acquired in two different
lines since such an early species as Quercysorex
primaevus shows a more advanced stage of con-
dylar division than the more recent Heterosorex
neumayriamis. The modern Soricinae developed a
divided mandibular condyle completely indepen-
dently from these phenomena in Heterosoricinae.

The increase in fusion of the hypocristid with the
entoconid in lower molars probably also developed
independently on both continents. This process
probably occurred twice, in North America and in
Europe, in two different evolutionary lines — in
North America, in the Orellan or earlier Pseudotri-
myliis compressus and in the Late Miocene Para-
domnina (see Fig. 4); in Europe Dinosorex zupfei
from Neudorf (Middle Miocene NMU 6) is the first
heterosoricine with a Modus A connection. A re-
newed tendency towards developing a Modus A
connection is displayed by the Late Miocene Di-
nosorex from Can Llobateres (NMU 9). Another
trend, which can be observed on both sides of the
Atlantic, is the increase in massiveness of the man-
dible. This bone is very slender in primitive forms
such as Quercysorex and Domnina, increases grad-
ually in massiveness and reaches extraordinary ro-
bustness in the geologically youngest forms such as
Pseudotrimylus mawbyi and Dinosorex pachygna-
thus. This increase in massiveness of the mandible
is accompanied by the increase in size of the lower
incisor, which reaches its largest dimensions in the
two latter forms.

Limnoecus

(Family Soricidae Gray, 1821)

Several authors (R. W. Wilson, James, Doben-
Florin, Baudelot) considered some small soricids
from European localities to represent the American
subfamily Limnoecinae (Repenning, 1967). This as-
signment is very doubtful because of the poor doc-
umentation and paucity of material. For example,
among American forms, the upper dentition is not
known.

The European forms repeatedly assigned to the
Limnoecinae are ^"Limnoecus" micromorphus

Doben-Florin from Wintershof-West, "Limnoe-
cus" grivensis Deperet from La Grive, "Limnoe-
cus" dehmi Viret and Zapfe from Vieux Collonges,
and Paenelimnoecus crouzeli from Sansan.

Doben-Florin (1964) assigned the species from
Wintershof-West to Limnoecus mostly because of
the slightly developed entoconid on the lower mo-
lars and the unicuspid talonid of M3. Chiefly be-
cause of the structure of the last lower antemolar
(P4?; "antemolars” are teeth between incisor and
M,, "Zwischenzahne” of Doben-Florin), Repen-
ning ( 1967:23-24) opposed this assignment and
united this shrew with the Crocidurinae? incertae
sedis. He also considered the distance between the
protoconid and metaconid in the species from Win-
tershof-West to be too large for a limnoecine.

Sorex grivensis and "Sorex" dehmi (Viret and
Zapfe, 1951) were also incorporated in Limnoecus
(James, 1963). According to James all small Ter-
tiary soricids from Europe (possibly including So-
rex antiquus Pomel) should be referred to Limnoe-
cus, based on the similarities of the mandible
(undivided condyle, large, triangular pterygoid fos-
sa) and tooth structure (reduced talonid of M3, mo-
lar protoconid and metaconid close to one another,
and others). Limnoecus is thus very broadly inter-
preted and unites almost all species that cannot be
assigned to Crocidura or Sorex. Repenning ( 1967)
defined the genus Limnoecus more strictly, placed
Mio sorex grivensis and "Sorex" dehmi in the
subfamily Crocidurinae, and included only two
species in Limnoecus — L. tricuspis Stirton and L.
niobrarensis. Sorex vireti was referred to a new ge-
nus, Angustidens , which, together with Limnoecus,
was placed in the Limnoecinae. The determining
factor for Repenning’s classification was the struc-
ture of the last lower antemolar (P4?) and M3, the
pigmentation and the position of the mental fora-
men.

1 concur with Repenning’s reasoning just to in-
clude the mentioned European forms in Limnoecus
or in the Limnoecinae. Limnoecus, interpreted as
broadly as James did. becomes a “wastebasket-ge-
nus,” in which all forms not identified as either Cro-
cidura or Sorex are assembled. The arguments by
which Repenning excludes Miosorex grivensis,
"Sorex" dehmi, and "Limnoecus" micromorphus
from the subfamily Limnoecinae seem convincing
to me. Although Baudelot ( 1972; 100) adopted Re-
penning’s definition of the subfamily Limnoecin-
ae— a strong entoconid on the lower molars — at the
same time (and on the same page!) she character-

20

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

3 be

Fig. 5. — Right mandibular condyle of a) Paenelimnoecus cron-
zell Baudelot, NMBS, Ss. 1002, Sansan. France. Middle Mio-
cene, 12. 5x; h) Limnoecus niohnirensis Macdonald, (after Re-
penning 1967) Nebraska, Barstovian, 5x; c) Angustidens vireti
(Wilson), (after Repenning 1967) Quarry A. Hemingfordian, 5x.

ized the new genus Paenelimnoecus by the loss of
the entoconid. Besides the loss of the entoconid,
the species from Sansan shows other differences
from American Limnoecinae, the most important
one being in the structure of the mandibular con-
dyle— the condyle of Paenelimnoecus (see Eig. 5a)
does not show any "labial emargination" as is typ-
ical in the Limnoecinae (Repenning, 1967:4), but a
"lingual emargination,” generally characteristic for
the Soricinae. In addition the horizontal part of the
condyle is distinctly wider in Paenelimnoecus com-
pared to that of the Limnoecinae. This condylar
structure implies the soricine affinities of Paenelim-
noecus. This tiny shrew from Sansan also shows
other differences from American soricids — the cor-
onoid process is distinctly higher; the mental fora-

Fig. 6. — Paenelimnoecus crouzeli Baudelot, LMi-Mg, NMBS,
Ss. 6706, Sansan (France), Middle Miocene, 25x.

men is located somewhat more posteriorly; the low-
er molars and the last antemolar (P4?) have a lingual
cingulum, lacking in American soricids (see Eig. 6);
the angle of the trigonid of M, is more open than in
Limnoecinae. Finally, there is the problem of pig-
mentation— of the seven jaw fragments of Paene-
limnoecus crouzeli in the Basel collection, none
shows pigmentation. Because the enamel in this
species is so delicate, the cavity at the base of the
teeth appears through the enamel, with the effect
that the tops of the teeth appear differently colored
than the base. Because of this I stated (1972:66) that
the teeth of soricid A from Sansan (=Paenelimnoe-
cus crouzeli) might have originally been pigmented.
I have not seen Baudelot’s specimen, but it seems
strange that only this specimen should show pig-
mentation, whereas seven others from the same lo-
cality do not. If Paenelimnoecus actually does not
show any pigmentation of the teeth, this would be
one more reason for separating it from the Limnoe-
cinae.

It seems that for the present there is no reason
to assign any of the European soricids treated here
with the Limnoecinae.

RELATIONSHIPS OF RODENTS

"SCIUROPTERUS”

(Family Sciuridae Gray, 1821)

History of Investigations

Fossil flying squirrels have been known for a long
time. In the last century a number of European
forms were described as different species of Sciu-
rus, but it was not until 1893 when Major connected
these finds with flying squirrels. From that time all
sciurid teeth showing secondary ridges and crenu-
lated enamel were united in the genus Sciuropterus.
The numerous Recent flying squirrels were split
into different genera much earlier, although they
resemble each other to some extent more than

many fossil forms. This difference in treatment was
perhaps motivated by the fact that this classification
of Recent forms was based not only on the dentition
but also on other morphological features (skeleton,
color of the fur, and others). On the other hand,
only the dentition of the fossil flying squirrels was
known. The scarcity of remains of flying squirrels,
although known from most Miocene localities of
Europe, seems to be an additional reason why these
animals were for so long united in the collective
genus ^'Sciuropterus Mein (1970) divided the for-
mer genus Sciuropterus into four genera — Crypto-
pterus, Forsythia, Miopetaurista, and Pliopetaur-
ista. The latter two names were introduced by

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

21

Kretzoi ( 1962), but because this author did not give
any diagnosis for these names, they were regarded
as nomina nuda. Henceforth, the name ^"Sciurop-
terus" must be regarded as a synonym of Pteromys,
the common Recent flying squirrel.

For a long time no Tertiary flying squirrels were
known from America. James (1963) described two
forms from the Barstovian and Clarendonian of the
Cuyama Valley, which he called Sciuroptems ma-
thewsi and S. uphaini. Also from the Upper Mio-
cene (Barstovian) Shotwell ( 1968) reported two iso-
lated teeth, found in the Red Basin of Oregon and
determined as Sciuroptems sp. Lindsay (1972) in-
troduced two new species, S. jamesi and S. mini-
mus, for remains from the Barstow Formation in
California.

James (1963) pointed out that remains of flying
squirrels from the Cuyama Valley did not show a
great similarity with the Recent Gloucomys of
North America. Instead he saw similarities between
North American dentitions and those of fossil flying
squirrels from Europe. He saw two possible expla-
nations for the phylogenetic relationships of his
finds — a faunal exchange with Europe, thus origin
from European forms, or a descent from certain
Early Tertiary paramyids from North America,
showing cheek teeth with crenulated enamel. James
preferred the former possibility. Shotwell (1968)
doubted if the two teeth from the Red Basin really
represent flying squirrels and referred to the occur-
rence of a similar tooth pattern in paramyids and
Marmoto nevadensis.

The earliest fossil flying squirrels are known from
the Lower Miocene of Europe. The earliest Amer-
ican remains generally considered as flying squirrels
are from the Upper Miocene. This fact led different
authors to see the European forms as ancestors for
the American and to postulate a Miocene “migra-
tion” of flying squirrels from Eurasia to North
America.

Regarding this interpretation I have some doubts.
First, in addition to numerous morphological simi-
larities between the forms from both continents (see
Mein, 1970:33), there are some important differ-
ences that make it impossible to trace the American
species back to any known form. In the following
I will go into details of these differences. Therefore,
if one wants to adhere to the migration theory, one
is compelled to postulate a hypothetical Eurasian
ancestor.

In addition, it can be assumed that in analogy to
Recent flying squirrels, the fossil ones were forest-

dwellers. Remains of fossil flying squirrels are
found frequently in lignite deposits (Goriach, Vieh-
hausen, Anwil). However, it is not known whether
they had already developed a patagium and were
able to glide from tree to tree. Moreover, the Mio-
cene flying squirrels were probably living in a trop-
ical or subtropical climate. We have many indica-
tions for such a climate in the Miocene of Europe,
and also James (1963:19) suggests a Neotropical
ecotone for the Cuyama Badland in the Neogene.
Also most species of Recent flying squirrels are liv-
ing in the tropics (Glaucomys and Pteromys seem
to be exceptions). Such dwellers of tropical or sub-
tropical forests may well be dependent upon special
ecologic conditions. Therefore, it is difficult to
imagine that such animals had spread over great
distances, without encountering any climatic or
ecologic barrier. In Miocene times, the only possi-
ble route between Eurasia and North America
seems to have been a land bridge across the Bering
Strait, but as Simpson ( 1947: 14) pointed out and
later authors confirm (see Beringia History, 1973),
there is no evidence that there was ever a tropical
or subtropical climate on this land bridge.

Mein (1970), in his study of the fossil flying squir-
rels from western Europe, concluded that not even
within this relatively small area was there a lively
faunal exchange. He suggests a polyphyletic origin
for the three groups he distinguished.

As an alternative to the “migration” theory ex-
plaining the dental similarities of forms from both
continents, the possibility of parallel evolution
seems more realistic to me. In this case, parallel
evolution seems even more probable because, as
has been mentioned several times, the whole group
of fossil flying squirrels is defined exclusively on
the dentition. Sciurid-like teeth showing enamel
crenulations usually are assigned to flying squirrels.
The fact that such teeth should not automatically
be assigned to flying squirrels is shown by the fact
that similar teeth are also known from completely
different groups (for example, paramyids Thisbe-
mys and Para my s).

In discussion with colleagues about interconti-
nental faunal relationships I heard several times:
“forms whose ancestors we do not know and which
suddently appear in a layer represent without doubt
immigrants.” In my opinion this point of view
greatly overestimates the actual state of our knowl-
edge. True, in the case of the fossil flying squirrels
of North America it is strange that no form from an
earlier level than Barstovian is known. However,

22

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

going further back in stratigraphy, one definitely
encounters forms coming into question as possible
ancestors of the American “sciuropteres.” As al-
ready mentioned, James considered a direct extrac-
tion of the American ‘'sciuropteres” from certain
paramyids, as an alternative to the derivation from
European species (Ctyptopterus from Wintershof-
West). As common features between these para-
myids and the Californian “sciuropteres” James
pointed out especially the subdivided lophs, the
'‘accessory loph” and the hypocone of the upper
molars distinctly separated from the protocone.

I was very surprised by the flying squirrel-like
dentition of some species of Prosciuriis (see Plates
9 and 10). Upper premolars and molars of many
species of Prosciuriis show well-developed proto-
and metaconules and distinct incisions in front of
and behind the protocone, otherwise typical for
flying squirrel teeth. In addition, the teeth of many
Prosciuriis species show a distinct crenulation.
Furthermore, some Prosciuriis species show fea-
tures typical of the Late Tertiary "Sciiiropteriis"
of North America — the long mesoconid, extending
almost to the labial margin of the tooth and the iso-
lated mesostylid of the lower molars. In the for-
mation of the masseteric fossa on the labial side of
the mandible I found a surprising resemblance, the
systematic significance of which I cannot evalu-
ate— a fragment of a Prosciuriis mandible from
Pipestone Springs (CM 9321) shows the same deep-
ly engraved, anteriorly rounded fossa as ^"Sciiirop-
teriis" mathewsi from Cuyama Valley (see James,
1963:Fig. 37a). The mandibles of other sciuropteres
known to me from the Old World show a much less
deep (Petciiirista, Petinomys) and anteriorly more
tapered (Miopetaiirista, lomys, Pteroinys, Hylo-
petes) masseteric fossa.

Also in the European Eocene and Oligocene
there are rodents showing a great similarity to Late
Tertiary sciuropteres. Several authors already re-
ferred to these resemblances. Thus, Schlosser
(1884), for example, pointed out the far-reaching
morphological correspondences of Sciiirodon from
the Quercy with the Recent Pteroinys. Major ( 1893)
emphasized the similarities of Sciiiroides and Pseii-
dosciuriis on the one hand and Sciiiropteriis {lo-
mys) horsfieldi and Sciiiropteriis (Belomys) pear-
soni on the other hand. To this list I would add
Plesiospermophiliis from the Phosphorites, whose
molars, especially the lower, might be determined
as belonging to a sciuroptere if found in a Late Ter-
tiary deposit (see Plate 10).

In total, in North America as well as in Europe
there are throughout the Early Tertiary forms
known which, based on our present knowledge,
cannot be excluded from the ancestry of the Mio-
cene forms, even though separated from the latter
by a long period. Therefore, it is quite possible that
the Late Tertiary American and European rodents
with flying squirrel-like dentition developed com-
pletely independently from each other, with the
European Miocene forms, for example, having orig-
inated from Plesiospermophiliis, the American
from Prosciuriis. However, our present knowledge
is too fragmentary to determine definitely an ances-
tral group. I only want to show that in both conti-
nents there are known Early Tertiary forms that
might have been ancestors of the later ones, without
the necessity of a Miocene migration. At any rate,
as an attempt to explain the striking similarities of
North American and European '^Sciiiropteriis,"
parallelism seems to me more probable than "mi-
gration.” In this opinion I am supported by the re-
sults obtained by Gorgas (1967) who, investigating
Recent Petauristinae, showed that the Recent flying
squirrels do not represent a homogeneous group.
Thus Glaiicomys, for example, corresponds in the
digestive tract with the Sciurinae rather than the
Petauristinae. Hence, possibly the Recent flying
squirrels also represent a polyphyletic group.

The generic name Sciiiropteriis has here been put
in quotation marks for two reasons. First, it seems
possible that the "Sciiiropteriis" from California
have nothing to do with flying squirrels, but perhaps
with rodents of a completely different group having
adapted their dentition to a similar manner of feed-
ing. (Maybe they are not even sciurids!) Thus,
many authors (McGrew, 1941; Stock, 1935; Wilson,
1949; Wood, 1962; Rensberger, 1975) have con-
nected Prosciuriis, in spite of its flying squirrel-like
dentition, not with the Sciuridae but rather with the
Aplodontidae.

The second reason is the following; as mentioned
above. Mein (1970) in his study of the Neogene
flying squirrels from western Europe divided the
genus "Sciiiropteriis" into four genera, and the
name "Sciiiropteriis" is only a synonym of Ptero-
mys. The four "Sciiiropteriis" species from Cali-
fornia (“5.” iiphami, mathewsi, jamesi, and mini-
mus) in my opinion represent a homogeneous group
that cannot be connected to any European genus.
Therefore, I propose to separate them as a partic-
ular genus, which I would call Petaiiristodon. This
separation is based on morphological differences

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

23

and not on the above-mentioned doubts about the
possibilities of an intercontinental dispersal.

Petauristodon, new genus

Derivatio nominis. — Pe t an ristodon:Pe tourist a
tooth-like; this name is intended to refer to the sim-
ilarities of the teeth of this genus with those of flying
squirrels. Thus, it is left open if Petauristodon is
really a flying squirrel or not.

Diagnosis. — Cheek teeth of a sciuroid pattern,
flying squirrel-like, with crenulated enamel. Upper
P and M — with distinct proto- and metaconules,
well developed mesostyle; in some specimens hy-
pocone distinctly separated from the protocone;
"accessory loph” between proto- and metaloph al-
ways present, often also an accessory loph in the
first syncline starting from the protocone; M® as far
as known (P. uphami and P. minimus) without
metaloph. Lower P and M — with short metalophid,
isolated mesostylid, long mesoconid mostly reach-
ing to the tooth margin, without any "anterosinu-
side" (Mein); enamel more crenulated than that of
the upper teeth.

Type species. — P. mathewsi James, 1963.

Type locality. — Remnant Hill, UCMP loc. V-5663,
Cuyama Valley, California.

Stratigraphic range. — Barstovian to upper Clar-
endonian.

Included species. — P. uphami James; P. ma-
thewsi James; P. jamesi Lindsay"*; P. minimus
Lindsay.

Differential diagnosis. — Prom Miopetaurista ,
Petauristodon differs in the following features: in
the mesostyle, the "accessory loph,” and the dis-
tinct hypocone of the upper molars; the lack of a
spur extending backward from the metaconule on
M*^; in the isolated mesostylid, the short mesolophid
and the large mesoconid, reaching to the tooth mar-
gin on the lower molars.

Prom Cryptopterus, Petauristodon is different in
smaller size; on the upper molars better developed
proto- and metaconule, the shorter distance of the
lophs from each other at their point of attachment
to the protocone; on the lower molars by: the lack
of an "anterosinuside” (Mein) and the long meso-
conid.

Prom Forsythia, Petauristodon is different in the
distinct mesostyle and the well developed "acces-

It seems possible to me that P. Jamesi represents a synonym of P. mathew si.
Outside of a small difference in size the two species show hardly any differences.
However, for a definite decision the material of P. jamesi is too fragmentary and
partly too poorly preserved.

sory loph” of the upper molars; by the lack of lat-
eral compression of Mj and M2, typical for For-
sythia, by the isolated mesostylid and the long
mesoconid of all three lower molars.

Prom Blackia, Petauristodon differs in lesser
amount of tooth crenulation, the proto- and meta-
conule, the "accessory loph,” the secondary crests
and the lack of a mesostylar crest of the paracone
on the upper molars; the mesoconid, the well de-
veloped posterior cusps, the isolated mesostylid
and the lack of an "anterosinuside” on the lower
molars.

Prom Pliopetaurista, Petauristodon is different
in the mesostyle, the "accessory loph” and the lack
of a metaconule-spur on the upper molars; by the
isolated mesostylid, the long mesoconid and the
higher metaconid on the lower molars.

Pamily Eomyidae Deperet
AND Douxami, 1902

The eomyids are one of those rodent families
whose history is known over a very long period. In
North America representatives of this family are
known from the Late Eocene until the Late Mio-
cene, in Europe from the Lower Oligocene until the
Pleistocene. It is curious that in this very long his-
tory no increase of size such as is known from a
number of other rodent families can be observed.
On the contrary, the youngest species of this family
are to some extent even smaller than the oldest.
Even more curious is that some forms do not
undergo any morphological changes during this long
history, at least as far as the dentition is concerned
(unfortunately at present skull and skeleton of the
eomyids are scarcely known). In the different ep-
ochs of this long history specialized forms were re-
peatedly developed, but they can be traced only
over relatively short periods. And all indications
suggest that these specialized forms often devel-
oped from the stock of those that remained primi-
tive, hence independently of each other. These de-
viations from the primitive type (or specializations)
almost all develop in one direction, to lophodonty,
to some extent combined with a certain hypsodon-
ty. As far as the dentition is concerned, these lo-
phodont forms resemble each other astonishingly,
although in some cases we are sure that they de-
veloped independently, thus in a parallel manner
(for example, Paradjidaumo, Pseudotheridomys,
and Rhodanomys).

In contrast to these specialized, lophodont forms,
there are bunodont forms, doubtless correctly con-

24

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

sidered as primitive, because the earliest known
representatives of the family — Protadjidaumo in
America, Eomys in Europe — show teeth of this
type. In addition, these bunodont forms look very
much like the sciuravids, considered to be the
ancestors of the eomyids. The relations between
North American and European eomyids were re-
cently studied by Eahlbusch in a very careful paper
(1973). Therefore, I will limit my present study to
those points on which I have a different opinion.

Pseudotheridomys
Historical Introduction

The genus Pseudotheridomys was established in
1926 by Schlosser for an eomyid from Haslach near
Ulm, which he had described in 1884 as Theridomys
parvulus. Eahlbusch (1969) described a second
species, P. pusillus, of this genus. The form from
Cournon called P. schauhi by Lavocat (1951)
should, according to Eahlbush (1968:231), be re-
garded as a synonym of P. parvulus. Because of
the great similarities of the dentition Wilson (1960)
incorporated an eomyid from Quarry A (Martin
Canyon, Colorado) in the genus Pseudotheridomys.
Shotwell ( 1967u) described a new species P. pagei,
from the Quartz Basin (Oregon).

Both Wilson ( 1960:72 and 74) and Black ( 1965:41)
explained the dental similarities of forms from both
sides of the Atlantic by a Burdigalian migration of
Pseudotheridomys from Eurasia to North America.
Eahlbusch ( 1973) concluded on the basis of an anal-
ysis of the evolutionary level of the oldest North
American Pseudotheridomys (from Quarry A) that
the immigration happened earlier, in the late Stam-
pian or early “Aquitanian.” All authors agree that
the genus Pseudotheridomys immigrated to North
America from Eurasia. As far as the chronological
correspondence is concerned, there is no reason to
object to this hypothesis. Eollowing Eahlbusch’s
study (1970) of the changes of eomyid populations
we can say with certainty that the genus Pseudo-
theridomys developed from the genus Eomys in the
Late Oligocene of middle Europe, and the hitherto
oldest North American Pseudotheridomys , P. hes-
perus from Quarry A, appears not before the Mar-
slandian, thus later. To this it can be added that
careful comparisons of the two American species
with those from Europe do not furnish any funda-
mental morphological differences in the dentitions.

On the other hand, regarding the astonishing sim-
ilarities that developed again and again within the

whole family Eomyidae during this long history, the
question arises if the resemblances of American and
European forms could not have been developed in-
dependently from each other, in a parallel way. It
is worth mentioning that all comparisons are nec-
essarily confined to the dentition, because neither
the skull nor the postcranial skeleton of these ani-
mals is known. In addition, just recently several
examples of other groups have showed that dental
features alone are often insufficient for recognizing
phylogenetic relationships.

There is one argument for the hypothesis that
those American and European eomyids now united
within the genus Pseudotheridomys could have
originated independently — now we can say with
great certainty that lophodont eomyids developed
from bunodont forms. These bunodont genera (Pro-
tadjidaumo, Adjidaumo, Eomys, Eeptodontomys)
are hardly distinguishable by the dentition (this is
also obvious in Eahlbusch’s new diagnosis for the
genus Eomys [1970:104] which can be transferred
easily to some members of the genus Adjidaumo).
With good reasons Wilson (1960) pointed out that
Pseudotheridomys hesperus shows no close rela-
tionship to any American eomyid. It has to be
added to this observation that the history of the
North American Eomyidae is very poorly docu-
mented after the Middle Oligocene and that new
connecting forms would not be surprising. Com-
paring the European species of Pseudotheridomys
with earlier eomyids, especially Eomys, the mor-
phological relations are not immediately apparent.
Not until the faunas of the fissure fillings from
southern Germany — especially Gaimersheim — were
known, were we informed that Pseudotheridomys
developed from Eomys. The morphological changes
which the dentition of Pseudotheridomys shows
compared with that of Eomys are not numerous:
( 1) labiad elongation of the last syncline of the lower
premolar and molars; (2) linguad elongation of the
first syncline of the upper molars; (3) development
of a first syncline on the upper premolar. As Plates
11 and 12 show, the North Arntncdin Pseudotheri-
domys species correspond very well in these three
features with the European. Thus there is not just
one characteristic but a combination of three fea-
tures in which Pseudotheridomys from the New
World differs from other eomyids of the same con-
tinent, while sharing these features with some Eu-
ropean forms. There are many reasons to believe
that the compared North American and European
forms of Pseudotheridomys are closely related. If

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

25

other mammal groups did not show even more as-
tonishing dental resemblance, which can be ex-
plained by convergence, I would consider the close
relationship as proved. With an example of Eocene
rodents I want to show how much derived forms,
whose ancestors resemble each other, can corre-
spond morphologically, without having any close
phylogenetic relationship as proved. With an ex-
ample of Eocene rodents I want to show how much
derived forms, whose ancestors resemble each oth-
er, can correspond morphologically, without having
any close phylogenetic relationship — certain Pseu-
dosciuridae from Europe (Protadelomys, Adelo-
mys\ see Stehlin and Schaub, 1951:Eigs. 26, 29, 31 1,
313) with their simple bunodont tooth pattern look
very much like some bunodont Eomyidae. From
these Pseudosciuridae there developed very prob-
ably lophodont forms of the Theridoniys group
(Theridomys, Trechomys, BlainviUimys', see Stehlin
and Schaub, 1951:Figs. 29-33, 316-320), which in
their form resemble very much the lophodont
eomyids, especially Pseudotheridomys, that origi-
nated from bunodont ancestors.

In the Eomyidae there are more examples of cer-
tainly independent development of similar, more or
less lophodont tooth patterns. In contrast to the
above-mentioned example, in the following two
cases the probably bunodont ancestral forms are
not known: the lower dentition of Centimanomys
major from the Chadronian of Colorado (Galbreath,
1955) shows great similarities to Pseudotheridomys ,
but is certainly not closely related to the latter. The
conclusive feature for this similarity, in which Cen-
timanomys is different from nearly all North Amer-
ican eomyids and which it has in common with
Pseudotheridomys , is the fourth internal syncline
of Ml and Mg extending far labially.

As a further example, Meliakrouniomys (Harris
and Wood, 1969) from the Chadronian of Texas
shows several features in common with the Euro-
pean genus Ritteneria (Stehlin and Schaub). The
similarities of these two genera are based essen-
tially on the complete interruption of the longitu-
dinal ridge (Langsgrat) and the direct connection of
the four main cusps by transverse lophs on the low-
er molars. As Emry (1972) demonstrated Melia-
krouniomys is probably not an eomyid but a het-
eromyid, and thus a close phylogenetic relationship
does not need to be discussed.

Comparing American and European representa-
tives of the genus Pseudotheridomys , it is interest-
ing that in the two continents different evolutionary

trends are found. Pseudotheridomys pagei (Plates
lid and 12d) from the Barstovian of Oregon — the
hitherto most recent Pseudotheridomys — is hardly
distinguishable from P. hesperus (Plates lie and
12e) of Quarry A; while in Europe, already in the
Aquitanian, tendencies can be observed in Pseudo-
theridomys, which lead over to Ligerimys. Also the
European genus Keramidomys, originating in the
Middle Miocene, has to be regarded as a descen-
dant of Pseudotheridomys. Thus we find the follow-
ing situation: whereas Pseudotheridomys in North
America does not continue differentiating, it splits
in Europe into two lines in which the molar pattern
is to some extent different from that of the ancestral
genus.

In summary, the history of this genus, especially
in North America, is insufficiently documented at
present. If for the present no fundamental differ-
ences between North American and European
species can be found, and therefore an interconti-
nental faunal exchange seems to be probable, I
nevertheless consider it not impossible that the
American and European forms developed indepen-
dently from bunodont eomyids.

Leptodontomys
History of Investigation

The genus Leptodontomys was established by
Shotwell (1956) on the basis of a single lower jaw
fragment with incisor and premolar from the Hem-
phillian of Oregon. Later Shotwell (1967a) reported
some isolated teeth from the Clarendonian of Ore-
gon belonging to this genus (Leptodontomys sp.).
Hugueney and Mein ( 1968) incorporated several
Miocene and younger eomyids from different Eu-
ropean localities in this genus (La Grive, Lissieu,
Can Llobateres, Manchones, Schernfeld bei Eichs-
tatt). In the meantime similar teeth were found also
in Neudorf a.d. March, Franzensbad, Vieux Col-
longes, and Anwil. Finally Fahibusch ( 1973) dis-
cussed exhaustively the relations of North Ameri-
can and European Leptodontomys. This author
preferred the possibility of a Middle Miocene "mi-
gration” of Leptodontomys from Europe to North
America to a derivation of the New World Lepto-
dontomys from older American eomyids.

Morphological Particularities of

Leptodontomys

As pointed out in the section on Pseudotheri-
domys, bunodont eomyids resemble each other

26

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

very much in tooth pattern. Thus Leptodontomys
differs from other bunodont forms, such as Adji-
daiimo and Eomys, only insignificantly. Among
North American eomyids Leptodontomys is out of
place in having a linguad anterior cingulum on the
upper molars, otherwise a rather rare feature. The
differences between Leptodontomys and Adjidau-
mo pointed out by Shotwell (“it [Leptodontomys]
differs from this genus[Adjidaumo]hy having a
poorly-developed posterior cingulum, a smaller me-
soconid, and widely separated protoconid and
metaconid”) all refer to the lower premolar and
turned out to be largely invalid when more complete
material from Black Butte was found. Also Eahl-
busch ( 1973) and Hugueney and Mein ( 1968) attach
great importance to the development of a linguad
anterior cingulum on the upper molars.*' The latter
authors might have been led at least in part by this
feature to unite American and European forms in
the same genus (see emended diagnosis, Hugueney
and Mein, 1968:200).

Discussion

Assignment of the European forms to Leptodon-
tomys was suspicious to me from the beginning and
after seeing Shotwell's original specimens from Or-
egon my doubts increased. We do not have any
indications of the dispersal from one continent to
the other. Supposition of a close relationship is
based exclusively on the correspondence of the
tooth pattern. Especially in cases like this, we must
be very careful because the tooth pattern of Lep-
todontomys is certainly very primitive. This be-
comes apparent in comparing the tooth pattern of
Leptodontomys with that of the earliest known
eomyids such as Protadjidaumo and Lomys. So far
as primitive forms are concerned, there is generally
a risk to assume close phylogenetic relationships on
the basis of morphological similarities to other
primitive forms, where actually no relation exists.
Comparing, for example, Protadjidaumo and Lep-
todontomys (see Plate 14b, e), it seems unlikely that
the forms now united in the genus Leptodontomys
represent in North America as well as in Europe
different lines of eomyids that remained primitive.

Fahlbusch (1973) considers probable the extrac-
tion of Leptodontomys from the Lomys rhodanicus
group. He assumes that Leptodontomys originated

' ' In this development of the linguad anterior cingulum Leptodontomys corresponds
very well with the genus Pseudadjidaumo. introduced by Lindsay ( 1972). The differ-
ences between the tw'o genera seem too vague, and 1 am of the opinion that Pseu-
dadjidaumo is a synonym of Leptodontomys.

in Europe and immigrated to North America in the
Middle Miocene. In my opinion there is nothing
against tracing back the descent of Leptodontomys
to the Lomys rhodanicus group, but for many rea-
sons it seems to me more likely that the American
Leptodontomys developed independently from the
European. In Protadjidaumo (see Plate 14e) we
have an American group to which, although sepa-
rated by a long time interval, Leptodontomys can
easily be traced morphologically. Among the Upper
Eocene eomyid materials of the Badwater area
(Wyoming) there is a form, possibly of the Protad-
jidaumo group, showing a well developed linguad
anterior cingulum on the upper molars. That this
cingulum is a primitive feature seems to be corrob-
orated by the very old Lomys teeth from Hoog-
buitsel clearly showing this feature too. Therefore,
it seems to me more probable that the dental simi-
larities of European and North American Lepto-
dontomys may be explained in the following way:
in the Early Tertiary the eomyids immigrated to
Europe (see Discussion), after having almost cer-
tainly developed in North America from sciuravid
ancestors. In both continents forms with a very
primitive dentition similar to the common ancestor
endured a long time (in America till the Upper Mio-
cene, in Europe perhaps till the Early Pleistocene).
Fahlbusch (1973) takes this possibility of two in-
dependent lines into consideration also but regards
the possibility of a migration as more likely.

Morphological Differences between
American and European Lorms

A number of differences between forms of the
two continents reinforced my opinion that in the
case of Leptodontomys we are dealing with two
largely independent lines, each of them confined to
its continent. These are differences in the dentition
but also in morphology of the lower jaw. I would
not have attached more importance to the dental
differences of the dentition than have other authors
if I had not noticed that exactly those features dif-
fering North American and European Leptodonto-
mys also separate the other eomyids of the two con-
tinents into two groups. Thus, the European
Leptodontomys in these features links up closely
with other eomyids from Europe, whereas the
North American genus in these features corre-
sponds with the eomyids of that continent. For ex-
ample, the exterior synclines of the upper molars
of most North American eomyids (especially of
Paradjidaumo and Adjidaumo) reach less far lin-

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

27

gually than those of European forms (most Eomys
and especially Ligerimys \ see Plate 13). Connected
with this, the junction of the anteroloph with the
protoloph in American forms lies generally in the
middle of the tooth or somewhat labiad of it, where-
as on upper molars from Europe this junction is
located more lingually. In these features both Lep-
todontomys from Europe and those from America
behave "continent-typically.” The fourth interior
syncline of the lower premolar and molars is a fur-
ther feature apparently splitting the eomyids of both
continents into two groups — so far as I have seen
this syncline is little developed in American
eomyids (for example, Adjidaumo, Panidjidaumo,
Aidolithomys, Nanmtomys , "Adjidaumo" quartzi),
but on the average distinctly better developed in
European forms (Plate 14). In this feature Lepto-
dontomys from North America are clearly different
from those of Europe. In the development of these
features the later European forms are possibly more
primitive, because the oldest known eomyids, those
from the Badwater Upper Eocene, show a similar
development in part.

An essential difference between Leptodontomys
from North America and Europe appears in the
lower jaw. Outside of the dentition the lower jaw
is for the moment the only element in which we can
compare the Leptodontomys of the two continents.
These differences, unlike those mentioned above,
are not found in the other eomyid genera, but they
do not seem to me to be less important systemati-
cally since the morphology of the lower jaw is oth-
erwise very uniform in this family.

Differences of the Lower Jaw

Eor these comparisons the type of L. oregonensis
from McKay Reservoir (Oregon) and two mandi-
bles from La Grive were at my disposal (see Eig.
7). It is immediately apparent that the American jaw
is more delicate than the European. Above all the
section between the tip of the masseteric fossa and
the upper margin of the horizontal ramus is distinct-
ly higher in the specimens from La Grive. In addi-
tion the upper edge of the diastema is more curved
in the La Grive jaws. Although the La Grive spec-
imens have the upper edge of the anterior part of
the diastema at the same level as that of the alveoli,
the same edge on the type specimen from Oregon
ranges somewhat lower (see projection Fig. 7). Be-
hind M;3 the La Grive jaws show a distinct swelling
caused by the end of the incisor. This swelling is
lacking on the Oregon specimen. On the other hand

the latter mandible shows a ledge extending from
the base of the ascending ramus to the condyle
which is lacking on the specimens from La Grive.
In addition the lower margin of the jaws is curved
in a different way: in the Oregon specimen the most
distinct curvature is below M3, in the specimens
from La Grive, further back.

Possibly the most important difference between
Leptodontomys from North America and Europe
is found in the lower incisor. Although the type
specimen from McKay Reservoir does not show the
slightest trace of a crenulation, one appears dis-
tinctly on the incisor of Leptodontomys from La
Grive (see Fig. 8 and Hugueney and Mein,
1968:196). Such enamel crenulations of the lower
incisors are known from many eomyids (for exam-
ple, Lomys, Pseudotheridomys ) and are, in my
opinion, important for systematics. In the case of
Leptodontomys this feature alone seems to justify
a generic separation for the European species pre-
viously referred to the genus.

Eomyops, new genus

Derivatio nominis. — Lomyops = Eo/ny.v-like.

Diagnosis. — Small eomyid with bunodont cheek
teeth and crenulated lower incisor. Upper molars —
with well developed linguad anterior cingulum; sec-
ond and fourth exterior syncline mostly extending
lingually past the middle of the tooth. Mesoloph
short, somewhat anteriorly directed. On P4, M,, and
M2, fourth interior syncline well developed.

Type species. — Lomys catalauniciis Hartenber-
ger, 1966, from Can Llobateres (Spain).

Included species. — Lomys catalauniciis Harten-
berger, 1966; Leptodontomys hodvanus Janossy,
1972.

Stratigraphic and geographic range. — Middle
Miocene (level of Neudorf a.d. March. NMU 6) to
Lower Pleistocene (Schernfeld bei Eichstiitt) of Eu-
rope, from Spain to Czechoslovakia.

Differential diagnosis. — From Leptodontomys
Shotwell. 1956, Lomyops is different in — the dis-
tinct crenulation of the lower incisor, lacking on the
incisor of Leptodontomys', the more robust and dis-
tinctly higher horizontal ramus of the mandible; in
addition the mandible of Lomyops shows a swelling
for the end of the incisor, lacking on the mandible
of Leptodontomys', the better developed fourth in-
terior syncline of the lower P, M,, and M2; the ex-
terior synclines of the upper molars extending more
lingually than on molars of Leptodontomys', the dif-
ferent course of the anterior branch of the proto-

28

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

Fig. 7. — a) Leptodontomys oregonensis Shotwell, L mandible (invers) with I and P4, (type) UO 3533, McKay Reservoir (Oregon),
Hemphillian, 15x; b) Eomyops aff. catalaunicus (Hartenberger), L mandible (invers) with M, and M.2, La Grive (France), Middle
Miocene, 15x.

conid and the metalophid, which in Eomyops ex-
tend more or less transversely and form a direct
junction of proto- and metaconid, whereas in Lep-
todontomys these two lophs extend somewhat for-
ward, joining only on the anterolophid or shortly
before it.

From Eomys Schlosser, 1926, Eomyops is differ-
ent in having a linguad anterior cingulum on the
upper molars; exterior synclines of the upper mo-
lars extending more lingually; the short mesoloph
of the upper molars, always anteriorly directed.

From Adjidaumo Hay, 1930, Eomyops is differ-
ent in having a linguad anterior cingulum on the
upper molars; the better developed fourth interior
syncline on lower premolar and molars; the position
of the foramen mentale, located distinctly higher in
Adjidaumo (see Black, 1965:37, Fig. 6b).

As to Pseudadjidaumo Lindsay, 1972, see foot-
note 1 1 .

COTIMUS, Leidymys, Eumyarion, and
Evcricetodon

(Family Cricetidae Stehlin and Schaub, 1951)
Historical Introduction

In his pioneer work on Tertiary cricetids, Schaub
(1925) described a hamster from the Swiss molasse
with the name '"Cricetodon helveticus." He gave
up the name “"Cricetodon medius'^ (Lartet, 1851),
established for a cricetid from Sansan, because no
type was preserved and because it was impossible
for him to identify “C. medius" among the three
medium-sized cricetids from Sansan using the poor

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

29

diagnosis of Lartet.’^ Eahlbusch (1964) split
Schaub’s genus Cricetodon and declared C. hel-
veticus a synonym of C. medius Lartet, after having
found a fragment of a maxillary from Sansan labeled
with Lartet’s own hand. In addition, he assigned C.
medius to the American genus Cotimus Black,
1961. Shortly after doubts were heard about this
assignment, probably essentially because only the
type mandible of Cotimus was known (Freuden-
thal, 1965:297). Thaler (1966) expressed these
doubts by introducing a new subgenus for the Eu-
ropean species — Eumyarion, later established as a
genus by Mein and Freudenthal (1971).

Two facts may contribute something to the so-
lution of the Cotimus problem — first of all (personal
communication, C. C. Black) the holotype of Co-
timus alicae is not from the Upper Miocene (middle
or late Barstovian) as initially stated, but from the
Upper Miocene (middle or late Barstovian) as ini-
tially stated, but from the Lower Miocene (Arika-
reean). The type originates from a layer possibly
representing an equivalent of the Gering Formation
(personal communication, D. Rasmussen). In ad-
dition, Rasmussen found new materials of Cotimus
at the type locality, among which there was an up-
per jaw. Judging from this jaw, Cotimus is without
doubt a synonym of Leidymys (Wood, 1936). Even
if these new discoveries shed new light on the dis-
cussion, the problem of the relations of North
American and European forms is far from being
solved. Hitherto lacking in the literature are exten-
sive comparisons of species from both continents.

Comparisons o/ Leidymys f=Cotimus Black)
and Eumyarion (Thaler)

In spite of the new materials from both continents
the comparisons are basically still confined to den-
titions.

Lower dentition. — On the basis of the lower in-
cisor the two forms are easily distinguishable — Lei-
dymys shows a distinctly flattened incisor anteriorly
with a triangular cross section (see Plate 17e) and
that of Lumyarion lacks the sharp exterior edge and
is rather kidney-shaped in cross section. The thin
enamel coating covers only the anterior surface of
the incisor of Leidymys (as in Lumys), whereas on
the incisor of Lumyarion the enamel seems to be
thicker and extends from the anterior surface nearly

According to Schaub (1925) at the Sansan locality besides C. sansaniensis and
C. minus, four species of cricetids are present — C. gaillardi, C. affinis. C. helveticus.
and C. brevis. As the extensive new materials from Sansan show. C. affinis is lacking
at this locality.

Fig. 8. — Comparison of the underside of the left lower incisor
of a) Leptodontomys oregonensis Shotwell, UO 3633, McKay
Reservoir (Oregon); h) Eomyops aff. cataluunicus (Hartenber-
ger), NMBS, G.A. 635, La Grive (France). Both figures 25x.

to the middle of the lateral side of the tooth. The
enamel stripes on the exterior face show differences
(see Plate 17e) — the incisor of Leidymys shows
three stripes of which two are close and somewhat
lateral of the middle of the anterior face. The incisor
of Lumyarion, on the other hand, shows only two,
more separated, stripes. The molars of the two gen-
era are very similar in size, but the lower incisor of
Lumyarion is distinctly more slender compared
with that of Leidymys.

Of the lower molars, M, shows the most distinct
differences between the two forms — although the
"Vorjochkante” (Schaub) in Leidymys always ex-
tends from the metaconid to the posterior branch
of the protoconid (Schaub’s "alte Vorjochkante”),
it extends in Lumyarion from the metaconid to the
anterior branch of the protoconid (Schaub’s “neue
Vorjochkante,” Plate 16a and b). Connected with
this the two anterior cusps of Lumyarion are more
alternating than in Leidymys, with the metaconid
more anteriorly located than the protoconid. The
two anterior cusps of M, of Leidymys are less al-
ternating. The great variation of connections of
both anterior cusps and the anteroconid in all
species of Lumyarion is very conspicuous (see En-
gesser, 1972:Figs. 109 and 111), whereas in Leidy-
mys these connections show little variation.

In M2 Leidymys is different from Lumyarion in
sometimes having both the anterior ‘ ‘Vorjoch-
kante” (the “new” one of Schaub) and the poste-
rior (the “old”) one. Such teeth show mostly a
strong genuine mesostylid spur (see Plate 17d). The

30

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

mesoconid is generally better developed in Leidy-
mys than in Eumyarion. Also the metaconid of M2
(as of M3) of Eumyarion is joined more closely to
the anterior cingulum than on M2 of Leidymys.

The M3 of Eumyarion and Eeidymys are not very
different, although that of Eumyarion is more re-
duced.

Schaub ( 1925) pointed out the importance of the
middle spur of the lower molars as a criterion for
determination of phylogenetic relationships — it
must be determined whether it is a genuine meso-
stylid spur originating from the mesoconid or an
elongated posterior branch of the protoconid
(“Pseudomesostylidsporn” of Schaub). In this ele-
ment Leidymys differs from Eumyarion only in
those specimens showing a posterior branch of the
protoconid extending to the metaconid. In this case
a strong, genuine mesostylid spur is developed,
often reaching to the lingual tooth margin (see Mi
of the holotype of Leidymys alicae, Plate 16b). In
specimens showing a freely terminating posterior
branch of the protoconid, thus having a "Pseudo-
mesostylidsporn” (mostly in M2 and M3), there is
in Leidymys as well as in Eumyarion no mesostylid
spur present or only a very small one (see Plate 16a
and b: E. medius and M2 of the type of L. alicae).

Some M, and M2 of Leidymys {L. alicae and L.
nematodon) show a freely terminating posterior
branch of the hypoconid in the fourth exterior syn-
cline. This spur is only rarely found in Eumyarion
medius from Sansan, but more frequently in the
species E. latior, E. bifidus, and E. leemanni.

Upper dentition. — The upper molars of Eumy-
arion are more different from those of Leidymys
than are the lower. Probably the most conspicuous
and, for systematic relationships, the most impor-
tant difference appears in the anterocone of M’. In
Eumyarion this anterocone is very wide, mostly
somewhat bipartite (distinctly bipartite in E. bifi-
dus) and its highest point is in the level of the ex-
terior cusps (Plate 15a). The anterocone of M' of
Leidymys, on the other hand, is distinctly narrow-
er, always one-part and its highest point is between
the top of the two anterior cusps (Plate 15b). In
addition, the M‘ of Eumyarion in most species (E.
latior, medius and leemanni) shows an anterior
transverse spur (“vorderer Quersporn”), a feature
which I did not observe in any tooth of Leidymys
(Plate 17a-c). However, most specimens of Leidy-
mys show a freely terminating anterior branch of
the protoconid. This detail is found in Eumyarion
too — in E. bifidus and, together with an anterior

transverse spur, in E. latior (see Engesser, 1972:
Eig. 106) — but never in E. medius. The connection
of the mesostyle spur with the metacone (Plate 15a)
is a peculiarity of E. medius and is found only rarely
in other species.

In Leidymys, the “Vorjochkante” of M'^ leads
directly to the protocone; in Eumyarion, on the oth-
er hand, it joins the posterior branch of the proto-
cone. The M^ of Leidymys (L. alicae, not L. ne-
matodon) has a lingual anterior cingulum; one is
lacking completely in Eumyarion (Plate 15a-c).

The M^ of Eumyarion is more reduced than that
of Leidymys. In Eumyarion the protocone covers
almost the whole lingual face and the interior syn-
cline is conspicuously shortened in favor of a cen-
tral crater. In Leidymys the M^ is, compared with
M^, only slightly reduced on its posterior end. It
has a lingual anterior cingulum in L. alicae (see
Plate 15b) and relatively well-developed metacone.

Lower jaw. — In the mandible, the only part of
these animals other than the dentition that we can
compare, differences between Eumyarion and Lei-
dymys can be observed — the anterior part of the
masseteric fossa, forming a short crista, in Eumy-
arion extends to below the anterior margin of Mi or
even farther anteriorly. As far as I observed in Lei-
dymys, it extends only somewhat beyond the mid-
dle of M,. On both lower jaws of Leidymys which
I could compare (L. alicae and L. nematodon) there
was a second, smaller foramen somewhat above
and in front of the mental foramen. I never ob-
served such a foramen in Eumyarion. The position
of the mental foramen seems to be of little use for
systematics, because the two mandibles of Leidy-
mys having this detail showed different positions of
it.

Discussion of the differences. — Most of these dif-
ferences between Leidymys and Eumyarion do not
seem to be basic, but rather represent different
levels of similar developments. The “modern" fea-
tures are without exception allotted to Eumyarion
(“neue Vorjochkanten," reduced lingual anterior
cingulum of M- and M®, slightly developed meso-
conid of the lower molars, reduced M^, and others).
Leidymys alicae is morphologically more primitive.

Comparison of Leidymys (=Cotimus) and
Eucricetodon

Differences between Eumyarion and Leidymys
seem to be due to different levels of development.
However, Leidymys (=Cotimus) and Eumyarion
are not here considered to be congeneric, as pro-

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

31

posed by different authors. To establish the general
differences between American and European forms
it is necessary to compare Leidymys with a more
or less contemporary cricetid such as Eucricetodon
collatus (see Plates 15c and 16c). In E. collatus fea-
tures are found again which showed Leidymys to
be more primitive as compared with Eumyarion —
“Alte Vorjochkanten” of M, and Ma, lingual ante-
rior cingulum of M“ and M^, well-developed meso-
conid on the lower molars. On the other hand, there
are differences between Leidymys and Eucriceto-
don too — as in most Oligocene cricetids of Europe
(exceptions, Paracricetodon and Heterocriceto-
don), E. collatus shows a distinctly more reduced
than Leidymys (see Plate 15b and c). The an-
terocone of M‘ in many species of Eucricetodon
(especially in E. atavus and E. huerzeieri) shows
a tendency to become wider and to develop two
peaks , not present in Leidymys. The lower incisor
of Eucricetodon is distinctly more slender than that
of Leidymys and its cross section is not triangular
but rather drop-shaped (see Plate 17e). The enamel
of the lower incisor of Eucricetodon extends more
to the lateral face than in Leidymys. In addition,
the incisor of Eucricetodon lacks the ledge on the
lateral face, typical for Leidymys (and Eumys) and
its enamel is thicker. In regard to these features of
the incisor, Eucricetodon fits closer to Eumyarion
than to Leidymys. The enamel stripes of the exte-
rior face of the incisor seem less significant for these
comparisons because they show a great variation in
the different species of Eucricetodon. The same
applies to the upper incisor showing, according to
Wood (1937:257), stripes in Leidymys iockingtoni-
anus. Certain species of Eucricetodon have striped
upper incisors (for example, E. collatus), whereas
others do not (for example, E. huberi).'^ In spite of
these differences the dental similarities of Eucrice-
todon and Leidymys are striking. Vianey-Liaud
(1972) raised the question whether Eucricetodon
and Leidymys should be united in the same genus.
Besides the differences mentioned above, I oppose
such a union for other reasons. Certainly by Ari-
kareean (or “ Aquitanian”) times, Leidymys as well
as Eucricetodon had had a long separate evolution;
Leidymys in North America being known since the
Whitneyan (L. vetus), Eucricetodon in Europe

The value of such features of the incisor for the systematics is for the moment
difficult to estimate. It is impossible to decide whether the incisor differences between
these species indicate heterogeneity of the genus Eucricetodon. or whether these
features vary even in closely related species and therefore are less useful for system-
atics. The heterogeneity of this group was pointed out by Thaler (1966:140) in intro-
ducing the new subgenus Eucricetodon.

since the lower Middle Oligocene (£. atavus).
Wood (in Clark et al., 1964) traces the descent of
Leidymys back to the Eumys group against which,
in my opinion, there are no objections, because a
form known as L. vetus seems to occupy an inter-
mediate position between Eumys and Leidymys.
The lower incisor of the latter with its nearly tri-
angular cross section, the distinct exterior edge and
the very thin enamel coat is very similar to that of
Eumys. Both Eumys and Leidymys represent prim-
itive forms, possibly resembling a common ances-
tor. This becomes evident by a comparison with
Simimys from the Uintan (Upper Eocene), the ear-
liest known cricetid (if it is really a cricetid!). Tak-
ing a look at Plates 15 and 16, it becomes obvious
how widespread this tooth pattern is and over what
a long time it held out, changing only in slight de-
tails. On these plates, forms are figured from the
range between Orellan (Middle Oligocene) and Mid-
dle Miocene which are intended to point out how
similar all these forms are, although coming from
different continents and ages. As observed by other
authors (among others, Schaub, 1925), some Recent
cricetids of Madagascar show a similar tooth pat-
tern. This resemblance easily could be dismissed as
pure convergence, if the lower incisors did not
show exactly the same two enamel stripes as, for
example, Eumyarion]

In summary, Eumyarion, Leidymys (=Cotimus),
Eucricetodon, and Eumys represent forms showing
a very primitive tooth pattern and are therefore dif-
ficult to differentiate from each other on the basis
of the dentition. The teeth may be throughout suf-
ficient for differentiation of species and for deter-
mination of the evolutionary level, but for recon-
struction of intercontinental relationships we need
more complete remains.

Democricetodon and Copemys

(Eamily Cricetidae)

History of Investigations

Eahlbusch (1964) divided the genus Cricetodon
and united Schaub’s species brevis, affinis, and
gaillardi in the new genus Democricetodon. Influ-
enced by the great dental similarity of Democrice-
todon and the American genus Copemys (Wood,
1936), Eahlbusch (1967) proposed that Democrice-
todon be regarded henceforth only as a subgenus of
Copemys. This generic assignment did not meet
with unanimous approval. In their new classifica-

32

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

tion of the Tertiary cricetids from Europe, Mein
and Freudenthal (1971) argued for a generic sepa-
ration of Copemys and Democricetodon. This sep-
aration was substantiated above all by differences
in the position and form of the incisive foramen.
The same authors (1971:31-32) assigned Democri-
cetodon to the subfamily Cricetinae (Murray, 1866),
whereas they considered possible an affiliation of
Copemys to the subfamily Hesperomyinae (Mur-
ray, 1866).

Comparisons

Because the dentitions of Copemys and Demo-
cricetodon were already compared extensively by
Fahibusch (1967), I can confine my statement to a
few details. The dentitions of some species of De-
mocricetodon and Copemys resemble each other to
such a degree that — as Lindsay (1972) wrote — even
a specific distinction would be difficult, if they were
found at the same locality. The only general differ-
ence between the forms of the two continents that
I see can be observed on M^ — in all species of Co-
pemys whose M^ I know, this tooth is more reduced
than in all species of Democricetodon (see Plate
18). This reduction affects not only the posterior
half of the tooth, as is usual on M^ of many other
cricetids, but all elements except the protocone.
Outside of this difference all details of the pattern
typical for Democricetodon appear in the different
species of Copemys — the short anteroconid of M,,
the anterior transverse spur of M‘, the spur in the
exterior syncline of the lower molars.

Authors such as Fahibusch (1967) and Lindsay
( 1972), arguing for a generic union of Copemys and
Democricetodon, were apparently guided by the
close similarities of the dentitions. But according to
Mein and Freudenthal (1971:30) — and I agree in this
respect with these authors — the dentition of the cri-
cetids is perfectly useful for the distinction of dif-
ferent species, but not for the proof of phylogenetic
relations above the species level. Mein and Freu-
denthal supported their opinion that the dental sim-
ilarities of Copemys and Democricetodon have to
be traced back to parallelism and that the two gen-
era very probably belong to two different subfami-
lies (Hesperomyinae and Cricetinae), based on dif-
ferences in the skull and skeleton. For the following
reasons I do not agree in all points in this argument:
these two authors attach great systematic impor-
tance to the position and shape of the incisive fo-
ramen. They state (1971:28) that Democricetodon,

with its short incisive foramen with a posterior mar-
gin located anterior to the anterior end of M’, is a
representative of the Cricetinae. The incisive fora-
men of Copemys, on the other hand, continues be-
tween the upper molars as typical for the Hesper-
omyinae. The systematic value of this foramen
seems to me somewhat doubtful, because in many
cricetids it exhibits considerable variation — on cer-
tain skulls of Eumys elegans (AMNH collections)
the posterior margin of this foramen occurs on the
level of the anteroconid of M‘, but in many speci-
mens assigned to the same species it is placed clear-
ly more anteriorly (see Wood, 1937:P1. 32). In the
only skull of Copemys known to me in which the
incisive foramen can be observed (Lindsay,
1972:Fig. 45), the foramen ends about at the level
of the anterior margin of Mb In Peromyscus, which
developed almost certainly from Copemys, the pos-
terior margin of this foramen is placed distinctly
anterior to the anterior end of M'. Mein and Freu-
denthal assign Democricetodon to the subfamily
Cricetinae and give at the same time a provisional
diagnosis of this subfamily based on the skeleton
and skull. According to this diagnosis the mandible
of Democricetodon is inclined lingually very dis-
tinctly, so that it is impossible to see the mental
foramen in occlusal view. In addition, the humerus
of Democricetodon shows an entepicondylar fora-
men. As to these two features, it can be said that
the mandible of Copemys is actually less inclined
lingually than that of Democricetodon, but also the
humerus of Copemys shows an entepicondylar fo-
ramen (this knowledge I owe to Everett Lindsay,
who kindly gave me a cast of a Copemys humerus).

For the moment it seems extreme to draw phy-
logenetic conclusions on the basis of these differ-
ences or similarities. According to Romer (1949) the
entepicondylar foramen is a primitive feature and
also occurs in other genera of cricetids (for exam-
ple, Fahlbiischia).

Although questioning the value of the criteria
mentioned by Mein and Freudenthal, for the follow-
ing reasons I still hold that Copemys and Democri-
cetodon, in spite of the similarities of the dentition,
need not be closely related: as we can suppose for
good reasons, probably all species of Democrice-
todon (as well as those of Rotundomys and Ko-
walskia) developed from a small and very primitive
Lower Miocene form, which could have been
closely connected with Demoricetodon minor fran-
conicus from Erkertshofen (see Plates 18b and

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

33

19b). Likewise it can be supposed that all North
American species of Copemys descend from a
primitive ancestral form, which might have been
similar to C. pagei or C. tenuis (see Lindsay,
1972:77).

These most primitive known forms of both con-
tinents— C. pagei or C. tennis, and D. minor — are
very similar dentally (see Plates 18 and 19). This
similarity led several authors to consider that both
genera must be regarded as immigrants, descending
from a common, probably Asiatic, ancestral form.

To me it is very strange that the evolved forms
from North America and Europe, which can be de-
rived with great probability from the primitive ones,
look again very similar to each other. This similarity
goes so far that, for example, Copemys rnsselli or
C. barstowensis show exactly the same details in
the dentition as are typical for Democricetodon
guUlardi — the long mesolophs or mesolophids, the
anterior transverse spur of the exterior syncline of
the lower molars, the start of a labial posterior cing-
ulum of M, and M.2, and others (see Plates 18c and
19c). Even the bend of the “Vorjochkante” and the
"Nachjochkante” of the lower molars, typical for
Democricetodon gaillardi, can be observed in cer-
tain forms of Copemys (see Shotwell, 19677>:Eig.
6A). This certainly parallel evolution of the De-
mocricetodon and Copemys groups continues in the
Pliocene — on both sides of the Atlantic forms de-
velop showing a bipartite division of the anterocone
of M’, a reduction of the posteriormost exterior
syncline of the upper molars, and distinctly ante-
riorly directed "Jochkanten” of the lower molars
(for example, Peromyscus pliocaenicus in North
America, Rotiindomys and Kowalskia in Europe).
How can these similarities be explained?

I) Migration? The fact that we do not know direct
ancestors either of Copemys or of Democricetodon
is for many authors proof that both genera were
“immigrants.” This statement is extreme because
there is no known representative from an earlier
deposit than Barstovian. The statement that the
morphological similarities of the evolved forms
from both sides of the Atlantic were caused by

'* According to Mein and Freudenthal (1971:28), Democricetodon from Viliafeliche
II A (NMU 4) and Erkertshofen (NMU 4) represent the oldest forms of the genus and
they comply with all requirements which could be made for an ancestor of all later
forms. The form of Viliafeliche seems to me to comply better with these requirements
because it shows an entoconid on M3, in contrast to that from Erkertshofen (see
Freudenthal. 1963:112). Nevertheless, I selected the form from Erkertshofen as an
example for a primitive species (see Plate 18b and 19b), because no original materia!
from Viliafeliche was available to me.

another “immigration wave” does not seem very
probable to me, because, as mentioned above, an
origin from the more primitive forms on both sides
of the Atlantic can be demonstrated with some de-
gree of certainty.

2) Parallel evolution? However closely related
Democricetodon and Copemys might be, wheth-
er— as Mein and Freudenthal believe — on the level
of the subfamily at best, or — as Fahibusch and
Lindsay state — on the level of the genus, certain
parallel developments, at least for the evolved
forms, can hardly be denied. To seek the reason for
these parallelisms in the same environmental con-
ditions is not reasonable at present, because the
functional significance and with it the selective val-
ue of these details of the tooth pattern are not
understood.

3) Completely independent development? It is
questionable whether Copemys could have devel-
oped completely independently of Democriceto-
don. Wood {in Clark et al., 1964) supposed Cope-
mys to have arisen from the Leidymys group.
However, Copemys was at that time much less
completely known than at present and its generic
union with Democricetodon had not yet been dis-
cussed. In any case, the possibility can not com-
pletely be excluded that both genera developed in-
dependently since the Oligocene, that Copemys
consequently could have descended from Leidymys
and Democricetodon from an Oligocene cricetid
from Eurasia.

4) Common ancestral form? I incline toward the
opinion that Democricetodon and Copemys de-
scended from a common, probably Asiatic, ances-
tral form, and developed for some time in parallel.
An interpretation of the similarities as pure conver-
gence seems to me to be hardly possible. Even if
we are not fully aware of the reasons leading to
parallelisms, we should in no case underestimate
the extent of similarities and morphological corre-
spondences caused by such parallelisms. As far as
systematic position is concerned, it is appropriate
to leave Democricetodon and Copemys as two dif-
ferent genera. For the solution of this problem it is
not so important in which geological stage the two
lines converge, whether in the Upper Miocene or
at an earlier level. Rather more important to me is
the fact that starting from primitive forms there ap-
peared evolutionary lines on both sides of the At-
lantic, even if they enter upon the same ways. This
fact justifies, in my opinion, a generic separation in

34

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

35

cases like this one, where for the present the dif-
ferent forms can hardly be distinguished.

Plesiosminthus

(Family Zapodidae Coues, 1875)

History of Investigations

The genus Plesiosminthus was introduced by
Viret (1926) for a form from St-Gerand-le-Puy {P.
schauhi). Schaub, having recognized one year prior
the affiliation to the Sicistinae of a tooth formerly
identified as cricetid, described in his monograph
( 1930) on fossil Sicistinae two new species from the
European Miocene and Oligocene (P. myarion from
Chavroches and P. promyarion from Rickenbach).
The first American sicistine was described as
Schaubeumys grangeri by Wood (I935u), but this
form was long regarded as a cricetid. Galbreath
(1953) described a Plesiosminthus-Vike form from
Quarry A (Martin Canyon, Colorado) but had some
doubts about the generic assignment and called the
form Plesiosminthus^ clivosus. Black (1958) had
the same doubts and proposed assigning P.l cli-
vosus to the American genus Schaubeumys as he
did the new species S. sabrae from the Miocene of
Split Rock (Wyoming). Wilson (1960), describing
the fourth American species {S. galbreathi from
Quarry A), considered Schaubeumys as a subgenus
of Plesiosminthus. He united the species S. gran-
geri, sabrae, and galbreathi in the subgenus Schau-
beumys, whereas he incorporated the species cli-
vosus together with the three European species in
the subgenus Plesiosminthus. Likewise he consid-
ered the Mongolian genus Parasminthus Bohlin,
1946, as a subgenus of Plesiosminthus. Finally,
Kllngener ( 1966) described a form from the Valen-
tine Formation of Nebraska (Upper Miocene)
which he assigned to the genus ^Plesiosminthus,
and introduced the giant form Megasminthus ti-
heni.

Comparisons of American and
European Forms

The fact that several species of Plesiosminthus
are known from each continent (three from Europe,
four from America) makes it difficult to establish
the common features of one group and to compare
them with the characteristics of the group of the
other continent. Larger sized teeth are common to
all American forms, compared with the European.
This is not surprising, since most American species
are distinctly younger than the European. How-
ever, Schaubeumys grangeri from the lower Rose-
bud— the earliest American zapodid — is, according
to the current stratigraphic correlations, the same
age or even somewhat older than the middle “ Aqui-
tanian” P. myarion from Chavroches (NMU 2a)
hut still distinctly larger than this one.

As several authors have shown, the forms of both
continents resemble one another morphologically.
In spite of extensive comparisons, I found no gen-
eral morphological differences, but only quantita-
tive ones. Thus, on M, of European species the
ridge extending anteriorly from the mesolophid al-
most always joins the protoconid, whereas in Amer-
ican forms this ridge extends straight anteriorly to-
ward the anteroconid and contacts the ridge
connecting protoconid and metaconid (see Fig. 9a-
g). Only in the type jaw of S. grangeri does M,
appear in this feature like the European forms (Fig.
9f). As an additional difference on M,, the hypo-
conulid and mesoconid of American species gen-
erally seem to be better developed than in Euro-
pean. Likewise the protoconule of M' of American
species seems to be more accentuated than in M'
from Europe, where this cusp is often completely
lacking (see Fig. 9h-l). On M' the “Vorjochkante”
of many American specimens extends from the
paracone to the mesoloph, whereas on all Euro-
pean specimens I compared, the "Vorjochkante"

Fig. 9. — a-h) Plesiosminthu.s myarion Schaub, LM,. NMBS, Chr. 761 and Chr. 759, Chavroches (France), Early Miocene (Aquitanian);
c-d) Plesiosminthus schauhi Viret, LM,. NMBS, Bst. 8838 and Bst. 8840. Coderet (France), Uppermost Oligocene; e) Schaubeumys
sabrae Black, LM,. CM 15854, Split Rock, Fremont Co., Wyoming, Hemingfordian; f) Schaubeumys grangeri Wood, LM,. AMNFI
13757 (invers). (type). Potato Creek, Pine Ridge, South Dakota, Lower Rosebud ( Arikareean); g) Megasminthus tiheni Klingener.
LM|, CM 18855 (invers). Verdigre Quarry, Knox Co. . Nebraska. Valentine Fm. (Late Miocene); h) Plesiosminthus myarion Schaub,
LM'. NMBS. Chr. 778 (invers), Chavroches (France). Early Miocene (Aquitanian); i) Plesiosminthus schauhi Viret, LM', NMBS.
Bst. 690 (invers), Coderet (France), Uppermost Oligocene;)) Schaubeumys sabrae Black. LM'. CM 15854 (invers). Split Rock, Fremont
Co., Wyoming, Flemingfordian; k) Schaubeumys grangeri Wood. LM'. AMNFI 13757, Potato Creek. Pine Ridge, South Dakota. Lower
Rosebud (Arikareean); I) Megasminthus tiheni Klingener. LM'. CM 18864, Verdigre Quarry. Knox Co., Nebraska. Valentine Fm.
(Late Miocene). All figures 25x.

36

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

a b

Fig. 10. — a) Plesiosminthus myarion Schaub, LM'. NMBS, Sau. .^901, Saulcet (France), Early Miocene (Lower Aquitanian); b)
Schauheumys sahnie Black, LM', CM 15854. Split Rock. Lremont Co.. Wyoming. Hemingfordian. Both figures 40x.

joined with the posterior branch of the protocone.
There seems to he a difference in the metacone of
M', and connected with this in the third exterior
syncline — in American forms the metacone is a ro-
tund cusp in occlusal view and on its lingual side
the “Nachjochkante" sets on abruptly. This ex-
tends as a posteriorly bent, delicate ridge to the
hypocone. The metacone of European forms is a
rather elongated cusp, gradually passing into the
hardly bent “Nachjochkante" (see Fig. 10). Final-
ly, shows some differences — European species
show only a wreath of little developed, small cusps,
whereas in American species a relatively distinct
anterior main cusp can be observed, surrounded by
two or three side-cusps (see Plate 20a-b).

Even if none of the differences mentioned alone
furnishes a criterion for distinguishing the groups of
both continents, the total differences seem to be of
some value for systematics. It is very interesting
too that all mentioned features of the American
forms appear again and in a higher degree in the
dentition of the Late Miocene zapodid Megasmin-
thiis tiheni from the Nebraska Valentine Formation
(see Fig. 9g and 1). This fact suggests a close phy-
logenetic connection among American forms.

Discussion

As the above comparisons show, the American
and European zapodids, united by many authors in
the genus Plesiosminthus, are so similar to each
other that — as Wilson (1960:86) pointed out — it is
difficult to find constant differences. Once more the
question arises how these similarities could origi-

nate. Wilson ( 1960:86) explained them by close phy-
logenetic relations, uniting the species of both con-
tinents in the same genus and stating that he was
guided chiefly by two considerations — he found no
distinct morphological differences to separate P.
clivosus, P. schiiubi, and P. myarion’, secondly, P.
clivosus is not the only mammal from Quarry A to
show close relationship with European forms. As
Wilson stated elsewhere (1960:21), the morpholog-
ical similarities are more readily explained by “in-
termigration" than by parallelisms.

At what point of time did this “intermigration"
take place? In a publication on the intercontinental
correlations of the Miocene Wilson (1968) suggests
for Plesiosminthus a “Burdigalian" immigration
from Eurasia. This is an interesting interpretation
considering that most American species first ap-
pear, according to the current correlations, some-
what after the latest European forms. Besides this
the American forms are more advanced than the
European. One fact, however, seems to contradict
a “Burdigalian" immigration — the earliest Ameri-
can species, S. grangeri, is known from the lower
Rosebud, which, according to most correlations, is
pre-“Burdigalian." It seems possible that the Ple-
siosminthus group could have divided very early,
perpaps in the Eocene or the Lower Oligocene and
that the different lines did not develop in their den-
tal morphology very far from the common ancestral
form. It does not seem likely to me that these sim-
ilarities between American and European Plesios-
minthus are the result of pure convergence. The
tooth pattern of this group is without any doubt a

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

37

Fig. 1 1. — The position of the posterior edge of the incisive foramen with regard to the P^. a) Plesiosminthus myurion Schaub, NMBS.
Chr. 797. Chavroches (France); b) Plesiosminthus schuubi Viret, NMBS, Bst. 8856, Coderet (France); c) Plesiosminthus myarion
Schaub. NMBS, Sau. 3899, Saulcet (France), if: posterior edge of the incisive foramen, Ph alveolus of P^. All figures 12x.

very primitive one, and even if parallelism were in-
volved, similarity in these primitive forms (Wilson,
1949:23; Schaub, 1934:24) would indicate much
more for the degree of relationship than in highly
specialized forms, which in certain circumstances
can reflect convergence from different evolutionary
lines.

Simimys from the Sespe Eocene of California,
whose systematic position is not yet clear, possibly
could furnish an indication for early division of the
zapodids. Even if Simimys cannot be considered as
an ancestor of Plesiosminthus because of certain
specializations, it still demonstrates that apparently
in the Eocene in America forms existed that had a
Plesiosminthus-Mke tooth pattern.

Another suggestion for an early division within
the zapodids is furnished by Parasminthus from the
Upper Oligocene of Mongolia (Bohlin, 1946), which
besides many dental features common with the Eu-
ropean Plesiosminthus shows some distinct differ-
ences.’® The more or less contemporaneous occur-
rence of Plesiosminthus and Parasminthus indicates

The most important difference between Parasminthus and Plesiosminthus seems
to be the lack of a groove in the upper incisors of the former. Wood ( 1935/7:225-226),
who considered it possible that there is a single gene governing grooving or absence
of grooving in heteromyid incisors, pointed out that the presence or absence of the
groove is a systematic character of generic rank.

that at least in the Upper Oligocene of Eurasia two
groups of forms existed.

As a final indication for early division within this
group, the great variation undergone by the incisive
foramen should be considered. Whereas in Schau-
beurnys sabrae (Black, 1 958: Fig: lA), Parasmin-
thus tangingoli (Bohlin, 1946: Fig. 2/8), and Plesios-
minthus myarion from Chavroches (see Fig. 11a),
the posterior edge of this foramen is on the level of
the anterior edge of P"*, in S. clivosus and S. gal-
breathi (Wilson, 1960: Figs. 126 and 128) it reaches
the level of mid-P'’, and in P. schaubi, the level of
the anterior edge of M' (see Fig. 1 lb). As far as the
position of this foramen is concerned, there are dis-
tinct differences between the forms from Chavroch-
es and from Saulcet, both assigned to the species
P. myarion by Schaub — in the form from Saulcet
the posterior edge of the foramen reaches the level
of the middle or the posterior edge of P'*; in the
Chavroches form the foramen ends distinctly an-
terior to P"* (Fig. 12a and c).”’ Could it be possible
that this variation indicates the existence of differ-

At first sight it seems possible to explain the different positions of the posterior
edge of the foramen incisivum in the forms from Branssat, Saulcet, and Chavroches
as different phases in one advancing process. Because in such Recent zapodids as
Sicista and Zapus. the foramen incisivum reaches farther back than in the Branssat
form (that is, to the level of the protocone of M'). this possibility becomes less prob-
able.

38

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

Fig. 12. — a) Plesiosminthus myarion Schauh, LMj, NMBS. Sau. 3983. Saulcet (France), Early Miocene (Lower Aquitanian); b)
Eumyarion medius (Lartet), LM2, NMBS, Ss. 6721. Sansan (France). Middle Miocene; c) Plesiosminthus myarion Schauh. LM^.
NMBS. Sau. 3913, Saulcet (France), Early Miocene (Lower Aquitanian); d) Eumyarion hifidus (Fahibusch), LM^. BSPM 1952 XIV 89,
Giggenhausen (Germany), Middle Miocene. All figures 40x.

ent lines, which are not distinguishable on the basis
of the dental morphology?

The tooth type of Plesiosminthus, let us call it
the “cricetoid” one, demonstrates for what long
spaces of time such primitive tooth patterns can be
maintained essentially unchanged and how slightly
significant for systematics it is, as shown even in
the Eocene with the dentition of Simimys, in the
Oligocene with Plesiosminthus , certain cricetids
and eomyids, and in the Upper Miocene, for ex-
ample, in Eumyarion. Certain teeth of Eumyarion
— especially M-, M^, M2, and M;, — are barely dis-
tinguishable from corresponding teeth of Plesios-

minthus (see Fig. 12, notice especially the double
“Vorjochkante” of M^ of both forms).

Summarizing, based on currently known speci-
mens, correspondence of American and European
Plesiosminthus can be explained as well as the re-
sult of an early zapodid division with little further
morphological development as by a Burdigalian fau-
nal exchange. Therefore and because of morpho-
logical differences, I propose to leave the Ameri-
can, Asiatic, and European species in the three
separate genera, until we know more about the phy-
logenetic relations within this group.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

39

DISCUSSION

In the preceding systematic section I tried to ap-
proach the problem of the relations between North
American and European Miocene mammal faunas
from a point of view different from that of most of
my colleagues. At present there is much discussion
of “migrations,” which centers on the question of
which groups passed from one continent to the oth-
er at what time and by which way. I have tried to
show that outside of the “migration theory” —
which actually is still a theory-there are other pos-
sible explanations for the often astonishing similar-
ities of forms between the northern continents. Of
course I am not the first to have envisaged other
possibilities than “migration.” Wilson ( 1960:21),
for example, in discussing insectivores of Europe
and North America, considered parallel evolution
as a possibility besides “migration,” although he
did prefer the latter interpretation. Simpson (1947:
622, footnote) mentioned the possibility of conver-
gence: “It is a conceivable alternative that some
taxonomic groups are moiphological but not phy-
logenetic and that they arose independently on the
two continents. This has been claimed for many
groups, but, if true at all, it must (in my opinion)
be highly exceptional and can be ignored in an over-
all view.”

I agree with Simpson that pure convergence, that
is, development of a similar charcter in groups not
closely related, is very unlikely, at least in the cases
examined here. Nevertheless, this possibility can-
not be excluded completely in forms with very poor
records on both continents. Thus, in the case of the
Miocene “sciuropterans” of California, it cannot
be ruled out that these animals do not belong to the
Sciuridae, and that we could be dealing with pure
convergence. But in the other genera investigated
it can hardly be contested that the forms were re-
lated at least at the family level.

Accordingly, the following alternatives remain to
explain morphological similarity of forms from both
continents:

1 ) the spread of similar forms to the other conti-
nent ( = “migration”);

2) the spread of ancestors of similar looking forms
from one continent to other, with

a) following parallel evolution over greater or
lesser time periods, or

b) further development and conservation of
certain primitive features over greater or lesser pe-
riods.

Common to all these alternatives is occurrence
in each case of intercontinental dispersal. Differ-
ences between the alternatives concern only the
point in phylogenetic history at which these spreads
occurred.

My investigations of the genera treated in the sys-
tematic section suggest that possibilities 2a and 2b
are more plausible than generally supposed.

Migration

The most popular current explanation for mor-
phological similarities of forms from different con-
tinents is that of “migration.” The use of the term
“migration” implies probably that one understands
the possibilities of a spread (what it is actually)
much more extensively and also considers regions
as passable, as long as they were more or less ice-
free and relatively dry. However, using the term
“migration” one has probably automatically in
mind the migration of lemmings and birds of pas-
sage, that is, situations in which the animals are
under a certain pressure and probably pass through
areas not completely corresponding with their eco-
logical needs.

In this connection it is much better to use the
term “spread” which implies that the animals ex-
pand their range and open up new living space, and
under normal conditions, they are subject to no
pressure. (It can happen of course, that, for ex-
ample by overpopulation, a certain pressure can
arise, which can induce the animals to colonize
areas not wholly suitable. However, such a pres-
sure in my opinion would not persist as long as
necessary, to overcome such enormous distances
as in the present case.)

As to Neogene faunal exchange, I have some
doubt with regard to the forms studied. Most cur-
rent authors consider the Bering land bridge as an
ideal spread route for land mammals (see Beringia
History, 1973). There is evidence from several
fields (paleobotany, malacology, sedimentology)
that during the Tertiary there were repeated land
connections across the Bering Strait. Investigations
of large mammals especially suggest that this land
bridge was utilized again and again (see H. Tobien.
L. K. Gabunia, 1. M. Novodvorskaya, N. M. Ya-
novskaya, and others, in Beringia History, 1973).
Recent investigations show that during the Neogene
certainly not a tropical or subtropical climate pre-
vailed in the Bering area, but a best a temperate

40

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

one. This is compatible with the view of S. F. Biske
and Yu. P. Baranova (/// Beringia History, 1973)
that in the Oligocene and Miocene a change from
a warm temperate to a cold temperate climate took
place. In the Asian part of Beringia, a forest tundra
appeared in the Middle Miocene and an alpine tun-
dra was characteristic of the close of the Miocene.
According to these and other investigations, essen-
tially coinciding with each other, 1 see some diffi-
culties for dwellers of tropical or subtropical forest
areas, as many of the small mammal forms studied
probably are, to colonize and pass through areas
like Beringia. Simpson (1947) took into consider-
ation the possibility that the spread of certain forms
undergoes a delay by such climatic and ecologic
barriers, that is, until a form adapted to the envi-
ronment in question is developed. This is accept-
able, but logically one should go beyond this, sup-
posing that the "migrants” redeveloped a form
adapted to the colonization of the warm areas in
Middle Europe and North America. This seems
very unlikely to me. With regard to the spread of
tropical or subtropical animals, Simpson (1947:651)
takes the view: "In any event, and contrary to what
seems to be a widespread impression (e.g.,
J. W. Gregory, 1930), there is no evidence from the
mammals that any truly tropical or subtropical an-
imals ever migrated between Eurasia and North
America and it is even dubious whether any spe-
cifically warm temperate groups made this jour-
ney.” Thus we have to determine whether the
forms studied here were actually dwellers of such
warm climates. The ecologic and climatic ties of
most of the 10 genera are not clear for the present.
However, if conclusions from modern faunas are
permissible, at least the fossil sciuropterans and
Lanthanotherium, having Recent relatives mainly
inhabiting tropical forest areas, may have lived un-
der similar conditions. In addition we know that in
the Miocene of Middle Europe there was a distinct-
ly warmer climate than today and the same is ap-
plicable to the corresponding areas of North Amer-
ica (for example, the "Gray-beds” of Cuyama
Valley in California, where four of the 10 genera
are found). Very probably, Lanthanotherium and
the sciuropterans were forest dwellers too, possibly
also Plesiosorex and the Heterosoricinae, for their
remains are mainly found together with typical for-
est faunas, and partly in lignites. In localities with
typical steppe faunas, they are either very rare or
are completely lacking. For such dwellers of warm
forest biotopes the Bering area with its tundras cer-

tainly represented a virtually unsurmountable bar-
rier, at least during the Neogene. However, this is
not incompatible with inhabitants of temperate or
cold climates having crossed the Bering area.

A further weak point in the theory of the Late
Tertiary faunal exchange across Beringia is that this
theory, at least as far as small mammals are con-
cerned, is not strengthened by any evidence from
eastern Asia. Certainly, there are well-documented
Oligocene faunas known from the Hsanda Gol For-
mation in Mongolia (Matthew and Granger, 1923;
Mellett, 1968) and from Taben-buluk in Kansu
(Bohlin, 1942-1946). These faunas contain forms
related to European and North American ones (for
example, Cricetops, Parasminthiis, Eumys, Am-
phechinus, Cylindrodontidae) but they do not fur-
nish any indication of an intercontinental faunal ex-
change, for they are largely endemic. It may be
argued that if migrations occurred, the migrants
may have passed south or north of Mongolia. In my
opinion, for climatic and ecological reasons the
southern route may have been more advantageous
for Oligocene and Miocene animals. If we want di-
rect mammal paleontological evidence for such ex-
change, we have to look for Oligocene and Miocene
mammals in southern Asia — in India, Burma,
southern China, or perhaps in Indonesia. However,
assuming such a southern distribution route, it be-
comes difficult to explain how the animals could
pass over a land bridge situated so far north as the
Bering bridge!

Doubts with respect to the Late Tertiary spread
of mammals across Beringia are also based on the
lack of fossil mammal remains in northern areas
such as Siberia, Alaska, and northern Canada. As
already mentioned, 1 cannot visualize such a spread
as unidirectional, therefore, 1 suppose that animals
able to pass over a land bridge as far north as the
Bering bridge would have been living also in other
northern areas.

No fewer problems arise assuming a post-Eocene
faunal exchange over a northern Atlantic bridge.
Because we have new evidence from continental
drift, this theory has lost ground, although some
authors still prefer the route across the northern
Atlantic to that across the Bering Strait. Recently
Strauch (1970) and Zhegallo (Beringia History,
1973) pleaded for the transatlantic route. In my
opinion, this theory has the same handicaps as that
of the spread over the Bering bridge — a northern
Atlantic bridge would be equally situated very far
north, and also from this area any fossil evidence

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

41

is lacking. Furthermore, according to the newest
paleomagnetic evidence, a direct land connection
between Europe and North America after the Low-
er Eocene (from about 47 million years ago) is very
doubtful.

The doubts advanced so far regarding intercon-
tinental faunal exchange referred mainly to the cli-
mate and ecology of the distribution routes. 1 have
also some doubts concerning the methods applied
in mammal paleontology — the assumption of close
relation of forms of both continents and the conclu-
sions deduced from this that “migrations” have oc-
curred are based above all on comparisons of den-
titions. In my own investigations, several examples
demonstrated that for determining phylogenetic re-
lations, the dentition alone cannot be trusted ab-
solutely. Of course, completely new problems arise
here, because many fossil mammals are known by
their dentition alone. As shown above, parallelism
and the conservation of primitive features over a
long period must considered. Therefore, the for-
merly generally used dictum — “What is not distin-
guishable morphologically is closely related” —
must be questioned. This principle may be valid in
cases in which we can compare more complete re-
mains, such as skulls or skeletons, but has to be
applied with extreme caution if we do not have any-
thing but teeth to compare. As in the case of "Co-
timus" close phylogenetic relations often have to
be disproved when more complete remains become
known. Therefore it is indispensable that in paleon-
tology of mammals, and especially of small mam-
mals, we try to compare other structures besides
the dentition.

Especially in earlier literature, the possibility of
rafting was considered (Schlosser, 1911; recently
also Lavocat, 1973). A typical example of this kind
of spread was the recolonization of the volcanic is-
land Krakatau. Rafting may be effective for cross-
ing of an arm of the sea or from one island to
another. For the spread from one continent to
another, this possibility can be dropped, in my
opinion, because the chance that a single individual
survives such a long transport is extremely small,
not to mention the number of individuals necessary
to build up a population.

Parallel Evolution

To show parallel evolution conclusively is in
most cases very difficult, and extensive fossil ma-
terial should be available. But as many animals
show, parallelism is not so rare and in many cases

can produce astonishing correspondence of form.
If morphological similarities can result from con-
vergence in not-related forms, logically even more
surprising similarities can be developed in related
forms. As Wood (1947) stated convincingly, the
chance of parallel evolution is greater in groups
having specialized very early — for example, ro-
dents or the Heterosoricinae — than in less special-
ized groups. With specialization, the spectrum of
evolutionary possibilities is restricted, and the
probability increases that the same development
occurs two or more times.

Subsequent Adaptation and Trends

It is commonly assumed that identical or similar
environmental conditions requiring a similar adap-
tation are important for the occurrence of parallel-
ism. So, for example, the increasing hypsodonty in
cricetids of the North American as well as of the
European Upper Miocene and Pliocene, is probably
connected with spread of the steppe environment
in these areas. Such functional adaptations are not
difficult to recognize as parallelism in most cases,
for example, the jumping mouse type within the ro-
dents (see Thenius, l972:Fig. 3) or the fossorial
type of rodents, insectivores, and marsupials.

When working on this paper. 1 was also faced
with cases that I interpret as parallelism and in
which the functional significance of a feature oc-
curring in different groups cannot be understood
easily. I found such strange phenomena in Demo-
cricetodon and Copeniys and in Plesiosorex . One
might ask again, are such parallelisms traceable to
evolutionary trends? (“Trends” in the sense that
the possibility for evolution in a certain direction
can be hereditary.) I am aware of the danger of
being branded as a partisan of an orthogenesis-the-
ory. However, I do not intend to interpret any di-
rection into the evolutionary process. I only want
to consider the possibility of evolutionary trends to
explain such cryptic parallelism in related groups,
for the sake of discussion.

Conservation of Primitive Features

It is conspicuous that most of the small mammals
from Europe and America studied here, which are
usually placed in the same genus, have a primitive
tooth pattern. In this category belong especially the
bunodont eomyids, Cotimus, Eumyarion, the sciu-
ropterans, Lanthanotherium, the Heterosoricinae,
and Plesiosorex. In some of these cases the possi-
bility of a “migration” cannot be completely ex-

42

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

eluded, but it seems more probable that the simi-
larities between the pairs from the two continents
result from conservation of certain primitive fea-
tures. The group of Eomys, Leptodontornys, Pro-
tadjidaumo, and Eomyops in which the same tooth
type was conserved from the Upper Eocene to the
Pleistocene, or that of Leidymys, Eumys, and Eu-
myarion, in which the time range of this tooth pat-
tern goes from the Middle Oligocene to the Upper
Miocene, demonstrate over what long time periods
a tooth pattern can be retained basically unchanged.
In my opinion this possibility of long-range conser-
vation of a tooth pattern is underestimated by col-
leagues, who consider as “immigrants" forms
“suddenly" appearing in a layer without immedi-
ately earlier ancestral forms being known. This
view overestimates very much the completeness of
the available fossil record. As innumerable exam-
ples demonstrate, the fossil documentation of most
mammal groups is incomplete. In addition, it should
be considered that for animals that were rare in
their lifetime or that lived in an area unsuitable for
fossilization, there is scarcely a chance that their
remains are ever found.

If one attempts to explain the similarities of Neo-
gene mammals of North America and Europe as 1

did in the preceding part by parallel evolution or by
conservation of primitive features, one also has to
consider how the ancestors of these similar forms
managed the trip from one continent to the other.
Supposing an Early Tertiary faunal exchange some
difficulties arise because according to present gen-
erally accepted opinion such an exchange across
the Atlantic was not possible after the Early Eocene
(McKenna, 1973). Because Europe and Asia were
separated by the Turgai Straits marine barrier
(McKenna, 1975) until the beginning of the Oligo-
cene, and Eocene faunal exchange between North
America and Europe via Asia was also hardly pos-
sible. However, it is conceivable that forms having
reached Asia during the Eocene got to Europe after
the disappearance of the Turgai barrier. Because
we know very little at present about Tertiary faunas
from Asia, it is also possible to suppose that some
of the groups studied here had their center of evo-
lution in Asia and spread from there to Europe and
North America. There is nothing to object to an
Early Tertiary spread across the Bering Strait be-
cause it can be considered certain that the Early
Tertiary climate of the Bering area was much warm-
er than in the Neogene.

CONCLUSIONS

Investigation of 10 small mammal genera osten-
sibly occurring in North America as well as in Eu-
rope demonstrates that generic equivalence is very
questionable, and that in three cases (Limnoecus,
Leptodontornys, "Cotimiis") is apparently incor-
rect, because important morphological differences
between American and European forms can be
found. In other cases these generic unions could

not be disproved, but the insufficient documenta-
tion renders them dubious. Dental remains alone
are often inadequate for illuminating phylogenetic
relationships. Eurther, it becomes apparent that not
all similarities between forms of both continents can
be explained by “migrations” and that other pos-
sibilities, especially parallelism and conservatism,
are more important than usually assumed.

LITERATURE CITED

Baudelot. S. 1972. Etude des Chiropteres, Insectivores et
Rongeurs du Miocene de Sansan (Gers). These Univ. Paul
Sabatier Toulouse, 364 pp.

Beringia History. 1973. The Bering Land Bridge and its role
for the history of Holarctic floras and faunas in the Late
Cenozoic. Theses of the reports of All-Union Symposium.
Acad. Sci. U.S.S.R. Khabarovsk.

Black, C. C. 1958. A new sicistine rodent from the Miocene
of Wyoming. Breviora, 86:1-7.

. 1961. Fossil mammals from Montana. Part I . Additions

to the Late Miocene Flint Creek local fauna. Ann. Carnegie
Mus., 36:69-76.

. 1965. Fossil mammals from Montana. Part 2. Rodents

from the Early Oligocene Pipestone Springs local fauna.
Ann. Carnegie Mus.. 38: 1-^8.

Blainville, H. M. Ducrotay de. 1839-1864. Osteographie
des Mammiferes. Tome I, Livraison 6-H. Insectivores.
Paris, 1 15 pp.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

43

Bohlin, B. 1946. The fossil mammals from the Tertiary deposit
of Tabenbuluk. western Kansu. Part II. Paleont. Sinica (new
ser. C). 8b: 1-259.

Butler, P. M. 1948. On the evolution of the skull and teeth in
the Erinaceidae. Proc. Zool. Soc. London, 118:446-500.

. 1969. Insectivores and bats from the Miocene of East

Africa: new material. Pp. 1-37, in Fossil vertebrates of Af-
rica (L. S. B. Leakey, ed.), Academic Press, New York and
London, Fix -i- 1-102.

Clark, J. B., M. R. Dawson, and A. E. Wood. 1964. Fossil
mammals from the Lower Pliocene of Fish Lake Valley,
Nevada. Bull. Mus. Comp. Zool., Flarvard Univ., 131:27-
63.

Cope, E. D. 1873. Third notice of extinct vertebrata from the
Tertiary of the plains. Paleont. Bull. (Philadelphia), 16:1-8.

Crochet, J. Y. 1968. Revision des Marsupiaux et des Insec-
tivores des Phosphorites du Quercy. These 3' cycle, Fac.
Sci. Paris, 119 pp. (not printed).

Crusafont, M., j. F. de Villalta, and J. Truyols. 1955.
El Burdigalense continental de la cuenca del Valles-
Penedes. Diputacion provincial de Barcelona. Mem. Co-
mun. Inst. Geologico, 12:1-272.

Dietrich, W. O. 1929. Referat iiber E. Stromers " Wirbeltiere
im obermiocanen Flinz Miinchens.” Neues Jahrb. f. Min-
eral. etc. 1929, 3:776-778.

Doben-Florin, U. 1964. Die Spitzmause aus dem alt-Burdi-
galium von Wintershof-West bei Eichstatt in Bayern. Bayer.
Akad. Wiss. Abh. Neue Folge, 117:1-82.

Emry, R. j. 1972. A new heteromyid rodent from the Early
Oligocene of Natrona County. Wyoming. Proc. Biol. Soc.
Washington, 85:179-190.

Engesser, B. 1972. Die obermiozane Saugetierfauna von AnwiI
(Baselland). Tatigkeitsber. der naturf. Ges. Baseband,
28:37-363.

. 1975. Revision der europaischen Heterosoricinae (In-

sectivora. Mammalia). Eclogae geol. Helv.. 68:649-672.

Fahlbusch, V. 1964. Die Cricetiden (Mamm.) der Oberen Siis-
swasser-Molasse Bayerns. Akad. Wiss. Math.-nat. Kl. Abh.
N.F., 118:1-134.

. 1967. Die Beziehungen zwischen einigen Cricetiden

(Mamm. Rodentia) des nordamerikanischen und euro-
paischen Jungtertiars. Palaeont. Z., 41:154-164.

. 1968. Neue Eomyiden (Rodentia, Mamm.) aus einer

aquitanen Spaltenfiillung von Wissenburg in Bayern. Mitt.
Bayer. Staatssaml. Palaont. Flist. Geol., 8:219-245.

. 1969. Pseudotheridomys pusillus n. sp. ein neuer

Eomyide (Rodentia, Mamm.) aus dem Oligozan Siid-
deutschlands. N. Jb. Geol. Palaont. Mn., 1969:673-679.

. 1970. Populationsverschiebungen bei tertiaren Nageti-

eren, eine Studie an oligozanen und miozanen Eomyidae
Europas. Abh. Bayer. Akad. Wiss. Math-naturwiss. Kl. N.
F., 145:1-136.

. 1973. Die stammesgeschichtlichen Beziehungen

zwischen den Eomyiden (Mammalia. Rodentia) Nordamer-
ikas und Europas. Mitt. Bayer. Staatssamml. Palaont. Flist.
Geol., 11:141-175.

Filhol, H. 1884. Note sur une nouvelle espece dTnsectivore
du genre Amphisorex. Bull. Soc. Philomathique de Paris
(7) 8, pp. 63-64.

. 1888. Description de quelques mammiferes nouveaux

trouves a Sansan. Bull. Soc. Philom. vol. 7.

Freudenthal, M. 1963. Entwicklungsstufen der miozanen

Cricetodontinae (Mammalia, Rodentia) Mittelspaniens und
ihre stratigraphische Bedeutung. Beaufortia, Zool. Mus.
Amsterdam, 10:5 1-157.

. 1965. Betrachtungen iiber die Gattung Cricetodon.

Koninkl. Nederl. Akad. Wetensch. Amsterdam, Proc.. Ser.
B., 68:293-305.

Gaillard, C. 1915. Nouveau genre de musaraignes dans les
depots miocenes de La Grive-Saint-Alban (Isere). Ann. Soc.
Linneenne de Lyon, 62:83-98.

Galbreath, E. C. 1953. A contribution to the Tertiary geology
and paleontology of north-eastern Colorado. Univ. Kansas
Paleont. Contrib., Vertebrata, 5: 1-120.

. 1955. A new eomyid rodent from the lower Oligocene

of north-eastern Colorado. Trans. Kansas Acad. Sci.,
58:75-78.

Gorgas, M. 1967. Vergleichend-anatomische Untersuchungen
am Magen- Darm- Kanal der Sciuromorpha, Hystricomorpha
und Caviomorpha (Rodentia). Z. Wiss. Zool., Leipzig,
175:237-404.

Gregory, J. W. 1930. The geological history of the Pacific
Ocean. Geol. Soc. London, Quart. J., 86:72-136.

FIall, E. R. 1929. A second new genus of hedgehog from the
Pliocene of Nevada. Univ. California Publ., Bull. Dept.
Geol. Sci., 18:227-231.

Harris, J. M., AND A. E. Wood. 1969. A new genus of eomyid
rodent from the Oligocene Ash Springs local fauna of Trans-
Pecos Texas. Pearce-Sellards Series, Texas Mem, Mus.,
14:1-7.

Hartenberger, J.-L. 1966. Les rongeurs due Valesien (Mio-
cene superieur) de Can Llobateres (Sabadell, Espagne):
Gliridae et Eomyidae. Bull. Soc. Geol. France (7), 8:596-
604.

Hugueney, M., and P. Mein. 1968. Les Eomyides (Mam-
malia, Rodentia) neogenes de la region lyonnaise. Geobios,
1:187-203.

Hutchison, J. H. 1966. Notes on some upper Miocene shrews
from Oregon. Bull. Oregon Univ. Mus. Nat. Hist., 2: 1-23.

. 1972. Review of the Insectivora from the early Miocene

Sharps Formation of South Dakota. Contrib. Sci. Nat. Hist.
Mus. Los Angeles, 235:1-16.

James, G. T. 1963. Paleontology and nonmarine Stratigraphy
of the Cuyama Valley Badlands, California. Univ. California
Publ., Geol. Sci., 45:1-154.

Klingener, D. 1966. Dipodoid Rodents from the Valentine
Formation of Nebraska. Occ. Papers Mus. Zool, Univ.
Michigan, 644:1-9.

Kretzoi, M. 1962. Fauna und Faunenhorizont von Csarnota.
A. Magy. All. Foldt Intez. Evi Jelent. Budapest, pp. 344-
382.

Lartet, E. 1851. Notice sur la colline de Sansan. Annuaire
Dep. Gers. 185 1 Auch. 47 pp.

Lavocat, R. 1951. Revision de la faune des mammiferes oli-
gocenes d’Auvergne et du Velay. Sciences et Avenir, Paris,
153 pp.

. 1973. Les Rongeurs du Miocene d’Afrique orientale.

Mem. Trav. Inst. Montpellier, 1 : 1-284.

Lindsay. E. H. 1972. Small mammal fossils from the Barstow
Formation, California. Univ. California Publ. Geol, Sci.,
93:1-104.

Macdonald, j. R. 1963. The Miocene faunas from the Wound-
ed Knee area of western South Dakota. Bull. Amer. Mus.
Nat. Hist., 125:139-238.

44

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

. 1970, Review of the Miocene Wounded Knee faunas of

southwestern South Dakota. Bull. Nat. Hist. Mus. Los An-
geles Co. Mus., 8:1-82.

McDowell. S. B. 1958. The greater Antillean insectivores.
Bull. Amer. Mus. Nat. Hist., 115:117-214.

McGrew, P. 1941. The Aplodontoidea. Field Mus. Nat. Hist..
Chicago, Geol. Ser., 9:1-30.

McKenna, M. C. 1973. Sweepstakes, Filters, Corridors,
Noah’s arks and beached Viking funeral ships in Paleogeog-
raphy. Pp. 295-308, in Implications of continental drift to
the earth sciences, Vol. I (D. H. Tarling and S. K. Runcorn,
eds.). Academic Press, London and New York. 622 pp.

. 1975. Fossil mammals and early Eocene North Atlantic

land continuity, Ann. Missouri Bot. Garden, 62:335-353.

Matthew. W. D., AND W. Granger. 1923. Nine new rodents
from the Oligocene of Mongolia. Amer. Mus. Novit., 102:1-
10.

Mawby, J. E. I960. A new occurrence of Heterosorex Gaillard.
J. Paleont., 34:950-956.

Major, E. 1893. On some Miocene squirrels, with remarks on
the dentition and classification of the Sciurinae. Proc. Zool.
Soc. London, pp. 179-215.

Mein, P. 1970. Les Sciuropteres (Mamm., Rod.) Neogenes
d’Europe occidentale. Geobios. 3(3): 1-53.

. 1976. Biozonation du Neogene mediterraneen a partir

des Mammiferes. Proc. 6th Neogene Congr., Bratislava,
September 1975.

Mein. P., and M. Ereudenthal. 1971. Une nouvelle classi-
fication des Cricetidae (Mammalia, Rodentia) du Tertiaire
de I’Europe. Scripta Geologica, 2:1-37 (Leiden. 1971).

Mellett, j. S. 1968. The Oligocene Hsanda Gol Eormation,
Mongolia: a revised faunal list. Amer. Mus. Novit., 2318:1-
16.

Palmowski, j.. and H. Wachendorf. 1966. Eine unteroli-
gozane Wirbeltierfauna aus einer Spaltenfiillung in Herrl-
ingen/Blau (Wiirtt). Mitt. Bayer. Staatssamml. Palaont.
Hist. Geol., 6:229-245.

Patterson, B.. and P. O. McGrew. 1937. A soricid and two
erinaceids from the White River Oligocene. Field Mus. Nat.
Hist., Geol. Ser., 6:245-272.

PoMEL, A. 1854. Catalogue methodique et descriptif des ver-
tebres fossiles. Paris 1854.

Rensberger, j. M. 1975. Haplornys and its bearing on the or-
igin of the Aplodontoid rodents. J. Mamm.. 56:1-14.

Repenning. C. A. 1967. Subfamilies and genera of the Soric-
idae. Geol. Survey Prof. Paper, 565:1-74.

Romer, a. S. 1949. The vertebrate body. Saunders Co,, Phila-
delphia and London, 643 pp.

ScHAUB. S. 1925. Die hamsterartigen Nagetiere des Tertiars
und ihre lebenden Verwandten. Abh. Schweiz. Palaont.
Ges., 45:3-112.

. 1930. Fossile Sicistinae. Eclogue Geol. Helv., 23:616-

637.

. 1934. Ueber einige fossile Simplicidentaten aus China

und der Mongolei. Abh. Schweiz. Palaont. Ges., 54:1-40.

ScHLOSSER. M. 1884. Die Nager des europaischen Tertiars
nebst Betrachtungen fiber die Organisation und die ges-
chichtliche Entwicklung der Nager fiberhaupt. Palaeonto-
graphica, 31: 1-143.

. 1911. Beitrage zur Kenntnis der oligozanen Landsau-

getiere aus dem Eayum: Aegypten. Beitr. Palaont. u. Geol.
Oesterr.-Ung. u. des. Orients, 24:1-167.

. 1926. Die Saugetierfauna von Peublanc (Dep. Allier).

Soc. Sci. Nat. Croatica 38/39 (Kramberger-Festband), pp.
372-394.

Seemann, 1. 1938. Die Insektenfresser, Fledermause und Na-
ger aus der obermiozanen Braunkohle von Viehhausen bei
Regensburg. Palaeontographica, 89, Abt. A., 55 pp.

Shotwell, j. a. 1956. Hemphilian mammalian assemblages
from north-eastern Oregon. Bull. Geol. Soc. Amer., 67:717-
738.

. I967u. Late Tertiary geomyoid rodents of Oregon. Bull.

Mus. Nat. Hist., Univ. Oregon, 9:1-51.

. 19676. Peromyscus of the late Tertiary in Oregon. Bull.

Mus. Nat. Hist.. Univ. Oregon, 5:1-35.

. 1968. Miocene mammals of southeast Oregon. Bull.

Mus. Nat. Hist., Univ. Oregon, 14:1-67.

Simpson. G. G. 1947. Holarctic mammalian faunas and conti-
nental relationships during the Cenozoic. Bull. Geol. Soc.
America, 58:613-688.

Stehlin, H. G. 1940. Zur Stammesgeschichte der Soriciden.
Eclogae Geol. Helv., 33:298-306.

Stehlin. H. G., AND S. ScHAUB. 1951. Die Trigonodontie der
simplicidentaten Nager. Schweiz. Palaeont. Abh., 67:1-385.

Stock. C. 1935. New genus of rodent from the Sespe Eocene.
Bull. Geol. Soc. Amer., 46:61-68.

Strauch, F. 1970. Die Thule-Landbrucke als Wanderweg und
Faunenscheide zwischen Atlantik und Skandik im Tertiar.
Geol. Rundschau, 60(1): 381-417.

Stromer, E. 1940. Die jungtertiare Eauna des Flinzes und des
Schweiss-Sandes von Mfinchen. Abh. Bayer. Akad. Wiss.
math.-naturw. Abt. Neue Folge, 48:1-102.

Thaler, L. 1966. Les rongeurs fossiles du Bas-Languedoc.
Mem. Mus. Nat. Hist. Nat. serie C, Sc. de la Terre, 17:1-
296.

Thenius, E. 1949. Zur Revision der Insektivoren des steiris-
chen Tertiars. Sitzungsber. Oesterr. Akad. Wiss., math.-
naturw. Kl. Abt 1, 158. Bd. 9. u. 10. Heft. pp. 671-693.

. 1969. Stammesgeschichte der Saugetiere (einschlies-

slich der Hominiden). Handb. Zool., Berlin, 8/2:1-722.

. 1972. Grundzfige der Verbreitungsgeschichte der Sau-
getiere. Eine historische Tiergeographie. Gustav Fischer
Stuttgart, 345 pp.

Trouessart, E.-L. 1909. /Veorraci« a new insectivore

of the Family Erinaceidae. Ann. Mag. Nat. Hist., serv. 8,
4:389-391.

Van Valen, L. 1967. New Paleocene insectivores and insec-
tivore classification. Bull. Amer. Mus. Nat. Hist. 135:217-
284.

Vianey-Liaud, M. 1972. Contribution a I'etude des Cricetides
oligocenes d'Europe occidentale. Palaeovertebrata. 5:1-43.

ViLLALTA, J. E., AND M. Crusafont. 1944. Nuevos insecti-
voros del Mioceno continental del Valles-Panades. Notas
Commun. Inst. Geol. Min. Espafia, 12:1-26.

Viret, j. 1926. Nouvelles observations relatives a la faune de
Rongeurs de Saint-Gerand-le-Puy. C.-R. Acad. Sc., 183:71-
72.

. 1940. Etude sur quelques Erinaceides fossiles (suite)

genres Plesiosorex, Lanthanotherium. Trav. Lab. Geol.
Eac. Sc. Lyon Ease., 39(28):33-70.

. 1946. Sur un nouvel exemplaire de Plesiosorex sorici-

noides Biainv. des argiles stampiennes de Marseille-St-
Andre. Eclogae Geol. Helv., 39:314-317.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

45

ViRET, J., AND H. Zapfe. 1951. Sur quelques Soricides mio-
cenes. Eclogae Geol. Helv., 44;41 1—426.

Webb, S. D. 1961. The first American record of Lanthanothe-
rium Filhol. J. Paleont., 35:1085-1087.

Wilson. R. W. 1949. Additional Eocene rodent material from
southern California. Carnegie Inst. Washington Publ.,
584:1-25.

— . 1960. Early Miocene rodents and insectivores from

Colorado. Univ. Kansas Paleont. Contrih., Vertebrata, 7: 1-
93.

. 1968. Insectivores, rodents and intercontinental cor-
relation of the Miocene. 23rd Int. Geol. Congr., 10:19-25.

Wood, A. E. 1935a. Two new genera of cricetid rodents from

the Miocene of western United States. Amer. Mus. Novit.,
789:1-3.

— . 1935^7. Evolution and relationship of the heteromyid
rodents. Ann. Carnegie Mus., 24:73-262.

— . 1936. The cricetid rodents described by Leidy and Cope
from the Tertiary of North America. Amer. Mus. Novit.,
822:1-8.

— . 1937. The mammalian fauna of the White River Oli-
gocene. Part 2. Rodentia. Trans. Amer. Philos. Soc., new
ser., 28:155-269.

— . 1947. Rodents— a study in evolution. Evolution, 1: 154-
162,

— , 1962. The early Tertiary rodents of the family Para-
myidae. Trans. Amer. Philos. Soc., new ser., 53:1-261.

46

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

CAPTIONS FOR PLATES

Plate 1. — a) Lanthanotheriiim sansaniense Filhol; LM'-M^, M': NMBS, Ss. 846, M*: NMBS, Ss. 862, M^;
NMBS, Ss. 6723; Sansan, France; Middle Miocene, b) Lanthanotherium sawini James; LM‘-M®, M‘: UCMP
50122, M“: UCMP 66779, M^: UCMP 54600 (invers); loc. V-5847, Big Cat Quarry, Ventura Co., California
(M' and M^); loc. V-5656, Hedgehog locality (M^); Early Clarendonian. c) Lanthanotherium sanmiqueli
Villalta & Crusafont; LM'-M^, M': NMBS, C. LI. 25, M“: NMBS, C. LI. 26 (invers); M^: Coll. E. Heiz-
mann, Stuttgart B 12; Can Llobateres, Spain; Late Miocene, d) Hylomys suillus dorsalis Thomas; LM'-M®,
NMBS 8892; Northern Borneo; Recent. All figures 15 x.

Plate 2. — a) Lanthanotherium sansaniense Filhol; LM,-M3, NMBS, Ss. 6727-29; Sansan, France; Middle
Miocene, b) Lanthanotherium sawini James; LM2-M3 (invers) UCMP 54594; loc. V-5666, Kent Quarry,
Cuyama Valley, California; Barstovian. c) Ocajila makpiyahe Macdonald; LM2-M3, SDSM 56105; loc.
V-5360, Wounded Knee Area, South Dakota; Sharps Em. (Early Arikareean). d) Lanthanotherium sanmi-
gueli Villalta & Crusafont; LM,-M3, NMBS, C. LI. 28-30; Can Llobateres, Spain; Late Miocene, e) Hy-
lomys suillus dorsalis Thomas; LMj-Mj, NMBS 8892; Northern Borneo; Recent. All figures 15x.

Plate 3. — a) Meterix sp.; LP4-M2 (invers), UNSM 1410-47, Ft-40; Frontier, Nebraska; Kimball Em. (Plio-
cene). b) Plesiosorex coloradensis Wilson; LP4-M1, KU 9989 (type); Quarry A, Martin Canyon, Colorado;
Hemingfordian. c) Plesiosorex schaffneri Engesser; LP4-M3 (invers), NMBS, Al. 149 (type); Anwil, Swit-
zerland; Middle Miocene, d) Plesiosorex cf. soricinoides (Blainville); LP3-M3, FSL 4428; Chaveroche,
France; Early Miocene (Middle Aquitanian). All figures 15x.

Plate 4. — a) Meterix sp.; LP4-M2 (invers), UNSM 1410-47, Ft-40; Frontier, Nebraska; Kimball Em. (Plio-
cene). b) Plesiosorex coloradensis Wilson; LP4-M,, KU 9989 (type) and LM2 (invers) KU 10001; Quarry
A, Martin Canyon, Colorado; Hemingfordian. c) Plesiosorex schaffneri Engesser; LM, -M3 (invers), NMBS,
Al. 149 (type); Anwil, Switzerland; Middle Miocene, d) Plesiosorex cf. soricinoides (Blainville); LP3-M3,
FSL 4428; Chaveroche, France; Early Miocene (Middle Aquitanian). All figures 15x.

Plate 5. — a) Plesiosorex styriacus (Hoffman); LM,, NMBS, OSM 312 (invers); Riimikon, Canton of Zurich,
Switzerland; Middle Miocene, b) Plesiosorex sp.; LM,, BSPM, 1926 I 81 (invers); Grosslappen, Germany;
Middle Miocene, c) Plesiosorex schaffneri Engesser; LM', NMBS, Al. 145 (invers); Anwil, Canton of
Baselland, Switzerland; Middle Miocene, d) Plesiosorex coloradensis Wilson; LM', KU 9990 (invers); Quar-
ry A, Martin Canyon, Colorado; Hemingfordian. e) Meterix sp.; LM', OU 22326 (invers); Ove, Quartz
Basin, Oregon; Barstovian. All figures 15 x.

Plate 6. — a) Dinosorex sansaniensis (Lartet); LP‘*-M'^, NMBS, P'': Ss. 901 (invers), M': Ss. 6687, M^: Ss.
899, M'^: Ss. 6724; Sansan, France; Middle Miocene, b) Pseudotrimylus roperi (Wilson); LP''-M'^, P^: KU
10013, M'-M'^: KU 10018, M": KU 10016; Quarry A, Martin Canyon, Colorado; Hemingfordian. c) Domnina
sp.; LM'-M'^ (invers), CM 10853; South of Reva Pass, Slim Buttes, Harding Co., South Dakota; Oligocene.
All figures 20x .

Plate 7. — a) Pseudotrimylus roperi (Wilson); LM,-M3, KU 10020; Quarry A, Martin Canyon, Colorado;
Hemingfordian. b) Dinosorex zcipfei Engesser; LM,-M3, (type), NHW, 1975/1712/1; Neudorf a.d. March,
CSSR; Middle Miocene, c) Dinosorex huerzeleri Engesser; LM,-M3, (type), MO, Rb. 101; Rickenbach,
Canton Solothurn, Switzerland; Late Oligocene. d) Pseudotrimylus sp.; LM,-M2, CM 10922; Lawson Ranch,
Goshen Co., Wyoming; Lower Orellan(?). e) Domnina sp.; LM,-M3, CM 10403; Pipestone Springs, Jef-
ferson Co., Montana; Chadronian. f) Quercysorex primaevus (Filhol); LM, -M2 (invers), (type), MHNP, Qu.
8681; Lamandine (Quercy); Middle Oligocene)?). All figures 20x.

Plate 8. — a) Dinosorex sansaniensis (Lartet); L mandible with I, M, -M3, NMBS, Ss. 603; Sansan, France;
Middle Miocene, b) Pseudotrimylus roperi (Wilson); L mandible with I, M,-M2, KU 10030; Quarry A,
Martin Canyon, Colorado; Hemingfordian. c) Pseudotrimylus roperi (Wilson); L mandible with M,-M3, KU
10020; Quarry A, Martin Canyon, Colorado; Hemingfordian. All figures 8x.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

47

Plate 9. — a) Petauristodon mathewsi (James); LM'-M^, UCMP 55514; loc. 5847, Cuyama Valley, California;
Early Clarendonian. b) Prosciums sp.; LP'*-M^, CM 9316; North of Pipestone Springs, Montana; Chadron-
ian. c) Forsythia gaudryi (Gaillard); LM''^, FSL 65433; La Grive, France; Middle Miocene; LM''“, NMBS,
Al. 225; AnwiI, Switzerland; Middle Miocene, d) Miopetaurista albanensis (Major); LP^-M', NMBS, Al.
215-217; AnwiI, Switzerland; Middle Miocene. All figures 15x.

Plate 10. — a) Petauristodon mathewsi (James); LP4-M2 (invers), UCMP 54516; loc. V-5662 Cuyama Valley,
California; Late Clarendonian. b) Prosciums sp.; LP4-M.J, CM 9321; Pipestone Springs, north of tracks,
Montana; Chadronian. c) Forsythia gaudryi (Gaillard); LM, and M3, Al. 222 and Al. 223, NMBS; AnwiI,
BL, Switzerland; Middle Miocene, d) Plesiospennophilus angustidens Filhol; LP4-M3, Q. U. 273, NMBS;
Quercy, France; Oligocene. All figures 15x.

Plate 11. — a) Pseudotheridomys sp.; LP^-M^, NMBS, P^: G.B. 251, Mb G.B. 266, M^: G.B. 272 (invers),
M®: G.B. 278; Estrepouy, France; Early Miocene, b) Pseudotheridomys parvidus Schlosser; LP^-M'b
NMBS, P^: C.G. 646, Mb C.G. 849, M^: C.G. 850, M^: C.G. 838; Laugnac, France; Early Miocene, c)
Pseudotheridomys aff. parvidus Schlosser; LP^-Mb NMBS, P^: Sau. 4271 , M'-M^: Sau. 3703, M^: Sau.
4162: Saulcet, France; Early Miocene, d) Pseudotheridomys pagei Shotwell; LP^-M^, P^: UO 22720, Mb
UO 22710, M^-M'b UO 24414; Quartz Basin, Oregon; Barstovian. e) Pseudotheridomys hesperus Wilson;
LP^-Mb KU 10200 (invers) and KU 10201 (invers); Quarry A, Martin Canyon, Colorado; Hemingfordian.
All figures 30x .

Plate 12. — a) Pseudotheridomys sp.; LP4-M3, NMBS, P4 : G.B. 221 , M,: G.B. 228, M2: G.B. 234, M3: G.B.
233; Estrepouy, France; Early Miocene, b) Pseudotheridomys parvidus Schlosser; LP4-M3, NMBS, P4-M2:
C.G. 436, M3: C.G. 861; Laugnac, France; Early Miocene, c) Pseudotheridomys aff. parvidus Schlosser;
LP4-M3, NMBS, P4-M2: Sau. 4974, M3: Sau. 4170 (invers); Saulcet, Erance; Early Miocene, d) Pseudothe-
ridomys pagei Shotwell; LP4-M2, U.O. 22708; Quartz Basin, Oregon U.O. loc. 2465; Barstovian. e) Pseu-
dotheridomys hesperus Wilson; LP4-M2, KU 10195 (type); Quarry A, Martin Canyon, Colorado; Heming-
fordian. All figures 30x.

Plate 13. — a) Eomyops catalaunicus (Hartenberger); LP^-M“, NMBS, P^: C. L 1 . 1 , Mb C. LI. 3 (invers),
M^: C. LI. 2; Can Llobateres, Spain; Late Miocene, b) Leptodontomys sp.; LP^-M^, P^: U.O. 24852, Mb
U.O. 24849, M^: U.O. 25285; Black Butte, Oregon; Clarendonian. c) Eomys zitteli Schlosser; LP^-M^, P^:
CM 24100, Mb CM 24101, M^: CM 24099; Gaimersheim, Germany; Late Oligocene. d) Adjidaumo minimus
(Matthew); LP^-M“ (invers), NMNH 186671; South Pork of Lone Tree Gulch, Natrona Co., Wyoming;
Early Oligocene. e) IProtadjidaumo typus Burke; LM'^“ (invers), CM 9955; Lapoint, West of Vernal, Utah;
Duchesne River Pm.; Late Eocene. All figures 30x .

Plate 14. — a) Eomyops catalaunicus (Hartenberger); LP4-M3, NMBS, P4: C. L 1 . 4, M,: C. LI. 5 (invers),
M2: C. LI. 6, M3: C. LI. 7 (invers); Can Llobateres, Spain; Late Miocene, b) P4: Leptodontomys oregonensis
Shotwell; UO 3633 (invers), type, McKay Reservoir, Oregon; Hemphillian. M,-M3: Leptodontomys sp.;
M,: UO 25283, M2: UO 25278, M3: UO 25274 (invers); UO loc. 2500, Black Butte, Oregon; Clarendonian.
c) Eomys zitteli Schlosser; LP4-M3, CM 24094; Gaimersheim, Germany; Late Oligocene. d) Adjidaumo
minimus (Matthew); LP4-M2 (invers), CM 9213; Pipestone Springs, Montana; Chadronian. e) Protadjidaumo
typus Burke; LP4-M2, CM 11931; Lapoint, West of Vernal, Utah; Duchesne River Pm., Late Eocene. All
figures 30x .

Plate 15. — a) Eiimyarion me dins (Lartet); LM1-M3 (invers), NMBS, Ss. 665; Sansan, Prance; Middle Mio-
cene. b) Leidymys (—Cotimus) alicae (Black); LM, -M3 (invers), KU 18395; Cabbage Patch, Montana; Early
Arikareean. c) Eucricetodon collatus (Schaub); LMj-M,, NMBS, Bd. 40; Boudry II, NE, Switzerland;
Early Aquitanian. All figures 30x.

48

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

Plate 16. — a) Eumyarion medius (Lartet); LMj-Mg, NMBS, Ss. 669; Sansan, Erance; Middle Miocene, b)
Leidymys (=Cotimiis) alicae (Black); LM1-M3, (type), CM 8868; Flint Creek, Montana; Early Arikareean.
c) Eucricetodon collatus (Schaub); LM,-Mg, NMBS, Pa. 13196; Paulhiac, France; Early Aquitanian. d)
Eumys sp.; LMj-Mg, CM 9711; Warbonnet Creek, Sioux Co., Nebraska; Oreodon Beds (Orellan). All figures
30 X.’

Plate 17. — a) Leidymys sp.; LM^ KU 1808; Cabbage Patch, Montana; Middle Arikareean; 30x. h) Eeidymys
alicae (Black); LM', KU, without number; Cabbage Patch, Montana, horizon 58-2; Early Arikareean; 30x.
c) Eumyarion medius (Lartet); LMb NMBS, Ss. 6722; Sansan, France; Middle Miocene, d) Eeidymys sp.;
LMg (invers), KU 1833; Cabbage Patch, Montana, horizon MV 6504 SQ; Middle Arikareean; 30x. e) Un-
derside of left lower incisors of (1) Leidymys alicae (Black) KU 18394 Cabbage Patch, Montana; (2) Eu-
myarion medius (Lartet), NMBS, Ss. 669; Sansan, France; (3) Eucricetodon collatus (Schaub) NMBS, Cod.
188 Coderet, France; I5x.

Plate 18. — a) Democricetodon gaillardi (Schaub); LM'-M^ (invers), NMBS, Ss. 676; Sansan, France; Middle
Miocene, b) Democricetodon minor franconicus Fahlbusch; LM'-M^, NMBS, Erk. 7-8; Erkertshofen, Ger-
many; Lower Miocene, c) Copemys russelli (James); LM’-M^ (invers), UCMP 55507; loc. V-5847, Big Cat
Quarry, Cuyama Valley, California; Early Clarendonian. d) Copemys tenuis Lindsay; LM‘, UCMP 75367;
loc. V-65134, Barstow Em., Southern California; Barstovian. All figures 30x.

Plate 19. — a) Democricetodon gaillardi (Schaub); LM,-Mg, NMBS, Mi-Mg: Ss. 682, Mg; Ss. 6730; Sansan,
France; Middle Miocene, b) Democricetodon minor franconicus Fahlbusch; LMj-Mg, NMBS, Erk. 4-6;
Erkertshofen, Germany; Lower Miocene, c) Copemys barstowensis Lindsay; LMj-Mg, M,: UCMP 74415,
M2: UCMP 74457; loc. V-5501 and V-5253 Barstow Em., California; Barstovian. d) Copemys tenuis Lindsay;
LM,, UCMP 7451 1; loc. V-65150 Barstow Em., California; Barstovian. All figures 30x.

Plate 20. — a) Schauheumys sabrae Black; LP^-M^, CM 13529; Split Rock, Fremont Co., Wyoming; Hem-
ingfordian. b) Plesiosminthus schaubi Viret; LP‘‘-M“, NMBS, Bst. 8832; Coderet, France; Uppermost Oli-
gocene. c) Schaubeumys sabrae Black; LM1-M2, CM 14774; Split Rock, Fremont Co., Wyoming; Heming-
fordian. d) Plesiosminthus schaubi Viret; LM,-M2, NMBS, Bst. 8837; Coderet, France; Uppermost
Oligocene. All figures 40x.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

49

Plate 1.

Plate 2.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

51

Plate 3

52

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY NO. 14

Plate 4.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

53

Plate 5.

54

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

Plate 6.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

55

Plate 7.

56

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

Plate 8

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

57

Plate 9.

58

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

Plate 10.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

59

Plate 1 1 .

60

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

Plate 12.

Plate 13.

62

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 14

Plate 14

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

63

Plate 15

64

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY NO. 14

Plate 16.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

65

CD

_Q

O

Plate 17

66

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

Plate 18.

1979

ENGESSER— MIOCENE INSECTIVORES AND RODENTS

67

Plate !9.

68

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 14

Plate 20

Copies of the foWowing Bulletins of Carnegie Museum of Natural History may be obtained at the prices
listed from the Publications Secretary, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pitts-
burgh, Pennsylvania 15213.

1. Krishtalka, L. 1976. Early Tertiary Adapisoricidae and Erinaceidae (Mammalia, Insectivora) of North

America. 40 pp., 13 figs $2.50

2. Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the
late Pleistocene paleoecology of the central Appalachian Mountains of Virginia. 88 pp., 21 figs. $12.00

3. Wetzel, R. M. 1977. The Chacoan peccary, Catagonus wagneri (Rusconi). 36 pp., 10 figs. . . $6.00

4. Coombs, M. C. 1978. Reevaluation of early Miocene North American Moropus (Perissodactyla,

Chalicotheriidae, Schizotheriinae). 62 pp., 28 figs $5.00

5. Clench, M. H., and R. C. Leberman. 1978. Weights of 151 species of Pennsylvania birds analyzed

by month, age, and sex. 87 pp $5.00

6. Schlitter, D. A. (ed.). 1978. Ecology and taxonomy of African small mammals. 214 pp., 48 figs.
$15.00

7. Raikow, R. J. 1978. Appendicular myology and relationships of the New World nine-primaried oscines

(Aves:Passeriformes). 43 pp., 10 figs $3.50

8. Berman, D. S, and J. S. McIntosh. 1978. Skull and relationships of the Upper Jurassic sauropod

Apatosaurus (Reptilia, Saurischia). 35 pp., 11 figs $3.00

9. Setoguchi, T. 1978. Paleontology and geology of the Badwater Creek area, central Wyoming. Part

16. The Cedar Ridge local fauna (Late Oligocene). 61 pp., 30 figs $4.50

10. Williams, D. F. 1978. Systematics and ecogeographic variation of the Apache pocket mouse (Roden-

tia: Heteromyidae). 57 pp., 23 figs $4.00

11. Guilday, J. E., H. W. Hamilton, E. Anderson, and P. W. Parmalee. 1978. The Baker Bluff Cave

deposit, Tennessee, and the Late Pleistocene faunal gradient. 67 pp., 16 figs $5.00

12. Swanepoel, P., and H. H. Genoways. 1978. Revision of the Antillean bats of the genus Brachyphyiia

(Mammalia: Phyllostomatidae). 53 pp., 17 figs $4.00

13. Schwartz, J. H., and H. B. Rollins, eds. 1979. Models and methodologies in evolutionary theory.

105 pp., 36 figs $6.00

BULLETIN

0/ CARNEGIE MUSEUM OE NATURAL HISTORY

QH

THE APPENDICULAR MYOLOGY AND PHYLOGENETIC
RELATIONSHIPS OF THE PLOCEiDAE AND
ESTRILDIDAE (AVES: PASSERIFORMES)

GREGORY DEAN BENTZ

NUMBER 15 PITTSBURGH, 1979

BULLETIN

of CARNEGIE MUSEUM OE NATURAL HISTORY

THE APPENDICULAR MYOLOGY AND PHYLOGENETIC
RELATIONSHIPS OF THE PLOCEIDAE AND
ESTRILDIDAE (AYES: PASSERIFORMES)

GREGORY DEAN BENTZ

Mount Vernon College,
Washington, D.C. 20007

NUMBER 15

PITTSBURGH, 1979

BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY
Number 15, pages 1-25, figures 1-5, tables 1-2

Issued 15 June 1979

Price: $2.00 a copy

Craig C. Black, Director

Editorial Staff: Hugh H. Genoways, Editor, Duane A. Schlitter,
Associate Editor, Stephen L. Williams, Associate Editor,
Barbara Farkas. Technical Assistant.

(c) 1979 by the Trustees of Carnegie Institute, all rights reserved.

CARNEGIE MUSEUM OF NATURAE HISTORY, 4400 FORBES AVENUE
PITTSBURGH, PENNSYEVANIA 15213

CONTENTS

Abstract 5

Introduction 5

Methods and Materials 7

Muscles of the Forelimb 8

M. latissimus dorsi 8

M. coracobrachialis cranialis 10

M. tensor propatagialis 10

M. deltoideus minor 11

M. pronator profundus II

Muscles of the Hindlimb 13

M. iliotibialis lateralis 13

M. iliotrochantericus medius 13

M. obturatorius lateralis 15

M. gastrocnemius 16

M. plantaris 16

M. flexor digitorum longus 17

M. extensor hallucis longus 18

Discussion 18

Cluster 1 20

Cluster 2 20

Cluster 3 20

Cluster 4 20

Cluster 5 20

Cluster 6 20

Autapomorphic Characters 21

Conclusions and Taxonomic Recommendations 22

Acknowledgments 24

Literature Cited 24

•^•7 “

21-

r

lin

V

1

♦i

r

1

ABSTRACT

The phylogenetic relationships of the Old World passerine
families Ploceidae and Estrildidae are analyzed mainly on the
basis of the structure of the forelimb and hindlimb muscles.
Monophyly of the assemblage is hypothesized on the basis of
common possession of a conical bill adapted to granivory, on
biochemical affinites, and in pterylographic similarities previ-
ously reported. The present study provided no myological syn-
apomorphies to cluster the entire group in support of this hy-
pothesis. Myological characters provide synapomorphies for all
but the first branching point of a cladogram, and autapomorphies

for most taxa. Analysis is at the subfamily level. The Passerinae
are the most primitive group myologically , and presumably the
sister group of the remainder of the assemblage. The Estrildidae
are more highly derived than are the Ploceidae. The Viduinae
are included among the Estrildidae rather than the Ploceidae.
Problems of classification are reviewed and a classification re-
flecting current understanding is presented. The family Ploceidae
includes the subfamilies Passerinae, Ploceinae and Bubalornith-
inae. The Estrildidae includes the Poephilinae, Viduinae. Lon-
churinae, and Estrildinae.

INTRODUCTION

The Old World finches, as the term is used
herein, form a group of approximately 46 genera
and 268 species (Moreau and Greenway, 1962;
Mayr et al., 1968) in the families Ploceidae and Es-
trildidae. Excluding introductions, they are Old
World in distribution and are especially numerous
in Africa. The bill is short, and typically rather thick
and sharply pointed to massive in adaptation to
seed cracking. There are ten primaries. One
subfamily, the Viduinae ( 10 species), is entirely
parasitic, often laying their eggs in the nests of es-
trildids.

The purpose of this study is to elucidate the phy-
logenetic relationships of the Old World finch as-
semblage by constructing a cladogram based on a
survey of the appendicular muscles. This will make
it possible to suggest answers to several related
taxonomic questions, some of which were summa-
rized by Sibley ( 1970).

Historically there has been discussion over
whether the Ploceinae and Estrildinae should be in-
cluded in one family or two. Are the Widow Birds
(Vidua) more closely related to the ploceines, the
estrildines, or to some other group? Is Passer more
closely related to the ploceines, the estrildines, the
fringillids, or some other group? Should a family
Passeridae be recognized? What are the closest rel-
atives of such genera as Buhalornis, Philetairus,
Plocepasser, and Sporopipesl Finally, does the Old
World finch assemblage constitute a monophyletic
group, or is it polyphyletic?

The taxonomic history of this group is long and
complicated and the following is only a brief sum-
mary; a thorough review was given by Sibley
(1970). Some authors have used obsolete generic
names. In those cases the current generic name as

it appears in the "Check-list of Birds of the World"
is given in parentheses.

In the past the principal character used to distin-
guish the Ploceidae (including Estrildidae) from the
Fringillidae was that in the former the tenth primary
is present and usually relatively large on the dorsal
side of the wing, whereas in the latter it is very
small and concealed ventrally.

The first modern classification of the Old World
finches ("Ploceidae") was given by Chapin (1917).
Before that the family was divided into two subfam-
ilies mainly on the length of the tenth primary — the
Ploceinae with a tenth primary longer than the up-
per primary coverts and the Viduinae with a small
and falcate tenth primary. Chapin believed that a
better idea of relationships would result if attention
was given to additional characters, such as song,
plumage, nest construction, bill and foot form, egg
color, and habits. One character, the mouth mark-
ings in estrildid nestlings, has been extremely use-
ful, and Chapin described two types. The "domi-
no" mouth has symmetrically arranged black spots
on a pale palate, whereas the "horseshoe" type
lacks spots on the palate but has one or two horse-
shoes or inverted U-shaped lines, a black line
around the tongue, and two crescents beneath it.
Chapin removed from the Estrildinae those forms
whose nestlings lack such markings. Species with
nestling mouth markings, even though they also
have a long outer primary, were placed in the Es-
trildinae.

Chapin placed "Te.xtor" (Buhalornis) and Dine-
niellia in a separate family, Textoridae (Bubalor-
nithidae), based on characters of the skull and ster-
num. In Buhalornis the fenestrae associated with
the orbital foramina differ in extent and number

5

6

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 15

from other ploceids. and have an obliquely ascend-
ing median bar that is lacking in the other genera
examined. According to Chapin (1917), Buhalornis
also differs from certain ploceids in the form of the
sternal rostrum, which is less forked and more
square in outline, and has a spina interna, as well
as a spina externa. Both Chapin ( 1917) and Sushkin
( 1927) note the presence of a phalloid organ in Bu-
halornis, which I have also observed. Sushkin sug-
gested that this structure served as an auxiliary cop-
ulatory organ.

Shelley ( 1905) placed all longtailed ploceids in a
subfamily Viduinae to which the Bishop-Birds {Eu-
plectes) were added. Reichenow (1914), however,
included them all in his Spermestinae (-1- Estrildi-
nae) but did not so closely associate Vidua with
Coliuspasser (Euplectes). Chapin (1917) placed
Vidua in the Estrildinae and Coliuspasser in the
Ploceinae. Vidua and three closely allied genera,
Tetraenura, Linura, and Steganura (currently all
considered Vidua \ Mayr et al., 1968) have only the
two median pairs of rectrices elongated, whereas in
Coliuspasser all twelve are lengthened. Chapin at-
tributed this to parallelism. The relationship be-
tween Hypoehera (Vidua), Vidua (sens, strict.),
and Steganura is further strengthened by a peculiar
condition in the braincase; the skull has a large clear
area in the frontal region, remaining throughout life,
whereas Coliuspasser has normal skull ossification.

Sushkin (1927) divided the Ploceidae into six
subfamilies — Bubalornithinae, Plocepasserinae,
Ploceinae, Sporopipinae, Estrildinae, and Passeri-
nae, creating the last subfamily by removing Pas-
ser, Petronia, and MontifringiUa from the Fringil-
lidae. He considered them ploceids because they
share a characteristic relief of the horny palatal sur-
face, molt the juvenile remiges and tail in the au-
tumn, and build a domed nest with a side entrance.
The unity of this group was supported by Bock and
Morony (1978) on the basis of the preglossale, a
unique skeletal element of the tongue.

Sushkin stated that Buhalornis and Dinemellia
are closely allied, but that Dinemellia lacks some
“primitive” features of Buhalornis, pointing more
in the direction of the “advanced” Ploceidae. He
used the terms “primitive” and “advanced” but
provided no basis for their usage. Further, the
group of Piocepasser, Philetairus, and Pseudoni-
grita fills the gap between the Bubalornithinae and
the Passerinae to a great extent, which serves to
make the separation of Buhalornis from the re-
mainder of the group less meaningful. Buhalornis

and Dinemellia, then, constitute the Bubalornith-
inae. The Passerinae are a close-knit group that is
nearer to the Ploceinae than to the Estrildinae. In
some osteological respects the Passerinae are more
“primitive” than either. In other respects, such as
specialized feathers at the base of the bill, the Pas-
serinae are more “advanced” than either the Plo-
ceinae or Estrildinae. The connection to the Bu-
balornithinae is established via Philetairus ,
Piocepasser, and Psuedonigrita or the Plocepas-
serinae. On the basis of osteology, Sushkin stated
that the Estrildinae are more advanced than the Plo-
ceinae. Vidua and Steganura appear to be the least
specialized of the Estrildinae and differ least from
the Ploceinae. Sushkin stated that Sporopipes was
halfway between the primitive Estrildinae and Pio-
cepasser, or even Buhalornis, and separated it as
a subfamily Sporopipinae.

The subfamily Estrildinae has often been divided
into two groups whose relationship has frequently
been debated. Chapin (1917) believed that the Vi-
duinae and Estrildinae were very close because of
the similar mouth markings of their young, and that
these markings were not acquired independently.
Beecher ( 1953) raised the group to the rank of fam-
ily. Estrildidae. He believed that the Viduinae arose
in Africa from the Ploceinae and only later became
parasitic on the estrildids, which came to Africa
from Australia. Delacourand Edmond-Blanc ( 1933-
1934) revised Euplectes and Vidua and proposed
that a separate subfamily, Viduinae, be recognized
in the Ploceidae for the Widow Birds. Delacour
(1943) revised the Estrildinae and concluded that
their nearest relatives are the Viduinae, and that
both groups evolved from the Sporopipinae. He
suggested that the Ploceidae were closer to the
Sturnidae than to the Fringillidae because of their
nesting habits. Roberts ( 1947) divided the Ploceidae
into eleven subfamilies. Neither Roberts’ nor Dela-
cour’s work has been universally accepted as they
represent the extremes of taxonomic philosophy —
Roberts as a “splitter” and Delacour as a “lump-
er.”

Tordoff (1954) transferred the Carduelinae from
the Fringillidae to the Ploceidae, based primarily on
the condition of the bony palate. Wolters (1949),
Steiner ( 1954), and Mayr (1955) also transferred the
Carduelinae to the Ploceidae; however, Bock ( 1960)
disagreed. In an exhaustive analysis, he concluded
that the palatine process had little value in showing
relationships among passerine families.

Stallcup (1954) argued that in hindlimb myology

1979

BENTZ— RELATIONSHIPS OF PLOCEIDAE AND ESTRILDIDAE

7

and serology the carduelines are most similar to the
estrildines, and placed the two subfamilies in a fam-
ily Carduelidae. Raikow (1978) included the Car-
duelinae in the Fringillidae based on the limb mus-
cles.

Friedmann (1960) reviewed the literature on the
parasitic weavers and on the basis of reflection
globules, nestling behavior, and plumage, conclud-
ed that the Widow Birds were most closely related
to the estrildines but were distinctive enough to be
considered a subfamily Viduinae.

In the widely followed "Check-list of Birds of the
World" (Moreau and Greenway, 1962; Mayr et al.,
1968; Traylor, 1968), two families (Ploceidae, Es-
trildidae) were recognized. The Ploceidae were di-
vided into the subfamilies Viduinae, Bubalornith-
inae, Passerinae, and Ploceinae. No subfamilies
were recognized in the Estrildidae, but the genera

are arranged in three groups of undesignated taxo-
nomic status — Estrildae, Poephilae, and Lonchur-
ae.

Sibley ( 1970) studied the egg-white proteins of
passerine birds. He concluded that the ploceids and
estrildids are related to one another more closely
than either is to any other group, and recommended
that they be placed in the same family. He also
suggested that a family Passeridae be recognized
until there is more information on Passer.

On the basis of pterylosis, Moiiion ( 1966, 1979)
found two groups, one consisting of Ploceinae, the
other of Viduinae and Estrildinae (which were not
separable). She concluded that recognition of two
families could not be supported on the basis of pter-
ylosis alone, but if the Ploceidae and Estrildidae
were upheld for other reasons, the Viduinae clearly
belonged to the Estrildidae, not to the Ploceidae.

METHODS AND MATERIALS

Variations in musculature were analyzed to determine ances-
tral and derived character states. Though logical and precise,
this methodology does not provide automatic answers, and in-
terpretive decisions are required in situations involving character
conflicts. For example, an individual species of a traditional
group may show an isolated ancestral character state. Precise
application of cladistic methodology might prevent this species
from being grouped with its presumed closest relatives because
sister groups must share derived states that have presumably
been acquired from a common ancestor. However, these isolated
variations are explainable and what is achieved is a cladogram
that represents the best fit with the data available. Character
conflict is discussed in further detail in the discussion.

Hennig (1966), Kluge (1971), Maslin (1952), and Ross (1974)
gave several methods for analyzing character states. The most
important method for this study is the outgroup comparison
between the group being analyzed and related groups. The out-
group comparison may be stated as follows: If a character varies
within a group and one of the variants is also found in a closely
related outside group, then the character state that occurs in
both groups is primitive within the group being studied. It is
supposed that the two groups arose by splitting from a common
ancestor, and that the character states that both groups share
are derived from that ancestor (Kluge, 1971:25-26; Ross,
1974:152-156).

Another method employed is the ingroup correlation. This
states (Kluge, 1971:26) that a character state restricted to the
group of organisms being studied, although only infrequently
exhibited, is primitive when it occurs in those individuals that
have the greatest number of primitive states as determined by
other methods. The probability that a character state is primitive
increases markedly with the increase in the number of primitive
characters with which it is positively correlated. This method is

of limited use, however, because there is no way of knowing
whether a character state is primitive or derived simply from its
frequency of occurrence.

Wiley ( 1975:234) has stressed that the determination of ances-
tral and derived character states is ultimately a question of ho-
mologies and that such homologies are not empirical facts but
hypotheses to be tested. Derived character states are hypothe-
sized to have been acquired from the immediate ancestral
species and to be absent in earlier common ancestors.

It is hypothesized herein that the Old World finches are a
monophyletic group and that the finch-type bill and seed-eating
habit arose only once, as explained below. The outgroups em-
ployed in this study are other birds in general and especially
other groups of passerines. Much of the anatomical data on these
groups has been summarized by George and Berger (1966). The
New World nine-primaried oscines have been analyzed by Rai-
kow ( 1978) and are also used for outgroup comparisons because
their myology indicates a close relationship to the ploceids and
estrildids.

The term “passerine” is an adjective referring to members of
the order Passeriformes and is not to be confused with the
subfamily Passerinae.

All of the hindlimb and forelimb muscles were dissected and
described in Ploceus cucullatus and 47 additional forms. This
species was chosen because of the number of specimens avail-
able and because Ploceus is the nominate genus of the Ploceidae.
Of the 46 genera listed in the "Check list of Birds of the Woi ld."
40 were available for this study. The species dissected are listed
below. Dissection was aided by a stereomicroscope at magnifi-
cations of 6x to 25 X . Visibility of small muscles and fiber ar-
rangements was improved by an iodine stain (Bock and Shear.
1972). Only one specimen of each species was dissected, with
the exception of P. cucullatus of which six specimens were ex-

8

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 15

amined. The range of myological variations among those six
specimens was so minute that a single specimen of each addi-
tional species was considered adequate.

Species dissected (nomenclatui e of ‘Check-list of Birds of the
World”) were as follows:

Ploceidae
Viduinae
Vidua parudisuea
Buhalornithinae
Bubalornis alhirostris
DinemeUia dine nielli
Passerinae
Plocepasser mahali
Pseudonigrita eahinisi
Philetairus socius
Passer domestieus
Petronia xanthocollis
Montifringilla nivalis
Sporopipes sp.

Ploceinae

Amhlyospiza alhifrons
Ploceus ocularis
Ploceus nigricollis
Ploceus cucullalus
Ma limbus cassini
Quelea quelea
Foudia madagascariensis
Euplectes afer
Anomalospiza iniberbis
Estrildidae
Estrildae

Parmoptila woodhouseii
Nigrita canicapilla
Pytilia sp.

Mandingoa nitidula
Cryptospiza reiehenovii
Pyrenestes sanguineus
Pyrenestes ostrinus

Spermophaga haematina
Spermophaga ruficapilla
Clytospiza monteiri
Hypargos niveoguttatus
Lagonosticta sene gala
Uraeginthus ianthinogaster
Estrilda paludicola
Estrilda astrild astrild
Estrilda astrild angolensis
Amandava amandava
Ortygospiza atricollis
Poephilae

Aegintha temporalis
Emblema guttata
Neoehmia phaeton
Poephila guttata
Poephila acuticauda
Lonchurae
Erythrura trichroa
Chloebia gouldiae
Lonchura striata
Lonchura punctulata
Padda oryzivora
Amadina fasciata

The myological nomenclature used is that employed by Rai-
kow ( 1976, 1977).

In the following section “Structure” describes the condition
found in P. cucullatus. Variations in other species are included
in "Comparisons,” and comments on derived versus primitive
character states appear under "Discussion.”

Drawings were made with the aid of a camera lucida micro-
scope attachment. Because the general pattern of musculature
is similar to that in the New World nine-primaried oscines as
described and illustrated by Raikow (1976, 1977), similar dia-
grams and descriptions of all muscles were not included here as
this would be repetitious. Instead, the descriptions and illustra-
tions in the present work show variations that are significant to
this study.

MUSCLES OF THE FORELIMB

The following muscles are present in the forelimb
of Ploceus but do not differ significantly from those
of Loxops virens as described by Raikow (1976):
M. rhomboideus superficialis; M. rhomboideus pro-
fundus; M. serratus profundus; M. serratus super-
ficialis; M. scapulohumeralis cranialis; M. scapu-
lohumeralis caudalis; M. subscapularis; M.
subcoracoideus; M. pectoralis; M. supracoracoi-
deus; M. coracobrachialis caudalis; M. sternocor-
acoideus; M. cucullaris capitis pars propatagialis;
M. deltoideus major; M. biceps brachii; M. triceps
brachii; M. expansor secundariorum; M. brachialis;
M. pronator superficialis; M. flexor digitorum su-
perficialis; M. flexor digitorum profundus; M. flexor
carpi ulnaris; M. ulnometacarpalis ventralis; M. ex-

tensor metacarpi radialis; M. extensor metacarpi
ulnaris; M. extensor digitorum communis; M. ect-
epicondyloulnaris; M. supinator; M. extensor lon-
gus digiti majoris; M. extensor longus alulae; M.
ulnometacarpalis dorsalis; M. abductor alulae; M.
adductor alulae; M. abductor digiti majoris; M. in-
terosseus dorsalis; M. interosseus ventralis; M.
flexor digiti minoris.

M. LATISSIMUS DORSI

Structure. — Pars cranialis arises by an aponeu-
rosis from the neural spines of the last cervical and
first dorsal vertebrae. The thin, strap-shaped, par-
allel-fibered belly passes laterally between the bel-
lies of M. scapulotriceps and M. humerotriceps of

1979

BENTZ— RELATIONSHIPS OF PLOCEIDAE AND ESTRILDIDAE

9

Table 1. — Major variations in forelimb myology of the Ploceidae and Estrildidae. M. latissimus dorsi pars caudaHs: + = muscle
present, — = muscle absent. M. corucobrachiulis cranialis: + = present, — = absent. M. tensor proputagiaUs pars brevis: Type
I = normal belly: Type 2 = elongated belly. M. deltoideus minor: Type I = normal condition; Type 2 = double-headed condition: Type
3 = e.xpanded origin. M. pronator profundus: Type I = single belly: Type 2 = double belly.

Species

M. latissimus
dorsi pars
caudalis

M. tensor

M. coracobrachialis propatagialis

cranialis pars brevis

M. deltoideus
minor

M. pronator
profundus

Vidua paradisaea

+

Family Ploceidae
Subfamily Viduinae

- Type 1

Type 1

Type 1

Bubalornis albirostris

+

Subfamily Bubalornithinae

+ Type 1

Type 1

Type 1

Dinemellia dinemelli

+

+ Type 1

Type 1

Type 1

Plocepasser mahali

+

Subfamily Passerinae

- Type 1

Type 1

Type 1

Pseudonigrita cabanisi

+

- Type 1

Type 1

Type 1

Philetairus socius

+

- Type 1

Type 1

Type 1

Passer domesticus

-

- Type 1

Type 3

Type 1

Petronia xanthocollis

-

- Type 1

Type 1

Type 1

Montifringilla nivalis

-

- Type 1

Type 3

Type 1

Sporopipes sp.

+

- Type 1

Type 1

Type I

Amblyospiza albifrons

+

Subfamily Ploceinae

- Type 1

Type 1

Type 1

Ploceus ocularis

+

- Type 1

Type 1

Type 1

Ploceus nigricollis

+

- Type 1

Type I

Type 1

Ploceus cucullatus

+

- Type 1

Type 1

Type 1

Malimbus cassini

+

- Type 1

Type 1

Type 1

Quelea quelea

+

- Type 1

Type 1

Type 1

Foudia madagascariensis

+

- Type 1

Type 2

Type I

Euplectes afer

+

- Type 1

Type 1

Type 1

Anomalospiza imberbis

+

- Type 1

Type 1

Type 1

Parmoptila woodhouseii

+

Family Estrildidae
Tribe Estrildae
- Type 1

Type 1

Type 1

Nigrita canicapilla

+

- Type 1

Type 1

Type 1

Pytilia sp.

-

- Type 1

Type 1

Type 1

Mandingoa nitidula

-

- Type 1

Type 1

Type 2

Cryptospiza reichenovii

-

- Type 1

Type 1

Type 1

Pyrenestes sanguineus

-

- Type 1

Type 1

Type 2

Pyrenestes ostrinus

-

- Type I

Type 1

Type 2

Spermophaga haematina

+

- Type 1

Type 1

Type 1

Spermophaga ruficapilla

+

- Type 1

Type 1

Type 1

Clytospiza monteiri

-

- Type 1

Type 1

Type 1

Hypargos niveoguttatus

+

- Type 1

Type 1

Type 2

Lagonosticta senegala

+

- Type 1

Type 2

Type 2

Uraeginthus ianthinogaster

-

- Type 1

Type 1

Type 1

Estrilda paludicola

-

- Type 1

Type 1

Type 1

Estrilda astrild astrild

-

- Type 1

Type 1

Type 1

Estrilda astrild angolensis

+

- Type 1

Type 1

Type 1

Amandava amandava

-

Type I

Type 1

Type 1

Ortygospiza atricollis

-

- Type !

Type !

Type 2

Aegintha temporalis

+

Tribe Poephilae
- Type 1

Type 1

Type 2

Emblema guttata

+

- Type 1

Type 1

Type I

Neochmia phaeton

+

- Type 1

Type 1

Type 2

Poephila guttata

+

- Type 1

Type 1

Type 1

Poephila acuticauda

+

- Type 1

Type 1

Type 2

10

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 15

Table 1. — Continued.

Species

M. latissimus
dorsi pars
caudalis

M. tensor

M. coracobrachialis propatagialis

cranialis pars brevis

M. deltoideus
minor

M. pronator
profundus

Erythrura Irichroa

+

Tribe Lonchurae

— Type 1

Type 1

Type 1

Chloehia gouldiae

+

— Type 2

Type 1

Type 1

Lonchura striata

+

- Type 2

Type 1

Type 1

Lonchura punctulata

+

— Type 2

Type 1

Type 1

Padda uryzivora

+

- Type 1

Type 1

Type 1

Amadina fasciata

+

— Type 2

Type 1

Type 1

M. triceps brachii to a fleshy insertion on the dorsal
surface of the humerus about 3 mm from the prox-
imal end of the bone.

Pars caudalis arises by an aponeurosis from the
third and fourth dorsal vertebrae. The thin, parallel-
fibered belly passes laterally superficial to M. rhom-
boideus profundus, tapers to a thin tendon, and in-
serts on the caudodorsal surface of the head of the
humerus. The insertion is deep and cranial to the
insertion of pars cranialis.

Comparison. — Pars caudalis was pr'esent in all
Ploceinae, Bubalornithinae, Lonchurae, Poephilae,
and Vidua. It was absent in many of the Estrildae
and in three species of Passerinae (Table 1).

Discussion. — Because pars caudalis is present in
many passerine and non-passerine birds (George
and Berger, 1966:288-292; Raikow, 1978) its pres-
ence here is clearly primitive and its absence de-
rived.

M. CORACOBRACHIALIS CRANIALIS

Structure. — This muscle is absent in Ploceus,
and the following description is based on Bubalor-
nis alhirostris. This parallel-fibered muscle arises
by tendinous fibers from the lateral surface of the
head of the coracoid. The fibers of the muscle are
embedded in thick fascia and pass distally to insert
fleshy on the ventral surface of the head of the hu-
merus, halfway between the coracohumeral liga-
ment and the belly of M. deltoideus minor.

Comparison. — This muscle is present only in Bu-
halornis and Dinemellia. In all other forms exam-
ined it is represented by a ligamentous band.

Discussion. — George and Berger (1966:313) stat-
ed that this muscle is absent in Agelaius phoeniceus
because no muscle fibers are visible. Ploceus cu-
cullatus exhibits a similar condition except that a
very few muscle fibers appear to be present. Be-
cause this muscle is present in many gr oups of birds

its presence in Buhalornis and Dinemellia probably
represents an ancestral state. Absence of the mus-
cle represents a loss and is therefore derived.

M. TENSOR PROPATAGIALIS

Structure. — M. tensor propatagialis pars longa is
a small parallel-fibered muscle about 9 mm long and
3 mm wide. It arises by both fleshy fibers and an
aponeurosis from the apex of the clavicle immedi-
ately pr oximal to the or igin of M. tensor pr opata-
gialis pars brevis. The belly of M. tensor propata-
gialis pars longa ends on a thin tendon that passes
distally in the cranial edge of the pr opatagium and
is joined by the tendon of M. cucullaris capitis pars
propatagialis. The tendon then passes superficial to
the tendon of insertion of M. extensor metacarpi
r adialis to inser t on the distal end of the radius and
on the palmar surface of the os radiale. The tendon
also fuses with thick fascia around the wrist and
hand.

A much lar ger head than M. tensor pr opatagialis
pars longa, the pars brevis has both a fleshy and
tendinous origin from the apex of the clavicle. The
13 mm spindle-shaped belly tapers to a str ong 9 mm
tendon that is joined by the pars pr opatagialis brevis
of M. pectoralis and fuses with the belly of M. ex-
tensor metacarpi radialis. The tendon of M. tensor
propatagialis pars brevis then passes proximad
along the dorsal surface of the belly of M. extensor
metacarpi radialis to insert on the ectepicondylar
process of the humerus.

Comparison. — In many of the Lonchurae the bel-
ly of pars brevis is elongated to within 1 mm of the
insertion on M. extensor metacarpi radialis (Fig. 1
and Table 1).

Discussion. — According to George and Berger
(1966:320) ther e has been emphasis on the taxonom-
ic value of the pattern formed by the tendon of in-
sertion of pars brevis. The condition in the Lon-

1979

BENTZ— RELATIONSHIPS OF PLOCEIDAE AND ESTRILDIDAE

RADIUS

HUMERUS

Fig. 1. — Dorsal view of the left shoulder. Left: derived condition of M. tensor propatagiaiis pars brevis (TPB) in Lonchuni punctulata.
Right: ancestral condition of the same muscle in Ploceus cucuUatus. Abbreviations: DMAC. M. deltoideus major caudalis; TS, M.
scapulotriceps; EMR, M. extensor metacarpi radialis.

churae closely resembles the condition characteristic
of some swifts, hummingbirds, and pigeons. How-
ever most birds, including the New World nine-pri-
maried oscines (Raikow, 1978), show a condition
similar to Ploceus. By outgroup comparison then
the described condition is primitive and the elon-
gated condition is derived.

M. DELTOIDEUS MINOR

Structure. — This small flat band of fleshy, nearly
parallel fibers is about 7 mm long and 0.5 mm wide.
It arises from the ventral and lateral edges of the
acromion process of the scapula. The belly passes
laterally and cranially superficial to the tendon of
insertion of M. supracoracoideus and inserts on the
craniodorsal surface of the deltoid crest just distal
to the insertion of M. supracoracoideus.

Comparison. — Most ploceids and estrildids ex-
hibit the condition described above. However, two
distinct variations occur. In Foudia and Lagono-
sticta this muscle arises by two independent heads
that fuse prior to insertion. In Passer and Monti-
fringilla the origin is from the scapula, the scapu-
locoracoidal ligament and the head of the coracoid.

Raikow ( 1978) found this latter condition in certain
genera of the New World nine-primaried oscines.

Discussion. — George and Berger (1966:236) state:
"M. deltoideus minor typically has a single head
. . . which has been found in most birds.” On the
basis of outgroup comparison then, the condition
described for Ploceus represents an ancestral state
and the two variations described above are derived
character states.

M. PRONATOR PROFUNDUS
Structure. — This muscle arises fleshy from the
humeroulnar pulley and by means of a short tendon
from the distal end of the humerus (between the
origins of M. pronator superficialis and M. flexor
digitorum superficialis). The fan-shaped belly
passes distally to insert on the caudal surface of the
proximal one-third of the radius.

Comparison. — In several Estrildidae there are
two distinct heads (Table 1). The muscle originates
as described above, but at its midpoint the belly
divides into two portions. The proximal belly in-
serts fleshy onto the ventral surface of the radius.
The distal belly tapers to a 5 mm wide aponeurosis

12

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 15

ULNA

Fig. 2. — A deep muscle of the forearm, M. pronator profundus, PP. Above; ancestral condition in Ploceus cucullatus. Below: derived
condition in Hypargos niveoguttatus.

and inserts on the radius approximately 5 mm distal
to the insertion of the proximal belly (Fig. 2). Ber-
ger (1968) described a similar condition in Den-
droica kirtlandii. Raikow (1978) also found two
heads in some of the New World nine-primaried
oscines.

Discussion. — The condition of this muscle in Plo-
ceiis is as it is in most birds (George and Berger,
1966:346). It therefore represents an ancestral char-
acter state. The two-headed condition is a derived
state within this group, by virtue of the outgroup
comparison.

1979

BENTZ— RELATIONSHIPS OF PLOCEIDAE AND ESTRILDIDAE

13

A

B

C

Fig. 3. — Lateral view of thigh musculature showing variation in the origin of the postacetabular portion of M. iliotibialis lateralis (IL).
A, condition as it exists in Ploveiis aicuUutus', B, Cryptospizu reichenovii', C, Plovepusser mahali. Abbreviation; 1C, M. iliotibialis
cranialis.

MUSCLES OF THE HINDEIMB

The following muscles are present in the hindlimb
of Ploceus but do not differ significantly from those
of Loxops Virens as described by Raikow (1977):
M. iliotibialis cranialis; M. iliotrochantericus cau-
dalis; M. iliotrochantericus cranialis; M. femorotib-
ialis internus; M. iliofibularis; M. flexor cruris lat-
eralis; M. caudoiliofemoralis pars caudofemoralis;
M. flexor cruris medialis; M. puboischiofemoralis;
M. ischiofemoralis; M. obturatorius medialis; M.
iliofemoralis internus; M. peroneus longus; M. pe-
roneus brevis; M. tibialis cranialis; M. extensor dig-
itorum longus; M. flexor perforans et perforatus
digit! Ill; M. flexor peiforans et perforatus digit! II;
M. flexor perforatus digit! II; M. flexor perforatus
digit! IV; M. flexor perforatus digiti III; M. flexor
hallucis longus; M. flexor hallucis brevis; M. lum-
bricalis.

M. ILIOTIBIALIS LATERALIS

Structure. — This broad, triangular muscle arises
by a large aponeurosis from the dorsal (anterior)
iliac crest and most of the dorsolateral (posterior)
iliac crest. The origin is fleshy for its caudal 2 to 3
mm. Cranially this aponeurotic origin obscures M.
iliotrochantericus cranialis and iliotrochantericus
caudalis. Indeed the entire muscle conceals most of

the deeper muscles of the lateral aspect of the thigh.
The distal half of this muscle consists of three dis-
tinct parts — the cranial and caudal edges are fleshy,
whereas the central part is aponeurotic. The fleshy
cranial and caudal parts become aponeurotic just
proximal to the knee. The common aponeurosis of
these three distal parts forms the outer or cranial
layer of the patellar ligament. The insertion is ten-
dinous on a line joining the cnemial crests of the
tibiotarsus.

Comparison. — In all forms studied this muscle
consists of well-developed preacetabular, acetabu-
lar, and postacetabular portions. There are, how-
ever, two variations from the condition described
above (Fig. 3). In Plocepasser and Montifrinpilla
the origin of the postacetabular portion is entirely
aponeurotic. In Sporopipes, Pseudonigrita and in
most Estrildidae examined the postacetabular por-
tion was entirely fleshy in origin. The estrildid ex-
ceptions to this were Aegintha and Lagonosticta,
in which approximately half of the postacetabular
portion was fleshy, and Spennophaga ruficapilla,
Parmoptila, Pyrenestes sanguineus , and Padda.
which were as described.

Discussion. — By outgroup comparison with most
other birds (George and Berger, 1966) the aponeu-

14

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 15

Table 2. — Major variations in the hindiimh myology of the Ploceidae and Estrildidae. M. gastrocnemius pars interna: Type I = anterior
head present, including patellar band: Type 2 = anterior head present but lacking patellar hand: Type J = anterior head absent. Patellar
band: size expressed as percentage of length of patellar ligament covered by muscular origin. M. ohturatorius lateralis pars dorsalis:
— = dorsal head absent: small, medium, and large size as defined in text. M. plantaris: + = present, — = absent. M. iliotrochantericus
medius: + = present. — = absent. M. iliotihialis lateralis: Type I = postacetabular portion entirely aponeurotic; Type 2 = condition
as described in text for Ploceiis; Type 3 = postacetabular portion entirely fieshy. M. flexor digitorum longus: variations in insertions

of accessory vincula as described in text.

Species

M. gastrocne-
mius

pars interna

Patellar

band

M. obturatorius
lateralis pars dorsalis

M.

plantaris

M.

iliotrochanteri-

cus

medius

M.

iliotibiaiis

lateralis

M. flexor
digitorum
longus

Vidua paradisaea

2

Family Ploceidae
Subfamily Viduinae

+

Type 2

ABB

Bubalornis albirostris

2

Subfamily Bubalornithinae

- - +

+

Type 2

ABB

Dinemellia dinemelli

2

-

-

+

+

Type 2

ABB

Plocepasser mahali

1

0.50

Subfamily Passerinae
small

+

+

Type 1

ABB

Pseudonigrita cahanisi

2

-

medium

+

+

Type 3

ABB

Philetairus socius

1

0.25

medium

+

+

Type 2

ABB

Passer domesticus

1

0.10

large

+

+

Type 2

ABB

Petronia xanthocollis

1

0.10

large

+

+

Type 2

ABB

Montifringilla nivalis

1

0.20

large

+

+

Type 1

ABB

Sporopipes sp.

1

1.00

small

+

+

Type 3

ABB

A mbiyospiza albifrons

2

Subfamily Ploceinae

+

+

Type 2

ABB

Ploceus ocularis

2

-

medium

+

+

Type 2

ABB

Ploceus nigricollis

-)

-

medium

+

+

Type 2

ABB

Ploceus cucuUatus

2

-

medium

+

+

Type 2

ABB

Mallmbus casslni

2

-

medium

+

+

Type 2

AAA

Quelea c/uelea

Y

-

small

+

+

Type 2

ABB

Foudia madagascariensis

2

-

medium

+

+

Type 2

ABB

Euplectes afer

2

-

-

+

+

Type 2

ABB

Anomalospiza imberbis

2

-

small

+

+

Type 2

ABB

Parmoptila n oodhouseli

2

Family Estrildidae
Tribe Estrildae

+

Type 2

ABB

Nigrita canicapilla

2

-

small

+

-

Type 3

ABB

Pytilia sp.

2

-

-

+

-

Type 3

ABC

Mandingoa nitidula

2

-

small

+

+

Type 3

ABB

Cryptospiza reichenovii

2

-

-

+

-

Type 3

ABB

Pyrenestes sanguineus

2

-

-

+

-

Type 2

ABB

Pyrenestes ostrinus

2

-

small

+

-

Type 3

ABB

Spermophaga haematina

2

-

-

+

-

Type 3

ABB

Spermophaga ruficapilla

3

-

-

+

-

Type 2

ABB

Clytospiza monteiri

3

-

-

+

-

Type 3

ABB

Hypargos niveoguttatus

i

0.20

-

+

-

Type 3

ABB

Eagonosticta sene gala

3

-

-

+

-

Type 3

ABB

Uraeginthus lanthinogaster

2

-

-

+

-

Type 3

ABC

Estrilda paludicola

3

-

small

+

-

Type 3

ABB

Estrilda astrild astrild

3

-

-

+

-

Type 3

ABB

Estrilda astrild angolensls

2

-

-

+

-

Type 3

ABC

Amandava amandava

3

-

-

+

-

Type 3

ABB

Ortygospiza atricollis

2

-

-

+

-

Type 3

ABB

Aegintha temporalis

2

Tribe Poephilae

+

+

Type 3

ABB

Emblema guttata

2

_

+

+

Type 3

ABB

1979

BENTZ— RELATIONSHIPS OF PLOCEIDAE AND ESTRILDIDAE

15

Table 2. — Continued.

Species

M. gastrocne-
mius

pars interna

Patellar

band

M, obturatorius
lateralis pars dorsalis

M.

plantaris

M.

iliotrochanteri-

cus

medius

M,

iliotibialis

lateralis

M. flexor
digitorum
longus

Neochmia phaeton

2

-

small

+

+

Type 3

ABB

Poephila guttata

3

-

-

+

+

Type 3

ABB

Poephila acuticauda

2

Tribe Lonchurae

+

Type 3

ABB

Erythrura trichroa

2

-

small

-

-

Type 3

ABB

Chloebia gouldiae

2

-

-

-

+

Type 3

ABB

Lonchura striata

3

-

-

-

+

Type 3

ABB

Lonchura punctulata

3

-

-

-

+

Type 3

ABB

Padda oryzivora

2

-

-

-

-

Type 2

ABB

Amadina fasciata

2

-

-

-H

-

Type 3

ABB

rotic and slightly fleshy origins appear to be the
ancestral conditions. The entirely fleshy postace-
tabular portion represents a derived character state.

M. ILIOTROCHANTERICUS MEDIUS

Structure. — Smallest of the three iliotrochanteri-
cus muscles, this band of muscle about 4 mm long
and 1 mm wide has a fleshy origin from the ventral
edge of the ilium just caudal to the origin of M.
iliotrochantericus cranialis. The parallel fibers pass
caudoventrally and insert tendinous on the lateral
surface of the femur between M. iliotrochantericus
caudalis and M. iliotrochantericus cranialis.

Comparison. — This muscle was absent in Vidua,
in most of the Estrildae examined except Mandin-
goa, and in most of the Lonchurae examined except
Lonchura punctulata, L. striata, and Chloebia. In
these three species the muscle was very small. It
was also absent in Poephila acuticauda and present
but very reduced in P. guttata, Emblema, and
Neochmia.

Discussion. — Designated by the letter “C” in leg-
muscle formulas (Hudson, 1937), this muscle is
present in many passerine and non-passerine
groups (George and Berger, 1966:392). It is univer-
sally present in the New World nine-primaried os-
cines, a group of families that is very close to the
Old World finches (Raikow, 1978). By the outgroup
comparison its presence in the Old World finches
therefore appears to represent an ancestral char-
acter state, whereas absence is due to loss and is
therefore derived.

M. OBTURATORIUS LATERALIS

Structure. — This muscle has two separate paral-
lel-fibered bellies, pars dorsalis and pars ventralis.

Pars dorsalis arises from the ischium between the
caudodorsal border of the obturator foramen and
the ventral border of the ilioischiatic fenestra. The
belly passes craniolaterally to a fleshy insertion on
the surface of the tendon of insertion of M. obtur-
atorius medialis and the trochanter of the femur.
Pars ventralis, the ventral belly, arises fleshy from
the cranioventral border of the obturator foramen.
The triangular belly passes laterally to insert fleshy
on the caudal surface of the femur just distal to the
insertion of M. obturatorius medialis. Fibers of pars
ventralis may extend dorsally deep to the tendon of
M. obturatorius medialis and should not be con-
fused with pars dorsalis.

Comparison. — Pars ventralis is present in all
forms studied but pars dorsalis may be absent.
When present, pars dorsalis shows considerable
variation in size. Raikow (1978) illustrates this vari-
ation and defines it as small if the area of origin is
not caudal to the obturator foramen, as medium if
the origin lies between the obturator foramen and
the midpoint of the ilioischiatic fenestra, and as
large if the origin lies caudal to the midpoint of the
ilioischiatic fenestra. Pars dorsalis was present in
all members of the Passerinae studied, and in most
of the Ploceinae except Amblyospiza and Eu-
plectes. It was absent in most of the Estrildidae,
Vidua, Bubalornis, and Dinemellia (Table 2).

Discussion. — Pars dorsalis is present in most pas-
serines (George and Berger, 1966; Raikow, 1978),
thus by outgroup comparison its absence appears
to be a derived state. No accurate statement can be
made as to the polarity of the phenocline exhibited
by the size of the muscle, as many factors may af-
fect muscle size.

16

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 15

M. GASTROCNEMIUS

Structure. — M. gastrocnemius originates by three
distinct bellies:

1) pars externa — This covers the caudolateral
surface of the crus and is intermediate in size be-
tween the other two heads. The muscle arises by a
short, strong tendon from a tubercle on the cau-
dolateral surface of the femur immediately proximal
to the lateral condyle. The tendon of origin is fused
with the distal arm of the biceps loop. The belly of
pars externa, basically unipennate in construction,
passes distally and ends in a well-developed tendon
that comprises the most lateral portion of the com-
mon tendo achillis of the gastrocnemius complex.

2) pars intermedia — The smallest of the three
heads, pars intermedia lies on the medial surface of
the crus. The belly of this unipennate muscle is sep-
arated from pars interna by the tendon of insertion
of M. flexor cruris medialis. Pars intermedia has its
origin by a short tendon from a tubercle on the cau-
doproxima! surface of the internal femoral condyle.
This origin is shared with the insertion of M. pub-
ischiofemoralis, pars caudalis. The short belly of
pars intermedia ends on an aponeurosis that passes
distally between pars externa and pars interna to
form the middle portion of the tendo achillis.

3) pars interna — The largest part of this complex,
pars interna covers most of the medial surface of
the crus and consists of two parts. The origin of the
cranial head is fleshy from the craniomedial surface
of the inner cnemial crest. The caudal head arises
from the caudomedial surface of the inner cnemial
crest and the head of the tibiotarsus. As noted by
Stallcup (1954) this origin is undivided in some
species. The belly of pars interna extends distally
and gives rise to a tendon that joins with the ten-
dons of pars intermedia and pars externa to form
the cranialmost portion of the tendo achillis. This
common tendon of insertion passes distally over the
tibial cartilage to which it is firmly bound. The in-
sertion is tendinous on the caudal surface of the
hypotarsus and along the caudolateral ridge of the
tarsometatarsus. The tendon is also bound in and
continuous with a fascia which forms a sheath
through which other tendons of this region pass.

Comparison. — In Ploceus both a cranial and cau-
dal head of origin of pars interna are present. When
present the cranial head arises, in part, from the
inner cnemial crest, whereas a band of muscle (the
patellar band) may proceed around the cranial sur-
face of the knee, arising from the patellar ligament.

When present this patellar band overlies the inser-
tion of M. iliotibialis cranialis. In this connection
three groups may be distinguished (Raikow, 1978).
In Type 1, the cranial head is present including a
patellar band; in Type 2, the cranial head is present
but lacks a patellar band (as in Ploceus)-, in Type
3, the cranial head is absent. These variations are
illustrated by Raikow (1978). Forms having Type 1
vary in the size of the patellar band. This size may
be expressed as a percentage of the length of the
patellar ligament, which is covered by the muscle
origin. For example, a value of 1.00 means that the
patellar band arises from the entire extent of the
patellar ligament. A value of 0.50 means that it aris-
es from only 50 percent of the patellar ligament
(halfway from the rotular crest to the patella). Most
of the Passerinae studied had Type 1, except Pseu-
donigrita. However, in Pseudonigrita a very few
fibers may have arisen from the patellar ligament.
All members of the Ploceinae, Viduinae, and Bu-
balornithinae exhibited Type 2. Most of the Estril-
didae had Types 1 or 3. One of the Estrildae (Hy-
pargos) displayed Type 1 (Table 2).

Discussion. — On the basis of the outgroup com-
parison with most other birds (George and Berger,
1966:423) and ingroup correlation. Type 1 is the an-
cestral state with Types 2 and 3 being derived from
it. Type 1 occurs mainly in the Passerinae, which
with rare exception exhibit no derived character
states in other appendicular muscles. Type 1 also
occurs in other groups of birds. Types 2 and 3 are
found in ploceids and estrildids, groups that exhibit
other derived character states such as a Type 2 M.
pronator profundus, loss of the patellar band, and
loss of M. iliotrochantericus medius with relatively
greater frequency.

M. PLANTARIS

Structure. — This small, triangular muscle lies on
the caudomedial side of the crus and has a fleshy
origin from the caudomedial surface of the proximal
end of the tibiotarsus, just distal to the internal ar-
ticular surface. The belly is about 6 mm long and
tapers to a slender tendon that inserts on the prox-
imomedial corner of the tibial cartilage. The muscle
lies deep to M. gastrocnemius pars intermedia.

Comparison. — M. plantaris was absent in all of
the Lonchurae examined except Amadina. In Eu-
plectes the belly was reduced to about 1 mm in
length (Table 2).

Discussion. — M. plantaris is designated by the
letter "F” in muscle formulas (Berger, 1959) and

1979

BENTZ— RELATIONSHIPS OF PLOCEIDAE AND ESTRILDIDAE

17

DIGIT II DIGIT III DIGIT IV

^

Fig. 4. — Diagram representing variations in the pattern of insertion of M. flexor digitorum longus.

was originally thought to be present in all passer-
ines. It occurs in many passerine and non-passerine
groups (George and Berger, 1966:442). Its absence
therefore represents a loss and is considered to be
derived, whereas its presence is an ancestral char-
acter state.

M. FLEXOR DIGITORUM LONGUS

Structure. — This large bipennate muscle lies
along the caudal surfaces of the tibiotarsus and fib-
ula. There are two separate heads of origin. The
lateral head arises fleshy from the caudal surface of
the fibula. The medial head arises fleshy from the
area between, but distal to, the articular surfaces of
the head of the tibiotarsus. These two heads fuse
at about the level of insertion of M. iliofibularis.
The common belly thus formed remains fused to
the tibiotarsus and the fibula for about two-thirds
of the distance of the crus and shortly thereafter
ends in a thick tendon. The tendon of insertion
passes through the medial half of the tibial cartilage
and then through the craniomedial canal of the hy-
potarsus. Just proximal to metatarsal I the tendon
trifurcates, sending branches to the plantar surface
of each of the foretoes. The branch to digit II per-
forates the tendon of M. flexor perforans et perfor-

atus digiti II and inserts on the proximal end of the
ungual phalanx. A single vinculum arises from the
deep surface of the tendon and inserts on the distal
end of the second phalanx of digit II.

The branch to digit III is the largest of the three
branches. It perforates the tendons of Mm. flexor
perforatus digiti III and flexor perforans et peifor-
atus digiti III and inserts on the proximal end of the
ungual phalanx. Two small vincula arise from the
deep surface of this tendon. The more proximal vin-
culum inserts in conjunction with the branches of
M. flexor perforans et perforatus digiti III on the
proximal end of the third phalanx of digit III. The
distal vinculum inserts on the distal end of the third
phalanx.

The branch-tendon to digit IV perforates M. flex-
or perforatus digiti IV and inserts on the proximal
end of the ungual phalanx. Two vincula arise from
the deep surface of this tendon also. The more prox-
imal vinculum inserts on the proximal end of pha-
lanx IV of digit IV, whereas the distal vinculum
inserts on the distal end of the fourth phalanx.

Comparison. — Variation in this muscle centers
on the pattern of insertion of the tendons to digits
II, III, and IV, the variation involving the number
and position of accessory vincula from the tendon

18

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 15

to the phalanges. These variations may be briefly
stated by means of a formula (Fig. 4) as used by
Raikow ( 1978).

The formula for Ploceus cucullatus is ABB, as
it is in all of the forms studied here with four ex-
ceptions. In Pytilia, Uraeginthiis, and Estrilda an-
golensis the pattern of insertion is ABC. In Mal-
imhus the pattern of insertion is AAA.

Discussion. — The Old World finches exhibit
greater uniformity in the arrangement of accessory
vincula than the New World nine-primaried oscines
(Raikow, 1978). As almost all forms studied here
are ABB this character is of little consequence
to this study.

M. EXTENSOR HALLUCIS LONGUS

Structure. — This muscle has two distinct parts.
The origin of the minute pars proximalis is fleshy
from the craniomedial edge of the proximal end of
the tarsometatarsus. The slender belly is about 15
mm long, approximately 0.75 mm wide, and ends

in a threadlike tendon of insertion. This tendon
passes over metatarsal I and on to the dorsal sur-
face of the hallux. It then passes through two bands
of fibroelastic tissue (the automatic extensor liga-
ment), and inserts on dense, fibrous connective tis-
sue immediately proximal to the base of the ungual
phalanx. Pars distalis consists of only a few fibers
extending fleshy from the distal end of the tarso-
metatarsus to insert by a short thin tendon onto the
tendon of pars proximalis at about the level of the
proximal end of phalanx one of the hallux.

Comparison. — In Bubalornis and Dinemellio ad-
ditional fibers arise from the craniomedial surface
of the tarsometatarsus. These pass medially to in-
sert on the belly of pars proximalis, all along its
tendon of insertion, and ultimately to blend in with
pars distalis.

Discussion. — Because the accessory fibers of M.
extensor hallucis longus occur only in the Bubalor-
nithinae they represent an autapomorphous char-
acter state.

DISCUSSION

Sister group relationships may only be deter-
mined on the basis of shared derived character
states (synapomorphies). All but one of the varia-
tions listed in Tables 1 and 2 were useful in deter-
mining relationships. The pattern of insertion of M.
flexor digitorum longus is presented purely for the
sake of describing that muscle completely.

The relationships determined will be presented in
the form of a cladogram. Strictly speaking, a dado-
gram is not a phylogeny, but a diagram of groups
clustered by synapomorphies. However, a clado-
gram may be hypothesized to represent the phylog-
eny of a group.

A common problem in the construction of a
cladogram is character conflict; that is, different
characters may indicate different cladistic branch-
ing patterns. These character conflicts arise be-
cause the complexity of evolutionary processes in
closely related groups is not amenable to an overly
simplistic view of cladistic procedure. Mayr (1974)
has pointed out that cladists often overlook the fre-
quency with which closely related groups indepen-
dently achieve derived states because of their com-
mon genetic background. Thus, parallelism,
convergence, and reversals are to be expected es-
pecially when dealing with structurally simple vari-

ations within a close-knit group. Therefore, al-
though it may not be possible to identify positively
the cause of each inconsistency individually, as a
group they are attributable to normal biological
causes, and can be accommodated in an overall hy-
pothesis of genealogical relationships. In these sit-
uations the convention is to adopt the most parsi-
monious explanation, although there is no biological
basis for assuming that the simplest explanation is
also the one most likely to reproduce the true his-
tory of the group. It may be best to state that par-
simony should be employed not because nature is
parsimonious but because only parsimonious hy-
potheses can be defended without resorting to
either authoritarianism or apriorism (Wiley,
1975:236). More recently, Farris (1977) demonstrat-
ed that the use of most parsimonious trees in phy-
logenetic analysis may be justified as a statistical
inference method.

I will now discuss the cladogram (Fig. 5), which
is similar in format to that of McKenna (1975).

It must first be determined whether or not the
Old World finch assemblage is monophyletic. The
only unambiguous way to do this would be to dem-
onstrate that the group shares some synapomorphy
not found in other birds. Unfortunately, this cannot

1979

BENTZ— RELATIONSHIPS OF PLOCEIDAE AND ESTRILDIDAE

19

PASSERINAE PLOCEINAE BUBALORNITHINAE POEPHILINAE VIDUINAE LONCHURINAE ESTRILDINAE

Fig. 5. — A cladogram indicating a hypothesis of phylogenetic relationships in the Ploceidae and Estrildidae. Points 2-13 indicate
apomorphic states; (1) noncladistic data supporting hypothesis of monophyly (conical bill and granivory; biochemical evidence; pter-
ylosis); (2) loss of patellar band of M. gastrocnemius; (3) loss of M. obturatorius lateralis pars dorsalis; (4) presence of mouth markings
and loss of M. coracobrachialis cranialis; (5) loss of M. iliotrochantericus medius; (6) fleshy origin of postacetabular portion of M.
iliotibialis lateralis; (7) loss of M. coracobrachialis cranialis; (8) independent loss of M. coracobrachialis cranialis; (9) presence of
accessory fibers of M. extensor hallucis longus pars proximalis; (10) M. pronator profundus Type 2, fleshy origin of postacetabular
portion of M. iliotibialis lateralis; (1 1) nest parasitism; (12) elongated belly of M. tensor propatagialis pars brevis, loss of M. plantaris;
( 13) M. pronator profundus Type 2. loss of M. latissimus dorsi pars caudalis.

be done on the basis of present knowledge. No
unique character states were found in the limb mus-
cles that would qualify as such characteristics.

The feeding mechanism is another possible
source of insight. All of the forms involved have a
conical bill used for cracking seeds, which is their
principal food. We may hypothesize that this adap-
tive complex arose once in a common ancestor of
the group, and that the Old World finch assemblage
thus represents an adaptive radiation paralleling
that of the Fringillidae in the New World. Present
understanding of the detailed structure of the feed-
ing apparatus does not allow this idea to be tested
critically. The idea that the ploceid-estrildid com-
plex is monophyletic has, however, generally been
accepted at least implicitly by most workers, be-
cause taxonomic problems within this group have
mainly centered on generic misplacements and var-
ious subgroup divisional difficulties. Delacour
(1943:69) and Beecher (1953:303) suggested that the
feeding specialization was acquired independently
in a number of different families, but neither pro-
vided evidence for this opinion. It is clear that this

question is unsettled and that detailed comparative
studies of the feeding mechanism in the Ploceidae
and Estrildidae are needed. However, on the basis
of our present understanding it appears both rea-
sonable and parsimonious to proceed on the tenta-
tive assumption that the seed-eating specializations
of the ploceid-estrildid complex represent a single
adaptive shift rather than a series of convergent de-
velopments. Utilizing an unrelated morphological
character, pterylosis, Morlion (1966, 1979) found
that the ploceids and estrildids shared basic pat-
terns of pterylosis to the extent that separation at
the family level was questionable.

Aside from morphological considerations, the
most compelling evidence for monophyly is from
the biochemical studies of Sibley (1970). Based on
electrophoretic studies of egg-white proteins, his
principal conclusion was that “the Ploceinae and
Estrildinae are related to one another more closely
than either is to any other group. Although each
seems to be a well-marked, readily defined group
they should be placed in the same family” (Sibley,
1970:96). Data of this type cannot be analyzed cla-

20

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 15

distically because the direction of evolutionary
changes in molecular structure cannot be deter-
mined from electrophoretic patterns. Therefore,
these data cannot be used to hypothesize cladistic
branching patterns within a group. However, they
do demonstrate the close genetic relatedness of the
assemblage, which supports the concept that the
assemblage constitutes a single radiation, or in oth-
er words, that it is monophyletic.

Cluster 1

On the basis of the foregoing discussion it is ten-
tatively concluded that the entire assemblage is
monophyletic, based on the common seed-eating
specialization, pterylosis, and biochemical indica-
tions of close genetic affinity.

Most members of the Passerinae (except Pseu-
donigrita) and only members of the Passerinae (ex-
cept Hypargos) possess a Type 1 gastrocnemius.
They are postulated to be the most primitive mem-
bers of this group, and are the sister group of the
remainder of this assemblage. They are character-
ized by a Type 1 gastrocnemius, an obturatorius
lateralis pars dorsalis of indeterminate size, a plan-
taris, an iliotrochantericus medius, an iliotibialis lat-
eralis with an aponeurotic postacetabular origin, no
mouth markings, an extensor hallucis longus with-
out accessory fibers, a tensor propatagialis with a
long tendon of insertion and no coracobrachialis
cranialis.

Poltz and Jacob (1974) analyzed uropygial secre-
tions biochemically in 18 species of passerines and
concluded that the Passerinae may be more closely
related to the fringillids and emberizine finches than
to the ploceids.

Cluster 2

The groups linked by character 2 are derived for
loss of the patellar band, except Hypargos (Estril-
dae), which is the only form outside of the Passer-
inae to exhibit a patellar band. Other evidence does
not indicate that Hypargos has been improperly
placed, and this inconsistency could be explained
by the secondary reappearance of the patellar band
in this form.

Cluster 3

Most forms grouped by character 3 are derived
for loss of M. obturatorius lateralis pars dorsalis. It
is also lost in Euplectes and Amblyospiza of the
Ploceinae. Estrilda paludicola, Nigrita, Pyrenestes
ostrinus, and Mandingoa of the Estrildae, Eryth-

rura of the Lonchurae, and Neochmia of the Poe-
philae possess the primitive state. The loss of this
muscle in Euplectes and Amblyospiza probably oc-
curred independently subsequent to the origin of
their group.

Cluster 4

All of the Poephilae, Viduinae, Lonchurae, and
Estrildae have some pattern of mouth markings
(character suite 4). Such markings surely represent
a derived state as they have never been reported in
any other passerine family. Delacour (1943:73) sug-
gested that mouth markings in viduines were ac-
quired by convergence to aid them in their nest par-
asitism of estrildids. However, Friedmann (1960)
argued convincingly for estrildid-viduine affinities.
On the basis of plumage, nestling behavior, and re-
flection globules Friedmann demonstrated that the
viduines were closer to the estrildids than to the
ploceines and that these markings were acquired
from a common ancestor. This position is now
strengthened by the fact that the viduines and es-
trildids are myologically very similar in derived
characters. For example, both have lost M. iliotro-
chantericus medius. This suggests that the viduines
are a subgroup of the estrildid radiation that has
become specialized for nest parasitism of other es-
trildids, rather than a distantly related group that
has converged extensively upon the estrildids.
These groups also share another derived state, the
loss of M. coracobrachialis cranialis (discussed be-
low). They are virtually identical in pterylosis and
differ collectively from the Ploceinae (Morlion,
1966, 1979).

Cluster 5

Most of the Lonchurae and Estrildae as well as
Vidua have lost M. iliotrochantericus medius, but
there are a few exceptions (Table 2). In the Estril-
dae, it occurs only in Mandingoa. Perhaps this ge-
nus is misplaced and should be included in the Poe-
philae in which the muscle occurs with greater
regularity. More probably it is a case of parallel loss
of the muscle, indicating an underlying genetic ten-
dency in the group. Possibly there are secondary
reappearances here also (Raikow et al., 1979). Mayr
(1974:80) discusses such cases.

Cluster 6

Most of the Estrildae and Lonchurae are derived
for a fleshy origin to the postacetabular portion of
M. iliotibialis lateralis. The exceptions are Sper-

1979

BENTZ— RELATIONSHIPS OF PLOCEIDAE AND ESTRILDIDAE

21

mophaga ruficapilla, Lagonosticta, Pannoptila,
Pyre nest es sanguineus, Padda, and Aegintha. The
Poephilae also exhibit this condition but for reasons
of parsimony are not grouped here. This is dis-
cussed below. Only two of the Passerinae (Ploce-
passer and Muntifringilla) exhibit the ancestral
condition of this muscle. Most likely the remaining
members of the Passerinae acquired the derived
state independently.

Autapomorphic Characters

Characters 7 through 13 are autapomorphic. That
is, they are derived character states that are not
shared with other groups.

Characters 7 and 8 are both loss of M. coraco-
brachialis cranialis. Separate numbers are given to
the same event because the loss of this muscle in
one lineage is apparently independent of its loss in
the other. Because this muscle is present in many
non-passerine birds, its presence in Bubalornis and
Dinemellia presumably is an ancestral condition.
This muscle has probably been overlooked in many
passerine birds due to lack of adequate staining
methods by earlier investigators. When the muscle
is not present it is usually represented by a liga-
mentous band with only a few muscle fibers and
imbedded in dense connective tissue as in Place us.
The loss of M. coracobrachialis cranialis in the Pas-
serinae, Ploceinae, viduines, and estrildids and the
retention of this muscle in the Bubalornithinae rep-
resents a case of character conflict. The parsimo-
nious choice, however, is to incorporate the syn-
apomorphies of the Bubalornithinae first, namely
the loss of the patellar band and loss of the dorsal
head of obturatorius lateralis. Independent loss of
a single muscle at several points (4, 7, and 8) then
becomes more probable than a single lineage ac-
quiring two derived character states independently,
especially because the loss of this muscle appears
to be a frequent occurrence in passerines (Raikow,
personal communication). An alternative possibility
is that this muscle was absent in the common ances-
tor of the group, and that it reappeared secondarily
in the Bubalornithinae. This explanation appears to
be less probable than multiple loss, which is a well
established and common phenomenon.

Character 9 is the presence of accessory fibers of
M. extensor hallucis longus pars proximalis in Bu-
balornis and Dinemellia only. Such fibers have not
been described before and their occurrence here is
clearly derived.

Character suite 10 includes two states here con-

sidered autapomorphous, the independent origin of
a fleshy postacetabular portion to iliotibialis later-
alis, and a Type II M. pronator profundus. Only
members of the Estrildae and Poephilae show the
latter condition (Table 2) and its occurrence is rath-
er sporadic. Raikow (manuscript) has interpreted
this condition as perhaps increasing the functional
versatility of the muscle because each belly could
act independently of the other. In any event the
double belly appears to be just becoming estab-
lished in these groups, and to have arisen indepen-
dently in several genera.

Character 1 1 is nest parasitism. Nest parasitism
is also practiced by the ploceine finch Anomalo-
spiza imberbis (Roberts, 1917). The viduine finches
all have this behavior and are specific parasites of
estrildids, whereas Anomalospiza parasitizes cisti-
coline warblers. Although nest parasitism occurs in
other groups of birds (Friedmann, 1929), its occur-
rence within the Old World finches is surely a de-
rived state.

Character suite 12 is elongation of the belly of M.
tensor propatagialis pars brevis and loss of the plan-
taris. Both of these derived states occur only among
the Lonchurae. Of all the Lonchurae examined only
Amadina retains the plantaris.

Character suite 13 is the presence of a Type II
pronator profundus and the loss of M. latissimus
dorsi pars caudalis. This latter muscle is one whose
pattern of occurrence is also difficult to interpret.
In this study the muscle is present in all Ploceinae,
Bubalornithinae, Poephilae, Viduinae, and Lon-
churae. It is absent in certain species of Estrildae
and also in Passer, Petronia, and Montifringilla of
the Passerinae (see Table 1). The muscle may even
occur in one species and be absent in another
species of the same genus (for example, Estrilda).
All of this suggests that this muscle may be lost and
subsequently regained in an evolving lineage. At the
very least the absence of this muscle in estrildines
further suggests that they are not closely related to
the ploceids.

Although a cladogram is not intended to be a phy-
togeny per se, an ideal cladogram should not pre-
sent any incompatibilities with other available in-
formation about the taxa included. One such source
of information is geographic distribution. Among
the taxa of the Old World finches, the only groups
for which the limb myology could support more
than one possible arangement in the cladogram are
the subfamilies of the Estrildidae. Of these, the Vi-
duinae and Estrildinae are endemic to Africa; the

22

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 15

Poephilinae are confined to Australia and New
Guinea; and the Lonchurinae are widely distributed,
with the genus Lonchura found from Africa to Aus-
tralia but its chief radiation in southeast Asia and
the East Indies, one endemic African genus (Ama-
dina), two monotypic Australian genera (Chloebia,
Aidemosyne), and one genus only tentatively placed
in the Lonchurinae by Mayr (in Mayr et al.,
1968:361), Erythrura, found from the Philippines
through the East Indies, New Guinea, and the Pa-
cific Islands, with a bare foothold in northern Aus-
tralia.

It seems apparent that, in view of the parasitism
of Viduinae on Estrildinae, of their present distri-
bution, and of their close resemblance in many mor-

phological characters, these two subfamilies must
have evolved together for some time, and that a
cladogram that separated them widely would rep-
resent a highly improbable interpretation of the evi-
dence. The common ancestor of all Estrildidae may
well have inhabited Africa, where the main radia-
tion of the family has taken place, or else southern
Asia. An early invasion of the Australian region ra-
diated into the present Poephilinae, with the lon-
churine inhabitants of Australia and vicinity repre-
senting a later invasion. Similarly, the lonchurine
inhabitants of Africa probably represent a later in-
vasion of a basically East Asian radiation. The
cladogram proposed here is compatible with such
a distributional history.

CONCLUSIONS AND TAXONOMIC RECOMMENDATIONS

We may now consider the problems posed at the
beginning of this study. First of all, should the plo-
ceines and estrildines be classified in one family or
two? This question is meaningful mainly from a
point of view emphasizing phenetic clustering. Es-
sentially, it asks whether the estrildids are suffi-
ciently different from the remainder of the group so
as to be given family rank. They differ in nest con-
struction and mouth markings (Chapin, 1917), in
osteology (Sushkin, 1927), and in egg-white pro-
teins (Sibley, 1970). The present study has also
demonstrated that they are myologically distinct. I
suggest therefore that it would be appropriate for
the Estrildidae to be given family rank if one does
not feel bound to classify according to the system
of Hennig (see Mayr, 1974, for a discussion of cla-
distic classification).

The above question may also be approached from
a cladistic viewpoint. The phylogeny of the group
(Fig. 5) shows that the subgroups of the Estrildidae
(including the Viduinae) share a common ancestor
that is not shared by the others. The estrildid por-
tion of the group is thus holophyletic and warrants
some categorical name. However, the remainder of
the group (Passerinae, Ploceinae, Bubalornithinae)
form a paraphyletic assemblage and cannot be a
coordinate sister group of the estrildids. Under
strict application of cladistic theory, if one makes
the Estrildidae a family then the entire Old World
finch assemblage would have to be given higher
taxonomic rank. One could also make the entire
assemblage a family based on the convention that

the category “family” usually represents a group
with a readily discernible adaptive niche — in this
case the seed-eating specialization. Basically, these
are matters of individual preference depending on
a worker's systematic philosophy.

The exact position of Passer has been a problem
for some time. It has never been conclusively dem-
onstrated whether the genus is more closely related
to the ploceines, estrildines, fringillids, or to
another group. There is little doubt that Passer is
a unique genus among the Old World finches in
terms of nest construction and geographic distri-
bution. Though no conclusive statement can be
made as to the exact status of Passer, it appears as
though it is myologically most similar to other mem-
bers of the Passerinae. However, the Passerinae are
seen to be the most primitive members of this group
(Fig. 5) and additional work could reveal that cer-
tain genera may be more closely related to the frin-
gillids, as suggested by Poltz and Jacob (1974) and
Sibley (1970) for Passer.

Sibley (1970) recommended that a family Passer-
idae be recognized for Passer and stated that the
relationships of Montifringilla were probably not
with Passer. Clench (1970) found that in pterylosis
Passer was similar to Pseudonigrita and Plocepas-
ser but differed significantly from Sporopipes. The
present study shows that the Passerinae are char-
acterized by a certain myological uniformity. The
above genera, as well as others, all possess a pa-
tellar band and retain the dorsal head of obturato-
rius lateralis (Table 2). Although some variation oc-

1979

BENTZ— RELATIONSHIPS OE PLOCEIDAE AND ESTRILDIDAE

23

curs in the origin of iliotibialis lateralis, the
members of the Passerinae are myologieally more
similar to each other than to any group investigated
here. However, these similarities are mostly ances-
tral character states and thus relatively weak indi-
cators of relationship. Passer and Montifringilla
are myologieally unique members of the Passerinae
in that they both share the derived condition of M.
deltoideus minor (Table 1). These two genera, along
with Petronia, also differ from other Passerinae in
the derived absence of M. latissimus dorsi pars cau-
dalis. It appears then that these three genera are
probably closely related. Bock and Morony (1978)
reached a similar conclusion based on a study of
the tongue skeleton.

Certain authors (Sushkin, 1927; Collias and Col-
lias, 1964) have recognized such subfamilies as the
Plocepasserinae and the Sporopipinae. Although
there may be an adequate osteological and behav-
ioral basis for such groupings, there is no myolog-
ical reason for the recognition of those groups. It
is recommended that the subfamily Passerinae be
retained and that the genera listed under that sec-
tion in Table 1 be included in that subfamily.

The relationships of the Widow Birds (Vidua)
have also been debated. On the basis of egg-white
proteins. Vidua most closely resembled Passer
(Sibley, 1970). Historically, however, the plo-
ceines, euplectines, and estrildines have been con-
sidered as possible relatives. On the basis of reflec-
tion globules in the mouth, nestling behavior,
plumage, and pterylosis, Friedmann (1960) and
Moiiion ( 1966, 1979) decided that the viduines were
most closely related to the estrildines. The present
study has demonstrated that in derived characters
the viduines are also myologieally more similar to
the estrildids than to the ploceids. If separate fam-
ilies are to be recognized, then the subfamily Vi-
duinae should be included within the Estrildidae
and not within the Ploceidae. This is in marked con-
trast with what has stood as the “preferred” clas-
sification for this assemblage (Sushkin, 1927). Also
within the “Check-list of Birds of the World” (Mayr
et al., 1968) the Viduinae are listed under the Plo-
ceidae. It is further recommended (if separate fam-
ilies are recognized) that the subgroups of the Es-
trildidae (Mayr et al., 1968, based on the tribes of
Delacour, 1943, but given the termination -ae in-
stead of the proper tribal termination -ini) be raised

to subfamily status because this is the first subdi-
vision of the family category, whereas the tribal cat-
egory is normally used as a subdivision of the
subfamily category. Thus, the Estrildae would be-
come the Estrildinae, the Poephilae would become
the Poephilinae, and the Lonchurae would become
the Lonchurinae.

On the basis of comparative myology it is also
possible to make some general conclusions as to the
closest relatives of certain other genera. It is gen-
erally agreed that Dinemellia is the closest relative
of Bubalornis, and that relationship is supported by
the present study. Sushkin ( 1927) believed that Bu-
balornis was more primitive than Dinemellia, and
was considered to be the most primitive of all plo-
ceids. However, only Bubalornis possesses an M.
expansor secundariorum that inserts on four sec-
ondaries. This, and the presence of a copulatory
organ, are derived states and suggest that Bubalor-
nis is derived relative to Dinemellia. It has been
demonstrated then that Bubalornis is derived in a
number of myological, osteological, and morpho-
logical traits and does not represent the most prim-
itive of ploceids. Therefore if there are sturnid af-
finities to the Ploceidae (Bartlett, 1889), they are
not through Bubalornis. Data from egg white pro-
teins did not support a stuvnid-Bubalornis affinity
(Sibley, 1970).

Philetairus , Sporopipes , and Plocepasser are
myologieally good members of the Passerinae but
appear to be more closely related to each other than
to Passer and the other members of the subfamily.

Do the Old World finches then represent a mono-
phyletic group? The myological evidence as well as
evidence from other disciplines discussed above
suggests that this group as a whole is probably
monophyletic.

On the basis of this study the following classifi-
cation is proposed:

Family

Subfamilies

Family

Subfamilies

Ploceidae

Passerinae

Ploceinae

Bubalornithinae

Estrildidae

Poephilinae

Viduinae

Lonchurinae

Estrildinae

24

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 15

ACKNOWLEDGMENTS

This study was completed as partial fulfillment of the require-
ments for the degree of Doctor of Philosophy at the University
of Pittsburgh. 1 would like to thank Dr. Robert J. Raikow for his
assistance throughout this project. Without his encouragement
and advice this work could not have been completed. 1 am also
indebted to the members of my committee: Dr. Kenneth C.
Parkes, Dr. Mary H. Clench, Dr. William P. Coffman, Dr. Mi-
chael Mares, and Dr. Jeffrey Schwartz. I am grateful to the
following for providing the specimens upon which this study is
based: Carnegie Museum of Natural History. Dr. Mary H.

Clench; National Museum of Natural History, Dr. Storrs L.
Olson; Peabody Museum of Natural History, Yale University,
Dr. Charles G. Sibley; American Museum of Natural History,
Dr. Walter J. Bock; Museum of Vertebrate Zoology, University
of California, Berkeley, Dr. Ned K. Johnson; Occidental Col-
lege. Dr. Luis F. Baptista; British Museum (Natural History),
Dr. Philip Burton.

Portions of this study were supported by NSF Grant No. BMS
74 18079 to Dr. Robert J. Raikow.

LITERATURE CITED

Bartlett. E. 1889. A monograph of the Weaverbirds, Ploce-
idae, and Arboreal and Terrestrial Finches, Fringillidae.
London, 193 pp.

Beecher, W.J. 1953. A phylogeny of the oscines. Auk, 70;270-
333,

Berger, A. J. 1959. Leg-muscle formulae and systematics.
Wilson Bull.. 71:93-94.

. 1968. Appendicular myology of Kirtland’s Warbler.

Auk. 85:594-616.

Bock, W. J. I960. The palatine process of the premaxilla in the
Passeres. Bull. Mus. Comp. Zool., 122:361^88.

Bock. W. J., AND J. J. Morony. Jr. 1978. Relationships of the
passerine finches (Passeriformes:Passeridae). Bonn Zool.
Beitr,. 29:122-147.

Bock, W. J,, AND R. Shear. 1972. A staining method for gross
dissection of vertebrate muscle. Anat. Anz.. 130:222-227.
Chapin, J. P. 1917. The classification of Weaver-Birds. Bull.

Amer. Mus. Nat. Hist., 37:243-280.

Clench. M. H. 1970. Variability in body pterylosis, with spe-
cial reference to the genus Passer. Auk, 87:650-689.
CoLLiAS, N. E., AND E. C. CoLEiAS. 1964. Evolution of nest-
building in the Weaverbirds (Ploceidae). Univ. California
Publ. Zool., 73:1-162.

Delacour, j. 1943. A revision of the subfamily Estrildinae of
the family Ploceidae. Zoologica. 28:69-86.

Delacour. J.. and F. Edmond-Blanc. 1933-34. Monogra-
phic des veuves (revision des genres Euplectes et Vidua.)
L'Ois., 3:519-562. 687-726. 4:52-110.

Farris, J. S. 1977. Phylogenetic analysis under Dollo’s Law.
Syst. Zool., 26:77-88.

Friedmann. H. 1929. The cowbirds, a study in the biology of
social parasitism. C. C. Thomas. Springfield, Illinois, 421

pp.

. I960. The parasitic Weaverbirds. Bull. U.S. Nat. Mus.,

223:viii + 1-196.

George. J. C., and A. J. Berger. 1966. Avian myology. Ac-
ademic Press. New York and London, xii -I- 500 pp.
Hennig, W. 1966. Phylogenetic systematics. Univ. Illinois
Press. Urbana. 263 pp.

Hudson, G. E. 1937. Studies on the muscles of the pelvic ap-
pendage in birds. Amer. Midland Nat., 18:1-108.

Kluge, A. G. 1971. Concepts and principles of morphologic
and functional studies. Pp. 3-51, in Chordate structure and
function (A. J. Waterman, ed.). Macmillan Company. New
York. 628 pp.

Maslin, T. P. 1952. Moi'phological criteria of phyletic rela-
tionships. Syst. Zool., 1:49-70.

Mayr, E. 1955. Comments on some recent studies of song bird
phylogeny. Wilson Bull.. 67:33-34.

. 1974. Cladistic analysis or cladistic classification? Z.

Zool. Syst. Evol.-forsch., 12:94-128.

Mayr, E., R. A. Paynter. Jr., and M. A. Traylor. 1968.
Eamily Estrildidae. Pp. 306-389, in Check-list of birds of
the World (R. A. Paynter. ed.), Mus. Comp. Zool., Cam-
bridge, Massachusetts. I4:x -I- 1^33.

McKenna. M. C. 1975. Toward a phylogenetic classification
of the Mammalia. Pp. 21^6, in Phylogeny of the Primates
(W. P. Luckett and F. S. Szalay. eds.). Plenum Publ. Corp.,
New York, xiv -r 483 pp.

Moreau, R. E., and J. C. Greenway, Jr. 1962. Family Plo-
ceidae. Pp. 3-74, in Check-list of birds of the World (E.
Mayr and J. C. Greenway. Jr., eds.). Mus. Comp. Zool.,
Cambridge, Massachusetts, 15:x -I- 1-315.

Morlion. M. 1966. Vergelijkende studie van de pterylosis in
enkele Afrikaanse genera van de Ploceidae. Unpublished
Ph.D. thesis, Rijksuniversiteit Gent, Belgium, 399 pp.

. 1979. Pterylosis as a secondary criterion in the taxon-
omy of the African Ploceidae and Estrildidae. Ostrich, in
press.

PoLTZ, J., AND J. Jacob. 1974. Burzeldrusensekrete bei Am-
mern (Emberizidae). Finken (Fringillidae) and Webern (Plo-
ceidae). J. Ornithol., 115:1 19-127.

Raikow, R.J. 1976. Pelvic appendage myology of the Hawaiian
honeycreepers (Drepandidae). Auk, 93:774-792.

. 1977. Pectoral appendage myology of the Hawaiian

honeycreepers (Drepandidae). Auk, 94:331-342.

. 1978. The appendicular myology and relationships of

the New World nine-primaried oscines ( Aves:Passeriformes).
Bull. Carnegie Mus. Nat. Hist.. 7:1-43.

Raikow, R. J., S. R. Borecky. and S. L. Berman. 1979. The
evolutionary reestablishment of a lost ancestral muscle in
the Bowerbird assemblage. Condor, in press.

1979

BENTZ— RELATIONSHIPS OE PLOCEIDAE AND ESTRILDIDAE

25

Reichenow, a, 1914. Die Vogel: Handbuch der Systematis-
chen Ornithologie. Vol. II. Verlag von Ferdinand Enke,
Stuttgart, 628 pp.

Roberts, A. 1917. Parasitism amongst finches. Ann. Transvaal
Mus., 5:259-262.

. 1947. Reviews and criticisms of nomenclatural changes.

Ostrich, 18:59-85.

Ross. H. H. 1974. Biological systematics. Addison-Wesley .
Reading, Massachusetts, 345 pp.

Shelley, G. E. 1905. The birds of Africa. R. H. Porter, Lon-
don. 6(part l):v + 1-511.

Sibley, C. G. 1970. A comparative study of the egg-white pro-
teins of passerine birds. Bull. Peabody Mus. Nat. Hist.,
Yale Univ., 32:1-131.

Stallcup, W. B. 1954. Myology and serology of the avian
family Eringillidae, a taxonomic study. Univ. Kansas Publ.,
Mus. Nat. Hist., 8:157-211.

Steiner, H. 1954. Das Brutverhalter der Prachtfinken. Sper-
mestidae, als Ausdruck ihres selbstandigen Familienchar-
akters. Eleventh Int. Cong. Orn. Basel, pp. 350-355.

SusHKiN. P. P. 1927. On the anatomy and classification of the
weaverbirds. Bull. Amer. Mus. Nat. Hist., 57:1-32.

Tordoff, H. B. 1954. A systematic study of the avian family
Fringillidae based on the structure of the skull. Misc. Publ.
Mus. Zool., Univ. Michigan, 81: 1-41.

Traylor, M. A. 1968. Family Ploceidae. subfamily Viduinae.
Pp. 390-397, in Check-list of birds of the World (R. A.
Paynter, ed.), Mus. Comp. Zool., Cambridge, Massachu-
setts, I4:x + 1-433.

Wiley, E, O. 1975. Karl R. Popper, systematics, and classifi-
cation: a reply to Walter Bock and other evolutionary tax-
onomists, Syst. Zool., 24:233-243,

WoLTERS, H. E. 1949. Beitrage Zur Gattungssystematik der
Einkvogel. Beit. Gattungs. Vogel. 1:3-17.

A^Jlf

/

J

I

r

f

...St

1

I

V

; :

ji-

Copies of the following Bulletins of Carnegie Museum of Natural History may be obtained at the prices
listed from the Publications Secretary, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pitts-
burgh, Pennsylvania 15213.

1. Krishtalka, L. 1976. Early Tertiary Adapisoricidae and Erinaceidae (Mammalia, Insectivora) of North

America. 40 pp., 13 figs $2.50

2. Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the
late Pleistocene paleoecology of the central Appalachian Mountains of Virginia. 88 pp., 21 figs. $12.00

3. Wetzel, R. M. 1977. The Chacoan peccary, Catagonus wagneri (Rusconi). 36 pp., 10 figs. .. $6.00

4. Coombs, M. C. 1978. Reevaluation of early Miocene North American Moropus (Perissodactyla,

Chalicotheriidae, Schizotheriinae). 62 pp., 28 figs $5.00

5. Clench, M. H., and R. C. Leberman. 1978. Weights of 151 species of Pennsylvania birds analyzed

by month, age, and sex. 87 pp $5.00

6. Schlitter, D. A. (ed.). 1978. Ecology and taxonomy of African small mammals. 214 pp., 48 figs.

$15.00

7. Raikow, R. J. 1978. Appendicular myology and relationships of the New World nine-primaried oscines

(Aves:Passeriformes). 43 pp., 10 figs $3.50

8. Berman, D. S, and J. S. McIntosh. 1978. Skull and relationships of the Upper Jurassic sauropod

Apatosaurus (Reptilia, Saurischia). 35 pp., 11 figs $3.00

9. Setoguchi, T. 1978. Paleontology and geology of the Badwater Creek area, central Wyoming. Part

16. The Cedar Ridge local fauna (Late Oligocene). 61 pp., 30 figs $4.50

10. Williams, D. F. 1978. Systematics and ecogeographic variation of the Apache pocket mouse (Roden-

tia: Heteromyidae). 57 pp., 23 figs $4.00

11. Guilday, J. E., H. W. Hamilton, E. Anderson, and P. W. Parmalee. 1978. The Baker Bluff Cave

deposit, Tennessee, and the Late Pleistocene faunal gradient. 67 pp., 16 figs $5.00

12. Swanepoel, P., and H. H. Genoways. 1978. Revision of the Antillean bats of the genus Brachyphylla

(Mammalia: Phyllostomatidae). 53 pp., 17 figs $4.00

13. Schwartz, J. H., and H. B. Rollins, eds. 1979. Models and methodologies in evolutionary theory.

105 pp., 36 figs $6.00

14. Engesser, B. 1979. Relationships of some insectivores and rodents from the Miocene of North Amer-
ica and Europe. 68 pp., 12 figs., 20 plates $5.00

BULLETIN

^ of CARNEGIE MUSEUM OF NATURAL HISTORY

CONVERGENT EVOLUTION AMONG DESERT RODENTS:

A GLOBAL PERSPECTIVE

p]... MICHAEL A. MARES

P NUMBER 16 PITTSBURGH, 1980

BULLETIN

of CARNEGIE MUSEUM OE NATURAL HISTORY

CONVERGENT EVOLUTION AMONG DESERT RODENTS:

A GLOBAL PERSPECTIVE

MICHAEL A. MARES

Research Associate , Section of Mammals;
Department of Biological Sciences, University of Pittsburgh,
Pittsburgh , Pennsylvania 15260; and
Pyniatuning Laboratory of Ecology,

Linesville, Pennsylvania 16424

NUMBER 16

PITTSBURGH, 1980

BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY
Number 16, pages 1-51, 25 figures, 19 tables

Issued 15 January 1980

Price: $3.50 a copy

Craig C. Black, Director

Editorial Staff: Hugh H. Gennoways, Editor. Duane A. Schlitter.

As.sociatc Editor: Stephen L. Williams, Associate Editor:
Nancy J. Parkinson, Technical Assitant.

© 1980 by the Trustees of Carnegie Institute, all rights reserved.

CARNEGIE MUSEUM OF NATURAL HISTORY, 4400 FORBES AVENUE
PITTSBURGH, PENNSYLVANIA 15213

CONTENTS

Abstract 5

Introduction 5

Desert Rodents: Colonization, Adaptation, and Convergent Evolution 5

A Three-desert Comparison of Convergent Evolution 12

Methods 13

Results and Discussion 14

An Analysis of the Desert Rodents of the World 28

Convergent Evolution of Desert Rodents 37

Acknowledgments 43

Literature Cited 43

Appendix 1 48

Appendix 2 50

s

-v

h: ... ■

!i-

>■'

?

i' ^

(

ii; :j

J!:-

4

■)-,

■* ■ -i

' .iili

ABSTRACT

The biogeographic histories of New and (3ld World deserts
are discussed from the viewpoints of small mammal adaptation
to aridity and the development of small mammal communities
in deserts. Morphoecological characteristics of rodents are com-
pared between deserts using multivariate statistical techniques.
In one analysis, Sonoran and Iranian (Kavir) rodents are shown
to be more similar to one another than are the more closely
phylogenetically related rodents of the Sonoran and Monte des-
erts of the New World. It is proposed that the rodents in the two
northern deserts have had a longer time to adapt to arid condi-
tions than have those of the Monte, thus convergent evolution
is more pronounced between them. Nevertheless certain suites
of traits (that is, locomotion and trophic characteristics) are as-
sociated in reversed ways among some species in the more eco-
logically similar deserts; parts of the “niche” of these species
have been switched.

A comprehensive analysis of most genera of desert rodents
from all major deserts of the world reveals that species of the
Argentine Monte Desert are quite distinct from those of the other
deserts, particularly in their lack of specialized traits for desert
life. Species in the deserts of North America and Africa are more
similar ecologically to each other than they are to species in the
deserts of Asia and Australia. Cluster analysis of all species
reveals that there are a limited number of roles filled by rodents
in deserts. Thus, despite varying genetic background and bio-
geographic histories, small mammals in deserts evolve in such
a manner that they can be placed into a very limited number of
guilds. Moreover, these limited guilds are largely repeated within
each major desert region. Thus, there is pronounced convergent
evolution evident in morphology and ecology of the world’s des-
ert rodents.

INTRODUCTION

The earth supports many and varied habitats
which are largely a result of the complex interac-
tions of isolation, wind and water currents, precip-
itation, topography, and the distribution of land
masses. The overriding factor, however, is that the
earth is an essentially spherical planet, tilted slight-
ly on its polar axis, rotating at roughly a fixed dis-
tance from the sun. Indeed, as Beaty ( 1978) pointed
out, given the earth’s shape and orbit, it was easier
for Wegener to hypothesize the movement of the
seemingly immutable continents than it was to sug-
gest a general shift in the earth’s climatic belts in
order to account for the obvious climatic changes
evident in the geological and paleontological re-
cord. While climate may vary from point to point
over time (and vary greatly), and while glacial pe-
riods may wax and wane, dry areas have probably
always been a part of the climatological mosaic of
the biosphere (Axelrod, 1950, 1972). Because xeric
areas occur in a disjunct manner around the world
on continents that have had varied geological and
biological histories (Fig. 1), they form a unique se-
ries of ecosystems sharing many climatological
traits that are ideal for studying various facets of
the evolutionary process.

In this paper I will examine how groups of largely
unrelated rodents have adapted to the various des-
erts of the world. In some cases, the similarities of
the adaptive strategies utilized are remarkable, con-
sidering the vastly different gene pools from which
they were independently derived, while in other
cases similar problems have been solved utilizing
different adaptations.

Desert Rodents: Coeonization, Adaptation,
AND Convergent Evoeution

During the Permian, a supercontinent (Pangaea)
existed which was composed of all of the earth’s
land masses. Over time this continent was fractured
until a number of isolated continents were formed,
with the current pattern of continents appearing
only in the Cenozoic, although present-day land
connections between North and South America
were not completed until the late Pliocene (Dietz
and Holden, 1970; Haffer, 1970; Molner and Tap-
ponnier, 1975). For the majority of mammals,
whose evolution was only just beginning in the Cre-
taceous, the breakup of Pangaea had little effect on
their biogeographic history (Cracraft, 1974). Pres-
ent-day mammals evolved from early ancestors
which were either isolated on one or more of the
continental sections which broke off of the first land
mass, or which had to colonize the continents as
they reached their current locations. This means
that adaptation by mammals to arid regions proba-
bly began only in the middle to later Cenozoic (see
for example, Simpson, 1961; Romer, 1966; Lunde-
lius and Turnbull, 1967; Riek, 1970; Keast, I972r/,
19726; Cooke, 1972; Patterson and Pascual, 1972).

The world’s deserts, as we know them today,
vary in topography and certain climatic features.
Generally, however, they attained their pronounced
levels of aridity with the orogenic activity of the
Miocene and Pliocene (see for example, Furon,
1941, 1960; Axelrod, 1950, 1956, 1957, 1958, 1967,
1970, 1972; Choubert, 1952; Vuilleumier, 1971;
Cooke, 1972; Bailey et al., 1977). As mountain mas-

5

6

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

sifs were lifted above the landscape, wind currents
were disrupted, rain shadows were created, water
runoff into enclosed basins led to leaching and the
formation of salt lakes, primary productivity de-
creased with a concomitant decrease in plant cover
(thus increasing the surface albedo resulting in a
further decrease in rainfall), and high deserts, such
as the puna of South America, were formed (Logan,
1968; Vuilleumier, 1971; Otterman, 1974; Charney
et al., 1975). In all cases these climatic changes
were initially gradual, occurring over millions of
years, and thus allowing gradual change by vege-
tational communities and their associated fauna. In
North and South America, for example, the tran-
sition from rather mesic forest through grassland
and thorn scrub to desert has been well documented
(Axelrod, 1958; Patterson and Pascual, 1972; Sol-
brig, 1976). Indeed, the major deserts of the world
appear to have developed their extreme aridity over
a time span of 20 million years or more. Although
it is known that there have been significant climatic
fluctuations in the past (particularly in the Pleisto-
cene), and while these have probably had a great
effect on mammalian speciation patterns, desert
plants and animals are largely the result of a slow
evolution through time from a period of greater
moisture to one of a moisture deficit (Martin and
Mehringer, 1965; Hubbard, 1974; Grenot, 1974;
Van Devender, 1977; Mares, 1979).

Deserts pose several challenges to small mam-
mals. The suite of adaptive strategies employed in
response to heat, aridity, and low vegetative cover
is becoming increasingly well understood (Schmidt-
Nielsen, 1967; MacMillen, 1972; Mares, 1973), and
illustrates the process of convergent evolution (that
is, the channeling of adaptations among distantly-
related organisms by similar selective pressures to-
ward a particular set or subset of similar morpho-
logical, physiological, or ecological characteristics).
Nevertheless, in some major desert areas small
mammal inhabitants seem to be acquiring adapta-
tions for desert life, but they have not yet reached
the pronounced levels of xeric adaptation exhibited
by counterparts living in other deserts. I will use
the colonization of Australia and South America by
rodents to illustrate this point.

Australia originally was a part of the southern
section of Pangaea (Gondwanaland) and was largely
tropical in climate. As Gondwanaland broke up in
the Cretaceous, Australia was connected to Ant-
arctica and shared faunal elements with that conti-
nent (Raven and Axelrod, 1972). Gradually the

movement of the Australian plate carried the Aus-
tralian land mass into its current position located
between 11° and 38° South Latitude and 112° and
153° East Longitude. This position places it over
the 30° Latitude high pressure area where descend-
ing, adiabatically-warmed air currents form sub-
tropical deserts (Logan, 1968). The only mammals
present on the continent during its earliest forma-
tion were monotremes and marsupials; because
monotremes were primitive mesic species (Keast,
19726), the first mammal species to adapt to the
newly developing arid area on the mountainless
continent were marsupials. In the Pliocene, the
Australian continent was colonized across water
barriers from southeast Asia by placental murid ro-
dents which subsequently underwent a great adap-
tive radiation resulting in a diverse array of ecolog-
ical types (which today includes 13 genera and over
60 species). Among these are members of five gen-
era (Leggadina, Notomys, Pseiidomys, Leporilliis,
Gyomys) which are components of the desert fauna
(Morton, 1979). Since the colonization route for
these island-hopping rodents was via tropical Asiat-
ic islands (Simpson, 1961), they could only begin
adapting to the extensive Australian desert in the
late Pliocene. Although several genera of rodents
have adapted to the Australian desert, species di-
versity and population density at any particular lo-
cality in the desert tends to be low (Watts, 1974;
Morton, 1979). Among those species which have
adapted to the desert, however, the most conspic-
uous adaptations are specializations in water con-
servation (production of an extremely concentrated
urine) and locomotion (bipedality) in species of the
genus Notomys (Walker, 1964; MacMillen and Lee,
1967, 1969; MacMillen et al., 1972; Purohit, 1974).
Since marsupials have had a longer period to adapt
to the Australian desert, it might be expected that
they would also exhibit specializations for life in an
arid region, and, indeed, such species as Dasycer-
ciis cristicatida , Sminthopsis crassicaiidata, S.
froggatti. and Setonix brachyurus are often com-
mon desert animals and possess appropriate phys-
iological and anatomical adaptations (Bartholo-
mew, 1956; Bentley, 1960; Schmidt-Nielsen and
Newsome, 1962; Walker, 1964; Crowcroft and God-
frey, 1968; Godfrey, 1968; Dawson and Brown,
1970; Purohit, 1971, 1976; Tyndale-Biscoe, 1973).

In South America, the situation is more complex.
South America and Africa were originally part of
the southern supercontinent and they separated
from one another in the early Cretaceous (Dietz and

1980

MARES— DESERT RODENT ECOLOGY

7

Fig. 1. — The deserts of the world (after Meigs, 1957). The three intensive study sites in the Monte Desert of Argentina, the Sonoran
Desert of Arizona, and the Dasht-e-Kavir Desert of Iran are indicated hy white dots and arrows.

Holden, 1970) as South America became a huge is-
land which drifted slowly northward and westward.
Possibly some of the early mammals (marsupials,
edentates, various ancient ungulates) were present
on the continent when it broke off of the larger land
mass, but subsequent mammalian evolution in
South America proceeded in isolation from the rest
of the world. In the latest Eocene or early Oligo-
cene a group of caviomorph rodents appeared on
the continent. These rodents may have come from
either North American stock (Simpson, 1950; Wood
and Patterson, 1970; Patterson and Pascual, 1972)
or from apparently closely-related African phio-
morph rodents (Lavocat, 1969; Hoffstetter and
Lavocat, 1970; Hershkovitz, 1972). At about the
same time, platyrrhine primates colonized over
water from North America. Bats were probably
present at this time, and procyonid carnivores ap-
peared in the Miocene (Savage, 1951; Patterson and

Pascual, 1972). No other mammal groups entered
the continent until after the completion of the Cen-
tral American land bridge in the later Pliocene. At
that time there was a great influx of diverse mammal
types (seven orders, 16 families), which gradually
began moving southward on the continent.

Since deserts first formed on the South American
continent in the Miocene-Pliocene period, the ear-
liest species to adapt to these arid areas were the
caviomorph rodents and the marsupials. Most no-
table of the latter group were members of the family
Argyrolagidae, a group of bipedal, rodent-like ani-
mals which were apparently ecological equivalents
of present-day jerboas or kangaroo rats. These
species were found in southern and northwestern
Argentina in areas that are today desert, xeric
serub, and grassland (Simpson, 1970). The recent
immigrants to South America probably only en-
countered the extensive Monte Desert of Argentina

8

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 16

in the latest Pliocene or early Pleistocene, at a time
when climatic events were greatly altering the ex-
tent and distribution of arid habitats (Mares, I975u,
1976; Solbrig, 1976). There is good evidence that
the deserts of South America were colonized in two
major waves from North America, particularly
since close fossil relatives of extant South Ameri-
can species have been found in Arizona (Baskin,
1978). The studies of Mares (I975n, 19756, 1976,
1977a, 19776, 1977c, \911d), Mares et al. (1977),
Mares and Rosenzweig (1978), and Williams and
Mares ( 1978) have all indicated that the Monte Des-
ert supports a depauperate rodent fauna; that
Monte rodents are not highly specialized for desert
life; that the most conspicuous faunal elements in
the Monte are caviomorph rodents, members of the
first wave of colonists to the continent; and that the
more recent colonists, muroid (cricetine) rodents,
have not evolved specialized desert species, per-
haps because there has not been enough time over
which such adaptations could have taken place.
Most cricetines which inhabit the Monte today
either live in patches of more mesic habitat within
a larger arid region, or are widespread throughout
the thorn scrub or dry montane habitats (for ex-
ample, puna) which border the desert. Many of
their adaptations to aridity (for example, ability to
exist with little free water, ability to utilize salt so-
lutions to obtain water, etc., Mares, 1977a, 19776,
1977c, \911d) could have evolved as responses to
aridity in the high Andean deserts as the animals
colonized the continent, and thus functioned as pre-
adaptations for life in the lowland Monte Desert
when they finally reached the southern third of the
continent. The only endemic non-caviomorph ro-
dent in the Monte, Andalgalomys olrogi, is closely
related to a Paraguayan species and appears to be
a relict from a previously widespread Chacoan
ancestor (Williams and Mares,. 1978).

Recently Marshall (1979), using data in Marshall
et al. (1979), Marshall and Hecht (1978), and Reig
(1979a, 19796) has suggested that cricetine rodents
may have entered South America as early as 7 Myr
BP, or at approximately the same time as the pro-
cyonids crossed the water barrier of the Bolivar
Trough. This suggestion corresponds roughly to
those of Hershkovitz (1966, 1972), and, while un-
supported by fossil evidence in northern South
America, is based on the appearance of modern cri-
cetine rodent genera in fossil beds of the Monte-
hermosa Fauna of southern Buenos Aires Province,
dated at about 3.5 Myr BP (Reig and Linares, 1969).

The neighboring Chapadmalal Fauna contained an
even greater number of modern genera and was dat-
ed at about 2.7 Myr BP; the latter date corresponds
to the suggested time period for the completion of
the Panama Land Bridge (see Marshall et al., 1979).
Marshall logically reasons that it is unlikely that this
rather high diversity of pastorally-specialized ro-
dent species could have developed from what were
probably sylvan ancestors, although Baskin (1978)
has given strong evidence that at least some South
American cricetines (Calomys, a generalized genus
suggested by Hershkovitz, 1962, to be ancestral to
some of the phyllotine rodents) evolved in North
America. If cricetines were able to enter the South
American land mass as early as suggested, some
mechanism to delay their appearance in fossil beds
of southern South America by about 3.5 Myr is
needed. Marshall proposes that Savanna-grassland
habitats did not allow a natural colonization route
for northern South American species to reach
southern South America until the combined activi-
ties of orogenic and glacial events disrupted the
major macrohabitats of the continent such that dry
habitats became contiguous north and south of the
Amazon; since the earliest fossils are already spe-
cialized for such habitats, it is suggested that such
specialization took place in northern South America
(for example, Venezuela, see Sarmiento, 1976).
Support for Marshall’s ideas concerning habitats
are available from various lines of evidence (for ex-
ample, Mercer, 1973, 1976; Van der Hammen, 1974;
Simpson, 1975; Webb, 1978). While this view is
markedly opposed to that of Simpson (1951) or Pat-
terson and PascLial (1972), it is still unlikely that
colonization of the Monte by cricetine rodents took
place much before the Pliocene-Pleistocene inter-
face; no fossil cricetines are known from beds lo-
cated in areas corresponding to either present-day
or suggested Plio-Pleistocene Monte Desert limits.

Mares (1975a, 1976, 1979), Mares and Hulse
(1977), and Mares, Enders et al. (1977) have dis-
cussed speciation patterns in the deserts of North
America and the Monte Desert of Argentina. Both
alpha and beta species richnesses are much greater
in the northern deserts, not only for rodents, but
for other groups of mammals as well. The overall
differences in the speciation patterns (and thus pat-
terns of species richness) between the North and
South American desert systems may be due to dif-
ferent effects of Pleistocene glaciation in each area.
In North America the Pleistocene probably frac-
tured a large, fairly continuous dry area into a num-

1980

MARES— DESERT RODENT ECOLOGY

9

Fig. 2. — A comparison of two desert rodent faunas (Monte and Sonoran sites) and a non-desert coniferous forest rodent fauna from
New Mexico utilizing a canonical analysis based on 28 morphoecological traits. The first two canonical variates account for essentially
all of the variance in dispersion. The two desert faunas, being plotted closely together on the first axis, are more similar to one another
than either is to the forest fauna, even though the latter is closely related, phylogenetically, to the Sonoran assemblage (after Mares,
1973). Individual mean species values are shown by a letter symbol, whereas the faunal mean is given by the large circle.

her of xeric refugia, thus forming a system condu-
cive to species multiplication via geographic
isolation, whereas in the Monte Desert, the Pleis-
tocene may have isolated the desert ecosystem into
a single, relatively small refugium in which some
species (including, perhaps, the argyrologid mar-
supials) went extinct (Mares, 1979). A system
whereby a single refugium formed repeatedly would
function as an extinction system leading to reduced
species diversity because of the island nature of
such a desert preserve (see MacArthur and Wilson,
1967; Diamond and May, 1976), while a system of
multiple desert refugia, such as what probably ob-
tained in North America during the Pleistocene,
would act as a species multiplication system
(Mares, 1979).

Obviously, the rodents of the North and South
American desert systems were quite distinct, phy-
logenetically, from each other as well as from those
species which were entering Australia at about the
same time. Yet each group of rodents encountered
elevated temperatures, great insolation, low and spo-
radic precipitation, and low productivity. Other ro-
dents were encountering similarly harsh environ-
ments during their independent evolution in Africa
and Asia.

Mares (1976) compared morphological and eco-
logical aspects of a locality in the Monte Desert
(near Andalgala, Catamarca Province, Argentina)
with one from a site in the Sonoran Desert (near
Tucson, Arizona, USA). The Sonoran Desert is
quite similar fioristically, climatologically, and geo-

10

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 16

morphologically to the Monte Desert (Orians and
Solbrig, 1977), although its rodent fauna has been
associated with developing arid, semiarid and grass-
land communities since at least the Oligocene
(Wood, 1935; Lidicker, 1960; Voorhies, 1975), and
thus they have had more time to adapt to aridity.
Mares also compared the rodent fauna of a non-
desert community (a Western Yellow Pine area in
central New Mexico, USA) with the two desert fau-
nas to determine if the latter were more similar to
each other morphologically than were the two
North American faunas. Phylogenetically, the two
northern faunas were more closely related than
were those of the North and South American des-
erts. By subjecting numerous morphological char-
acteristics, which were selected because they pre-
sumably reflected ecological functions, to several
multivariate mathematical techniques (principal
components analysis, cluster analysis, discriminant
function analysis, and canonical analysis), Mares
was able to show that the rodents from the two
desert areas were more similar to one another, de-
spite the fact that they were more distantly related,
than were the two North American faunas (Fig. 2).
Convergent evolution had occurred between the
desert rodents and similarities were particularly ap-
parent in several traits associated with a desert ex-
istence (for example, various dental characteristics,
inflated tympanic bullae, light-colored pelage, and
others). These similarities might have been even
greater had the historical biogeographic histories of
the two deserts been more comparable.

In these earlier papers (Mares, 1975^, 1976), I
attempted to assess the degree of convergent evo-
lution existing between two disjunct desert rodent
faunas as indicated by the mathematical techniques.
Before I utilized morphological measurements in
the comparative analyses, however, I spent much
time and effort studying the distribution, natural
history, reproductive biology, physiology, and pop-
ulation ecology of many of the Monte Desert
species; Sonoran Desert rodents had been well-
studied by others for more than a half-century. My
familiarity with many aspects of the biology of the
Argentine species allowed me to arrive at a rough
determination of the degree of difference or simi-
larity between the two amphitropical desert faunas.
This determination was subjective in the sense that
any ecologist who had spent a number of years
working with the mammals of the Monte, and who
possessed familiarity with the northern species,
would very likely have developed some opinions

regarding which species might fill similar niches in
each area. Although some proposed examples of
convergent pairs were obvious (for example, Cten-
omys and Dolichotis of the Monte versus Tliomo-
inys and Lepus of Arizona), other suggested equiva-
lents might have found less acceptance among
investigators (for example, Octomys and Microca-
via of the Monte versus Neotoma and ground-
dwelling sciurids of Arizona). Thus the problem I
faced was to encounter some method of objectively
comparing at least portions of the niche space of
each species, that is, its ecological position within
each desert.

Form and function have long been known to be
associated (for example, Darwin, 1859, discussing
Galapagos Island finches; Thompson, 1917). An or-
ganism’s morphology is the product of synergistic
interactions of its environment and its genetic
makeup, but various functional attributes can gen-
erally be deduced from disparate morphological
traits. Many investigators (for example, Shoener,
1965; Tamsitt, 1967; Cody, 1975; Karr and James,
1975; Findley, 1976) have pointed out that numerous
ecological characteristics of organisms are strongly
correlated with morphological traits (see also Hes-
penheide, 1973). I found that by analyzing a number
of morphological traits I was able to arrive at a sta-
tistical comparison between rodents of both deserts
that seemed logical from the ecological point of
view. Some of these traits, such as the long hind
legs and tail of desert rodents which are associated
with bipedal locomotion, readily suggest a function;
others, such as the width across the zygomata have
a less obvious functional relationship. In quantify-
ing these traits, and comparing them between fau-
nas, I distinguished between these two types of
morphological measurements (Appendix 1). In fact,
both sets of measurements gave a reasonable inter-
pretation of similarities between distantly-related
species, but it is sometimes easier to see the func-
tional applicability of one or more traits, because
they strongly reflect ecological function. To use an
extreme example, it would be quite surprising to
find an herbivorous felid possessing the shearing
teeth of a carnivore; the basis for assigning function
to form is well-grounded in comparative zoology.
Thus, I assigned an ecological function to diverse
morphological traits and termed these traits “mor-
phoecological” characteristics (Mares, 1975/?); the
subordinate position of the prefix eco to morpho is
intentional and points out that the traits are, above
all, morphological and may be quite labile in their

1980

MARES— DESERT RODENT ECOLOGY

II

Fig. 3. — A site in the North American desert near Needles, California, where low shrubs (especially creosotebush, Larrea tridentata)
and bursage (Franseha ) predominate.

ecological implementation. Nevertheless, I fee! that
it is possible to arrive at a first approximation of
locomotor, trophic, physiological, and habitat sim-
ilarities using nothing more than these types of mea-
surements. Also, since most desert rodents of the
world are poorly studied ecologically, such analy-
ses should yield an impression of the ecological
mosaic formed by the various rodent species com-
prising a particular fauna, as well as an indication
of the evolutionary forces that seem to mould fairly
predictable sets of species in widely scattered xeric
regions.

The limitations of this method are obvious. If we
begin with a well-studied fauna, such as the rodents
of the Sonoran Desert of Arizona, and compare
these with species which are essentially ecological-
ly unknown, then we arrive at a comparison of mor-
phoecometrics of the two groups. The underlying
assumption is that a member of the unknown fauna
sharing many functional traits with a species in the
well-studied group probably fills a similar role in its
own ecosystem. Thus a species in the Monte Des-
ert, such as Eligmodontia typus, might be closely
allied morphoecologically with Peromysciis ereini-
cus of Arizona, meaning that such traits as overall
body size and dimensions, ear length, omnivorous
food habits, scansorial locomotion, pelage color

and auditory bullae size, are shared between the
species. E. typus differs in having longer hind feet
and a longer tail, which reflects its propensity to
inhabit fairly open sandy habitats. In this respect it
is more similar to members of the genus Perog-
natluis, and this fact can be inferred from the anal-
yses. Such traits of Elignodontia such as its ability
to extract water from cacti or to process solutions
of extreme salinity (Mares, \915a , \911a) are not
apparent from the measurements analyzed. Thus,
regardless of the degree of sophistication of a mor-
phological study, many important attributes of a
species will only be discovered through intensive
field and laboratory investigations.

Up to now comparative multivariate morphomet-
ries have been limited to analyses within higher taxa
(Chiroptera, Rodentia, small birds, etc.) because of
the difficulty in utilizing sets of measurements
across morphologically unlike groups. Hence such
important competition studies of widely differing
taxa, such as those concerning the interactions of
granivorous desert ants and mammals (Brown and
Davidson, 1977), do not lend themselves to these
methods. However, the multivariate Monte-Sono-
ran desert rodent comparison did suggest that gran-
ivory was an important attribute of rodents of Ari-
zona and unimportant among those of the Monte.

12

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 16

Fig. 4. — A view of the Campo Arenal in the Monte Desert of Catamarca Province, Argentina, with Larrea cuneifolia as the dominant.

These observations led directly to a comparison of
the patterns of granivory of birds, ants, and rodents
in the two deserts (Mares and Rosenzweig, 1978),

With the limitations of these techniques in mind
(Appendix 2), and with a belief that a first compar-
ison of the ecological aspects of the desert rodent
faunas of the world is at least desirable, I will pro-
ceed to an analysis of morphoecological traits of the
numerous species of rodents which have managed
to successfully inhabit one of the earth’s most chal-
lenging regions.

A Three-Desert Comparison of
Convergent Evolution

The rodent faunas of the Sonoran and Monte des-
erts were not as similar as might have been ex-
pected given the great similarities in their environ-
ments. If the lack of more pronounced convergent
evolution is a result of a time factor (perhaps be-
cause the Monte rodents have not had sufficient
time to specialize for desert life because of their
late arrival to the Monte), then one can presumably
test this hypothesis by comparing both deserts with
a third, environmentally-similar area with an evo-
lutionary history more like that of the North Amer-
ican desert. In the following example, the third des-
ert area chosen was the Dasht-e-Kavir of Iran.

The Iranian Desert has a winter-rain Mediterra-
nean climate much like the Great Basin Desert of
North America, and not like the bimodal, summer-
winter precipitation characteristic of the northern
Sonoran Desert, or the summer rainfall of the north-
ern Monte (Ganji, 1955; Jaeger, 1957; Morello,
1958). Basically, however, all three areas are warm,
subtropical deserts. All have a basin and range to-
pography, with the two northern deserts being most
similar in this regard (Zohary, 1963; Lustig, 1968;
Logan, 1968). The two New World deserts have a
pronounced tree and tail-cactus component, while
the Kavir is a low shrub desert, but extensive areas
of physiognomically similar habitats can be found
in all three deserts (Figs. 3-5).

The rodents of the Dasht-e-Kavir have had a long
history of association with arid areas, either in Iran
proper or in more northern Old World deserts, and
they are phylogenetically less closely related to the
rodent faunas of the two New World deserts than
the latter are to each other (Simpson, 1945; Eller-
man, 1949; Hall and Kelson, 1959; Dawson, 1967).
Thus any similarities between the rodent faunas of
the two deserts which have had a similar period of
time for desert adaptations to develop would have
to override those characteristics which might be
due to common inheritance of the Sonoran and
Monte desert rodents (that is, parallelism).

1980

MARES— DESERT RODENT ECOLOGY

13

Fig. 5. — A locality in the Dasht-e-Kavir Desert of Iran about 100 km southeast of Tehran, chenopodeaceous shrubs predominating.

Methods

The basic statistical techniques utilized in the following anal-
yses are stepwise discriminant function analysis, canonical anal-
ysis, and cluster analysis. The first technique is a fairly straight-
forward method of comparing the ability of a number of variables
to distinguish between groups that have been previously delin-
eated. Thus, if variable a is some morphological measurement
that best distinguishes between the groups, and variable is a
separate variable, uncorrelated with a. and that is the second
best variable to distinguish between the groups, then these two
variables are weighted and combined to form a linear discrimi-
nant function. The number of discriminant functions formed may
be one less than the number of groups, or may be equal to the
number of variables, depending on the significance with which
the groups are distinguished. After the discriminant functions
are formed, the groups are classified and the individual members
comprising each group are compared to the overall group vari-
ables to assess the statistical validity of their inclusion within a
particular group. This allows for a determination of the precision
of group assignations, as well as a method of calculating the
probability that a group member actually belongs in another of
the assigned groups rather than in the group under consideration.

The second technique, canonical analysis, proceeds from the
first. Here, the original variables are used to form a separate set
of canonical variables which are themselves uncorrelated with
one another. The members of each group are then plotted along
each canonical axis (often the first two axes explain the greatest
amount of variance in the data), and the distances in /(-dimen-
sional space (Mahalanobis distance), where // = the number of
variables, are given for each group member to its group mean
value, and to every other individual being examined. It is thus
possible to get a 2-dimensional visual impression of /(-dimen-
sional spatial relationships. For example, if there is great overlap

between groups with many of their component members being
similar to others in separate groups, the plot would indicate ap-
proximately the same space being occupied by all groups. Sim-
ilarly, if only one or two group members are generally similar to
those of another group, the former may be plotted more closely
to members of the latter group (see also Heyck and Klecka,
1973; Klecka, 1975; Cody, 1978). Cluster analysis has been
widely discussed in the literature and will not be explained here
(for example, Cooley and Lohnes, 1971; Sneath and Sokal,
1973).

The species analyzed in the three-desert comparison are listed
in Table I . while Table 2 lists the morphoecological traits utilized
in the analyses. The rationale for measurements is discussed at
length in Appendix I. In addition to the rodents used in the
earlier North American-South American comparison, represen-
tatives of rodent species from a locality located approximately
100 kilometers southeast of Tehran, Iran, were included in the
present analysis. Individuals were measured and the data were
subjected to various multivariate techniques including stepwise
discriminant function analysis and canonical variate analysis
(BMDP and SPSS programs), and the cluster analysis techniques
of the NT-SYS computer program package of F. James Rohlf of
the State University of New York at Stony Brook. The un-
weighted arithmetic pair-group method with averages was used
to generate phenograms, and both distance and correlation pho-
nograms were plotted.

I separated the North and South American rodent assemblages
into a number of loose functional groups. These included the
following categories; Dipodomys (the bipedal heteromyids); Per-
ognathus (the quadrupedal heteromyids); Sonoran (the remain-
der of the Arizona rodents); “gophers” (the highly fossorial Tho-
momys of Arizona and Ctenoniys of Argentina); and Monte (all
Monte rodents excepting Ctenomys). Each rodent species from
the Iranian Desert was entered into the multivariate analysis as

14

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 16

Table I. — Rodent species in one locality each of the Monte Desert of South America, the Sonoran Desert of North America, and the

Dasht-e-Kavir Desert of Iran.

Sonoran (Tucson)

Monte (Andalgaia)

Iran (near Tehran)

Family Heteromyidae

Family Cricetidae

Family Cricetidae

Perognathus baileyi

Subfamily Cricetinae

Subfamily Cricetinae

Perognathiis intermedins
Perognathus penicillatus
Perognathus flavus
Dipodomys merriami
Dipodomys spectabilis

Tribe Hesperomyinae
Eligmodontia typus
Phyllotis griseofiavus

Family Caviidae

Tribe Cricetini

Calomyscus bailwardi
Subfamily Gerbillinae
Meriones lihycus

Dipodomys ordii

Microcavia australis

Meriones crassus

Family Geomyidae

Family Octodontidae

Meriones persicus
Gerbillus nanus

Family Dipodidae

Allactaga elater
Jaculus blandfordi

Thomomys bottae
Family Cricetidae
Subfamily Cricetinae
Tribe Hesperomyinae

Octomys mima.x
Family Ctenomyidae

Ctenomys fulvus

Neotoma albigula
Peromyscus eremicus
Peromyscus maniculatus
Onychomys torridus
Reithrodontomys megalotis
Reithrodontomys fidvescens
Sigmodon hispidus

Family Sciuridae

Ammospermophilus harrisi

being of “unknown affinity”, and the computer procedure was
to assign each unknown to one of the previously described major
groups, the assignation being based on the number and interre-
lationships of shared character states.

Results and Discussion

The rodent fauna of the Kavir site is not nearly
as rich in species as is the Tucson site, rather it
resembles that of the Andaigala locality in the
Monte Desert (Table 1, Figs, 14-16). Rodents seem
to be abundant in the Kavir, however (Lay, 1967;
Mares, personal observation), and the low densities
of rodents and other small mammals in the South
American desert is an uncommon observation for
a major desert area; this is also the case in Australia
(for example, Schall and Pianka, 1978; Morton,
1979). Possibly the fact that Monte rodents are less
specialized for desert life helps account for the dif-
ferences in relative abundances evident among
the three deserts, although the Pleistocene history
of extinction and the coevolutionary relationships
of plants, ants, and granivorous rodents might also
be an important factor in the smaller population
sizes of rodents supported in the Argentine desert
(Mares and Rosenzweig, 1978; Mares, 1979; Mor-
ton, 1979).

The positioning of members of the Iranian, So-
noran, and Monte rodent faunas on the first two
canonical axes when only external traits were uti-
lized is shown in Fig. 6, while Table 3 gives the
loadings of each variable making up the first two
canonical variates. The expectation that the simi-
larities between Iran and Sonora would override the
phylogenetic relationships of the Sonoran and
Monte deserts was not entirely realized. The So-
noran and Monte rodents are closer to one another
than either is to the Iranian fauna. The mean Ma-
halanobis Distance (D^) of the Sonoran bipedal Di-
podomys species to the Iranian species is 9.8 units,
while the mean distance of each Sonoran species to
its faunal mean is 18.5 units. The distance of the
Sonoran bipeds to the Monte is 14 units. In fact,
two Sonoran species, the bannertail kangaroo rat,
Dipodomys spectabilis, and the grasshopper mouse,
Onychomys torridus, were assigned to the Iranian
desert. Two other species, the white-throated
woodrat, Neotoma albigula, and the gopher, Tho-
momys bottae, were placed with the Monte ro-
dents. The assignments were undoubtedly based on
the Iranian desert having few species, but two of
these are bipedal forms (like Dipodomys) and one
is a cricetine with coronally-hypsodont dentition

1980

MARES— DESERT RODENT ECOLOGY

15

Table 2. — Morphological measurements, ratios, and categories used in the various multivariate determinations. Measurements denoted
by an asterisk(*) are traits which reflect ecological aspects quite strongly (Figs. 2! , 23). Those marked with a ( ') were used in the analyses
given in Figs. 6, 10; those with a were used in computing Figs. 7, II , 19; those with a were utilized in Figs. 20. 24 whereas all traits

(except weight) were used in Figs. 18. 22.

External Crania!

1) Head-body length (HBL)’--'® 15)

2) Tail length (TL)*-^ 16)

3) Hind foot length (HFL)'’® i7)

4) Height of ear from notch (EL)^’^ 18)

5) Length of longest vibrissae (VL)*’^ 19)

6) Length of hair between shoulders 20)

*7) Tail length/head-body length (TL/HBL)* 2!)

*8) Hind foot length/head-body length (HFL/HBL)' 22)

*9) Ear length/head-body length (EL/HBL)' *23)

*10) Vibrissae length/head-body length (VL/HBL)' *24)

*1!) Hair iength/head-body length (HL/HBL)' 25)

*12) Weight (W)> *26)

*13) Foot bristles (FB)*’^ *27)

*14) Tuftiness of tail (TT)’’^ *28)

*41) Vibrissae density (VD)*-^ *29)

*30)

*31)

*32)

*33)

*34)

*35)

*36)

*37)

*38)

39)

*40)

Basal length (BL)^’“

Incisor-molar length (IML)''*

Bullar length (UL)^

Bullar width (UW)-*

Width across molariform tooth rows (mouth width) (MW)-’
Zygomatic breadth (ZB)’’*

Incisor width
Incisor length (IL)-'^

Incisor-molar length/basal length x 100(1ML/BL)-
BuMar index (Ul)‘

Length of molar tooth row (TRL)^"’

Tooth row length/incisor-molar length x iOO (TRL/IML)^
Width across tooth rows/basal length x 100 (TRL/BL)^
Zygomatic breadth/basal length x 100 (ZB/BL)

Incisor width/basal length x 100 (IW/BL)^

Incisor length/basal length x 100 (IL/BL)^

Incisor angle (lA)-”’

Seizer-digger incisors (SZ)-’^

Triturator incisors (TI)--^

Molar planation (MP)-’^

Molar complexity (MC)^’^

Tubercular hypsodonty (TH)-"’

Coronal hypsodonty (CH)-'^

Molar triangulation (MT)'’-’

Molar tooth row width (TW)-’^

Relative molariform surface area (SA)-

(like Onychomys). The Monte, with even fewer
species, has a rodent {Octomys mi max) which
strongly resembles a woodrat, externally, and a fos-
sorial tuco-tuco (Ctenomys) which is practically in-
distinguishable from a North American geomyid.

The location of the three desert faunas plotted on
the first two canonical axes when only dental traits
are utilized is given in Fig. 7; Table 4 shows the
weightings of the various characteristics on the ca-
nonical variates. As before, the Sonoran and Monte
rodents are closer to each other than either is to the
Iranian group, while the latter is most similar to the
North American desert. This same pattern was ob-
tained using 25 ecological variables.

Correlation and distance phenograms were com-
puted using 25 morphoecological variables (Figs. 8
and 9). In the correlation phenogram (Fig. 8), a
number of points are of interest. The first cluster is
composed of a large group of North American des-
ert specialists, the heteromyids (pocket mice and
kangaroo rats, here termed “K-rats”), which are
grouped fairly tightly, and this cluster is loosely
linked to one containing the two bipedal dipodids.

AUactaga and Jacidus, of the Iranian Desert. No
Monte species are included within this large cluster.
The second grouping includes most Sonoran
Desert small scansorial micro-omnivores ("pero-
myscines”), as well as the leaf-eared mouse,
Phyllotis (Graomys) griseoflavus, of the Monte.
The carnivorous-insectivorous grasshopper mouse
of Arizona, Onychomys torridus, is closely clus-
tered with the hamster of Iran, Calornysciis bail-

Table 3. — The coefficients for each of the original variables
forming the first two canonical variates in Fig. 6.

Original

variable

Canonical
variate 1

Canonical
variate 2

TL/HBL

0

0

HFL/HBL

0

_ ^2

EL/HBL

-.!

.1

VL/HBL

0

0

HL/HBL

-.1

0

W

0

0

FB

0

1.5

TT

-.6

-.3

VD

2.5

.5

16

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 16

Fig. 6. — A canonical analysis of external morphological traits comparing the Monte. Sonoran, and Iranian rodent faunas. Individual
species means are shown by the appropriate letter symbol, whereas the faunal mean is denoted by an asterisk (*).

wardi. The next clusters involve Arizona and Ar-
gentine species only, with no Iranian species being
grouped with (that is, being ecological equivalents
of) the “wood rats” (scansorial, medium-sized her-
bivores), the “ground squirrels” (scansorial, me-
dium-sized omnivores and/or herbivores which bur-
row extensively), or the “gophers” (highly fossorial
root, tuber, and above ground vegetation feeders).
The final cluster, loosely joined to the “gopher”
group, is comprised of the Iranian jirds, Meriones,
and the Monte caviomorph, Octomys mimax.

The distance phenogram (Fig. 9) presents a some-
what different view of the three desert rodent fau-

nas. Now there is a major cluster of desert rodents
made up largely of the heteromyids, the small Ger-
hillus nanus and the Meriones species of Iran, and
the Monte caviomorph, O. mimax. The other clus-
ters are similar to those of the correlation pheno-
gram, except that the bipedal dipodids are only
loosely grouped with the rest of the desert rodents.
The inclusion of Octomys within the category
which I have termed “desert specialists” is not sur-
prising considering that this is a highly desert adapt-
ed caviomorph which inhabits extremely arid areas
within the Monte subsisting largely on cacti and
other vegetation and, in other respects as well.

1980

MARES— DESERT RODENT ECOLOGY

17

Fig. 7. — A canonical analysis utilizing 23 dental traits in a comparison of three desert rodent faunas. Individual species mean values
are denoted by the appropriate letter symbol, whereas the faunal mean is indicated by an asterisk (*).

being similar to the wood rats (Neotoma) of North
America (Mares, 1976). However, the failure of the
dipodids to be clustered, even tenuously, with the
heteromyids was not expected. Both of these
groups of desert rodents have long been considered
ecological equivalents that have strongly converged
in their morphology and ecology in becoming highly
adapted desert specialists (Howell, 1932; Vaughn,
1972; Gunderson, 1976).

In order to determine why such a priori examples
of convergent evolution were not clustered together
as ecological equivalents, I performed a series of
stepwise discriminant function and canonical anal-
yses on two subsets of the data. In examining the
characteristics that were utilized in the cluster anal-
yses, it became apparent that two major ecological
parameters were being measured — food habits and
food procurement (certain dental and cranial mea-
surements); and locomotion and overall external
appearance (most external measurements). When
these two groups of traits were analyzed separately

(Table 2), each analysis produced a markedly dif-
ferent pattern.

The canonical analysis based on external char-
acteristics is given in Fig. 10. The Mahalanobis Dis-
tance (D-) values for the group means and for the
various unknowns are given in Table 5, while the
approximate weightings of each trait forming the
canonical variates are shown in Table 6. Both Al-
lactaga and Jacuhis were assigned to the Dipodo-
inys group, along with Gerbillus nanus. Callomys-
ciis and the two jirds. Me none s crassus and M.
persicus, were assigned to the Sonoran assemblage,
while the remaining jird, M. lihycus, was assigned
to the “gopher” group. From the loadings on the
first two canonical variates, it can be seen that Ca-
nonical Variate 1 is composed primarily of variables
describing foot bristles, relative ear, hair, and hind
foot lengths, and vibrissae density. On Canonical
Variate 2, vibrissae density, tail tuftiness, and rel-
ative ear and hind foot lengths are weighted fairly
heavily. Thus it appears that species located on the

18

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

CORRELATION

0

•5

1

DIPODOMYS

DIPODOMYS

DIPODOMYS

PEROGNATHUS

PEROGNATHUS

PEROGNATHUS

PEROGNATHUS -

GERBILLUS

OCTOMYS

MERIONES

MERIONES

MERIONES

ALLACTAGA

JACULUS

NEOTOMA

SIGMODON -

PHYLLOTIS.

AMMOSPERMOPHILU

MICROCAVIA

THOMOMYS'

CTENOMYS”

PEROMYSCUS

PEROMYSCUS

REITHRODONTOMYS

REITHRODONTOMYS

ELIGMODONTIA

ONYCHOMYS '

CALOMYSCUS

Desert Specialists

"Pack Rats"

"Ground Squirrels"
"Gophers"

"Peromyscines"

Insectivores"

Fig. 8. — A correlation phenogram of the rodents of three deserts based on 25 morphoecological traits. Allactaga and Jaculus. thought
to be Old World equivalents of kangaroo rats (Dipodoinys), are not clustered with the heteromyids. Most Sonoran and Iranian species
fall into the “desert specialist" category.

Table 4. — The coefficients for each of the original variables
forming the first two canonical variates in Fig. 7.

Original

variable

Canonical
variate 1

Canonical
variate 2

W

-.1

0

IML/BL

.1

-.1

TRL/IML

0

— .2

MW

-.4

.6

ZB/BI.

.1

-.1

IW/BL

2.2

1.4

IL/BL

-1,0

-.7

lA

-1,8

-.7

SZ

-6.1

-2.3

MP

1.5

1.0

MC

3.5

-.2

IH

.4

.3

CH

-2.6

.7

MT

2.0

-2.5

SA

31.8

-6.2

left half of the figure would possess dense vibrissae
and have fairly long hind feet, while those located
in the upper part of the figure would have tails with
conspicuous tufts, as well as relatively long hind
feet and ears. It is evident from Table 5 that the D“
distances from the Kavir species to each of the as-
signed group means were often quite close and final
determination of the most similar group was based
on small differences in //-dimensional space. Ger-
billiis nanus, for example, was located 54.5 units
from the Dipodomys mean and only 54.9 units from
the Sonora mean value. The slightly greater dis-

1980

MARES— DESERT RODENT ECOLOGY

19

DISTANCE

-8 -I

Diponnws ■

DIPODOflYS
DIPODOMYS
OCTOMYS ,
NEOTOMA
SIGMODON -
PHYLLOTIS ,
MERIONES
MERIONES .

HER I ONES

PEROGNATHIIS'

PEROGNATHIIS

PEROGNATHIIS ‘

PEROGNATHUS

GERBILLIIS ,

PEROMYSCIIS

PEROMYSCIIS

REITHRODONTOMYS

REITHRODONTOMYS '

EUGfiODONTIA

ONYCHOWS

CALOMYSCIJS

AmOSPERriOPHiUJS'

mCROCAVIA

THOMOMYS ].

CTENOMYS J

ALLACTAGA

JACULUS

"K-Rats

"Pack Rats"

".JiRDS

"Pocket Mice

"Peromysc I NES"

"Ground Squirrels

"Gophers"

Di PODins

Fig. 9. — A distance phenogram of the rodents of three deserts based on 41 morphological traits. Dipodids are greatly separated from
the remainder of the rodents.

Table 5. — Square of the Mahalanohis Distance (D-) of each of the groups and unknowns examined in an analysis of external char-
acteristics (Fig. 10). Mean ± SD.

Group

Mean D* values

Dipodomys

Perognathus

Sonora

“Gophers’*

Monte

Dipodomys

4.3 ± 3.0

40.8 ± 23.6

170.9 ± 25.3

335.7 ± 39.2

216.1 ± 18.2

Perognathus

39.5 ± 5.6

3.0 ± 1.2

138.0 ± 20.9

285.7 ± 26.7

216.3 ± 26.7

Sonora

175.6 ± 29.9

144.0 ± 26.6

8.9 ± 3.0

69.4 ± 17.5

37.2 ± 9.0

“Gophers”

335.4 ± 39.5

286.7 ± 32.1

64.4 ± 10.0

3.95 ± 0

84.1 ± 5.4

Monte

221.8 ± 23.7

223.3 ± 36.0

38.2 ± 15.7

89.8 ± 6.9

10.0 ± 1.3

Calomyscus

121.1

146.8

42.6

158.9

60.8

Gerhillus

54.4

62.7

54.9

164.9

107.8

Meriones lihycus

323.2

315.1

48.4

42.1

46.8

Meriones crassus

203.7

189.1

17.0

40.9

46.7

Meriones persicus

143.9

147.8

39.1

124.1

78.7

Allactaga

447.5

631.7

628.3

853.6

479.7

Jaculus

131.4

264.5

485.2

691.5

462.2

20

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

Canonical Variate I

Fig. 10. — A canonical analysis of rodents from three desert localities based on external traits. Iranian rodents have been assigned as
unknowns, whereas the North American and Monte rodents have been broken into a number of functional groups. Bipedal Allactaga
is placed with Dipodomys: quadrupedal Meriones is placed with the "gophers.” x equals group means; dots equal cages.

tance from the Sonora mean is almost certainly due
to the disparate nature of the species comprising
The Sonoran assemblage, including medium-sized
ground squirrels (Ammospennophihis) and cotton
rats (Sipmodon), as well as small scansorial crice-
tines. It should be noted that the distance of Ger-
hillus to the Peropnatliiis mean was only 64.4 units.
Allactapa was over three times farther from the
Dipodomys mean than was Jacidns (447.5 versus
131.4 units, respectively).

In the analysis of dental characteristics (Eig. 11,
Tables 7 and 8), the three species of Meriones were
all assigned to the Dipodomys category, with M.
lihycns being the most similar to the Dipodomys
group. Gerhillns and Calomyseus were also as-
signed to the kangaroo rat assemblage. The two di-
podids were assigned differently, with Jacnlns
being placed with the kangaroo rats, while Alloc-
taga was grouped with the “gophers.” The weight-
ings of the various dental characteristics on the first

1980

MARES— DESERT RODENT ECOLOGY

21

o

o

5

a

o

o

O

Canonical Variate I

Fig. II. — A canonical analysis of desert rodents from three localities based on dental traits. Iranian rodents have been assigned as
unknowns, whereas North and South American species have been broken into a number of functional groups. Bipedal AUavtaga is
placed with the “gophers”; quadrupedal Merioiies lihycus is assigned to the Dipodomys groups. Symbols as in Fig. 10.

canonical axis (Table 8) show that the variables
which made up that axis were primarily the molar-
iform surface area, the possession of seizer-digger
incisors, coronal hypsodonty, molar planation pat-
terns, molar cusp patterns (for example, molar com-
plexity), relative incisor width and incisor angle.
Canonical Variate 2 was composed primarily of the
molariform surface area, seizer-digger incisors, mo-
lar planation and complexity, coronal hypsodonty,
relative incisor-molar length, and relative incisor
width. Thus species with pronounced seizer-digger
incisors or large molariform surface areas would
have low values (strongly negative) on Canonical
Variate 1, and the value of the canonical coefficient
would increase as the molariform surface area, or
the degree to which the incisors indicated a seizer-
digger function, decreased. The great differences in
the dental characteristics, and the great variability

of these traits, is apparent when the Mahaianobis
Distances are examined — most are quite large when
out-group comparisons are made.

The above analyses suggest that, within the So-

Table 6. — The coefficients for each variable forming the first two
canonical variates in Fig. 10.

Original

variable

Canonical
variale 1

Canonical
variale 2

TL/HBL

0.02

-0.04

HFL/HBL

-0.64

0.32

EL/HBL

0.71

0.49

VL/HBL

-0.27

-0.01

HL/HBL

0.28

-0.09

W

0.01

0.01

FB

3.86

0.20

TT

0.13

0.49

VD

-2.08

0.81

22

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

Fig. 12. — Bipedal desert rodents are found in both North America {Dipodomys) and Iran (AUactaga, upper right), whereas a highly
fossorial group {Tliomoinys. lower left) inhabits North American deserts, and a less fossorial, but extensively burrowing species
{Meriones lihycns, lower right) is found in Iran. The food habits of the externally similar species are reversed, however. Niche segments
(food and locomotion) have been switched.

Table 7. — Square i

of Mahalanohis Distance

(£)-) of each of the groups and unknowns
(Fig. II). Mean ± SD.

examined in an analysis

of dental characteristics

Mean D“ values

Dipodomys

Perognathus

Sonora

“Gophers”

Monte

Dipodomys

10.7 ± 0

*

*

*

*

Perognathus

*

12.0 ± 0

*

*

*

Sonora

*

■¥

14.0 ± 0

7572.5 ± 57.6

1863.9 ±91.1

“Gophers"

*

*

7566.5 ± 22.6

8.0 ± 0

5072.4 ± 22.7

Monte

*

*

1862.0 ± 46.7

5076.4 ± 20.3

12.0 ± 0

Calomyscus

9131.5

*

*

*

*

Gerhillus

1384.2

*

*

*

*

Meriones lihycns

1765.8

*

*

*

*

Meriones crassus

*

*

*

*

M eriones persicus

**

*

*

*

*

AUactaga

*

*

*

**

*

Jaciilus

**

*

*

*

*

* Distance exceeded program printing capabilities.

** Assigned to this group although program printing capabilities were exceeded.

1980

MARES— DESERT RODENT ECOLOGY

23

Fig. 13. — Serir habitat in Wadi Natroun, about 100 km NW of Cairo, where Pachyuromys diiprasi, Gerhilliis perpallidiis (quadrupedal
species), and Jandus Jaculus (a bipedal species) co-occur.

noran and Iranian deserts, both seed and root-and-
tuber eating niche parameters are represented, and
that bipedal, externally similar rodents are present
in each desert, but food preferences and locomotor
adaptations are associated in reversed ways (Fig.
12; that is, parts of each of these niche parameters
have been switched). The Monte Desert, which
contained bipedal, possibly granivorous rodenti-
form marsupials as recently as the late Pliocene or
early Pleistocene, currently lacks obligate granivo-
rous rodents or species which are bipedal (Figs. 14-
16).

The causes of bipedality in desert rodents are not
well understood, but a number of major factors
(which are not necessarily mutually exclusive) have
been suggested as possible causative agents, in-
cluding predator avoidance and the freeing of the
forelimbs for seed gathering (Bartholomew and
Caswell, 1951; Vaughan, 1972; Hildebrand, 1974).
One of the problems with invoking predator avoid-
ance as a factor in the evolution of bipedality is that
throughout the world most desert rodents are not
bipedal. While it may be argued that bipedal forms
live in areas of lower plant cover than quadrupedal
species (for example. Price, 1978; Wondolleck,
1978), I have not found this to be the case. In Egypt,
for example, bipedal Jaculus jaculus inhabits the
same sparsely vegetated gravelly serir habitat as the
quadrupedal Pachyuromys duprassi and Gerbillus

perpallidus (Fig. 13). In the southwestern United
States, bipedal Dipodomys deserti and/or D. mer-
riami, as well as the quadrupedal pocket mice, Per-
ognathus ampins or P. longimemhris, co-occur in
exceedingly sparse habitats. Some bipedal species
(for example, Dipodomys agilis, of the southwest-
ern United States, which live under a dense canopy
of chaparral vegetation, or Microdipodops mega-
cephalus, near Mono Lake, California) may inhabit

Table 8. — The coefficients for each of the original variables
forming the first two canonical variates in Fig. II.

Original

variable

Canonical
variate 1

Canonical
variate 2

w

1.3

0.1

IML/BL

-5.2

4.5

U1

4.8

-2.1

TRL/IML

-17.6

-0.1

TRL/BL

20.7

4.0

ZB/BL

-2.8

-0.3

IW/BL

- 16.7

8.0

IL/BL

1.2

-3.0

lA

11.4

-7.0

SZ

-135.4

-42.5

MP

-45.1

-19.6

MC

31.7

17.0

TH

-16.5

-2.6

CH

71.1

18.2

MT

-37.2

0.8

SA

-408.4

-57.0

24

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 16

Fig. 14. — Representation of microhabitat selection of small mammals in the Dasht-e-Kavir Desert southeast of Tehran, a low scrub
desert. In this and the following two figures, no attempt is made to designate abundance of a particular species within a particular
microhabitat, nor are animals drawn to scale.

Fig. 15. — Microhabitat selection of rodents in the Monte Desert of northwestern Argentina, a floristically complex plant community.

1980

MARES— DESERT RODENT ECOLOGY

25

localities supporting dense vegetation. The recent
paper by Wondolleck (1978) presents data of for-
aging microhabitats of one bipedal species (D. mer-
riami) and three quadrupedal coexisting species of
Perognathus. These data suggest that the bipedal
rodent forages in large open spaces devoid of vege-
tation while the three pocket mice forage in and
around perrenial shrubs. Possibly bipedal species
will be found to preferentially forage in areas where
predator attacks would have a greater probability
of success (open areas), while quadrupedal species
will primarily forage near cover, scampering quick-
ly across intervening patches of open desert. Sub-
strate does not seem to be related to the bipedal
habit. Some species, such as D. agilis of the chap-
arral scrub of California, occur on hard, clayey hill-

sides, while other species may be found on soils
ranging from sand to gravel (for example, M. me-
gacephalus on sand, D. merriami on some gravel
slopes). Many quadrupeds are obligate granivores,
and some bipeds eat vegetable material and insects.

Recently Reichman and Oberstein (1977) pro-
posed that bipedality is a morphophysiological re-
sponse for efficient and rapid locomotion to better
exploit patchy, widely-spaced resources (see also
Dawson, 1976; Price, 19786). The niche reversal
situation described here implies that among desert
rodents granivory and bipedality may not be as ob-
ligately associated as has been supposed. Although
it is possible that roots and tubers are more widely
spaced in the sparse Iranian Desert than in the more
vegetated Sonoran Desert, thus selecting for a “go-

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

26

SONORAN KAVIR MONTE

E

S

Fig. 17. — A subjective designation of the major food categories of small mammals in the Sonoran, Iranian, and Monte deserts. Note
that granivory is lacking in the Monte, whereas herbivory predominates there. Insectivory is a minor trophic category in all deserts,
being filled by a rodent and a shrew in both Arizona and Iran, and a marsupial in Argentina. The relative sizes of the circles within a
food category are strictly my subjective estimate of the importance of that category in each desert. Thus, herbivory seems most
important in the Monte, whereas seeds are about equally as important in the two northern deserts.

pher” of high mobility (that is, a bipedal gopher),
there is almost no hard evidence to support the sup-
posed correlation of bipedality and the strategy of
foraging on clumped resources. I have evidence
from experimental field studies on desert rodents of
the southwestern United States that bipedal and
quadrupedal species forage on patchy and fine-
grained resources in much the same manner, and
that both groups concentrate on the finely-distrib-
uted seeds (Mares, unpublished). Eor the moment,
the major selective forces affecting the evolution of
bipedality among desert rodents are unknown.

There are many similarities between the Iranian
and Sonoran desert rodent faunas, particularly in
some of the overall morphological adaptations and
the food habit specializations. Generally there were
great similarities in the ecological relationships de-
picted by the distance and correlation phenograms,
particularly the rather close clustering of Onycho-
mys and Calomyscus. Although little is known
about the natural history of the long-tailed hamster
of Iran, Lay (1967) reports that some individuals
fed primarily on seeds, whereas Walker (1964)
notes that they readily consume animal material.

1980

MARES— DESERT RODENT ECOLOGY

27

Table 9. — Phylogenetic listing of desert rodents of the world
used in multivariate analyses. Appro.ximate location of capture
given. Classification based on Simpson (1945) and Wood (1955).

Suborder Sciuromorpha
Family Sciuridae
Xerus (Ghana: Damango)

Geosciurus { Bechuanaland )

Spermophllopsis leptodactylus (USSR, Turkmen)
Ammospermophilus harrisi (southwestern U.S.)
Family Geomyidae

Thomomys hottae (southwestern U.S.)
Puppogeomys castanops (southwestern U.S.)
Family Heteromyidae

Dipodomys merriami (southwestern U.S.)
Dipodomys ordii (southwestern U.S.)

Dipodomys spectahllis (southwestern U.S.)
Dipodomys microps (southwestern U.S.)
Dipodomys agilis (southwestern U.S.)

Dipodomys desert i (southwestern U.S.)
Microdipodops megacephalus (southwestern U.S.)
Perognathus haileyi (southwestern U.S.)
Perognathus intermedins (southwestern U.S.)
Perognathus penicillatus (southwestern U.S.)
Perognathus flavus (southwestern U.S.)

Suborder Myomorpha
Family Cricetidae
Subfamily Cricetinae
Tribe Hesperomyini

Reithrodontomys megalotis (southwestern U.S.)
Reithrodontomys fulvescens (southwestern U.S.)
Peromyscus eremicus (southwestern U.S.)
Peromyscus maniculatus (southwestern U.S.)
Peromyscus crinitus (southwestern U.S.)

Baiomys taylori (southwestern U.S.)

Onychomys torridus (southwestern U.S.)
EUgmodontia typus (Argentina)

Phyllotis griseoflavus (Argentina)

Sigmodon hispidus (southwestern U.S.)

Neotoma lepida (southwestern U.S.)

Neotoma alhigtila (southwestern U.S.)

Tribe Cricetini

Calomyscus haihvardi (Iran)

Phodopus rohorowskii (China)

Mystromys alhicaudatus (South Africa)

Cricetulus harahensis (China)

Cricetulus curtatus (Mongolia)

Subfamily Gerbillinae
Gerhillus nanus (Iran)

Gerhillus campestrls (Morocco)

GerhiUurus paeha (Southwest Africa)

Sekeetamys calurus (Egypt)

Taterillus harringtoni (Kenya)

Desmodillus auricularls (?)

Pachyuromys duprasi (Egypt)

Meriones llhycus (Iran)

Meriones crassus (Iran)

Meriones persicus (Iran)

Psammomys ohesus (Libya)

Rhomhomys opimus (China)

Table 9. — Continued.

Family Spalacidae

Spala.x ehrenhergi (Egypt)

Family Muridae
Subfamily Murinae

Thallomys nigricaudata (British East Africa)
Leggadina delicata (Australia)

Notomys ale.xis (Australia)

Notomys carpentarius (Australia)

Subfamily Dendromurinae

Malacothri.x typicus (Kalahari)

Petromyscus harbour (South Africa)

Steatomys athi (British East Africa)

Family Dipodidae
Subfamily Dipodinae

Dipus sowerhyi (Mongolia)

Paradipus ctenodactylus (U.S.S.R.)

Eremodipus lichtensteini (U.S.S.R.)

Stylodipus andrewsi (Mongolia)

./aculus hlandfordi (Iran)

Jaculus orientalis (Morocco)

Jaculus deserti (Morocco)

Scirtopoda telum (U.S.S.R.)

Allactaga mongolica (Mongolia)

Allactaga elater (Iran)

Pygeretmus shitkovi (U.S.S.R.)

Subfamily Cardiocraniinae

Cardiocranius paradoxus ( Mongolia)

Suborder Caviomorpha
Family Caviidae

Microcavia australis (Argentina)

Family Octodontidae
Octodon degus (Chile)

Octomys mima.x (Argentina)

Octodontomys simonsi (Bolivia)

Family Ctenomyidae

Ctenomys fulvus (Argentina)

Suborder Bathyergomorpha
Family Bathyergidae

Bathyergus janetta (South Africa)

Georhychus capensis (South Africa)

Cryptomys darlingi (Southwest Africa)

Heterocephalus glaher (Kenya)

Suborder .’Sciuromorpha, Hystricomorpha, or Myomorpha
Family Ctenodactylidae

Ctenodactylus gundi (Morocco)

The clustering of the hamster with the principally
insectivorous-carnivorous Onychomys suggests
that the propensity to eat animal matter may be
more important than has previously been realized.
Although Calomyscits bears a strong resemblance
to Peromyscus (Osgood, 1947), the multivariate
analyses were usually able to separate the hamster
from the Peromyscus assemblage. Jaculus hland-
fordi is an uncommon species that is very poorly

28

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

Table 10. — Regional listing of the genera comprising each desert rodent fauna used in the multivariate analyses.

North America

South America Australia

Northern Africa

Southern Africa

China

U.S.S.R.

Iran

Dipodomys

Ctenomys Leggadina

Heterocephaliis

Thallomys

Cricetuius

Stylodipus

Calomyscus

Microdipodops

Octomys Notomys

Xerus

Geosciurus

Phodopus

Dipus

Gerhillus

Perognathus

Octodon

Pachyuromys

Steatomys

Dipus

Spermophllopsis

Meriones

Neotoma

Octodontomys

Sekeetumys

DesmodiUus

Cardiocranius

Pygeretmus

AUactaga

Onychomys

Microcavia

Psanunomys

Mukicothrix

AUactaga

Paradipus

Jacidus

Peromyscus

Phyllotis

Spala.x

Petromyscus

Rhomhomys

Eremodipus

Rhomhomys

Baiomys

Eligmodontia

Taterillus

Georhychus

Scirtopoda

Reithrodontomys

Ctenodactylus

Bathyergus

Alactagidus

Sigmodon

Jacidus

Cryptomys

AUactaga

Ammospermophi-

Gerhillus

Mystromys

Rhomhomys

ins

Gerhillurus

Thomomys

Pappogeomys

known ecologically. Other species of Jacttlns are
known to eat seeds as well as other plant material
(Fetter, 1961; Ognev, 1963; Eisenberg, 1975), and
the particular diet probably varies greatly from
species to species. At least one species, J. turc-
menictts of the U.S.S.R., is almost totally herbiv-
orous (Naumov and Lobachev, 1975). The multi-
variate analyses suggest that J . hlandfordi is closer
ecologically to Dipodomys than is AUactaga. Ger-
hillits nanus is probably quite similar ecologically
to Perognathiis and was clustered with the pocket
mouse group in both distance and correlation anal-
yses. Schematic representations of some of the
broad niche categories of rodents at the Kavir, So-
noran, and Monte desert localities are given in Eigs.
14-17.

The fact that only one rodent from the Monte
Desert was clustered with Iranian and Sonoran
species within the desert specialist category sup-
ports the contention that the South American
species have not had time to evolve a high degree
of adaptation to desert conditions. This is particu-
larly true for the cricetines, descendants of the most
recent colonizers of the South American continent.

An Analysis of the Desert
Rodents of the World

The previous analyses yielded a number of re-
sults that, in retrospect, either appeared logical
when I had an intimate knowledge of the ecology
of the various species being examined, such as in
the Monte-Sonora-New Mexico Eorest compari-
son, or were counterintuitive when the fauna was
less well known ecologically, for example in the
Sonora-Monte-Iranian study. The anomalous rela-
tionship of the bipedal herbivorous AUactaga finds
at least some support in the literature, whereas the
similarities between Meriones and Dipodomys in
overall food habits have been shown in a number
of studies. I decided to measure individuals of as
many species of desert rodents as possible in order
to determine whether or not these same multivari-
ate techniques utilizing various suites of
morphoecological characteristics could be extended
to the world desert system. If the information pro-
duced by such analyses reflects aspects of the bi-
ology of the species, it should allow one to deter-
mine which species in the various disjunct deserts
might fill similar roles, as well as which species are

Table 1 1. — Mean Mahalanohis Distances (D-) of each desert fauna to all other faunas when 40 traits are utilized (Fig. 18).

U.S.

South America

Australia

North Africa

South Africa

China

Russia

Iran

U.S.

25.0

55.0

110.2

43.4

39.2

76.2

76.8

62.4

South America

—

25.5

113.9

70.8

72.6

86.1

83.9

68.1

Australia

—

—

27.4

1 17.7

101.9

97.1

103.3

100.7

North Africa

—

—

—

29.8

40.5

85.7

83.3

68.5

South Africa

—

—

—

—

30.3

73.4

77.9

70.6

China

—

—

—

—

—

29.7

46.8

71.8

Russia

—

—

—

—

—

—

28.1

70.1

Iran

—

—

—

—

—

—

—

26.7

1980

MARES— DESERT RODENT ECOLOGY

29

Fig. 18. — Canonical analysis of desert rodent faunas from eight regions based on 40 morphological traits. Symbols: number I and letters
A indicate the group mean and individual cases respectively, for United States desert rodents; 2 and B = South America; 3 and C =
Australia; 4 and D = North Africa; 5 and E = South Africa; 6 and F = China, 7 and G = Russia; 8 and H = Iran. One species per
genus occurring in each desert region (and listed in Table 9) was used in the analysis so as not to weight any particular genus more
than any other.

unique to a particular desert and seem to fill a niche
which is not even loosely repeated in any other des-
ert of the world. Further, it might be possible to
compare the degree of desert adaptation of the var-
ious rodent faunas and correlate this with the time
span over which desert adaptations have occurred.
Although I was not able to examine every known
species of desert rodent, most genera and many
species are represented in the analyses that follow
(Tables 9 and 10). Basically, rodent faunas from
eight major desert regions were examined. In many

cases the results are only tentative (and essentially
predictions) since the majority of species are very
poorly known ecologically.

When the genera and species are grouped by des-
ert region and compared with one another (one
species/genus) utilizing all 40 morphological char-
acteristics, the South American desert is widely
separated from all other desert regions along the
second canonical axis (Fig. 18, Tables II and 12).
The North American, North African, and South
African deserts are closely clustered, as are the

30

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

Table 12. — Coefficients for the first two canonical variates
shown in Fig. 18.

Original

variable

Canonical
variate 1

Canonical
variate 2

HBL

0

0

TL

0

0

HFL

2

.1

EL

^2

2

VL

0

.1

HL

-.3

— .2

TL/HBL

0

0

HFL/HBL

-.1

-.1

EL/HBL

-.4

-.3

VL/HBL

0

0

HL/HBL

.5

2

FB

-1.6

_ 2

TT

-.3

-.3

UL

-.1

-.3

UW

.1

-.3

MW

-1.0

.6

IL

.8

.7

UI

.4

.4

TRL

2

-1.3

TRL/IML

_ 2

-.1

TRL/BL

.5

-.1

ZB/BL

-.1

2

IL/BL

.1

-.3

lA

.7

.8

SZ

1.6

.1

MP

.9

0

TH

-.1

2

CH

-2.0

__ 2

MT

-.4

-.8

VD

.9

-.4

rw

-3.5

-1.4

SA

_ 2

0

China and Russian assemblages, while the Iranian
desert rodent fauna seems to bridge a gap between
the African and Asian faunas. Australia is located
between Iran and the China-Russia cluster. One
North American species, Baioinys taylori, was as-
signed to the South African fauna. The mean Ma-
halanobis Distances of each species within a partic-
ular fauna to the faunal mean values of the other
groups (Table 12) show that the highly desert spe-
cialized groups are close to one another (for ex-
ample, the United States-North Africa distance,
versus the United States-South America distance),
while South America is closest to its phylogenet-
ically most similar assemblage. North America. The
South American fauna is located low on Canonical
Variate 2 where such variables as molariform tooth-
row width, molar triangulation, proodont or ortho-
dont incisors, and molariform toothrow length, are
weighted heavily. The preponderance of herbivory

Table 13. — Coefficients for the first two canonical variates
shown In Fig. 19.

Original

variable

Canonical
variate 1

Canonical
variate 2

BL

2

-.5

IML

.2

.5

MW

.5

.5

iW

0

1.6

TRL

-2.1

-.9

TRL/IML

.1

.3

ZB/BL

.1

-.1

IW/BL

.1

-.9

IL/BL

-.1

0

lA

.1

-.7

SZ

.1

-.9

T1

-.7

-.1

MP

-.7

-.1

MC

-.4

.2

TH

.5

-.1

CH

1.3

.9

MT

-1.0

-.1

TW

1.0

.7

SA

.1

0

as opposed to granivory or insectivory among the
Monte rodents (Eig. 17) thus seems to distinguish
them from the other desert faunas.

The same desert groups were analyzed using 23
dental traits (Fig. 19, Tables 13 and 14), 27 non-
ratio traits (Fig. 20, Tables 15 and 16), and 25 mor-
phoecological characteristics (Fig. 21, Tables 17
and 18). The basic pattern shown in the 40 trait
analysis is repeated in that the South American ro-
dents are always separated from the other groups,
and Russia and China are usually plotted closely
together. North America is generally located quite
close to North Africa, and Australia, which con-
tains only three rodents, is either placed with the
China-Russia groups, or with the Iranian fauna. The
Russia and China groups contain the most bipedal
forms (Dipodidae), with herbivores such as Dipus,
Ereinodipiis, Stylodipus, and Allactaga, and qua-
drupedal herbivores such as Cricetidus , Phodopus,
or Rhomhomys. Herbivory and bipedality are the
major strategies among rodents of these two re-
gions. Iran actually is more like Africa (and thus
North America) as far as the overall rodent assem-
blage is concerned, although it shares some generic
affinities with the Asian deserts to the north and
east. Nevertheless, its location in most analyses as
either being similar to the African-North American
groups, or positioned between the African and
Asian faunas, seems logical, and probably reflects
both its phylogenetic affinities (to the Asian faunas).

1980

MARES— DESERT RODENT ECOLOGY

31

Fig. 19. — Canonical analysis of desert rodent faunas from eight regions based on 23 dental traits. Symbols as in Fig. 18.

as well as its ecological affinities (to the North Af-
rican and North American faunas).

I showed earlier that cluster analysis is a useful
technique for examining evolutionary convergence.
Presumably, those species sharing a large number

of traits will be grouped closely together. If they
are only distantly phylogenetically related, then
those that are clustered together may be considered
species that are convergent. Since many of the
traits utilized in these analyses are morphoecolog-

Table 14. — Mean Mahalanohi.^ Distances (D-) of each desert fauna

to all other faunas when 23 dental traits are utilized {Fig. 19).

u.s.

South America

Australia

North Africa

South Africa

China

Russia

Iran

u.s.

15.9

30.7

65.7

37.3

23.7

29.5

30.9

25.6

South America

—

13.9

75.9

40.9

41.2

49.5

54.2

46.5

Australia

—

—

27 2

76.5

69.0

69.1

74.3

78.7

North Africa

—

—

—

18.4

24.8

32.1

38.7

29.7

South Africa

—

—

—

—

16.7

24.9

37.0

27 2

China

—

—

—

—

—

13.3

27.0

27.1

Russia

—

—

—

—

—

—

17.6

32.2

Iran

—

—

—

—

—

—

—

14.6

32

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

ical ones, thus strongly implying some ecological
function, species forming a cluster can be consid-
ered ecological equivalents for those ecological
traits reflected in their morphology. Although both
distance and correlation phonograms (Sneath and
Sokal, 1973) were determined for each group of
measurements, I will use only one or the other of
these to illustrate the resultant clusters, particularly
since differences between the two clustering tech-
niques were relatively minor. I will briefly describe
the clusters in each phonogram before discussing
my overall impressions of convergent evolution in
desert rodents derived from the various analyses.

The 40-character distance phonogram listing all
of the 78 species of desert rodents is shown in Eig.
22; the cophenetic correlation coefficient is 0.83,
indicating that the 2-dimensional representation of
the 40-dimensional relationship is not greatly dis-

torted. There are eight major clusters composed of
numerous smaller clusters. The first includes kan-
garoo rats of the genus Dipodoinys (Eamily Het-
eromyidae) of the Sonoran, Mojave, and Great Ba-
sin deserts and the chaparral scrublands of
California. Sekeetamys calunis (Cricetidae) of the
Egyptian Sahara Desert is closely clustered with
Dipodoinys. The dipodids, Jacnlns orientalis and
J. deserti of the Sahara, Stylodipiis andrewsi of
Mongolia, and Scirtopoda telnm and Eremodipiis
lichtensteini of the U.S.S.R., form a second small
cluster. The caviomorph octodontids Octomys mi-
max of the Argentine Monte, and Octodontomys
simonsi of the Bolivian altiplano, are loosely clus-
tered with Rhombomys opimus (Cricetidae) from
China and Iran. Completing the first major cluster
is a loose grouping of the dipodids. Dipus sowerhyi
of Mongolia and Paradipus ctenodactyliis of the

1980

MARES— DESERT RODENT ECOLOGY

33

Canonical Variate I

Fig. 21. — Canonical analysis of desert rodent faunas from eight regions based on 25 morphoecological traits. Symbols as in Fig. 18.

U.S.S.R. Basically species in this assemblage are
highly desert specialized, medium-sized rodents,
having a pronounced inflation of the auditory bullae
and reduced pinnae; most are bipedal.

The second cluster is a small group composed of
the caviomorphs Microcavia australis (Caviidae) of
the Argentine Monte, and Octodon degas (Octo-
dontidae) of the arid region of Chile, and the cten-
odactylid Ctenodactylas gandi from the Sahara of
Morocco. These animals are similar in body size
and overall proportions; all frequent rock piles and
other rocky areas, with M. australis apparently
being the most labile in habitat requirements (see
Walker, 1964; Mares, 1973; Glanz, 1977; Meserve
and Glanz, 1978). All three species are herbivorous.

The third major cluster is a large one composed
of numerous smaller clusters. The first of these con-
tains the small granivores, Perognathus (Hetero-
myidae) of the United States, and Gerhillus nanus

(Cricetidae) of Iran, which are loosely clustered
with the small heteromyid bipedal granivore, Mi-
crodipodops megacephalus of North America. Be-
cause many traits used in the 40-character analysis
are correlated with overall body size, this parame-
ter has a great influence on the final depiction of
relationships. Interestingly however, Microdipo-
dops is grouped with Perognathus, rather than Di-
podoniys (which it resembles, externally); this is in
accordance with suggested phylogenetic relation-
ships (Hafner, 1978).

The next small cluster is comprised of the North
American grasshopper mouse, Onychomys torri-
dus, the Chinese dwarf hamster, Pliodopus roho-
rowskii, the Mongolian rat-like hamsters, Cricetulus
curtatus and C. barahensis, and the white-tailed rat
of South Africa, Mystroinys alhicaudatus. All of
these are rather small quadrupedal cricetids; most
apparently are seed eaters (Walker, 1964).

34

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

CDipodomys merviami
D. ordi

JH

D. speetabilis
■ Sekeetamys calurus
D. micvops
D. agilis
D. deserti
Jaculus orientalis
J. deserti
Stylodipus andrewsi
•Scirtopoda telum
‘Eremodipus lichtensteini
Octomys mimax
Octodontomys simonsi
Rhombomys opimus
R, opimus
■ Dipus sowerbyi
Paradipus ctenodactylus
Microcavia australis
Oatodon degus
Ctenodactylus gundi
Microdipodops megacephalus
PerognathuB bailey i
P. intermedius
P. penicillatus
P. flavus
Gerbillus nanus
Onychomys torridus
Cricetulus barabensis
Phodopus roborowskii
Mystromys albicaudatus
Cricetulus curtatus
Peromyscus eremicus
P. maniculatus
Reithrodontomys fulvescens
R. megalotis
Eligmodontia typus
Petromyscus barhour
Baiomye taylori
Steatornjs athi
Peromyscus crinitus
Taterillus harringtoni
Gerbillus campestris
Gerbillurus paeba
Notomjs alexis
Notomys carpentarius
Malacothrix typicus
Calomyscus bailwardi
Cardiocranius paradoxus
Neotoma albigula
N. lepida
Sigmodon hispidus
Phyllotis griseoflavus
Ammospermophilus harrisi
Pachyuromys duprasi
Desmodillus auricularis
Thallomys nigricaudata
Psarmorrys obesus
Meriones libycus
M. crassus
Allactaga mongolica
Pygeretmus shitkovi
Alactagulus pumilio
Allactaga elater
Meriones persicus
Jaculus blandfordi
Thomomys bottae
Ctenomys fulvus
Spalax ehrenbergi
Pappogeomys castanops
Bathyergus Janetta
Georhyohus capensis
Cryptomys darlingi
Heterocephalus glaber
Xerus

Geosciurus

Spermophilopsis leptodactylus
Leggadina delicata

DISTANCE

Table 15. — Coefficients for the first two canonical variates
shown in Fi)’. 20.

Original

variable

Canonical
variate 1

Canonical
variate 2

HBL

0

.1

TL

0

0

HFL

-.1

-.1

EL

0

0

VL

.1

0

HL

— 2

-.3

FB

1.4

-.5

TT

.1

— .2

BL

.1

0

IML

0

0

UL

.1

0

UW

-.3

.5

MW

.4

-.1

ZB

.6

0

IW

_ 2

-.4

IL

-.5

0

TRL

-1.5

-1.7

lA

.4

.8

SZ

-.3

2.3

Tl

-1.2

-.1

MP

-.1

— .2

MC

-.6

-.9

TH

.4

.1

CH

-1.2

-.1

MT

-.9

-1.1

VD

-1.1

.3

TW

1.0

.8

Small, quadrupedally-scansorial omnivores form
the next small cluster. Included are Peromyscus
and Reithrodontomys of North America; Eligmo-
dontia typus of Argentina; Petromyscus harbour of
South Africa; the North American (Chihuahuan
Desert) pygmy mouse, Baiomys taylori \ and the
South African fat mouse, Steatomys athi. All of
these are cricetids.

The last subcluster composing the third major
cluster is comprised of three groups. North Amer-
ican Peromyscus crinitus, and the North African
gerbils (Taterillus harringtoni, Gerbillus campestris,
and G. paeba) are grouped together and attached
to the second subcluster which includes the murid
Australian hopping mice, Notomys alexis and TV.
carpentarius. The South African gerbil mouse, Ma-
lacothrix typicus, and the Iranian hamster, Calo-
myscus bailwardi, comprise the third subcluster.

Fig. 22. — Distance phenogram resulting from a cluster analysis
of 78 species of desert rodents utilizing 40 morphological char-
acteristics.

1980

MARES— DESERT RODENT ECOLOGY

35

Table 16. — Mean Mahalanobis Distances (D-) of each desert fauna to all other faunas when 27 non-ratio traits are utilized (Fig. 20).

u.s.

South America

Australia

North Africa

South Afnca

China

Russia

Iran

u.s.

22 3

49.0

70.0

42.. S

41.2

61.4

60.2

47.5

South America

—

24.0

81.6

71.0

73.2

81.3

78.6

74.6

Australia

—

—

21.1

55.3

54.6

67.8

71.8

56.4

North Africa

—

—

—

25.2

31.7

59.2

54.4

44.9

South Africa

—

—

—

—

25.2

49.4

54.5

44.0

China

—

—

—

—

—

23.2

36.2

50.1

Russia

—

—

—

—

—

—

23.6

49.0

Iran

—

—

—

—

—

—

—

20.2

Finally, the five-toed dwarf jerboa, Cardiocranius
paradoxus, is very tenuously included within the
third major cluster.

The fourth major cluster is composed of medium-
sized, quadrupedal forms only loosely associated
between subclusters. The North American wood
rats and cotton rats, Neotoma and Signiodon, re-
spectively, are clustered with the South American
leaf-eared mouse, Phyllotis (Graoinys) griseofla-
vus. The North American sciurid, Atnmosper-
mophilus liarrisi is not closely allied with other
members of this cluster. Two smaller clusters, one
containing the Sahara fat-tailed sand rat, Pachyii-
roinys dtiprasi ; the South African Cape short-eared
gerbil, Desmodillus auricidaris ; and the South Af-
rican acacia rat, Thallomys nigricaudata ; and the
other including the Sahara sand rat, Psammomys
obesiis', and the two jirds, Meriones libycus and M.
crass us, complete the major cluster.

The fifth large cluster is made up of two distinct
subclusters. The first contains the jerboas — Allac-
taga mongolica of China; Pygeretmus shitkovi of
the U.S.S.R.; Allactaga elater of Iran; and Alac-
tagulus pumilio of the U.S.S.R. The second in-
cludes the large jird, Meriones per sic us, and the
dipodid, Jaculus blandfordi.

The sixth major cluster includes fossorial species
of all types. Thomomys bottae, a North American
gopher, is closely allied with the tuco-tuco of the
Argentine Monte {Ctenomys fulvus), and these are
connected to the Sahara mole rat, Spalax ehren-
bergi. A second group, loosely joined to the first,
is composed of another North American gopher,
Pappogeomys castanops, and the South African
bathyergid mole rat, Bathyergus Janetta . The other
two bathyergids, Georhychus capensis and Cryp-
tomys darlingi, are clustered together and joined to
the aforementioned fossorial species, while the na-
ked mole rat of eastern Africa, Heterocephalus gla-
ber, is loosely clustered with the other fossorial
species, thus completing this major cluster.

The final cluster is a small one composed of sci-
urids — Xerus from North Africa; Geosciurus from
South Africa; and Spennopliilopsis leptodactylus
from the U.S.S.R. Finally, the phenogram is com-
pleted with the inclusion of the Australian murid,
Leggadina delicata.

When only 25 morphoecological traits are utilized
in cluster analysis, a somewhat different picture of
relationships is obtained. Six major clusters, many
composed of a number of loosely associated sub-
clusters, are evident in the correlation phenogram
of Fig. 23 (cophenetic correlation coefficient =
0.763).

Table 17. — Coefficients for the first two canonical variates
shown in Fig. 2L

Original

variable

Canonical
variate 1

Canonial
variate 2

HBL

0

0

TL/HBL

0

0

HF/HBL

0

0

EL/HBL

0

.1

VL/HBL

0

0

HL/HBL

-.1

0

FB

1.2

-.1

TT

.6

0

TRL/BL

0

.1

U1

0

0

TRL/IML

.2

2

TRL/BL

-.1

0

ZB/BL

.1

-.1

IW/BL

.1

-.1

IL/BL

_ 2

2

lA

-.5

-.4

SZ

-1.4

-1.5

T1

.4

-.5

MP

.2

.2

MC

0

.9

TH

.2

-.4

CH

.6

-.4

lA

2

.7

VD

-.9

.4

SA

0

0

36

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

Dipodomys mevriami
D. ordi

D. epeotabilis
Pevognathue haileyi
P. penicillatus
P. intermedins
P. flavus
Gevhillus nanus
Octomys mimax
Sekeetconys calurus
Meviones arassus
Dipodomys microps
D. agilis
D. deserti

Microdipodops megacephalus
Paradipus ctenodactylus
Eremodipus tichtensteini
Jaculus ovientalis
J. deserti
Sairtopoda telian
Stylodipus andrewsi
Neotoma alhigula
Sigmodon hispidus
Neotoma lepida
Phyllotis griseoflavus
Meriones libycus
Psarmomys obesus
A I lactaga mongo liaa
Pygeretmus shitkovi
Alactagulus pumilio
Meriones persicus
Allactaga elater
Jaculus bland f ordi
Ammospertjupkilus harrisi
Geoaciuims

Spermophilcpsis leptO'iaetylue
Xerus

Thomomye bottae
Ctencmys fulmrs
Pappogeomys oastanops
Bathyergus Janetta
Spalax ehrenbergi
Heteroaephalus glaber
Georhyahus aapensis
Cryptomys darlingi
Ootodontomys simonsi

— Microoavia australis

- Octodon degus

- Ctenodactylus gundi

- Rhombomys opimus
— P. opimus
•^Onychomys torridus
-Malacothrix typicus
— CalomysGus bailwardi

Peromyscus arinitus
Baiomys taylori
Gerhillus campestrns
Taterillus harringtoni
Gerbillurus paeba
Notomys alexis
N. oarpentarius
Steatomys athi
.^Mystromys albicaudatus
Cricetulus barabensis
Phodopus roborowskii
Cricetulus curtatus
Pachyuromys duprasi
Desmodillus auricularis
Thallomys nigrioaudata
Peromyscus eremicus
P. maniculatus

CReithrodontomys fulvescens
R. megalotis
^-^Eligmodontia typus

Petromyscus barbour

Cardiocranius paradoxus

Dipus souerbyi
—^Leggadina delicata

p 5 1

CORRELATION

The first major cluster includes three subclusters.
The first of these is composed of species of Dipod-
omys and Perognathus of North America, Gerhillus
nanus of Iran, Octomys from Argentina, Sekeeta-
mys calurus of Egypt, and Meriones crassus from
Iran. The second subcluster includes heteromyids
in the genera Dipodomys and Microdipodops, and
the dipodids Paradipus ctenodactylus and Eremo-
dipus liclitensteini. The third subcluster includes
only dipodids — Jaculus, Stylodipus, and Scirtopo-
da.

The majority of species included within the first
major cluster are bipedal (13/21), or have rather
long hind feet (5/21). Most are seed eaters (19/21),
although some, such as Stylodipus, are reported to
take roots and tubers as well as seeds (Walker,
1964). Octomys from the Argentine Monte eats
cacti, green vegetation, and, perhaps, large seeds
(Mares, 1973), but its inclusion with what are main-
ly heteromyids is largely based on cranial (bullar
inflation) and dental characteristics (Mares, 1976).
Most of the species in the first major cluster have
relatively short ears and long tails. Nevertheless,
the first cluster is composed mainly of bipedal seed
eaters with simple dentition, inflated bullae, rela-
tively short ears, long tails, long hind feet, and
which possess foot bristles. They are principally in-
habitants of flatlands varying from sand to gravel,
although such species as Sekeetamys calurus and
Octomys mimax (which are clustered together), are
rock dwellers.

The second major cluster generally includes me-
dium-sized, quadrupedal herbivores. Thus, Neo-
toma and Sigmodon of North America are clus-
tered with Phyllotis of Argentina, Meriones libycus
of Iran, and Psammomys obesus of Egypt. The sim-
ilarities of the North American species and P. gri-
seoflavus have been discussed earlier (Mares, 1973,
1976). Meriones libycus feeds almost entirely on
seeds (Naumov and Lobachev, 1975) and is in-
cluded within this cluster largely on the basis of
body size and body proportions. Psammomys ohe-
sus inhabits salty-clayey flats and builds extensive
burrows under green vegetation in hummocks; it
feeds on green vegetation (Walker, 1964; Wassif,
1972).

Fig. 23. — Correlation phenogram resulting from a cluster anal-
ysis of 78 desert rodent species utilizing 25 morphoecological
characteristics.

1980

MARES— DESERT RODENT ECOLOGY

37

Table 18. — Mean Mahalanohis Distance ID'-) of each desert fauna to all other faunas when 25 inorphoecological traits are utilized

(Fig. 21).

u.s.

South America

Australia

North Africa

South Africa

China

Russia

Iran

u.s.

19.9

35.0

81.2

29.0

31.0

48.0

47.0

38.3

South America

—

19.7

80.7

47.3

51.5

54.3

49.0

49.3

Australia

—

—

27.2

83.3

85.4

77.1

77.4

83.2

North Africa

—

—

—

23.5

28.8

52.3

51.6

41.2

South Africa

—

—

—

—

20.0

49.1

52.2

45.0

China

—

—

—

—

—

22.8

31.3

51.0

Russia

—

—

—

—

—

—

22.8

53.6

Iran

—

—

—

—

—

—

—

19.1

The third major cluster includes the dipodids
{Allactaga, Alactagiilus, Pygeretmiis, and Jacitlus
hlandfordi) and the jird, Meriones persiciis. With
the exception of Meriones, all are bipedal. All are
apparently herbivorous.

The fourth major cluster is composed of sciurids,
and includes Ammospennophilits of North Ameri-
ca, Geosciiirus of South Africa, Xerus of North
Africa, and Spermophilopsis of the U.S.S.R. All
are medium-sized, burrowing, scansorial herbi-
vores-omnivores.

The fifth major cluster is composed of two dis-
tinct subclusters. The first contains the fossorial
species clustered together in Fig. 22; Octodontomys
simonsi is loosely joined to the fossorial group. The
other subcluster includes medium-bodied herbi-
vores, such as Microcavia, Octodon, Ctenodacty-
liis, and Rliomhomys.

The sixth major cluster is comprised of eight
smaller clusters — the first includes Onychomys,
Calomyscits and Mcdacothrix, which may be qua-
drupedal insectivores, or at least micro-omnivores
(Mares, 1976); the second contains Peromyscus,
Baiomys, Gerbillus, Taterillus, Notoinys, and Stea-

toinys, largely omnivorous quadrupeds which may
eat significant amounts of seeds; the third subgroup
contains Mystromys, Cricetidns, and Phodopus\
the fourth includes Pachyitromys and Desmodillns,
while Thcdlomys is only loosely included within this
subcluster; the fifth subcluster contains Peromys-
ciis, Reithrodontomys, EUgmodontia, and Petro-
inyseiis. The final group of rodents completing the
major cluster includes Cricetidiis barahensis, Dipiis
sowerbyi, and Leggadina delicata.

The 27-character distance phenogram utilizing
only non-ratio traits is shown in Fig. 24; the cophe-
netic correlation coefficient is 0.798. I will not dis-
cuss the individual OTU’s at length. Basically this
technique divides the rodents into four major
groups — bipedal forms; quadrupedal species; fos-
sorial species; and ground squirrels. In many ways
there were no significant deviations in the 27-trait,
non-ratio analysis, from those derived from the 40-
and 25-traits analyses, although I believe that hav-
ing all three analyses is an aid to understanding var-
ious aspects of the comparative biology of numer-
ous species.

CONVERGENT EVOLUTION OE DESERT RODENTS

Although the various deserts examined in this
paper contain a diverse array of desert rodents, the
computer analyses indicate that one particular re-
gion is not as ecologically distinct from another dis-
tant region as one might have expected. The most
diverse desert examined, from the viewpoint of the
total number of broad adaptive categories of ro-
dents it supports, is that found in the United States
and northern Mexico (Table 19). Basically, my anal-
yses indicate that there are nine major niche types
(or guilds) represented among the world’s desert
rodents, including bipedal granivores, quadrupedal

granivores, miero-omnivores, medium-sized omni-
vores, small insectivores, fossorial herbivores, me-
dium-sized herbivores, larger herbivores, and bi-
pedal herbivores. It is important to note that,
although most of these categories are represented
within each desert region, they are not always filled
by a rodent. It is not uncommon to find at least one
member of each category present in any partic-
ular locality, and coexistence of species within one
particular category is often seen (for example, Hoff-
meister and Goodpaster, 1954; Chew and Chew,
1970; Rosenzweig et al., 1975; Brown, 1975). The

38

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

r~

r

... r~

1

r

, H —

J

, ^

1

Dipodomys merriconi
D. ordi

Miarodipodops wegacephalus
Dipodomys microps
Dipodomys agilis
D. spectabilis
D. deserti
Octomys mimax
Sekeetamys aalurus
Octodontomys simonsi
Rkombomys opimus
R. opimus

Jaoulus orientalis
J. deserti
Stylodipus andrewsi
Evemodipus liahtensteini
Scirtopoda telian
Paradipus ctenodactylus
Jaoulus blandfovdi
Neotoma atbiguta
N. tepida
Sigmodon hispidus
Phyllotis griseoflavus
Psammomys obesus
Meriones libyaus
M. arassus
.M. persious
-Dipus sowerbyi
^icTocavia australis
-Oatodon degus
-^tenodaotylus gundi

Allactaga mongolica

ThjQeyyefrmi^i shitkovi
-^—.Alactagulus pimilio
—.Allactaga elater

Pprnn/7-nnthiJfi baileyi

I —P. intermedius
\ I. P. penicillatus
,mp. flavue
— nanus
Onuchamue torridue

Calomyscus bailwardi

Perorryacus cirinitus

Jialaoothrix typiaus
Phodopus roborowskii
Cricetulus barabensis
•C. curtatus
'otomys alexis
jV. carpentarius
J'aterillus harringtoni
Gerbillus campestris
^erbillurus paeba
'eromyeaus eremicus
P. manioulatus
^eithrodontomys fulvescens
megalotis
—€teatomys athi
^etromyscus harbour

•Eligmodontia typus
Saiomys taylori
Jjcggadina delicata

Cardioaraniue paradoxus

—Ammospermophilus harrisi
.Jtystromys albicaudatus

S'hallomys nigricaudata
'achyuromys duprasi
Desmodillus auricularis
-Thomomys bottae
-JDtenomys fulvus
Spalax ehrenbergi

' Pappogeomys castanops
^^^athyergus janetta
Georhychus capensis
Cryptomys darlingi
Reterocepha lus glaher
—J(erus

-Geosciurus

-Ppermophilopsis leptodactylus

DISTANCE

great diversity of desert rodents in North America
may be at least partially explained by the enormous
fluctuation, fracturing, and reformation of xeric
habitats in the Pleistocene {see Van Devender,
1977; Mares, 1979). Indeed, the only major niche
type that is lacking in the New World is that of
bipedal herbivore, a category that is important in
the Old World, particularly in Russia and Australia.

The Monte, as has been noted, is quite depau-
perate in both number of species inhabiting the des-
ert, and in abundance of individuals at a particular
locality (Mares, 1976; Mares and Rosenzweig,
1978). Granivorous mammals are lacking entirely,
although the overall array of niche types is not ex-
ceedingly narrow. The small insectivore niche,
which is filled in North America by a rodent {Ony-
cliomys) and a shrew (Notiosorex) is represent-
ed by a marsupial mouse, Marmoset pusilla,
which is rare, but regular, over much of the
Monte (Mares, 1973). Ctenomys, fossorial cavio-
morphs, are close ecological analogues of the
gophers of North America, while DoUchotis pata-
gona is a large, cursorial rodent quite similar in
morphology and ecology to the leporids of North
America (Mares, Blair et al., 1977). Microcavia
bears some ecological similarities to ground squir-
rels, and very likely fills a part of this niche cate-
gory. It is strictly herbivorous, however, and thus
not a perfect ecological equivalent, although in
overall body proportions, habitat requirements,
time of activity, and, perhaps, behavior, the two
groups (ground squirrels and Microcavia) are sim-
ilar (see Hawbecker, 1947; Hudson, 1962; Mares,
1973). A potential candidate for the medium-sized
omnivore niche in the Monte is the armadillo, Cliae-
tophractus vellerosus. This species eats a wide va-
riety of plant and animal matter, and is a conspic-
uous element of the Monte Desert (Greegor, 1975;
Mares, Blair et al., 1977).

The Australian desert, while supporting a low
diversity of rodents, nevertheless has a rich mam-
mal fauna, although most species are of low density
(Watts, 1974; Morton, 1979). Eive of the nine major
guilds are represented in Australia, with two others
perhaps partially represented. N otomys seems to
fill the bipedal granivore niche, even though it was
not clustered with bipedal rodents except when

Fig. 24. — Distance phenogram resulting from a cluster analysis
of 27 non-ratio morphological characteristics.

Table 19. — Suggested categorization of niche types of small mammals represented within each of the desert regions included in the preceding analv
Although many species in the table were not included in the computer analyses, they are listed here for completeness.

1980

MARES— DESERT RODENT ECOLOGY

39

■S-

-5

I

CO ^ a. § 05

o

u

<3 5:

c ^
0:: U

C) -ir-
'a

U Co

C3 5
Oo

S

.Cl.

Q

S

a r*'

a,

C!

§■■2
S S
^ §■

s ^

<3 Ci.
C 2

5

as CO

o -ie-

^ Co

.a. !3 -2

Q ^ ^ cC'

2 ^

o

a:

Q

o ^ c:)

o

-c>

53

Q a, Q c^)

2 §

2 ^

'3 -s:

a, h.

3;

3!

a.

C 35
«o .5:

u E2 I ‘S §'

■S o ^
^ c

05 U

'ij "2 S'
a a

Si a, O Co

•5 K

35

«0

o

2

Oo (i;

4*5

s s

s

s S a "a

•= l| L

^ s. ^

2 - 2 o
s: 2 ^ >

^ CO ^ ^

35 3>
O C;

~ S ®«

r® cu C!

U 03 Kj

O

§■

o «

5 Oo

c2" ^

^ 2

a, u

<3 ^

t 2
§ O

2

C

o

Q

§■

C'

2 .a

•2

^ 2

.a >

Q §

C* i 2

^ 3:

2 “S 2
i 'S
S eg

a c=
a

. &
^ s

c ^

35 O

o S

^ oc
<

o I

2 ^

I §:

R la

Co >-J

■= o

w QS
C 0>

Cs5

■S C

*33 ^

•u ^

<U o

M

■o

g Sb

O'

-o

<U </)
N <L>

•?

II

0>

> w

T3 o c

OJ C/3 _D

.N 2^ =« a

C)

^ > — "O
E .a m c

3 X) ac 3 i£

•— t- b o ^

(D Cd ^ cy

(u x: “ cj) a.

— o
>

■R 5

O

>“ Xi

V U
00 (U

- ° %

® 3;' M a.

“• .« (i>

^ X pa ^ e

<U (D C3 c .iS

a X ” zi D.

5

40

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

analyses were limited to external traits alone
(Mares, unpublished). Most analyses placed Noto-
inys with various gerbils or micro-omnivores. It is
likely that Notomys may not be as obligately gra-
nivorous as surmised (Morton, 1979) and may, in
fact, exhibit a catholic diet including insects (for
example. Watts, 1973). Medium-sized herbivores
may be lacking in the Australian Desert, although
Leporillus might fill this niche. It is known to
take some animal matter, although most informa-
tion regarding this species is anecdotal (Walker,
1964; Morton, 1979). Small insectivorous mar-
supials (some bipedal forms) are common in
Australia, as are smaller, bipedal herbivores and
larger herbivores. These latter two categories are
represented by small rat kangaroos and wallabies.
The fossorial marsupial mole, Notoryctes, is listed
as a desert animal by Morton (1979) without data
on its distribution or abundance; the species is ap-
parently wholly insectivorous (Walker, 1964).

The North African fauna is the second most com-
plex assemblage of desert rodents in the world.
Many of the species may be quite abundant in par-
ticular localized areas (Mares, personal observa-
tion), and, in this respect as well as in overall di-
versity, this fauna resembles that of North
America. The high diversity across this immense
region may be due in part to a Pleistocene history
of formation of desert refugia paralleling that which
occurred in the southwestern United States (Schlit-
ter, 1976). Bipedal granivores are represented by
Jaculus, while various genera of quadrupedal gran-
ivores are common in the desert. As more is learned
about the biology of the various species comprising
the quadrupedal category, it would not be surpris-
ing to find that some are actually micro-omnivores;
various analyses suggest that Pachyuromys and
Desinodillus may include animal matter in their
diet. The highly desert specialized sciurid, Xems,
fills the medium-sized omnivore niche that, in des-
erts, is apparently limited to squirrels. Psammomys
and Ctenodactylus seem to fill a role similar to the
wood rats of North America and the cavy of the
Monte; indeed, there is a pronounced morphologi-
cal resemblance between the cavy and the cteno-
dactylid. Two genera of fossorial herbivores (Spa-
lax and Heterocephalus), strongly convergent upon
other species filling a similar role in other deserts,
inhabit northern and eastern Africa. The morpho-
logical convergence associated with extreme fos-
soriality is well known (McNab, 1966), and similar-

ities probably extend to behavioral and other
ecological attributes as well. The small, bipedal her-
bivore category is represented by Allactaga tetra-
dactyla which occurs in far northern Africa in a
very limited geographical area (Setzer, 1956; Hoog-
straal, 1966; Ranck, 1968; Harrison, 1972). Finally,
as in North America, the large rodentiform herbi-
vore of North Africa is the hare, Lepus, a role filled
by a wallaby in Australia and a rodent in the Monte.
Even the gaits of these hopping-cursorial herbi-
vores in the widely-separated deserts are somewhat
similar. The shrew, Crocidiira, is known to fre-
quent xeric habitats, and is the small insectivore of
the Sahara (Wassif and Hoogstraal, 1953; Setzer,
1956; Hoogstraal, 1962). No small insectivorous ro-
dents are known in the Sahara, although larger in-
sectivores, hedgehogs of the family Erinaceidae
(Paraechinas) are common in the desert (Setzer,
1956; Hoogstraal, 1962), and seem similar to the
Monte armadillos, trophically, in their burrowing
habit, and in overall body proportions.

Southern Africa is the only desert, besides the
Monte, which lacks bipedal granivores; small, bi-
pedal herbivores are also missing in this desert. All
other categories are represented, however. Gerhil-
lus (and, perhaps, Desmodillus) are the quadrupe-
dal granivores, while Petromysciis, Dendromiis ,
Rhahdomys and perhaps, Desmodillus , are mi-
cro-omnivores (see Walker, 1964; Nel and Rauten-
bach, 1975; Nel, 1978). Geosciurus is the medium-
sized omnivore, and Paratomys and Thallomys are
the medium-sized herbivores; the latter with its ar-
boreal tendency, recalls Phyllotis griseoflavus of
the Monte, and Neotoma of the United States (see
Walker, 1964; Mares, 1973; Nel and Rautenbach,
1975; Nel, 1978). Mystromys is suggested in a num-
ber of computer analyses as possibly filling a small
insectivore role, or that of a micro-omnivore, in
southern Africa. Various insectivores are known
from xeric regions in southern Africa, however, in-
cluding Crocidura and two genera of elephant
shrews (Family Macroscelididae), Elephantulus
and Macroscelides (Bigalke, 1978; Nel, 1978); the
latter two genera are saltatorial (Walker, 1964).
Golden moles (Family Chrysochloridae), fossorial
insectivores, are also known from xeric habitats in
this region (Walker, 1964; Rautenbach, 1978). The
larger herbivore of southern Africa is the spring-
hare (Family Pedetidae), Pedetes, feeding on both
above- and below-ground plant parts (Walker, 1964).
Sciurids are present in southern Africa, and fosso-

1980

MARES— DESERT RODENT ECOLOGY

41

rial herbivores (Family Bathyergidae) are also com-
mon (Walker, 1964; Rautenbach, 1978).

The desert of Iran presents one of the poorest
deserts from the number of genera of small mam-
mals occurring in arid areas; nevertheless, seven of
the nine niche categories are represented. Species
of Jaculns are the bipedal granivores, while Ger-
hilliis and Meriones are quadrupedal granivores
(Naumov and Lobachev, 1975). Computer analyses
imply that Calomyscus is a micro-omnivore, while
Rhomhomys is the medium-sized herbivore (Lo-
bachev and Khamdamova, 1972). AUactaga is the
bipedal herbivore, while fossorial herbivores and
medium-sized omnivores are lacking in the arid
areas (Lay, 1967). Lepus is the large herbivore, as
it is throughout Asia.

The ecology of the small mammal fauna of the
Russian desert has been greatly clarified by Nau-
mov and Lobachev (1975). This desert is quite com-
plex faunistically, rivaling both the Sahara and
North American deserts. Three genera are included
as bipedal granivores, although two of these, Par-
adipiis and Eremodipus, may actually be more her-
bivorous than granivorous (Walker, 1964; Naumov
and Lobachev, 1975). Various species of Jaculus,
however, are primarily granivorous. Many species
of Meriones are present over this desert as qua-
drupedal granivores. Cardiocranius paradoxus, a
species which is poorly known ecologically, is pos-
sibly a bipedal micro-omnivore, although Naumov
and Lobachev (1975) suggest seeds are the only
food. Salpingotus, a near relative of Cardiocranius,
and also a bipedal species, may fill the role of small
insectivore. The ground squirrel, Sperinophilopsis,
fills the medium-sized omnivore niche, while Rhom-
boniys and, perhaps, Scirtopoda are the medium-
sized herbivores (Lobachev and Khamdamova,
1972; Naumov and Lobachev, 1975). No fossorial
herbivores are present in the Russian desert, while
Lepus is the large herbivore. The most diverse guild
in this desert is that of the small bipedal herbivores,
having no fewer than four genera represented (Di-
pus, Alactagiilus, AUactaga, Pygeretnms), and per-
haps as many as six, if Paradipus and Eremodipus
feed largely on roots and tubers.

The Chinese desert system is largely unstudied
(Lowe, 1968), but generally it is quite similar to the
Russian desert. Seven of the adaptive categories are
represented with only fossorial herbivores and
ground squirrels being absent. There are fewer gen-
era of dipodids than in Russia, but they neverthe-

less contribute bipedal granivores (Jaculus, Stylo-
dipus), a micro-omnivore (Cardiocranius), an
insectivore (Salpingotus), and bipedal herbivores
(AUactaga, Euchoreutes, Dipus) to this desert.

Clearly, there are great similarities in the types
of niche categories represented in each desert. No
desert supports all nine categories although the des-
erts of North America, North Africa, and Russia
(Turkestan Desert), each have eight. Usually these
are filled by rodents, but a perusal of Table 19 in-
dicates that there is great plasticity present in the
small mammals evolving in a particular desert re-
gion, whatever their taxonomic composition. Thus,
for example, the small insect eater niche is filled by
a shrew and a rodent in North America, a marsupial
mouse opossum in the Monte, various marsupials
in Australia, elephant shrews, a soricid, and per-
haps a rodent in southern Africa, a shrew in North
Africa, and a shrew and a rodent in the Turkestan
and Gobi deserts. Some of these insectivores are
quadrupedal, two are burrowing (a marsupial and
a golden mole), and some are bipedal (elephant
shrews and a rodent).

Bipedal and fossorial herbivores seem to be mu-
tually exclusive within a desert region, with the ex-
ception of AUactaga in northern Africa (where Spa-
lax is present). AUactaga is a fairly recent colonizer
of Africa, however, and has a very small geographic
distribution (Ranck, 1968). Apparently, as a desert
develops, a group of rodents can opt for fossoriality
and a diet of underground plant parts, or can spe-
cialize on the same food resource by being bipedal
(a strategy also employed by marsupials). Typical
examples of the former adaptation are geomyids in
North America, or bathyergids in southern Africa,
while dipodids are the only bipedal root-eating ro-
dents. Only one or the other strategy will predom-
inate in any one desert, however.

Large herbivores in deserts are all specialized for
rapid, rather long-distance, locomotion; some are
cursorial, others bipedal. Each desert generally
supports only one genus and species per locality
within this category; coexistence of species is the
exception (Sylvilagus and Lepus in North America)
rather than the rule.

Medium-sized herbivores are found in each of the
deserts of the world. They are often arboreal or
limited to localized areas of dense (often halophyt-
ic) vegetation. They are almost always quadrupe-
dal, with the exception of Scirtopoda, whose inclu-
sion in this category is uncertain.

42

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

Micro-omnivores and medium-sized omnivores
are present in most deserts. The latter are usually
sciurids, and the inclusion of Chaetophractus of the
Argentine Monte is only to illustrate that a medium-
sized omnivorous armadillo may fill part of the
squirrel niche in that desert. Iran and China lack
desert squirrels and this may be because either in-
sects and/or the usual seed and fruit resource is in
short supply. Certainly the arboreal diversity of
Iranian arid areas is exceedingly low.

Quadrupedal and bipedal granivores are usually
present within each desert, although the Monte and
Australia lack the former, and the Monte, southern
African and, perhaps, Australian deserts may lack
the latter. Bipedal granivores are always rodents
and usually medium-sized species (in the 30-150 g
body size range), whereas quadrupedal granivores
may include smaller species (for example, Gerhil-
lus, Perognothus).

Micro-omnivores are always rodents and almost
always quadrupedal (with Cardiocraniiis being a
possible bipedal exception). They are present in
probably all major desert areas and often repre-
sented by more than one genus within a desert.

Bipedality occurs in all deserts except the Argen-
tine Monte. It may be found in any food category,
although it is most common in species eating seeds,
insects, and underground plant parts. It seems to
be associated with sparse vegetation cover, al-
though bipeds and quadrupeds usually live sym-
patrically in most deserts. The Turkestan
(U.S.S.R.), Chinese (Gobi), and Australian deserts
have the greatest diversity of bipedal small mam-
mals, with four of the major niche types represented
by bipedal species.

From the preceding analyses, I suggest that, giv-
en time and opportunity, rodents will evolve to fill
particular guilds in desert ecosystems. In deserts
where it is known that rodents were late colonizers
(Monte, Australia), these same guilds were filled by
marsupials. Few deserts have unoccupied guilds,
however, and it is likely that future evolution within
these deserts will be that of refinement within a
guild, rather than the development of a new, unex-
ploited adaptive zone. Only in the Monte and in
Australia do there seem to be vacant zones awaiting
exploitation by rodents. Given the evenness with
which such zones have been filled in other desert
areas, I feel that the evolution of such species in
these deserts is highly probable.

Schall and Pianka (1978) discussed convergent
evolution in general terms and concluded that there

is much evidence, both theoretical and empirical,
suggesting that communities do not converge, al-
though they point out that spectacular incidents of
convergent evolution do occur occasionally. I be-
lieve, however, that there is much evidence, both
theoretical and empirical, that communities do con-
verge (for example, Cody, 1974; Puentes, 1976;
Orians and Solbrig, 1977; Cody et al., 1977). The
similarities in the plant communities of the Sonoran
and Monte deserts, or of the Chilean and Califor-
nian chaparral scrub areas are indeed remarkable;
most deserts have similar physiognomies. Also,
there are many examples of broad similarities in
vertebrate and invertebrate communities in xeric
areas. Ecologists have often looked for the differ-
ences in communities around the world rather than
the similarities. Thus, one with experience in the
Sonoran Desert is apt to notice a Rhea in the Monte
as a remarkably odd species, and not reflect on the
great similarities between the rodent Dolichotis and
Sonoran jackrabbits. Also, the presence of a small
mammalian desert insectivore in various deserts,
whether it be a marsupial, such as Sminthopsis in
Australia, Mannosa in Argentina, a shrew (No-
tiosorex) in the Sonoran Desert, a rodent {Onych-
omys or Calomyscus) in the Sonoran or Iranian des-
erts, or even an elephant shrew (Macroscelididae) in
Africa, suggests that there are niches in deserts that
are best filled by mammals, if mammals have had
sufficient evolutionary time to evolve the necessary
adaptations. The preponderance of granivory and
bipedality among small desert mammals needs no
further demonstration. These adaptive strategies
have evolved repeatedly in disjunct deserts in un-
related groups. Future research should center on
the selective forces that channel adaptations into a
broad suite of adaptive strategies. Deserts are se-
vere environments for a small mammal. For those
species which make an evolutionary hurdle into a
new xeric-adapted mode, immense areas and abun-
dant food resources supporting few other mammals
become available for exploitation.

Convergent evolution should not be examined
with the goal of finding a one-to-one correspon-
dence of species (that is, niches) in disjunct areas,
rather the entire array of adaptations employed by
organisms inhabiting a biome should be examined.
Three enormously diverse mammalian orders, the
Marsupialia, Insectivora, and Rodentia, have de-
veloped numerous desert species that fit within a
relatively few adaptive moulds; these species are
often uncannily-similar morphologically, trophical-

1980

MARES— DESERT RODENT ECOLOGY

43

ly, ecologically, and physiologically. Unique adap-
tations or ways of life are to be expected in species
which often differ enormously in their genetic
stock, but it is indeed interesting that, given these
genetic differences, so many adaptations have
evolved to endpoints that are so similar. Selection
of one particular ecological zone early in the history
of a taxon seems to canalize the adaptations that
will develop along a relatively narrow evolutionary
pathway. Such an occurrence suggests overriding

similarities in selective forces within similar eco-
systems, and suggests that the evolutionary options
open to organisms inhabiting a desert (and, presum-
ably, other ecosystems as well) are limited. As in-
vestigations of convergent evolution continue, and
they are only in their infancy at the moment, we
should begin to get a much deeper understanding of
the evolutionary process, and the possible limita-
tions to the diversity of modes of existence.

ACKNOWLEDGMENTS

This paper is the result of a number of years of research as
well as many personal interactions. My Argentine research was
supported under the International Biological Program’s Struc-
ture of Ecosystems Subprogram and a Research Fellowship from
the United States Fulbright Commission. My work in the south-
western United States was funded by a grant from the National
Council on Chicanos in Higher Education (NCCHE). Iranian
research was supported by an award from the Provost's Devel-
opment Fund, University of Pittsburgh. Field time in Egypt was
supported by a Research-Development award from The Smith-
sonian Foreign Currency Program under PL 480. I am grateful
to all of these agencies for their generous financial support, i
thank his Excellency Eskandar Firouz, Mr. Robert Tuck, and
the courteous staff of the MMTT Natural History Museum in
Tehran, for making my Iranian field research possible. The cu-
rators of The American Museum of Natural History. Dr. Sydney
Anderson; the United States National Museum, Drs, Charles
Handley and Henry Setzer; and the British Museum (Natural
History), Mr. Iain Bishop, very kindly allowed me to examine
specimens in their collections. The facilities of The Museum of

Northern Arizona were made available to me and I thank the
administration, investigators, and staff (particularly Dr. Jim
Reichman) for their courtesy. Drs. Susan Berman, Tom Lacher,
Jr,, Beryl Simpson, Duane Schlitter, and Dan Williams, were
helpful in various stages of the preparation of this manuscript,
as were my students Karl Streilein and Mike Willig. Drs. James
Brown and Michael Rosenzweig made very careful readings of
the penultimate version of this manuscript and their comments
are greatly appreciated. The final form of this manuscript is
largely a result of their suggestions, although they may not agree
with all of the points made in the paper. The length of this report
and the rudimentary artwork are strictly my responsibility. Dr.
Mike Bogan has helped me with many aspects of the data anal-
ysis and manuscript preparation; his patience is most gratefully
acknowledged. Finally, 1 thank the many people in various coun-
tries who have shared time, expertise, camaraderie, heat, and
spirits with me in the desert, notably Drs, Kamal Wassif and
Duane Schlitter, Abbas Ahmedi, Ricardo Ojeda, Ruben Bar-
quez, Brian O’Reagan, and my wife. Lynn.

LITERATURE CITED

Albrecht. G. H. 1978. Some comments on the use of ratios.
Syst. Zool., 1 7; 67-71.

Atchley. W. R. 1978. Ratios, regression intercepts and the
scaling of data, Syst, Zool., 27;78-83,

Atchley. W. R., and D. Anderson. 1978. Ratios and the
statistical analysis of biological data. Syst. Zool., 27;71-78.
Atchley, W. R.. C. T. Gaskins, and D. Anderson. 1976.
Statistical properties of ratios. I. Empirical results. Syst.
Zool., 25; 137-148.

Axelrod, D. I. 1950. Evolution of desert vegetation in western
North America. Carnegie Inst. Publ.. 590; 2 17-306.

. 1956. Mio-Pliocene floras from west-central Nevada.

Univ. California Publ.. Geol. Sci., 33; 1-321.

. 1957. Late Tertiary floras and the Sierra Nevadan

uplift. Bull. Geol. Soc. Amer.. 68; 19-46.

. 1958. Evolution of the Madro-Tertiary geoflora. Bot.

Rev., 24;433-509.

. 1967. Drought, diastrophism, and quantum evolution.

Evolution. 2I;201-209.

. 1970. Mesozoic paleogeography and early angiosperm

history. Bot. Rev., 36;277-3 19.

. 1972. Ocean-floor spreading in relation to ecosys-

tematic problems. Pp. 15-70, in A Symposium on ecosys-
tematics (R. T. Allen and F. C. James, eds.). Occas. Paper
No. 4, Univ. Arkansas Mus., Fayetteville, Arkansas.

Bailey, H. P., B. B. Simpson, and F. Vervoorst. 1977. The
physical environment; the independent variable. Pp. 13-49.
in Convergent evolution in warm deserts (G. H. Orians and
O. T. Solbrig, eds.), Dowden, Hutchinson and Ross, Publ.,
East Stroudsburg. Pennsylvania.

Bartholomew, G. A. 1956. Temperature regulation in the
macropod marsupial, Setonix hmchyurns. Physiol. Zool.,
29; 26-40.

Bartholomew. G. A., and H. H. Caswell. 1951. Locomo-
tion in kangaroo rats and its adaptive significance. J.
Mamm., 32; 155-169.

Baskin, J. A. 1978. Bensonomys. Caloinys. and the origin of
the phyllotine group of Neotropical cricetines (Rodentia:
Cricetidae). J. Mamm., 59; 125-135.

Beaty, C. B. 1978. The causes of glaciation. Amer. Sci.,
66;452-459.

Bennett, S., and D. Bowers. 1976. An introduction to multi-

44

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 16

variate techniques for social and behavioural sciences. John
Wiley and Sons, Inc., New York.

Bentley, P. J. I960. Evaporative water loss and temperature
regulation in the marsupial Setony.x hrachyurus. Australian
J. Exp. Biol. Med. Sci., 38:301-306.

Bigalke, R. C. 1978. Mammals. Pp. 981-1048, in Biogeogra-
phy and ecology of southern Africa (M. J. A. Werger, ed.).
Junk (h.v.). The Hague, The Netherlands.

Blackith, R. E., and R. A. Reyment. 1971. Multivariate
morphometries. Academic Press, New York.

Brown, J. H. 1975. Geographical ecology of desert rodents.
Pp. 315-341, in Ecology and evolution of communities (M.
L. Cody and J. M. Diamond, eds.), Belknap Press of Har-
vard Univ., Cambridge, Massachusetts.

Brown, J. H., and D. W. Davidson. 1977. Competition be-
tween seed-eating rodents and ants in desert ecosystems.
Science, 196:880-882.

Charney, j., P. H. Stone, and W. J. Quirk. 1975. Drought
in the Sahara: a biogeophysical feedback mechanism. Sci-
ence, 187:435-436.

Chew, R. M., and A. E. Chew. 1970. Energy relationships of
the mammals of a desert shrub {Larrea trident at a) com-
munity. Ecol. Monogr., 40:1-21.

Choubert, G. 1952. Histoire geologique du domaine de I’Anti-
Atlas. 19. Congr. Geol. Intern. Monographic Regionales,
Geologic du Maroc, 1:77-194.

Clark, D. 1975. Understanding canonical correlation analysis.
Concepts and techniques in modern geography. Geo. Abstr.
Ltd.. Univ. of Anglia, Norwich, England.

Clifford, H. T., and W. Stephenson. 1975. An introduction
to numerical classification. Academic Press, New York.

Cody, M. L. 1974. Competition and the structure of bird com-
munities. Princeton Univ. Press, Princeton, New Jersey.

. 1975. Towards a theory of continental species diversi-
ties: bird distributions over Mediterranean habitat gradients.
Pp. 214-257, in Ecology and evolution of communities (M.
L. Cody and J. M. Diamond, eds.), Belknap Press, Cam-
bridge, Massachusetts.

. 1978. Habitat selection and interspecific territoriality

among the sylviid warblers of England and Sweden. Ecol.
Monogr., 48:351-396.

Cody, M. L., E. R. Fuentes, W. Glanz, J. H. Hunt, and A.
R. Moldenke. 1977. Convergent evolution in the consum-
er organisms of Mediterranean Chile and California. Pp.
144-192, in Convergent evolution in Chile and California
(H. A. Mooney, ed.), Dowden, Hutchinson and Ross, Inc.,
East Stroudsburg, Pennsylvania.

Cooke, H. B. S. 1972. The fossil mammal fauna of Africa. Pp.
89-139, in Evolution, mammals, and southern continents
(A. Keast, F C. Erk, and B. Glass, eds.). State Univ. New
York Press, Albany, New York.

Cooi EY, W. W.. AND P. R. Lohnes. 1971. Multivariate data
analysis. John Wiley and Sons, Inc., New York.

Corruccini, R. S. 1977. Correlation properties of morpho-
metric ratios. Syst. Zook, 26:211-214.

Cracraft, j. 1974. Continental drift and vertebrate distribu-
tion. Ann. Rev. Ecol. Syst., 5:215-261.

Crovello, T. j. 1970. Analysis of character variation in ecol-
ogy and systematics. Ann. Rev. Ecol. Syst., 1:55-98.

Crowcroft, P., and G. K. Godfrey. 1968. The daily cycle
of activity in two species of Sminthopsis (Marsupialia:
Dasyuridae). J. Anim. Ecol., 37:63-73.

Darwin, C. 1859. On the origin of species. John Murray, Lon-
don.

Daultrey, S. 1976. Principal components analysis. Concepts
and techniques in modern geography. Geo. Abstr. Ltd.,
Univ. Anglia, Norwich, England.

Dawson, M. R. 1967. Fossil history of the families of Recent
mammals. Pp. 12-53, in Recent mammals of the world (S.
Anderson and J. K. Jones, Jr., eds.), Ronald Press, New
York.

Dawson, T.J. 1976. Energetic cost of locomotion in Australian
hopping mice. Nature, 259:305-307.

Dawson, T. J., and G. D. Brown. 1970. A comparison of the
insulative and reflective properties of the fur of desert kan-
garoos. Comp. Biochem. Physiol., 37:23-38.

Diamond, J. M., and R. M. May. 1976. Island biogeography
and the design of natural preserves. Pp. 163-186, in Theo-
retical ecology, principles and applications (R. M. May,
ed.), W, B. Saunders Co., Philadelphia, Pennsylvania.

Dietz, R. S., and J. C. Holden. 1970. Reconstruction of Pan-
gaea: breakup and dispersion of continents, Permian to
present. J. Geophys. Res., 75:4939-4956.

Dodson, P. 1978. On the use of ratios in growth studies. Syst.
Zook, 27:62-67.

Eisenberg, J. F. 1975. The behavior patterns of desert rodents.
Pp. 189-224. in Rodents in desert environments (I. Prakash
and P. K. Ghosh, eds.). Junk (b.v.). The Hague, The Neth-
erlands.

Ellerman, j. R. 1949. The families and genera of living ro-
dents. Brit. Mus. Nat. Hist. Pubk. London.

Findley, J. S. 1972. Phenetic relationships among bats of the
genus Myotis. Syst. Zook, 21:31-52.

. 1976. The structure of bat communities. Amer. Nat.,

110:129-139.

Fuentes, E. R. 1976. Ecological convergence of lizard com-
munities in Chile and California. Ecology, 57:3-17.

Furon, R. 1941. Geogolie du plateau iranien (Perse, Afghani-
stan, Beloutchistan). Mem. Mus. Nation. Hist. Nat., n.s.,
7:177-414.

. I960. Geology of Africa. Hafner Pubk Co., New York,

New York, 370 pp.

Ganji, M. H. 1955. The climates of Iran. Bulk Soc. Geog.
Egypte, 28: 195-299.

Glanz, W. 1977. Comparative ecology of rodent communities
in California and Chile. Unpublished Ph.D. thesis, Univ.
California, Berkeley.

Godfrey, G. K. 1968. Body-temperatures and torpor in Smin-
thopsis crassicaiiJata and S. larapinta (Marsupialia: Dasy-
uridae). J. Zook London, 156:499-511.

Gnanadesikan, R. 1977. Methods for statistical data analysis
of multivariate observations. John Wiley and Sons, Inc.,
New York.

Goodman, M. M. 1972. Distance analysis in biology. Syst.
Zook, 21:174-186.

Goodman, M. M., and E. Paterniani. 1969. The races of
maize: III. Choices of appropriate characters for racial clas-
sification. Econ. Bot., 23:265-273.

Greegor, D. H., Jr. 1975. Renal capabilities of an Argentine
desert armadillo. J. Mamm., 56:626-632.

Grenot, C. j. 1974. Physical and vegetational aspects of the
Sahara Desert. Pp. 103-164, in Desert biology, Vok II (G.
W. Brown, Jr., ed.). Academic Press, New York, London.

1980

MARES— DESERT RODENT ECOLOGY

45

Gunderson, H. L. 1976. Mammalogy, McGraw-Hill Puhl.,
New York, New York.

Haffer, J. 1970. Geologic-climatic history and zoogeographic
significance of the Uraba region in northwestern Colombia.
Caldasia, 10:603-636.

Hafner, J. C. 1978. Evolutionary relationships of kangaroo
mice, genus Microdipodops. J. Mamm., 59:354-365.

Hall, E. R., and K. R. Kelson. 1959. The mammals of North
America. Ronald Press, New York, l:xxx -i- 1-546 + 79;
2:viii + 547-1083 + 79.

Harrison, D. L. 1972. The mammals of Arabia. Vol. 3. La-
gomorpha, Rodentia. Ernest Benn Ltd., London.

Hawbecker, a. C. 1947. Food and moisture requirements of
the Nelson antelope ground squirrel. J. Mamm., 28: 1 15-125.

Hershkovitz, P. 1962. Evolution of neotropical cricetine ro-
dents (Muridae) with special reference to the phyllotine
group. Fieldiana. Zool., 46:1-524.

. 1966. Mice, land bridges and Latin America faunal in-
terchange. Pp. 725-747, in Ectoparasites of Panama (R. L.
Wenzel and V. J. Tipton, eds.). Field Mus, Nat. Hist., Chi-
cago, Illinois.

. 1972. The Recent mammals of the neotropical region:

a zoogeographic and ecological review. Pp. 311-431. in
Evolution, mammals, and southern continents (A. Keast, F.
C. Erk, and B. Glass, eds.). State Univ. New York Press,
Albany, New York.

Hespenheide, H. a. 1973. Ecological inferences from mor-
phological data. Ann. Rev. Ecol. Syst., 4:213-230.

Heyck, T. W., and W. Klecka. 1973. British Radical M.P.’s,
1874-1875: new evidence from discriminate analysis. J. In-
terdis. Hist., 4:161-184.

Hildebrand. M. 1974. Analysis of vertebrate structure. John
Wiley and Sons. Inc., New York.

Hills, M. 1978. On ratios — A response to Atchley, Gaskins,
and Anderson. Syst. Zool., 27:61-62.

Hoffmeister, D. F., and W. W. Goodpaster. 1954. The
mammals of the Huachuca Mountains, southeastern Ari-
zona. Illinois Biol. Monogr., Univ. Illinois Press. Urbana.
24:1-152.

Hoffstetter. R., and R. Lavocat. 1970. Decouverte dans
le Deseadien de Bolivie de genres pentalophodontes appuy-
ant les affinities Africaines des rongeurs caviomorphs.
Comptes. Rend. Acad. Sci. (Paris), 271:172-175.

Hoogstraal, H. 1962. A brief review of the contemporary
land mammals of Egypt (including Sinai). 1 . Insectivora and
Chiroptera. J. Egypt. Ptibl. Hlth. Assoc., 37(4): 143-162.

. 1966. A brief review of the contemporary land mam-
mals of Egypt (including Sinai). 2. Lagomorpha and Roden-
tia. J. Egypt. Publ. Hlth. Assoc., 38( I): 1-35.

Howell, A. B. 1932. The saltatorial rodent Dipodoniys: the
functional and comparative anatomy of its muscular and
osseus systems. Proc. Amer. Acad. Arts and Sciences,
67:377-536.

Hubbard, J. P. 1974. Avian evolution in the aridlands of North
America. The Living Bird. 12th Annual, 1973: 155-196.

Hudson. J. W. 1962. The role of water in the biology of the
antelope ground squirrel Citellns leiici/rns. Univ. California
Publ.. Zool., 64:1-56.

Ito, K. 1969. On the effect of heteroscedasicity and non-nor-
mality upon some multivariate procedures, in Multivariate
analysis II (P. R. Krishnaiah, ed.). Academic Press, New
York.

Jaeger, E. C. 1957. The North American deserts. Stanford
Univ. Press, Stanford, California.

Karr, J. R., and F. C. James. 1975. Eco-morphological con-
figurations and convergent evolution in species and com-
munities. Pp. 258-291, in Ecology and evolution of com-
munities (M. L. Cody and J, M. Diamond, eds.), Belknap
Pi •ess. Harvard Univ., Cambridge, Massachusetts.

Keast. A. 1972</. Australian mammals: zoogeography and evo-
lution. Pp. 195-246, in Evolution, mammals, and southern
continents (A. Keast, F. C. Erk, and B. Glass, eds.). State
Univ. New York Press, Albany, New York.

. 19727>. Continental drift and the evolution of the biota

on southern continents. Pp. 23-87, in Evolution, mammals,
and southern continents (A. Keast, F. C. Erk, and B. Glass,
eds.). State Univ. New York Press, Albany, New York.

Klecka, W. R. 1975. Discriminant analysis. Pp. 434-467, in
SPSS statistical package for the social sciences (N . H. Nie,
C. H. Hull. J. G. Jenkins, K. Steinbrenner, and D. H. Brent,
eds.), 2nd ed., McGraw-Hill. New York.

Lavocat, R. 1969. La systematique des rongeurs hystrico-
morphs et la derive des continents. Comptes. Rend. Acad.
Sci. (Paris), 269: 1496-1497.

Lay, D, M. 1967. A study of the mammals of Iran resulting
from the Street Expedition of 1 962-63 . Fieldiana, Zool.,
54:1-282.

. 1972. The anatomy, physiology, functional significance

and evolution of specialized hearing organs of gerbilline ro-
dents. J. Morphol., 138:41-118.

Layne, j. N. 1972. Tail autotomy in the Florida mouse Pero-
tnyscus ftoridanns. J. Mamm., 53:62-71.

Lidicker, W. Z.. Jr. I960. An analysis of intraspecific varia-
tion in the kangaroo rat Dipodoniys merrianii. Univ. Cali-
fornia Publ., Zool., 67:125-218.

Lobachev, V. S., AND T. U. Khamdamova. 1972. Food hab-
its of the great gerbil. Byull. Mosk. O-va. Ispyt. Prir. Otd.
Biol., 77(5):40-54. In Russian with English summary.

Lockard, R. B., and D. H. Owings. 1974. Seasonal variation
in moonlight avoidance by bannertail kangaroo rats. J.
Mamm., 55: 189-193.

Logan, R. F. 1968. Causes, climates, and distribution of des-
erts. Pp. 21-50, in Desert biology (G. W. Brown, Jr., ed.).
Academic Press, New York and London.

Lowe, C. H. 1968. Fauna of desert environments, with disease
information. Pp. 567-645, in Deserts of the world ( W. G.
McGinnies, B. J. Goldman, and P. Paylore, eds.), Univ.
Arizona Press, Tucson.

Lundelius, E. L., Jr., AND W. D. Turnbull. 1967. Pliocene
mammals from Victoria, Australia. Australian-New Zealand
Assoc. Adv. Sci. 39th Congr., Melbourne, Jan. 1967. Abstr.
Sec. G:k-9.

Lustig, L. K. 1968. Geomorphology and surface hydrology of
desert environments. Pp. 93-283. in Deserts of the world
(W. G. McGinnies, B. J. Goldman, and P. Paylore, eds.),
Univ. Arizona Press, Tucson.

MacArthur, R. H., and E. O. Wilson. 1967. The theory of
island biogeography. Princeton Univ. Press, Princeton.
New Jersey.

Mares, M. A. 1973. Climates, mammalian communities, and
desert rodent adaptations: an investigation into evolutionary
convergence. Unpublished Ph.D. thesis, Univ. Texas, Aus-
tin, Texas.

. I975u. South American mammal zoogeography: evi-

46

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 16

dence from convergent evolution in desert rodents. Proc.
Nat. Acad. Sci. USA, 72:1702-1706.

. 19756. Observations of Argentine desert rodent ecol-
ogy, with emphasis on water relations of Eligmodontia ty-
piis. Pp. 155-175. in Rodents in desert environments (I.
Prakash and P. K. Ghosh, eds.). Junk (h.v.). The Hague,
The Netherlands.

. 1976. Convergent evolution of desert rodents: multi-
variate analysis and zoogeographic implications. Paleobiol-
ogy, 2:39-63.

. 1977«. Water economy and salt balance in a South

American desert rodent. Eligmodontia typns. Comp. Bio-
chem. Physiol., 56A:325-332.

. 19776. The comparative water relations and habitat re-
quirements of three South American phyllotine rodents. J.
Mamm., 58:514-520.

. I977r. Aspects of water balance of Oryzoinys longi-

candatns from northwest Argentina. Comp. Biochem. Phys-
iol., 57A: 237-238.

. \911d. Water independence in a non-desert South

American rodent. J. Mamm., 58:653-656.

. 1979. Small mammals and creosotebush: patterns of

richness. In Larrea: A vast resource of the American des-
erts (E. Campos Lopez and T. J. Mabry, eds.), Centro.
Invest. Quim. Aplicadas (CIQA), Saltillo, Coahuila, Mexi-
co, in press.

Mares, M. A., W. F. Blair, F. A. Enders, D. Greegor, Jr.,

A. C. Hulse, j. H. Hunt, D. Otte, R. D. Sage, and C.
S. Tomoff. 1977. The strategies and community patterns
of desert animals. Pp. 107-163. in Convergent evolution in
warm deserts (G. H. Orians and O. T. Solbrig, eds.), Dow-
den. Hutchinson and Ross, Publ., East Stroudsburg, Penn-
sylvania.

Mares, M. A., F. Enders, J. Kingsolver, J. Neff, and B.

B. Simpson. 1977. Prosopis as a habitat component. Pp.
123-149, in Mesquite: its biology in two desert scrub eco-
systems (B. B. Simpson, ed.), Dowden, Hutchinson and
Ross, East Stroudsburg. Pennsylvania.

Mares. M. A., and A. C. Hulse. 1977. Patterns of some ver-
tebrate communities in creosotebush deserts. Pp. 209-226,
in Creosote bush: biology and chemistry of Larrea in New
World deserts (T. Mabry, J. Hunziker, and D. R. DiFeo,
eds.), Dowden, Hutchinson and Ross. East Stroudsburg,
Pennsylvania.

Mares. M. A., and M. L. Rosenzweig. 1978. Granivory in
North and South American deserts: rodents, birds, and ants.
Ecology, 59:235-241.

Marshall, L. G. 1979. A model for paleobiogeography of
South American cricetine rodents. Paleobiology, 5: 126-132.

Marshall, L. G., R. F. Butler. R. E. Drake, G. H. Curtis,
AND R. H. Tedford. 1979. Calibration of the Great Amer-
ican Interchange. Science. 204:272-279.

Marshall, L. G., AND M. K. Hecht. 1978. Mammalian faunal
dynamics of the Great American Interchange: an alternative
interpretation. Paleobiology, 4:203-206.

Marshali , L. G., R. Hoffstetter, and R. Pascual. 1979.
Geochronology of the continental mammal-bearing Tertiary
of South America. In Vertebrate paleontology as a disci-
pline in geochronology (M. O. Woodburne, ed.), Univ. Cal-
ifornia Press, in press.

Martin, P. S., AND P. J. Mehringer. 1965. Pleistocene pollen
analysis and biogeography of the Southwest. Pp. 433-451,

in The Quaternary of the United States (H, E. Wright and
D. G. Frey, eds.), Princeton Univ. Press, Princeton, New
Jersey.

McNab, B. K. 1966. The metabolism of fossorial rodents: a
study in convergence. Ecology, 47:712-732.

Meigs, P. 1957. Arid and semiarid climatic types of the world.
Pp. 135-138, in International Geophysical Union, 17th Con-
gress, Washington, D.C., 8th General Assembly, Proceed-
ings.

Mercer, J.H. 1973. Cainozoic temperature trends in the south-
ern hemisphere: Antarctic and Andean glacial evidence. Pp.
87-1 14, in Palaeoecology of Africa, the surrounding islands
and Antarctica (E. M. van Zinderen Bakker, ed.), Balkema,
Capetown.

. 1976. Glacial history of southernmost South America.

Quat. Res., 6:125-166.

Meserve, P. L., AND W. E. Glanz. 1978. Geographical ecol-
ogy of small mammals in the northern Chilean arid zone. J.
Biogeog., 5:135-148.

Miller, C. R., and E. G. Butler. 1966. Anomia and eunom-
ia: a methodological evaluation of Srole’s Anomie Scale.
Amer. Soc. Rev., 31:400-406.

Molnar, P. . AND P. Tapponnier. 1975. Cenozoic tectonics of
Asia: effects of a continental collision. Science, 189:419-
426.

Morello, j. 1958. La provincia fitogeografica del Monte. Op-
era Lilloana, 2: 1-155.

Morton, S. R. 1979. Diversity of desert-dwelling mammals: a
comparison of Australia and North America. J. Mamm.,
60:253-264.

Naumov, N. P., AND V. S. Lobachev. 1975. Ecology of desert
rodents of the U.S.S.R. fjerboas and gerbils). Pp. 465-598,
in Rodents in desert environments (I. Prakash and P. K.
Ghosh, eds.). Junk (b.v.). The Hague, The Netherlands.

Net, J. A. J. 1978. Habitat heterogeneity and changes in small
mammal community structure and resource utilization in the
southern Kalahari. Pp. 118-131, in Ecology and taxonomy
of African small mammals (D. A. Schlitter, ed.). Bull. Car-
negie Mlis. Nat. Hist., 6:1-214.

Nel, j. a. j., and I. L. Rautenbach. 1975. Habitat use and
community structure of rodents in the southern Kalahari.
Mammalia, 39:9-29.

Nevo, E. 1973. Adaptive variation in size of cricket frogs.
Ecology, 54:1271-1281.

Ognev, S. J. 1963. Mammals of the U.S.S.R. and adjacent
countries. Vol. 6, Rodents. Israel Program for Scientific
Translations.

Orians, G. H., and O. T. Solbrig (eds.). 1977. Convergent
evolution in warm deserts. Dowden. Hutchinson and Ross,
Publ., East Stroudsburg, Pennsylvania.

Osgood, W. H. 1947. Cricetine rodents allied to Pliyllotis. J.
Mamm.. 28:165-174.

Otterman, j. 1974. Baring high-albedo soils by overgrazing:
a hypothesized desertification mechanism. Science, 186:531-
533.

Oxnard, C. E. 1978, One biologist's view of morphometries.
Ann. Rev. Ecol. Syst., 9:219-241.

Patterson, B., and R. Pascual. 1972. The fossil mammal
fauna of South America. Pp. 247-309, in Evolution, mam-
mals, and southern continents (A. Keast, F. C. Erk, and B.
Glass, eds.). State Univ. New York Press, Albany, New
York.

1980

MARES— DESERT RODENT ECOLOGY

47

Pearson, K. 1897. On a form of spurious correlation which
may arise when indices are used in the measurement of
organs. Proc. Royal Soc. London, 60:489-502,

Petter, F. 1961 . Repartition geographique et ecologie des ron-
geurs desertique. Mammalia, 24: 1-219.

Price, M. V. 1 978r; . The role of microhabitat in structuring
desert rodent communities. Ecology, 59:910-921 .

. I9787>. Seed dispersion preferences of coexisting desert

rodent species. J. Mamm., 59:624-626.

PuROHiT, K. G. 1971 . Absolute duration of survival of tarn mar
wallaby (Macropus eugenii, Marsupialia) on sea water and
dry food. Comp. Biochem. Physiol.. 39A:473-48I .

. 1974. Observations on histomorphology of kidneys and

urine osmolarities in some Australian desert rodents. Zool .
Anz., Jena, 193:221-227.

. 1976. Effects of drinking sea water on urinary and plas-
ma electrolytes concentrations and osmolarities in the tarn-
mar wallaby (Macropus eugenii, Marsupialia). Comp. Phys-
iol. Ecol., 1:136-140.

Ranck. G. L. 1968. The rodents of Libya: taxonomy, ecology,
and zoogeographical relationships. Bull. U.S. Nat. Mus.,
275: 1-272.

Rautenbach, 1. L. 1978. A numerical re-appraisal of the
southern African biotic provinces. Pp. 175-187, in Ecology
and taxonomy of African small mammals (D. A. Schlitter,
ed.). Bull. Carnegie Mus. Nat. Hist., 6:1-214.

Raven, P. H.. and D. 1. Axelrod. 1972. Plate tectonics and
Australian paleobiogeograph y . Science, 176: 1379-1386.

Reichman, O. J.. and D. Oberstein. 1977. Selection of seed
distribution types by Dipodoniys inerriciini and Perognathus
ampins. Ecology. 58:636-643.

REtG, O. A. 1979n. Roedores cricetidos del Plioceno Superior
de la Provincia de Buenos Aires (Argentina). Publ. Mus.
Munic. Cienc. Nat. “Lorenzo Scaglia,” in press.

. 19796, A new fossil genus of South American cricetid

rodents allied to Wiedomys, with an assessment of the Sig-
modontinae. Publ, Mus. Munic. Cienc. Nat. “Lorenzo
Scaglia," in press.

Reig, O. a., and O. j. Linares. 1969. The occurrence of
Akodon in the upper Pliocene of Argentina. J. Mamm.,
50:643-647.

Riek, E. F. 1970. Lower Cretaceous fleas. Nature, 227:746-
747.

Robinson, J. W., and R. S. Hoffmann. 1975. Geographical
and interspecific cranial variation in big-eared ground squir-
rels (Spennophilus): a multivariate study. Syst. Zool. ,
24:79-88.

Rohlf, F. j., and R. R. Sokal. 1965. Coefficients of corre-
lation and distance in numerical taxonomy. Univ. Kansas
Sci. Bull.. 45:3-27.

Romer, a. S. 1966. Vertebrate paleontology. Univ. Chicago
Press, Chicago. 3rd ed., 468 pp.

Rosenzweig, M. L., B. Smigel, and A. Kraft. 1975. Pat-
terns of food, space and diversity. Pp. 241-268, in Rodents
in desert environments (I. Prakash and P. K. Ghosh, eds.).
Junk (b.v.). The Hague. The Netherlands.

Sarmiento, G. 1976. Evolution of arid vegetation in tropical
America. Pp. 69-99. in Evolution of desert biota (D. S.
Goodall, ed.). Univ. Texas Press, Austin.

Savage, D. E. 1951 . A Miocene phyllostomatid bat from Co-
lombia, South America. Univ. California Publ., Geol. Sci.,
28:357-366.

ScHALL, J. J., AND E. R. PiANKA. 1978. Geographical trends
in numbers of species. Science, 201:679-686.

Schlitter, D. A. 1976. Taxonomy, zoogeography, and evo-
lutionary relationships of hairy-footed gerbils, Gerhillus
(GerhiUus), of Africa and Asia, (Mammalia: Rodentia). Un-
published Ph.D. thesis, Univ. Maryland, College Park.

Schmidt-Nielsen, K., T. J. Dawson, H. T. Hammel, D.
Hinds, and D. C. Jackson. 1965. The jack-rabbit — a
study in desert survival. Hvalradets Skrifter. 48:125-142,

Schmidt-Nielsen, K., and A. E. Newsome. 1962. Water bal-
ance in the mulgara (Dasycen us cristicauda), a carnivorous
desert marsupial. Australian J. Biol. Sci., 15:683-689.

ScHNELL, G. D. \91()a. A phenetic study of the suborder Lari
(Aves). 1. Methods and results of principal components
analyses. Syst. Zool.. 19:35-57.

. 19706. A phenetic study of the suborder Lari (Aves).

11. Phenograms, discussion, and conclusions. Syst. Zool.,
19:264-302.

Schoener, T. W. 1965. The evolution of bill size differences
among sympatric congeneric species of birds. Evolution,
19:189-213.

Setzer, H. W. 1956. Mammals of the Anglo-Egyptian Sudan.
Proc. U.S. Nat. Mus., 106:447-587.

Simpson, B. B. 1975. Pleistocene changes in the flora of the
high tropical Andes. Paleobiology, 1:273-294.

Simpson. G. G. 1945. The principles of classification and a
classification of mammals. Bull. Amer. Mus. Nat. Hist.,
85:1-350.

. 1950. History of the fauna of Latin America. Amer.

Sci., 38:361-389.

. 1961. Historical zoogeography of Australian mammals.

Evolution, 15:431-446.

. 1970. The Argyrolagidae, extinct South American mar-
supials. Bull. Mus. Comp. Zool., 139: 1-86.

Snea TH, R. H. A., AND R. R. Sokal. 1973. Numerical tax-
onomy. W. H. Ereeman and Co., San Francisco.

SoLBRiG, O. T. 1976. The origin and floristic affinities of the
South American temperate desert and semidesert regions.
Pp. 7-49, in Evolution of desert biota (D. W, Goodall, ed.),
Univ. Texas Press, Austin.

Tamsitt, J. R. 1967. Niche and species diversity in neotropical
bats. Nature. 2 13:784-786.

Tarling, D. H. 1971. Gondwanaland. paleomagnetism and
continental drift. Nature, 229: 17-21.

Tatsuoka, M. M. 1971. Multivariate analysis: techniques for
educational and psychological research. John Wiley and
Sons, Inc., New York.

Thompson, W. D’Arcy. 1917. On growth and form. Cambridge
Univ. Press. London.

Tyndale-Biscoe, H. 1973. Life of marsupials. American El-
sevier Publ., New York.

Van der Hammen, T. 1974. The Pleistocene changes of vege-
tation and climate in tropical South America. J. Biogeogr.,
1:3-26.

Van Devender, T. R. 1977. Holocene woodlands in the south-
western deserts. Science, 198:189-192.

Vaughan, T, A. 1972. Mammalogy. W. B. Saunders, Phila-
delphia.

VooRHiES, M. R. 1975. A new genus and species of fossil kan-
garoo rat and its burrow. J. Mamm., 56: 160-176.

VuiLLEUMiER, B. S. 1971. Pleistocene changes in the fauna and
flora of South America. Science, 173:771-780.

48

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 16

Walker, E. P. 1964. Mammals of the world. John Hopkins
Press, Baltimore, Pxlviii + 1-646; 2:viii + 647-1500.
Wassif, K. 1972. The present distribution of rodents in Egyp-
tian deserts and its bearing on future agricultural projects.
Pp. 55-60, in Proc. 1st. Sci. Symp. on Rodents and their
control in Egypt, Cairo, Gen'l. Org. for Govt. Printing Of-
fice.

Wassif, K., and H. Hoogstraal. 1953. The mammals of
South Sinai, Egypt. Proc. Egypt. Acad. Sci., 9:63-79.
Watts, C. H. S. 1973. The Australian rodents Notoinys ale.xis
and Pseudomys australis as laboratory animals. Pp. 179-
185, in Experimental animals, Vol. 22, Suppl., 1973, Proc.
ICLA Asian Pacific Meeting of Laboratory Animals, 20-25
September 1971, Tokyo and Inuyama.

. 1974. The native rodents of Australia: a personal view.

J. Australian Mamm. Soc., 1:109-116.

Webb, S. D. 1978. A history of savanna vertebrates in the New
World. Part 11: South America and the Great Interchange.
Ann. Rev. Ecol. Syst., 9:393-426.

Webster, D. B. 1962. A function of the enlarged middle-ear
cavities of the kangaroo rat, Dipodoinys. Physiol. Zook,
35:248-255.

Williams, D. F., and M. A. Mares. 1978. A new genus and
species of phyllotine rodent (Mammalia: Muridae) from
northwestern Argentina. Ann. Carnegie Mus., 47:193-221.

WoNDOLLECK, J. T. 1978. Forage-area separation and overlap
in heteromyid rodents. J. Mamm., 59:510-518.

Wood, A. E. 1935. Evolution and relationships of the hetero-
myid rodents with new forms from the Tertiary of western
North America. Ann. Carnegie Mus., 24:76-262.

. 1955. A revised classification of rodents. J. Mamm.,

36:165-187.

Wood, A. E., and B. Patterson. 1970. Relationships among
hystricognathous and hystricomorphous rodents. Mamma-
lia, 34:628-639.

ZoHARY. M. 1963. On the geobotanical structure of Iran. Res.
Counc. of Israel, Suppl. to Vol. 1 1 D: l-l 13.

APPENDIX 1

Annotated list of morphological characters utilized in multivariate analyses

1. Head-body length. — Distance from tip of snout to begin-
ning of tail, generally derived by subtracting length of tail (ver-
tebrae) from total length (tip of snout to tip of tail).

2. Tail length.

3. Hind-foot length. — From back of heel to tip of longest toe-
nail.

4. Height of ear from notch.

5. Length of longest vihrissae. — A millimeter ruler was placed
at base of vibrissae and the longest was measured to the nearest
millimeter.

6. Length of hair between shoulders. — Measured by placing
millimeter ruler against skin and noting length of majority of
hairs.

7. Tail/head-body ratio. — The tail functions as an organ of
balance, and would be expected to be particularly important to
bipedal species. Arboreal species would also have need of a long
tail for balance while climbing over branches.

8. Hind foot/head-body ratio. — Long hind feet relative to
body size are often adaptations for a bipedal habit, as evidence
by kangaroos, to cite an extreme example. Bipedality is a desert
rodent adaptation, thus desert species would be expected to have
a high hind foot to head-body ratio.

9. Earlhead-body ratio. — Mammals in deserts have two ways
of increasing audial acuity. First, pinnae can be greatly enlarged,
resulting not only in greater hearing ability, but in an efficient
organ for radiating body heat to the environment (see Schmidt-
Nielsen, 1965, regarding Lepus). Secondly, species can evolve
inflated tympanic bullae to increase sound reception. It is pos-
sible that non-fossorial species would tend toward larger ears,
particularly if they were diurnal in activity. Species which live
in burrows might find long ears a handicap in narrow tunnels
and would thus tend toward inflated bullae (for example, Web-
ster, 1962; Lay. 1972). Lepus (jackrabbits and hares) is an ex-
ample of non-fossorial species having long ears, and Dolichotis
(the Patagonian “hare”) a species which lives in burrows and
possesses shorter ears.

10. Vibrissaelhead-body ratio. — Possibly long vibrissae are

associated with desert living for a number of reasons. Desert
rodents, known to be very nocturnal even to the point of avoid-
ing moonlit nights (for example. Lockard and Owings, 1974),
could facilitate moving about on a pitch-black evening by utiliz-
ing long and dense vibrissae. The open habitat of a desert would
be conducive to long vibrissae, whereas an animal such as a
microtine which lives in dense vegetation or a fossorial animal
might not reap selective advantage by having exceedingly long
vibrissae which would be in constant contact with vegetation or
burrow walls.

11. Hair/head-body ratio. — Ratio of the length of hair be-
tween shoulders to head-body length. 1 would not expect desert
species to have either particularly long or short hair, whereas
species from colder localities, such as a coniferous forest, might
possess fairly long pelage for greater insulation.

12. Weight. — Measured to 0.1 grams.

13. Foot bristles. — Coded none (0), somewhat (I), many and
well developed (2), stiff and specialized (3), brush-like (4). This
character offers an example of the distinction between an eco-
logical character and a taxonomic one. The fossorial rodents
Thomoniys and Ctenoniys possess stiff bristles between the toes
which facilitate soil movement. Thomoniys pushes soil from its
burrow by a forward motion of the body and forefeet and pos-
sesses bristles between the toes of the front paws. Ctenoniys,
on the contrary, pushes soil with the posterior parts of the body
and by rapid backward kicks with the hind feet whose toes have
stiff bristles. Were one interested in taxonomic characters, oc-
currence of bristles at different ends of the body could differ-
entiate the species. Being interested primarily in ecological func-
tion, however, in this study only the presence of these
functionally similar structures was noted, not their location on
the body.

14. Tuftiness of tad. — Coded from no tuft (0) to large, well-
developed. conspicuous tuft (5). Why many desert rodents have
tufts at the ends of their tails is unclear. Possibly it functions as
a balancing structure, although 1 doubt the weight of the tuft is
sufficient to function in this manner. Increased resistance as it

1980

MARES— DESERT RODENT ECOLOGY

49

moves through the air could make the tuft act as a rudder to
allow the animal more easily to flick its body sideways in mid-
air. More plausible perhaps is a predator-distraction function of
a large white tuft. If the predator's attack could be deflected to
this point, and if the tail were easily autotomized (as Layne,
1972 has shown for Peromyscus floridanus ) , the possibilities of
an animal’s escaping would be greatly increased.

15. Basal length (modified). — From the anterior inferior bor-
der of the foramen magnum to the anterior parts of the premax-
illary bones, not necessarily in the midline of the skull (Fig. 25).

16. Incisor-molar length. — Length of line connecting poste-
rior margins of alveoli of upper incisors with posterior margin
of molariform tooth row occlusal surface. Such a measure gives
an idea of overall length of the “masticating area’’ of the mouth
(Fig. 25).

17. BuUar length. — Length of straight line connecting anterior
point of insertion of bulla into basilar region of skull with pos-
terior point of bulla evident when the skull’s basilar region is
facing upward (Fig. 25).

18. BuUar width. — Straight-line distance approximately per-
pendicular to bullar length line connecting widest points of up-
ward-facing bulla (Fig. 25).

19. Width across molariform tooth rows. — Length of straight
line connecting right and left labial margins of tooth rows at their
midpoints. This character gives an indication of mouth width
much as character number 16 measured mouth length (Fig. 25).

20. Zygomatic breadth. — Greatest distance across zygomatic
arches, at whatever point along the arch at which distance was
maximal. The line is perpendicular to the long axis of the skull.
This measure gives some idea of width of skull, greater width
often being associated with animals which are heavy and pow-
erful (Fig. 25).

21. Incisor width. — Width measured across both incisors just
above point where tapering begins, or, on those that do not
taper, at tip of incisors. Hershkovitz ( 1962) noted the tendency
of triturating incisors to be thick and powerful (Fig. 25).

22. Incisor length. — Straight-line distance connecting distal
portion of incisor with its point of exit from the premaxillary
bone. Species having seizer-digger incisors often have long, slen-
der incisors (Fig. 25).

23. Incisor-molar lengthihasal length x 100. — Relative mouth
length, size removed as a confounding factor.

24. Bullar index = hidlar length x bullar widthihasal length. —
Index of bullar inflation.

25. Length of molar tooth row (TRL). — Length of occlusal
surface of molariform teeth, from anteriormost to posteriormost
points (Fig. 25).

26. TRLIincIsor-molar length x 100. — Relative proportion of
mouth composed of grinding teeth. I expect that a species such
as the vole Microtus. for example, which consumes such sili-
cacious materials as grass, might need larger molariform tooth
rows to allow for a larger grinding area over which to crush this
material. Seeds, which are eaten by many desert specialists,
would not necessitate a large grinding surface.

27. Width across tooth rows/basal length x 100. — Relative
width of mouth.

28. Zygomatic breadth/basal length x 100. — Relative width
of zygomata, an indication of relative breadth of skull.

29. Incisor widthihasal length x 100. — Relative incisor width.

30. Incisor length/basal length x 100. — Relative incisor length.

31. Incisor angle. — Coded from very proodont (0) to very op-
isthodont (5). Hershkovitz (1962) discussed many rodent incisor

Fig. 25. — Cranial measurements used in the various multivariate
analyses.

types. 1 would surmise that incisor angle can reflect diet some-
what. Grass-clipping Sigmodon and Microtus possess orthodont
incisors, whereas granivorous heteromyids have incisors which
are markedly opisthodont. Inflexion of the incisors may increase
biting force on the tip such that greater efficiency at husking
seeds results.

32. Seizer-digger incisors. — Hershkovitz (1962) defined this
type of incisor as being generally slender and proodont, and
functioning as a tool for digging up worms, insects, or roots.
Some Thomomys possess such incisors and use them in digging
burrows and roots. Coding: not a seizer-digger incisor ( I ) to very
much so (3).

33. Triturator incisors. — Coded from not a triturator (I) to a
very pronounced triturator function (3). Hershkovitz (1962) not-
ed these incisors are used in gnawing and chopping and as hoes
in digging. They are generally heavy and well-pigmented.

34. Molar planation. — Coded from crested molars (1) to pla-
nar molars (4). Hershkovitz (1962) remarked on the evolutionary
stages involved in the change from primitive crested molars to
specialized planar types. He also suggested such planation is
correlated with increasing hypsodonty. Planar molars have a
grinding or crushing function rather than the tearing and cutting
inherent in the occlusion of the ridges, cusps and valleys of
crested molars.

35. Molar complexity . — Coded from complex pentalophodont
molars (0) to simple cylindrical teeth (6). Grassland and desert
species have tended to evolve simple, cylindrical molariform
molars. It would be expected that a species specializing on

50

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 16

tough, fibrous material would need not only planar teeth, but
quite a complex system of enamel ridges in order to expose the
hardest tooth material to the grinding of the vegetation. Com-
plexity might even be increased by the formation of many enamel
triangles (see triangulation, below) which would make the grind-
ing of vegetation even more efficient. Granivorous desert
species, on the other hand, could minimize enamel surface and
simplify molariform teeth such that a basin is formed in which
soft, husked seeds could be crushed.

36. Tubercular hypsodonty. — Coded from none (0) to pro-
nounced (4). As Hershkovitz (1962:89) defined tubercular hyp-
sodonty, it is the ", . . elongation of the coronal tubercle, or
tubercles, at the expense of the remainder of the tooth, including
the root. This type of hypsodonty is an adaptation for seizing,
grasping, cutting, chopping or cracking.” When this occurs
among molariform teeth, it is often an indication of an insectiv-
orous diet ( vespertilionid bats, for example). Presence of tuber-

cular hypsodonty in Onychomys is associated with its insectiv-
orous diet.

37. Coronal hypsodonty . — Coded from none (0) to pro-
nounced (3). The grazing habit is characterized by these grinding
and crushing teeth, whereas species such as generalized forest
dwellers (for example, Peromyscus) have not developed this
molariform type.

38. Molar triangulation. — Coded from none (0) to pronounced
(3), See Molar Complexity above.

39. Molar tooth row width. — Width of one molar tooth row
(occlusal surface) at midpoint of the row.

40. Relative inolariforin surface area. — (Two times the molar
tooth row width x molar tooth row length)/basal length. If in-
deed a grazer needs more surface area than a seed eater because
of the tough dietary regimen, then the index should reflect food
habits to some extent.

41. Vihrissae density. — Coded from low density (1) to very
dense (4). See Vibrissae/head-body ratio above.

APPENDIX 2

Multivariate Analyses and Ratio Traits

Recently a number of questions have been raised regarding
the use of ratios in certain multivariate tests (Atchley et al.,
1976; Atchley and Anderson, 1978; Atchley, 1978; and others).
Opposing points of view have been rendered by Corruccini
(1977). Dodson (1978). Albrecht (1978), and Hills (1978). The
comments of Atchley and his colleagues regarding the use of
ratios in such techniques as principal components analysis, ca-
nonical variates analysis, canonical correlation analysis, and so
on. are based on the underlying assumption that the data used
in such analyses are multivariate normally distributed (m.n.d.).
Ratios are usually employed in an attempt to scale the data for
some common allometric relationship such as body size effects,
which may be an overriding element in the different data sets
(for example, Goodman and Paterniani, 1969; Goodman, 1972;
Findley. 1972; Nevo, 1973; Karr and James. 1975; Mares. 1976;
and many others). Part of the problem is that some of the math-
ematical steps leading up to the multivariate analysis are based
on the assumption of a normal distribution, particularly such
analyses as the formation of a similarity matrix based on product
moment correlations (Pearson’s r), or when data are standard-
ized prior to being treated for conversion to such values as ca-
nonical loadings, etc. (Clark, 1975). Daultrey (1976:41) states,
■"Principal components does not require the data to be normally
distributed; the use of Pearson's r does." He suggests that other
correlation coefficients (for example, Spearman’s rank correla-
tion coefficient) be used where the m.n.d. of data is questionable.
Cooley and Lohnes ( 1971:38) note that although a m.n.d. of data
is necessary for many significance tests, and that the marginal
distributions of the various data sets can be checked, normal-
izing the non-normal data sets may help, but caution that '"nor-
mal marginals do not themselves guarantee an m.n.d., and we
do not know of any useful test for multivariate normality.” Var-
ious involved tests of normamy have been outlined (Gnanade-
sikan. 1977), but I am not familiar with any paper dealing with
a multivariate analysis of complex data sets that has first tested
for complete multivariate normality of all of the data.

The arguments of Atchley and his coworkers are compelling,
particularly if the level of significance of the various multivariate
tests is important. Ito (1969) notes that violation of the multi-
variate normality requirement may be compensated for by a
large sample size as far as testing hypotheses about mean vec-
tors, but is not compensated for when data are employed in a
variance-covariance matrix. Tatsuoka (1971) notes that the
m.n.d. is a requirement for the strict validity of significance tests,
but Blackith and Reyment (1971) consider significance testing of
little value in biological data, and, indeed, suggest that most
multivariate techniques are sufficiently robust to allow their ba-
sic assumptions to be violated to an extent (see also Crovello.
1970; Klecka, 1975; Robinson and Hoffmann, 1975). As far as
transforming non-normal data to a normal distribution, Clifford
and Stephenson (1975) suggest that the statistical transforma-
tions required to conform to strict normality may result in the
loss of the "'ecological sense” of the data.

Rohlf and Sokal (1965) and Sneath and Sokal (1973:147, 153)
suggest the use of ratios to scale data, even though the frequent
departure of ratio data from normality has been known for many
years (for example, Pearson, 1897). Presumably they feel that
violations of the m.n.d. assumption in multivariate analyses can
be tolerated to a degree. Schnell (197(i/, 19706) used ratios in
principal components analysis and found that analyses based on
ratios reflected earlier classical taxonomic assumptions about
the particular taxon he was studying (the suborder Lari). He
found (I97(T;:48) that "As before” (when non-ratio traits were
used) "the gulls, terns, and skimmers are separated by a fairly
distinct gap. However, dividing by the Sternum Length had the
additional effect of separating the skuas from the gulls.” Further
(Schnell, 19706:294), "When correlated characters are used”

. . . one should transform the "'. . . character space (such as by
the use of ratios) before clustering to reduce the effect of a gen-
eral trend in characters, such as a general size factor . . .” which
". . . would make the resulting phenogram a possible candidate
for a general phenetic classification. ” Although Atchley has

1980

MARES— DESERT RODENT ECOLOGY

51

shown that in some cases ratios are actually nu>re correlated
with the factor supposedly being removed via the use of the
ratio, other workers (Corruccini, 1977; Lemen and Findley,
manuscript) have not found this to be the case. Undoubtedly
more work remains to be done in this area (see also Oxnard,
1978).

Multivariate analyses apparently have been successfully per-
formed on data which are qualitative in nature and which ob-
viously violate the assumption of adherence to the m.n.d. ( Miller
and Butler, 1966), but as Bennett and Bowers (1976:118) point
out "If the purpose of a particular factor analysis on qualitative
data is simply to identify clusters of similar variables, then anal-
ysis of such matrices may be satisfactory.”

There is little doubt that Atchley and his coworkers are correct
in their strict interpretation of having the data conform to all of
the assumptions of the various multivariate procedures (some-
thing that is seldom, if ever, done), and this is certainly neces-
sary if an investigator is interested in attaching a precise level
of significance to his results, but it does not seem to be true for
a general understanding of the interrelationships of the various
factors that are included in a multidimensional analysis of the

data. As Cooley and Lohnes (1971:38) point out, " fhe hazards
of overfitting in multivariate analysis are great. Although signif-
icance tests, when appropriate, can help to protect against re-
porting results that can never be replicated, we tend to treat our
multivariate models as primarily heuristic rather than inferential
procedures.”

In earlier papers (Mares, 1975/r, 1976), I have attempted to
use multivariate analyses of numerous morphological traits (in-
cluding some ratios) as a heuristic tool to attain a preliminary
assessment of convergent characteristics among disparate desert
rodent taxa. I continue that line of reasoning in this paper bearing
in mind that the use of ratio characters may limit the degree of
precision of the data in certain analyses. In the past, results of
such analyses had fit quite well with my interpretations of the
ecological relationships of the rodents which were obtained from
various other research techniques (for example, comparative
physiological investigations, natural historical studies, food hab-
its analyses, etc.). As is evident in this report, the use of such
techniques has also led to counterintuitive interpretations that
appear to have merit.

Copies of the following Bulletins of Carnegie Museum of Natural History may be obtained at the prices
listed from the Publications Secretary, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pitts-
burgh, Pennsylvania 15213.

1. Krishtalka, L. 1976. Early Tertiary Adapisoricidae and Erinaceidae (Mammalia, Insectivora) of North

America. 40 pp., 13 figs $2.50

2. Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the
late Pleistocene paleoecology of the central Appalachian Mountains of Virginia. 88 pp., 21 figs. $12.00

3. Wetzel, R. M. 1977. The Chacoan peccary, Catagonus nagneri (Rusconi). 36 pp., 10 figs. .. $6.00

4. Coombs, M. C. 1978. Reevaluation of early Miocene North American Moropus (Perissodactyla,

Chalicotheriidae, Schizotheriinae). 62 pp., 28 figs $5.00

5. Clench, M. H., and R. C. Leberman. 1978. Weights of 151 species of Pennsylvania birds analyzed

by month, age, and sex. 87 pp $5.00

6. Schlitter, D. A. (ed.). 1978. Ecology and taxonomy of African small mammals. 214 pp., 48 figs.

$15.00

7. Raikow, R. J. 1978. Appendicular myology and relationships of the New World nine-primaried oscines

( AvesiPasseriformes). 43 pp., 10 figs $3.50

8. Berman, D. S, and J. S. McIntosh. 1978. Skull and relationships of the Upper Jurassic sauropod

Apatosaurus (Reptilia, Saurischia). 35 pp., 11 figs $3.00

9. Setoguchi, T. 1978. Paleontology and geology of the Badwater Creek area, central Wyoming. Part

16. The Cedar Ridge local fauna (Late Oligocene). 61 pp., 30 figs $4.50

10. Williams, D. E. 1978. Systematics and ecogeographic variation of the Apache pocket mouse (Roden-

tia: Heteromyidae). 57 pp., 23 figs $4.00

11. Guilday, J. E., H. W. Hamilton, E. Anderson, and P. W. Parmalee. 1978. The Baker Bluff Cave

deposit, Tennessee, and the Late Pleistocene faunal gradient. 67 pp., 16 figs $5.00

12. Swanepoel, P., and H. H. Genoways. 1978. Revision of the Antillean bats of the genus Brachyphylla

(Mammalia: Phyllostomatidae). 53 pp., 17 figs $4.00

13. Schwartz, J. H., and H. B. Rollins, eds. 1979. Models and methodologies in evolutionary theory.

105 pp., 36 figs $6.00

14. Engesser, B. 1979. Relationships of some insectivores and rodents from the Miocene of North Amer-
ica and Europe. 68 pp., 12 figs., 20 plates $5.00

15. Bentz, G. D. 1979. The appendicular myology and phylogenetic relationships of the Ploceidae

and Estrildidae (Aves: Passeriformes). 25 pp., 5 figs $2.00

'^\THS07vZ?^

ju'j ni9i

J-ZBRARIES,

THE COMPARATIVE SOCIAL BEHAVIOR OF
KERODON RUPESTRIS AND GALEA SPIXII AND THE
EVOLUTION OF BEHAVIOR IN THE CAVIIDAE

THOMAS E. LACKER, JR.

NUMBER 17

PITTSBURGH, 1981

f

5a,

|j

ak .J

• ‘ '. }•,

BULLETIN

of CARNEGIE MUSEUM OE NATURAL HISTORY

THE COMPARATIVE SOCIAL BEHAVIOR OF
KERODON RUPESTRIS AND GALEA SPIXII AND THE
EVOLUTION OF BEHAVIOR IN THE CAVIIDAE

THOMAS E. LACHER, JR.

Department of Biological Sciences, University of Pittsburgh,
Pittsburgh, Pennsylvania 15260 and
Pymatuning Laboratory of Ecology ,

Linesville, Pennsylvania 16424

NUMBER 17

PITTSBURGH, 1981

BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY
Number 17, pages 1-71, 40 figures, 24 tables

Issued 9 June 1981

Price: $6.00 a copy

Craig C. Black, Director

Editorial Staff: Hugh H. Gennoways, Editor: Duane A. Schlitter,
Associated Editor; Stephen L. Williams, Associate Editor;
Nancy J. Parkinson, Technical Assistant.

© 1981 by the Trustees of Carnegie Institute, all rights reserved.

CARNEGIE MUSEUM OF NATURAL HISTORY, 4400 FORBES AVENUE
PITTSBURGH, PENNSYLVANIA 15213

CONTENTS

Abstract 5

Introduction 5

General Habitat Description 8

Materials and Methods 9

Habitat Analysis 9

Comparative Behavior and Ecology 9

Colony Studies 9

Data Analysis 12

Field Studies 14

Results 15

Habitat Analysis 15

Maintenance Behavior 16

Ingestive Behavior 16

Locomotion 17

Resting Postures 18

Grooming 19

Marking 20

Comfort Movements 22

Digging 22

Elimination 22

Attend 22

Social Behavior 23

Contactual Patterns 23

Agonistic Patterns 28

Sexual Patterns 31

Vocalizations 33

Galea 33

Kerodon 33

Social Organization and Reproduction 34

Reproductive Behavior 34

Reproduction and Growth 35

Behavioral Development 37

Agonistic Behavior 38

Use of Space 46

Time Budgets and Analyses 47

Discussion 55

Evolution of Behavior in the Caviidae 55

Trends in Behavioral Evolution 56

Mating Systems 60

Morphological and Behavioral Adaptations to Microhabitat 63

Acknowledgments 66

Literature Cited 66

Appendix I — Vegetation of Fazenda Batente 69

Appendix II — Behavioral Protocols 70

,1

'i

f

ABSTRACT

1 conducted an 18 month study on the behavior and ecology
of two species of sympatric caviid rodents (Kerodon rupestris
and Galea spixii) in northeastern Brazil. Preliminary observa-
tions indicated that Kerodon was a habitat specialist, occurring
only in large boulder piles, whereas Galea appeared to be a
habitat generalist, occurring in a variety of open habitats ex-
cepting the boulder piles inhabited by Kerodon. This situation
presented an ideal field experiment to compare the social struc-
tures in these two closely related genera.

I first established breeding colonies of both in order to describe
their behavioral displays and to discern their function. Complete
behavioral repertoires, including vocalizations, are presented for
both Kerodon and Galea.

Reproduction and growth, behavioral development, sexual
behavior, agonistic behavior, and use of space were all examined
both quantitatively and qualitatively in the colonies and in the

field. Time budgets were calculated and analyzed for both gen-
era. Differences in rates of growth and behavioral development
between the two genera are probably related to ecological as-
pects of their significantly different microhabitat preferences.
Data on sexual and agonistic behavior collected in the colonies
suggested that Kerodon exhibited resource defense polygyny,
whereas the Galea mating system approximated male domi-
nance polygyny. Field data supported the colony observations.
These differences in mating systems may be related to the dif-
ferent habitat preferences observed. Kerodon is compared to
other resource defense polygynists.

Finally, a model for the evolution of behavior in the family
Caviidae is presented. The social organizations of the various
genera seem to be very responsive to ecological requirements.
The importance of social organization in ecological adaptation
is discussed.

INTRODUCTION

Morphology, physiology, and behavior all play a
role in the adaptation of an organism to its environ-
ment. Morphological and physiological adaptations,
as well as behavioral displays, seem to have
evolved in response to average long term environ-
mental influences. Skeletal morphology, for exam-
ple, seldom shows seasonal variations; basic be-
havioral gestures are generally quite consistent
within a species from one geographical locality to
another (Eisenberg, 1967). Social behavior, how-
ever, is far more labile in its response to the envi-
ronment (Stacey and Bock, 1978), and in conjunc-
tion with mating strategies, may well represent the
means by which organisms keep their reproductive
success high in the face of short term environmental
fluctuations. One of the most effective techniques
for assessing the adaptive value of social organi-
zation is the analysis of behavior of closely related
species occupying markedly different habitats.

Eisenberg’ s (1963, 1967) work on the behavior of
heteromyid, murid, and dipodid rodents is an ex-
cellent example of the use of a comparative study
to explain behavioral repertoires from an adaptive,
evolutionary perspective. Two major concepts that
emerged from this work which should be consid-
ered in any comparative behavioral study are: 1)
“discrete behavior patterns exhibit a profound sim-
ilarity”; and 2) “differences in the frequency of oc-
currence rather than in the form of the movement
have proved to be the most effective criterion for
delineating taxon-specific differences.” Subsequent
comparative studies on mammals have reinforced

these concepts both for behavioral repertoires
(Rood, 1972; Kleiman, 1974; Wilson and Kleiman,
1974) and for vocal repertoires (Eisenberg, 1974).

Behavioral research on mammals has most re-
cently been concentrated in two groups — the higher
primates in general, and the South American ca-
viomorph rodents. The interest in the caviomorph
rodents is due in part to their substantial morpho-
logical variability (Walker, 1974) and the wide va-
riety of habitats that they occupy (Osgood, 1943;
Moojen, 1952; Cabrera and Yepes, I960). The ca-
viomorphs diversified greatly from the Oligocene
through the Pleistocene when other rodent compet-
itors were absent (Simpson, 1945; Landry, 1957;
Mares, 1975). Eleven families and forty-six genera
are currently found in South America (Simpson,
1945), and their habitats range from fossorial (Cten-
omys and Clyomys) to semi-aquatic ( Hydrochoerus
and Myocastor) to arboreal (Coendou, Kerodon
and various echimyids). Caviomorphs are found in
all major South American habitat types including
Tropical Eorest, Savannah, Cerrado, Caatinga,
Temperate Forest, Chaco, Llanos, Desert and Puna
biomes. This morphological variability, coupled
with the diversity of habitats, has led to the evo-
lution of many interesting ecological and behavioral
adaptations (Eisenberg, 1974; Kleiman, 1974). The
caviomorphs, for these reasons, offer excellent op-
portunities for behavioral studies of a comparative
nature.

The family Caviidae contains six genera and ap-
proximately twelve species and has undergone an

5

6

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

interesting radiation. There are two distinct subfam-
ilies— the Dolichotinae, which includes the genera
Doliciwtis and Pediolagus\ and the Caviinae,
which contains four genera, Kerodon, Galea, Ca-
via, and Microcavia.

The dolichotines are hare-like in their external
morphology, and inhabit open semiarid grasslands,
xeric thornscrub, and temperate steppe in Argen-
tina and southwestern Paraguay. The basic social
unit of the dolichotines is the pair, and after par-
turition a temporary family unit is formed that per-
sists from 6 to 8 months. In addition to possessing
most of the typical caviomorph behavioral reper-
toires, the dolichotines also show many behavioral
adaptations similar to those of the large African her-
bivores (Cabrera, 1953; Dubost and Genest, 1974;
Kleiman, 1974).

In contrast to the dolichotines, the subfamily Ca-
viinae contains the more familiar “Guinea-pig” like
rodents. Microcavia, Cavia, and Galea are all quite
similar in morphology, resembling the common lab-
oratory Cavia porcellus in body form. The genera
differ primarily in size and in characteristics of the
pelage. All three genera have four clawed digits on
the manus and three clawed digits on the pes, and
have a plantigrade foot posture. Galea is easily dis-
tinguishable from the other two genera by the pres-
ence of yellow incisors. These three genera occupy,
in general, more open formations, ranging from
Larrea flats (Microcavia) to open pampas (Cavia),
or thornscrub forest and high Andean grasslands
(Galea). All possible pairs of these three genera
occur sympatrically in certain parts of their distri-
butions. In certain areas of northeastern Argentina,
all three genera may be sympatric (Cabrera, 1953;
Contreras, 1965; Rood, 1972; Vaughan, 1972).

Kerodon has the same basic caviine body form
as the above three genera, but possesses some mor-
phological and cranial characteristics unique to the
subfamily. The manus and pes are padded with a
leather-like epidermis (very similar to the feet of
hyraxes) and claws are absent. The feet have sub-
cutaneous nails on all digits but the innermost digit
of the pes, where the nail has been modified as a
small grooming claw used to comb the pelage.

Kerodon is restricted in distribution to terrestrial
islands of granitic boulders which occur throughout
the Caatinga, a semi-arid region in Brazil. Animals
use fissures and hollows in the rocks for shelter and
to escape from predators, and emerge throughout
the day and night to forage in the adjacent trees and
shrubs. They are extremely agile climbers, which

is unusual considering that they have neither claws
nor a tail, two adaptations normally associated with
arboreality. The skull, especially the rostrum, is
longer and narrower than in other caviines, and the
distance between the incisors and the premolars is
proportionately greater (Moojen, 1952; Walker,
1974).

The phylogenetic relationships of the caviids are
still unclear (Patterson and Pascual, 1972; Spotor-
no, 1979). The family is first known from the mid-
Miocene, and by late Miocene the two subfamilies
had separated (Landry, 1957). The current genera
had probably all evolved by the mid-Pleistocene
(Rood, 1972). Pascual (1962) diides the subfamily
Caviinae into four groups based on molar charac-
teristics— an extinct Allocavia group; a Cavia
group, which includes Cavia and the extinct genus
Paleocavia', a Microavia group; and a Galea
group, which also includes the genus Kerodon. Of
the current genera of Caviinae, only Galea and
Kerodon are classified in the same group. Kerodon
is seemingly more closely related to Galea than to
any other caviid.

The caviids have been fairly well studied behav-
iorally. Both the dolichotines (Smythe, 1970; Du-
bost and Genest, 1974; Wilson and Kleiman, 1974)
and the caviines (King, 1956; Kunkel and Kunkel,
1964; Rood, 1970, 1972) have been observed under
at least seminatural conditions, and complete be-
havioral repertoires are present for all caviid genera
except Kerodon. An examination of the behavior
and ecology of Kerodon is therefore essential in
any attempt at understanding the evolution of social
behavior in the Caviidae.

In this study the behavior of Kerodon is com-
pared with the behavior of Galea spixii spixii (Wag-
ler), a species sympatric with Kerodon in certain
areas of northeastern Brazil. Kerodon is restricted
to the semiarid Northeast, however, whereas Galea
has a far more extensive distribution, occurring in
Caatinga, the semiarid habitat of the Northeast; in
the more mesic Cerrado savannahs; in Agreste,
transition between Caatinga and Atlantic forest;
and in the more open sections of Atlantic rain for-
est. In comparison with Galea, therefore, Kerodon
is a habitat specialist, restricted entirely to a poten-
tially physiologically taxing environment. Galea is
clearly a habitat generalist, occurring in a variety
of open formations. This study focuses on the be-
havioral repertoires, foraging behavior, patterns of
aggression, mating systems, and social organization
of both species. I then attempt to relate differences

1981

LACHER— CAVIID SOCIAL BEHAVIOR

7

A/

Fig. 1. — Major vegetational formations in the eight states that comprise northeastern Brazil. Modified from Carvalho ( 1973).

in social behavior to differences in ecology, specif-
ically in relation to the different habitat require-
ments of the two species.

Many important parameters that we use to define
an organism’s niche are primarily behaviorally de-
termined (for example, feeding preferences and for-

aging behavior, choice of burrows or nesting sites,
territoriality, reproduction and courtship, activity
patterns, and microhabitat selection). Simpson
(1958) stated that behavior is “the actual means of
interaction between physical organization and the
environment, hence the direct and visible expres-

8

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

48® 42° 36°

Fig. 2. — Annual rainfall in northeastern Brazil. Values are reported in mm. The limits of the Caatinga (dotted line) approximately follow
the 1,000 mm isohyet. Modified from Souza Reis (1976).

sion of the relationship that is adaptation.” This
study offers the opportunity to evaluate the adap-
tive importance of social behavior in two closely-
related organisms with differing ecological require-
ments.

Data collected on the ecology and social behavior
of Kerodon are, in addition, compared to certain
aspects of the behavior of the other caviids. Einally
I attempt to clarify the trend of evolution of social
behavior for this family.

GENERAL HABITAT DESCRIPTION

The Caatinga (Eig. 1) covers an area of 650,000
km- between 3° and 16° South Latitude and 45° and

35° West Longitude in the northeastern corner of
Brazil (Erota-Pessoa et al., 1971 ; Souza Reis, 1976).

1981

LACHER— CAVllD SOCIAL BEHAVIOR

9

The annual rainfall of the Caatinga varies from
slightly over 1 ,000 mm to less than 400 mm (Fig. 2).
The region is not dry enough to be classified as a
desert, but because of anomalies in yearly precipi-
tation, it is susceptible to serious droughts and
flooding (Eidt, 1968)

Physically, the Caatinga is dominated by three
major elements — 1) a basement of pre-Cambrian
crystalline rocks which, when exposed, form a sur-
face of gentle slopes; 2) many groups of hills, and
ranges of low mountains of granitic rock (serras)
which are the products of erosion and denudation;
and 3) a cover of sandstone strata which at one time
covered most of the crystalline basement, but is
now partly stripped off by post-Cretaceous erosion.
This sandstone layer, when eroded, forms the
mesa-like plateaus known as chapadas (James,
1942; Ab'Saber, 1970). More information on the cli-
mate and geology of the Caatinga can be found in
Vasconcelos Sobrinho ( 1971), Markham ( 1972), and
Carvalho ( 1973).

The zoogeographic relationships of the Caatinga
are poorly understood. Sick (1965) and Vanzolini
(1974, 1976) have indicated that the Caatinga, as
well as the intervening Cerrado, have very low rates
of endemism. Both areas appear to have been col-
onized by tropical elements from the Amazon Basin
(north and west) and the Atlantic rainforest (east),
and by scrub and semitropical elements from the
belt of Caatinga, Cerrado, and Chaco, which runs
southwesterly to Paraguay and Argentina. Guima-

raes (1972) reports that the mammalian faunas of
the Caatinga and Cerrado have many species in
common. Vanzolini ( 1976) has compared the lizard
faunas of the Cerrado and Caatinga, and reported
that species compositions were virtually identical.
The one endemic Caatinga lizard, PUitynotus semi-
taeniatus, is associated with peculiar granitic rock
formations that occur extensively throughout the
Caatinga, and are known as lajeiros (extensive ex-
posed tables of the crystalline basement, which
often have large piles of boulders scattered upon
their surface). Data on mammal distribution pat-
terns suggest that Kerodon rapes tris, which occurs
in the same habitat, is also a Caatinga endemic
(Cabrera, 1953; Walker, 1974).

My field site. Fazenda Batente, is a privately
owned farm located in the municipality of Exu, 6
km SE of the town of Exu, Pernambuco, Brazil.
Although Exu is located near the center of the Caa-
tinga, it receives more rainfall than areas 20 to 30
km to the south, being situated at the southern base
of the extensive plateau system of the Chapada do
Araripe, which is one of the major centers of oro-
graphic rainfall in the Caatinga (Markham, 1972).
The amount of rainfall received in the low areas
decreases as one moves south from Exu. The vege-
tation of the Exu region, therefore, contains some
Cerrado elements not normally found in Caatinga
associations, while it lacks various species common
in the more arid regions to the south.

MATERIALS AND METHODS

Habitat Analysis

The study site contains a series of large boulder piles and
lajeiros (rock faces) separated by patches of scrub forest, cacti,
and second growth vegetation (Figs. 3 and 4). During the peak
of the rainy season, samples of leaves, buds, and flowers of all
species of plants present on the area were collected. These in-
cluded not only trees and shrubs, but also cacti and all grasses.
These were taken to the Instituto de Pesquisas Agronomicas in
Recife, Pernambuco, for positive identification. The various mi-
crohabitats present on the study areas were then visually sepa-
rated as to their most common plant form and species. These
classifications were subsequently used to examine microhabitat
associations of the mammal fauna via mark and recapture data.

A maximum-minimum thermometer was set up in the middle
of the study area and a year-long record of weekly temperature
minima and maxima was made. In addition, the relative humidity
was recorded with a sling psychrometer on the morning of every
trapping day. General weather conditions were also noted daily.

Comparative Behavior and Ecology
Colony Studies

Breeding colonies were set up for both Kerodon riipestris and
Galea spixii spi.xii. Animals for both colonies were captured in
the wild at a variety of localities in northwestern Pernambuco.

The Kerodon colony was established 4 February 1977 with
the introduction of two adult males, three adult females, and one
subadult female. The first litter was born in late July 1977. The
colony increased in size to a maximum of 17 in March 1978, and
contained 14 animals on the date of termination, 26 March 1978.
The colony was maintained in a large (25 m by 10 m) windowed
laboratory room of an abandoned agricultural school in Exu.
Prior to the introduction of the animals, the room was divided
into four subareas, each of which was modified to mimic a mi-
crohabitat available to Kerodon on the study site. Two areas
(15 m by 5 m each) were altered to resemble a boulder strewn
rock-face and a dry, shallow-soil scrub forest. The other two

10

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

CORN FIELDS

Fig. 3. — Map of the Fazenda Batente study area, indicating major rock piles. Study area subdivisions are indicated by roman numerals.
Arabic numbers indicate trap sites and microhabitat type. The area was bordered on the east by cornfields, the west by a fenced corral,
the north by a small stream and the south by a dirt road. Microhabitat code: 1 = Croton thickets; 2 = level shrub and grass areas; 3 =
boulder, cacti areas; 4 = cultivated area; 5 = Cnidoscolus flat; 6 = Jacobinia, Rnellia flat. Distance from fence to beginning of corn-
fields is about 80 m.

areas (10 m by 5 m each) were modified to resemble a grassy
field and a brush tangle, two microhabitats that are fringe habi-
tats for Kerodon, but preferred habitats for Galen. These areas
were included in the Kerodon colony room to observe the degree
of utilization of these habitats by Kerodon during normal daily
activity.

Throughout the course of the study, Kerodon were presented
a diet of green vegetation and pineapple, occasionally supple-
mented with dried corn and Brazil nuts; water was provided ad
libitum. All of the plants selected were common in Kerodon
habitat, and included Cassia excelsa (Leguminosae), Ziziphns
Joazeiro (Rhamnaceae), Solanuni paniculatus (Solonaceae),
Croton campestris (Euphorbiaceae), and Brachiaris nnitica
(Graminae).

The Galea colony was established on 3 September 1978 with
the introduction of four adult females, two subadult females,
three adult males, and two subadult males. The two subadult
males were the progeny of one of the adult females. The first
birth in the colony occurred in late November, and reproduction
continued until the end of the study on 26 March 1978. Colony
size peaked twice at 15, in early February and in late March.
The Galea were also maintained in a windowed laboratory room
(10 m by 8 m) of the agricultural school. The room was altered
primarily to mimic Galea habitat; however, rocks and trees were
added to determine if they were utilized in the absence of Ker-
odon. The animals were provided with water, pineapple, and
green vegetation, primarily Brachiaris, and, occasionally, Cro-
ton or Cassia (Brachiaris was highly preferred).

1981

LACHER— CAVIID SOCIAL BEHAVIOR

Fig. 4. — View of the major rock pile of subdivision II and associated vegetation. Kerodon dwell in the cracks between the boulders.

General observations were taken in the same manner on both
colonies. The animals were all individually marked with a com-
mercial fur dye, Jamar D. Observations were made from outside
the colony through an open shutter. Because both species are
active throughout the day, with a slight depression in the level
of activity during the mid-day hours, observations were concen-
trated during the morning and evening hours, and, in the case
of Gcdeu, at night. All behaviors observed were recorded, many
postures being sketched upon observation. Additional sketches
were made from numerous still and motion picture films. Both
8-mm and 16-mm movies were taken during general colony ac-
tivity and in encounter situations. These general observations
were used to establish the basic behavior postures and displays,
and to clarify the overall patterns of social behavior of each
species.

Throughout the period that the colonies were maintained, re-
productive and growth data were collected on all animals. Ani-
mals were captured, weighed, checked for reproductive condi-
tion, and redyed once each month. When females neared the
end of their gestation, they were captured and palpated daily.
On the day of birth, the female was captured, weighed, and
checked for signs of copulation. Numerous copulations occurred
on the day of birth, and in situations where copulations were not
directly observed, the females were found to have either copu-
latory plugs or semen present in the vagina. As both species
exhibit a postpartum estrous, gestation periods were calculated
from the day of birth. The newborn juveniles were also captured.

sexed, measured, and weighed at five-day intervals until 30 days
of age, and then at iO-day intervals until 120 days of age. Beyond
this, animals were handled only during the monthly weighings.
The growth data on juveniles were then plotted and a regression
line fitted to the data in order to calculate a growth equation
(Lacher, 1979).

An additional aspect of the comparative behavioral study was
an examination of the interspecific interactions, which occurred
in a mixed colony of the two genera. Three separate mixed col-
onies were observed for a period of 4 to 6 days each — one in
which both Kerodon and Galea were introduced simultaneous-
ly; a second where Galea were introduced into the existing Ker-
odon colony; and a third where Kerodon were released into an
existing Galea colony. Animals were again dyed for individual
identification. Density of animals per m^ of colony floor was kept
constant for all three situations so as not to alter aggressive
levels artificially. Observations of one hour's duration were con-
ducted from one to three times daily, during which all interspe-
cific and intraspecific aggressive and sexual interactions were
recorded. The number of aggressive encounters was then stan-
dardized to a per hour rate, so that the day to day trends in inter-
and intraspecific interactions could be interpreted.

Patterns of aggression were also examined in the individual
colonies. For all aggressive encounters observed, the aggressive
and submissive individuals were identified, and the type of ag-
gressive display, as well as the corresponding subordinate re-
sponse, were noted. These data were then used to compare the

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

12

overall patterns of aggression within each species in relation to
its social organization. The data used in this comparison were
collected simultaneously for both colonies between September
1977 and April 1978.

Time budget data were collected on six separate groups — adult
male Kerodon ; adult female Kerodon ; juvenile male Kerodon ;
juvenile female Kerodon ; adult male Galea ; and adult female
Galea. Time budgets were obtained in the same manner for both
Kerodon and Galea excepting the time of observation. Kerodon
time budgets were obtained in the morning, between 7 and 10
AM, whereas Galea time budgets were calculated at dusk and in
the early evening hours. This was done to minimize disturbances
on the Galea colony which was located near a footpath used
during the day by farmers. I observed Kerodon and Galea in
both the field and in the colony for over a year before beginning
the time budget observations. In the wild, Kerodon and Galea
were active throughout the day, with peaks in activity during the
crepuscular hours. Both genera were active 24 h a day in the
colony, with a depression in the level of activity during the mid-
afternoon. Although these observations were not quantified to
the degree the time budgets were, I observed no qualitative dif-
ferences in the kind and frequency of activities exhibited during
either the early morning or the late afternoon activity peaks.
Animals were fed once daily, and the hour of feeding was set so
that time budgets were calculated for both species approximately
12 h after the addition of food. This was done to eliminate pos-
sible differences in the allocation of time that might have oc-
curred if the animals were at different levels of satiation.

For each time budget, a single animal was observed inten-
sively for 0.5 h. The determination of which animals would be
observed was random. Each animal was observed from between
one to four times, giving a total of 56 Kerodon trials and 24
Galea trials. All movements of the animal were described and
recorded on cassette tape. After the observations, the tape was
played back, and the time spent in each behavioral category was
marked by stopwatch and recorded in 3-min intervals, so that
each individual data sheet contains what is essentially "a com-
plete record” of the behavior for each animal observed, allowing
for the analysis of both frequencies and durations of all events,
as well as sequence analysis of series of events (Slater, 1978).

Data Analysis

Multivariate techniques have been used primarily for the ex-
amination of morphological (Blackith and Reyment, 1971) or
morphoecological (Karr and James, 1975; Mares, 1975, 1976,
1980; Cody, 1978) traits. Recently multivariate techniques also
have been applied to behavioral data (Svendsen and Armitage,
1973; Bekoff et al., 1975; Hazlett, 1977; Colgan, 1978; Davies,
1978; Ringo and Hodosh, 1978). Two techniques, which are of
particular value to the ethologist, are principal components anal-
ysis and discriminant analysis. The basic methodology of these
analyses and their application to behavioral data have been de-
scribed in detail by various authors (Cooley and Lohnes, 1971;
Aspey and Blankenship, 1977; Frey and Pimentel, 1978; Pimen-
tel and Frey, 1978; Lacher, 1980).

A few precautions about statistical inference with these anal-
yses need to be mentioned. Two basic assumptions of discrim-
inant analysis are 1) group variance-covariance matrices are
equal, and 2) all samples were drawn from a multivariate-nor-
mally distributed population. Although no program was available
to test equality of the variance-covariance matrices, the large
differences present among the within-group variances for certain

variables leads me to doubt that this assumption was met. As
for the assumption of multivariate normality, Cooley and Lohnes
(1971) stated that “we do not know of any useful test for multi-
variate normality.”

I will assume that the above two assumptions have not been
met and discriminant analyses (D.A.) will be treated as a de-
scriptive heuristic process only. The only assumption necessary
for the use of D.A. in this context is that all initial samples are
potential members of the predefined populations (Neff and
Smith, 1979). There is no difficulty in correctly classifying ani-
mals as Kerodon and Galea, or as males and females. Classifi-
cation of animals as juveniles was based on external reproduc-
tive characteristics. No individual sample could potentially
belong to a population other than those used in the construction
of the discriminant axes.

When used as a descriptive process, discriminant analysis in-
dicates the relative importance of the variables in separating the
populations and indicates the relative distances between the cen-
troids of each population. The influence of the inequality of vari-
ance-covariance matrices on classification of individuals is poor-
ly known. Studies have shown that classification by Geisser
classification probabilities performs well when the assumptions
are not met (Pimentel and Frey, 1978). Geisser probabilities are
more robust to departures from equality than are classification
functions, the D.A. method employed in this study (SPSS DIS-
CRIMINANT package; Nie et al., 1975). Discriminant analysis
is, in general, especially robust, and has been shown to give
close to optimal results even when variance-covariance matrices
are significantly different (Lachenbruch, 1975; Neff and Smith,
1979). Thus probabilities of classification will be assumed to re-
flect the true biological affinities between samples and popula-
tions.

Although discriminant analysis assists in the separation of
these groups, it tells one very little about the nature of the vari-
ation within groups. This is best accomplished by the calculation
of principal components for each of the homogeneous groups.
The computation of principal components requires no assump-
tions about the structure of the data, as long as no statistical
inferences are to be drawn (Neff and Smith, 1979).

The ability to use quantified aspects of an animal's ethogram
as a “system of measurement" in order to compare closely-re-
lated species is especially appealing to the ethologist. In this
study, all recorded aspects of the behavioral repertoire of each
species as well as certain acts which gave some indication of the
importance of olfactory signals (sandbathing, scent-marking)
were used in the quantitative analyses (Table 1). A quantitative
approach was chosen in order to facilitate the examination of
species-specific differences by the criteria proposed by Eisen-
berg ( 1967).

The raw data used in the multivariate analyses were taken
from the time budget sheets. Both number of occurrences for
each trait and total time spent in each behavioral act were re-
corded, but mean duration was chosen as the variable form to
be analyzed. Duration is widely used in behavioral research, as
it gives an estimate of the degree of variability or stereotypy of
behavioral acts. It also has advantages over both frequency of
behavioral acts and total time spent in the various behaviors
(Bekoff, 1977; Fagen and Young, 1978). Mean duration is the
only variable form which allows the observer to treat all behav-
iors as a unit phenomenon possessing a variable temporal com-
ponent.

Mean durations were calculated from the time budget data for

1981

LACKER— CAVIID SOCIAL BEHAVIOR

13

all displays observed. For each trial the total time an animal was
observed performing a given behavioral act was divided by the
total number of times the act was performed. Each 0.5-h trial
presents mean durations for all aspects of the behavioral rep-
ertoire executed during the observations. Displays not executed
during a given trial were, by definition, of zero duration. Each
trial was considered one sample unit. The total sample size for
each group on which time budget data were collected was there-
fore equal to the number of trials performed on animals of that
group.

Durations have been used as a behavioral phenotype in a va-
riety of comparative studies (see Bekoff, 1977, for a review),
particularly to compare mean durations and their coefficients of
variation. Although the same is being done for Kerodon and
Galea (Lacher and Mares, in preparation), the objective in these
analyses was to determine if a population could be defined in
terms of the mean durations of the various displays present in
the behavioral repertoire. In addition, I wished to determine how
consistently a given animal could be correctly classified into a
population, based on a single observation period for which mean
durations were calculated.

One inherent difficulty was related to the degree of common-
ness of certain acts. Not all acts were observed in a single ob-
servation period. This especially created difficulties when at-
tempting to compare populations.

For example, a given display may have exactly the same mean
duration in three different populations, however, may occur in
quite different frequencies in each. Referring to Eisenberg (1967)
the above discrete behavior patterns exhibited a profound sim-
ilarity in duration. A simple comparison of durations, however,
neglects a second important concept which is invaluable in any
attempt at classification: "differences in the frequency of oc-
currence, rather than in the form (or mean duration, in this case)
of the movement have proven to be the most effective criteria
for delineating taxon-specific differences" (Eisenberg, 1967).

I therefore weighted all measurements of mean duration for
a given population by the frequency of occurrence within that
group. Mean population-durations for all displays in the behav-
ioral repertoire (Table 1) were calculated by summing the mean
durations for all the trials, and dividing by the total number of
trials in that population. Mean durations of zero were thus in-
cluded in the calculations. Individual cases (each 0.5-h trial)
were then assigned a probability of classification within each
population, based upon the comparison of the suite of mean
durations with the corresponding mean population-durations of
each group.

The analysis of durations may reveal additional information
of biological relevance. In a discussion of behavioral act dura-
tions, Fagan and Young (1978) note that there are two possible
types of behavioral acts. One type of act has a termination point
which is independent of the time at which the act begins (that
is, interrupted by another). Such acts exhibit a negative expo-
nential probability distribution. A second type of act tends to
terminate with increasing probability as the performance time
increases. Their performance will begin when the benefit result-
ing from the act outweighs the cost of its enactment, and will
end when the cost exceeds the benefit. This kind of act would
represent what the ethologist commonly refers to as a behavioral
display (for example, the relatively uniform postures and move-
ments discussed by Eisenberg, 1967, in his report on compara-
tive rodent behavior). The coefficient of variation of the temporal
duration of these acts would consequently by smaller than the

Table 1 . — The various categories of behavior used in the time
budget analyses. There are four functional groupings of acts,
each of which contains a number of variables. A number of
relatively infrequently observed acts that were not recorded dur-
ing the time-budgets are not included.

Maintenance and locomotion

a) Inactive (out of sight in rocks)

b) Sitting, rocks

d) Sitting, ground

d) Sitting, tree

e) Sitting, sand

f) Running

g) Climbing

h) Walking

i) Sandbathing

j) Grooming

k) Suckling

l) Nursing

m) Foraging/drinking

Aggression

a) Lunging

b) Chasing

c) Fleeing

d) Grappling

e) Curved-body posture

f) Submit

Reproduction

a) Follow (Chin-rump)

b) Followed

c) Tail-up

d) Repulsed

e) Mounting

Contact-promoting behavior

a) Crawl-overs

b) Allogrooming

c) Scentmarking

Estimate of degree of activity
a) Activity-ratio

coefficient of variation for “interrupted" act durations. One type
of act would therefore exhibit a fairly limited temporal variability
(that is, stereotypy), whereas the other type would exhibit the
opposite property (that is, variability).

Using these criteria, it is possible to construct classes of be-
havior that can be separated by appropriate statistical analyses
(Table 2). In this framework. Class I behaviors are those which
are important in separating groups (for example, species or
sexes). Class II behaviors are those that are of little importance
in group separation, but which indicate acts that are important
in maintaining higher-order group integrity (for example, acts
shared at the generic level in an interspecific analysis). Class 111
behaviors are those that function as ‘ ‘interrupted-behaviors' '
(that is, behavioral acts which are not duration specific and oc-
cupy random length time periods that are disrupted at variable
intervals by other acts or displays). Class 111 behaviors may also
have an importance which is not related to duration, thus they
should be cautiously interpreted.

14

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

Table 2. — The classes of behavior, based on stereotypy and variability. Once groups are defined, the three behavioral classes can be
distinguished through discriminant analysis and the examination of coefficients of variation. See text for a more detailed description.

Class

Type of variability

Level

1

Low intragroup variability
High intergroup variability

Group-specific behaviors

11

Low intragroup variability

Behaviors common to

Low intergroup variability

higher group levels

111

High intragroup variability

“Filler-behaviors” or not

High intergroup variability

duration specific

Discriminant analysis proved to be of particular value in dis-
tinguishing behavioral acts by these criteria. Variables, which
were indicated as differing significantly between groups by dis-
criminant analysis, were subsequently analyzed by a variety of
univariate tests (Table 3). Once groups were defined, factors,
which were important in explaining significant percentages of
the intragroup variance, were constructed by a principal com-
ponents analysis.

Field Studies

Population and reproductive data were collected simulta-
neously for both Kerodon and Galea, which were sympatric at
the study site. Animals were live-trapped for a one-year period
beginning in October 1976. In addition, Kerodon were handcap-
tured and marked beginning in January 1978 until April of the
same year.

For live-trapping and observations of movements, the study
area was divided into six subdivisions. Each subdivision con-
tained a major rock pile, and was trapped for a minimum period
of 8 days. The traps were then rotated to another subdivision.
In place of a grid, traps were placed in an irregular pattern
throughout the site. Because of its physiognomy, the site was
difficult to accurately grid. Piles of large boulders (3 to 6 m high)
were separated by crevasses, vine tangles and thorny vegetation.
A second disadvantage of a grid was related to the daily move-
ments of the animals. Galea extensively utilize well-marked run-
ways, and a regular grid would have potentially missed many

runways. Kerodon are extremely habit specific, being associated
with the boulder piles, which were located throughout the study
area, and a grid would therefore have placed too few traps in
Kerodon habitat. I thus chose to place traps selectively on run-
ways and in rockpiles to increase trapping and marking success
of the animals. Intertrap distances were then measured to cal-
culate home range sizes.

Trapped animals were weighed, sexed, and checked for re-
productive status. All new animals were toe-clipped for future
identification. In addition, all Kerodon were marked with a com-
mercial fur dye, Jamar D, for easy individual recognition in the
field. Observations of these marked animals were used to sup-
plement the trapping data on home range and movements.

As Kerodon proved to be extremely difficult to trap, beginning
in January of 1978, the animals were captured with the assistance
of dogs and a number of young children. Kerodon were cornered
in the rocks by the dogs, and subsequently captured with nooses
by the children. These animals were weighed, sexed, toe-clipped
and dyed, and their subsequent movements followed visually.
These observations on movement were taken from a series of
points on the study area, an equal amount of time being spent
at each point.

General behavioral observations in the field were concentrated
at two particular localities on the site (III and VI in Fig. 3),
although observations were taken on various occasions through-
out the study site. Observations were made either from behind
a blind or in the shelter of some vegetation. Watch periods varied
both in length (30 min to 3 h) and in the time of day observations

Table 3. — A summary of all statistical procedures used, indicating the source, and the groups compared. Additional procedures used
for special situations are not included, and are referenced in the text when applied.

Test

Source

Groups compared

Mann-Whitney U-test

BMDP3S (Dixon, 1975)
Siegal (1956)

Kerodon sexes
Galea sexes

One-way F-test

SPSS ONEWAY (Nie et al., 1975)
Sokal and Rohlf (1969)

All groups
All adults

Kruskal-Wallis test

BMDP3S (Dixon, 1975)
Siegal (1956)

All groups
All adults

Non-parametric

multiple comparisons

Hollander and Wolfe (1973)

All groups
All adults

Discriminant analysis

SPSS DISCRIMINANT (Nie et al., 1975)
Pimental and Frey (1978)

All groups
All adults

Principal component analysis

BMDP4M (Dixon, 1975)
Frey and Pimental (1978)

All Galea Adults
Kerodon males
Kerodon females

1981

LACHER— CAVIID SOCIAL BEHAVIOR

15

occurred (ranging from dawn to dusk), with emphasis during
crepuscular hours. All field observations were compared to ob-
servations made in the more controlled colony situation in order
to evaluate possible differences that may have existed.

The total observation time, for both field and colony studies,
was approximately 1,600 h. In the following descriptions of the
behavioral repertoires, 1 use the terms “display” and “gesture”
as synonyms.

RESULTS

Habitat Analysis

The record of maximum temperatures shows a
cycle with high temperatures September through
December, followed by a cooling trend (Fig. 5). The
temperature remains low until mid-July, when it
again climbs to the September levels. The minimum
temperatures follow the same trend, but tend to lag
behind the maximum temperatures. The coolest
temperatures of the year coincided with the period
of the most regular rainfall, late April through Au-
gust. These months are generally considered the
rainy season in the Exu area, but during the year
that these data were collected there was no well
defined rainy season, and rain occurred periodically
throughout the year. Conversations with local
farmers and the Service de Peste in Exu indicated
that 1976-1977 was an especially wet year. Indeed,
with the exception of a few species, there was no
period of deciduousness that year; vegetation is
usually sparse during the dry season in the Caatin-
ga.

Although the study site was a relatively small
area, the high structural diversity present allowed
for a fairly high species richness of plants (Table 4);
29 families, 55 genera, and 70 species are repre-
sented. There are a number of different microhab-
itats present that tend to be dominated by one or a
few species. The relatively level, rock-free areas
were dominated by low canopy second growth
thickets of Croton jacobinensis, Cordia globosa,
and, in areas where the soil became somewhat
rocky, Croton argyrphylloides. More recently cut
level areas were dominated by various species of
Malvaceae (Bogenhardia tiubae, Sida panicidata),
Compositae {Centrathenim punctatimi, Blainvillea
rhomboidea), the vine-like Leguminosae {Phaseo-
Ins and Macroptilium), and, on the more open
areas, various grasses (Gramineae, Cyperaceae).

Cacti, especially the two species of Pilosocerens,
were most common in the boulder areas. The com-
mon trees were Cordia insigna and Cedrela sp.,
and occasionally Erythroxyhim and Rhamnidium.
A heavy vine layer of Cissus simsiana, Cissus si-
cyoides, Serjania caracasana, Serjania sp., and

Vismia guionensis blanketed the trees and low
vegetation. In addition to the above three habitats,
which were the most extensive, there were three
additional areas of relatively uniform vegetation on

Table 4. — Dates of the first observation of selected traits and
gestures in Kerodon rupestris juveniles. All values in days.

Individuals

Trait

n

JR

JF

J5

Bl

B2

B3

Maintenance and

locomotion

Frisky-hops

2

13

Running

1

2

7

3

3

Climbing

14

9

21

Basking

14

36

Stretching

14

Yawning

14

Grooming

14

6

29

2

3

Refection

6

29

Foraging

14

17

12

7

13

3

Drinking

15

51

Aggression

Lunge

45

101

48

28

Chase

28

34

51

Flee

98

15

12

21

32

58

51

Grappling

70

15

26

26

Grappling stops

175

80

57

54

71

71

Submit

76

51

Reproduction

Follow

Followed

84

17

12

105

30

Circling

210

Mounting

70

30

34

Copulation

Estrous

84

71

60

Copulatory Plug

83

Pregnancy

81

Partum
Testes down

119

156

113

114

114

Contactual behavior

Crawl-overs

14

3

9

4

2

2

Allogrooming

Nose-nose

70

12

46

51

Vocalizations

Alarm call

84

2

Weaning

35

48

35

23

25

45

16

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

100-

90 ■

80-

UJ

cr

70 -

=)

K

60 -

<

q:

50 -

Ul

Q.

40 -

5

UJ

30 -

i-

20 -

10 -

0

1 1 1 1 1 1 1 1 1 1 1 h 1 1 1 1 1 h -r- 1 1 1 1 1 1

NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV

Fig. 5. — Weather data collected at the Fazenda Batente study site for November 1976 to November 1977. Dashed lines above the graph
indicate days with rainfall. Upper graph represents 15 day maxima and lower graph 15 day minima in degrees Fahrenheit.

the study site. One area had been previously culti-
vated, and was still dominated by the domesticated
fruit-bearing plants Manioniica clwrantia, Ciicii-
mus anguria, Lycopersicon esculentum, and Sola-
num americanmn. A small rocky area between two
large boulder piles was almost entirely occupied by
Cnidoscolum urens. A third locale, which was a flat
sandy area separating the study site from an adja-
cent stream bank, primarily contained the shrubs
Jacohinia sp. and Riiellia asperula, with a mixed
grass ground cover. This area was separated from
the stream bank by a line of trees of Cassia excelsa
and Pterogyne nit e ns.

Although no statistical technique of habitat anal-
ysis was used to define the above six microhabitats,
the combination of structural and floristic differences
that existed made each area quite easily distinguish-
able from the others. Trap sites could therefore be
assigned to a given microhabitat with confidence.

Maintenance Behavior
Ingestive Behavior

Feeding. — Although foraging behavior is basical-
ly the same in both genera, there were a number of
notable differences. Galea walked with the body
held close to the ground, the head also held low and
extended forward. The animals proceeded forward
interspersing slow, cautious steps with series of
quick hops. They would periodically hold the head
high, looking around and sniffing the air. Galea

rarely manipulated food with the forepaws. Kerodon
also moved forward with the same low-body pos-
ture, holding the nose just slightly above the level
of the ground (Eig. 6c). Like Galea, they also in-
terchanged bouts of slow walking and hop-like run-
ning while exploring. When Kerodon paused, they
generally sat upright, with their front paws off the
ground, very similar to a dog begging (Fig. 6a, b).
While in this position, they sniff the air and look
around. Kerodon in the field, however, spent very
little time foraging on the ground, and when they
did, it was almost always on a rock surface. The
majority of foraging bouts observed in the field oc-
curred in trees. Animals would emerge from rock
fissures and move about on the tops of boulders.
From there they would climb into adjacent trees
and forage on leaves, buds, flowers, and bark.
When disturbed, animals would leap from the trees,
breaking their fall by grasping at branches and vines
on the way down. Kerodon used the forepaws ex-
tensively while foraging, commonly sitting upright
and holding a leaf in both hands while eating. In the
field, both genera foraged day and night.

Drinking. — Animals were only observed drinking
in the colony. Both genera lean over the water
source (dish or puddle) and lap the water without
raising the head. Animals would occasionally pause
and look about, then continue drinking.

Refection. — Both genera were observed nibbling
the anal region, then lifting the head and chewing
(Fig. 10a).

1981

LACHER— CAVIID SOCIAL BEHAVIOR

17

Locomotion

Walking and running. — Both Kerodon and Galea
walk quadrupedally with alternating steps. Walking
was most commonly observed while animals were
casually feeding. Animals would also accelerate
into a run, using the same alternate foot pattern,
although rapid locomotion more often consists of
series of hops. During aggressive chases and when
following a female. Galea would often exhibit a
bounding locomotion (stotting), similar to Dolicho-
tis (Dubost and Genest, 1974). When hop-running,
Kerodon sprung simultaneously off its hind-feet,
then alternated the left and right forepaws, then
sprung off the hind feet, and so on. Kerodon are
extremely agile while running in the rock piles, and
would often run full speed towards a rock, leap
against the surface, then twist the body up to 180°
while in the air, running in a completely different
direction upon landing. They possess considerable
spring in their hind limbs, and can easily leap from
a run up onto a 3-ft high boulder. At times (for

example, exploring a strange area) they would run
short distances in quick bursts, then stop and re-
orient their body in jerky 90° and 180° turns.

Climbing. — Although Rood (1972) reported
climbing behavior in Galea musteloides , Galea
spixii were never observed climbing. Kerodon, how-
ever, were extremely agile climbers (Eig. 7). When
ascending a tree, Kerodon used an alternating foot
pattern much like a man climbing a ladder. Once in
the tree, Kerodon moved along the branches with
the body held close to the branch, again using a left
forepaw, right hindfoot, right forepaw, left hindfoot
pattern. For animals of their relatively large bulk,
Kerodon were capable of navigating extremely del-
icate branches, and were observed moving about in
the field and in the colony on branches of less than
1 cm in diameter. When moving from branch to
branch, they generally reached out with the fore-
paws, secured the branch, and then stepped or
hopped over with the hindfeet. When descending,
Kerodon either stepped down, as if descending a

18

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

Fig. 7. — Kerodon adult sitting on a low branch shortly after emerging from the rocks.

ladder headfirst, or slid down backwards or side-
ways. When frightened, they would leap 10 to 20 ft
to the ground, breaking their fall by grasping at
branches during the descent.

Frisky hops. — Both Kerodon and Galea juve-
niles exhibited frisky hops. The animal would leap
into the air, twist the body and shake the head to
and fro, often landing in a different orientation.
Frisky hops were occasionally accompanied by
bouts of exaggerated running. The various compo-
nents present in frisky hops correspond quite
closely to the components of Locomotor-Rotational
movements described in Wilson and Kleiman
(1974). The importance of olfactory stimulation in
eliciting frisky hops, however, has not been de-
fined.

Resting Posture

Galea spi.xii resting postures were the same as
described for G. mnsteloides (Rood, 1972). Galea

would commonly form a small nestlike depression
in the dry grass in which to sit (Fig. 8). Kerodon
exhibited a variety of resting postures (Fig. 9).
There were two basic sitting postures — one in
which the animal maintained the forelegs erect,
with the chest off the ground and the head up; and
a second more relaxed posture with the venter in
contact with the substrate, the forelimbs bent, and
the head drooped with the eyes partly closed. An-
imals were commonly observed in the latter posi-
tion sitting in the sun. Kerodon would often shift
from this second sitting position to a semiprostrate
posture, with both hindlegs extended out to one
side. In very hot weather, Kerodon would lie flat
on the ground with the hindlegs splayed out behind.
A common gesture in both Kerodon and Galea as
they sat was the head twitch, a rapid shaking of the
head, tilted to one side. Its apparent function was
to shoo small flies and mosquitoes away from the
eyes and ears.

1981

LACHER— CAVIID SOCIAL BEHAVIOR

19

Fig. 8. — Adult Galea in typical resting posture.

Grooming

Face wipes. — These were exhibited in both gen-
era (Figs. 10b, c; Fig. II; Fig. 12a). The animal
licked the forepaws, either together or separately,
then using the inner side of the paws, wiped the
face from behind the ears to the nose. The cheek
areas were wiped most often. Both paws were used
simultaneously, alternately, or one at a time. Galea
generally sat on their haunches while face wiping,
whereas Kerodon more commonly sat in a “beg-
ging dog” posture.

Scratching. — Both genera used the hindfeet to
scratch the ventrum, side, back, and head, partic-
ularly behind the ear. Kerodon possesses a groom-
ing claw on the innermost digit of the hind foot,
which it used to groom the pelage (Fig. 12b). The
mouth was often used to clean the claws after
scratching.

Nibbling and licking. — Both genera nibbled the
fur with the incisors. The areas groomed included
all areas posterior to and including the shoulders
and forelimbs. In addition to licking the paws during
face wipes, animals were observed to lick the anal
and genital regions. Adult females would often lick
the anal region of young juveniles, apparently to
stimulate defecation.

Nosing. — Both genera also groom by rubbing the
nose through the fur of the sides and belly.

B

20

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

A

Fig. 10. — Adult Kerodon exhibiting refection (a) and two different postures of an adult grooming by face wipes (b and c).

Combing. — When nosing the pelage, Kerodon
would often sit upright, and simultaneously comb
the fur with the forepaws. As in nosing, the sides
and ventrum were combed. Galea were not ob-
served combing.

Rolling and sandbalhing. — This aspect of groom-
ing behavior was observed only in Galea, and
formed an integral part of this species' sandbathing
behavior. Kerodon, although provided with sand-
bathing arenas, did not sandbathe. There were two
basic components to sandbathing in Galea — side
rubs, during which the animal would roll onto its
side, generally in a sandy or dusty area, and vig-
orously kick backwards with both hindfeet; and
ventral rubs, in which the forelimbs were held un-
der the body, the belly was flat against the sub-
strate, and the hindlimbs were again vigorously

kicked backwards. When siderubbing, the fore-
limbs and chest were usually in contact with the
ground, but occasionally the animal would roll com-
pletely onto its back. In addition, animals frequent-
ly reoriented their body position during ventral
rubs, kicking a few times, then turning 180° (that is,
head now in rump position), and kicking again. This
sequence was often repeated several times. As roll-
ing and sandbathing were highly integrated with
marking behavior, they will be discussed further in
the next section.

Marking

Kerodon in the colony were observed dragging
the perineum, and apparently marking, only three
times in 20 months of observation. This behavior
was not seen in the field. This indicates that olfac-

1981

LACHER— CAVIID SOCIAL BEHAVIOR

21

Fig. 11. — Adult female Kerodon face wiping.

tory communication may not be as important in

Kerodon as in Galea , a species which exhibits fre-
quent scentmarking behavior. Galea scentmarking
occurred both separately and in conjunction with
sandbathing. Marking during sandbathing occurred
primarily during ventral rubs, when the animal
would drag the perineum forward in between kicks.
Occasionally the animal would pause in the ventral
rub position and rub the perineum from side to side,
then continue sandbathing. Galea were also ob-
served to urinate while performing a perineal drag.

Sandbathing and marking were observed in a va-
riety of situations. All animals marked and sand-
bathed intensively upon introduction to the colony.
When a strange male was introduced to the already
extant colony, he urinated, marked, and sandbathed
extensively within 3 min after his introduction. On
one occasion the dominant male was observed to
scentmark and sandbathe on the same spot where
a subordinate male had done the same a few min-
utes before. Marking was also observed in more
specific aggressive situations. The dominant male
was observed marking after chasing a subordinate.

A

Fig. 12. — Adult female Kerodon face wiping while a juvenile
nurses (a), and an animal grooming the head and neck with the
hind-toe grooming claw (b).

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

TO

Fig. 13. — Juvenile Kerodon suckling from the forward position.

and the same behavior was also observed in the
dominant female. The spot marked was the point at
which the chased animal had initially been sitting.
In a reproductive context, a female was observed
scentmarking after urinating on a male who was fol-
lowing her.

Comfort Movements

Both genera often stretched and yawned while
resting. When stretching, they would commonly
stretch the body, then stretch one hind leg out be-
hind, followed by the other. Kerodon was observed
stretching the hindlimbs by kicking back in a man-
ner very similar to that employed by Galea while
sandbathing. The strong correspondence between
the postures and movements suggests that such
comfort movements as stretching may be the origin
of more complicated series of movements such as
those used in sandbathing. Within Galea, the move-
ments used while stretching are essentially sand-
bathing movements in slow motion.

Digging

Digging was not observed in either Kerodon or
Galea.

Elimination

When urinating, both genera would elevate the
hindquarters slightly. Kerodon also would occa-

sionally lift one hindleg and extend it to the side. In
the field. Galea feces were generally encountered
in runways. Kerodon, however, defecated in a
number of selected areas in each rockpile. Great
quantities of feces would accumulate in these areas,
in piles 1 to 3 ft in diameter and often an inch or
two deep. This same behavior is exhibited by
another group of rock dwelling animals, the rock
hyraxes (Hoeck, 1975).

Attend

In response to an external stimulus, Galea would
sit upright, forelimbs extended, with the head
raised. The animal would freeze in this position,
and would subsequently flee for cover or resume its
earlier activity, depending upon the presence or ab-
sence of subsequent stimuli. The dominant male
was commonly observed in an attend position, ap-
parently observing the movements of other animals.
He would generally sit on top of a large rock while
in this position to facilitate his observations. Ker-
odon normally sat in a relaxed posture, hunched
slightly forward. As birds moved or called in the
forest, Kerodon occasionally watched them, but
didn’t change posture. When an unexpected rus-
tling occurred in the trees, the animal would move
its body forward onto its toes, extend the forelimbs,
and stretch its upper body and head forward (Eig.
6d). If the stimulus were sufficiently intense, Ker-

1981

LACHER— CAVIID SOCIAL BEHAVIOR

23

odon would abruptly turn towards the adjacent
crevice and ready itself for fleeing. When in this
position, animals would commonly raise one foot
up and sniff at the air. They would occasionally go
into a complete upright posture. In addition, the
animal often made rapid 90 and 180 degree pivots
of the body while in the attend position maintaining
the rump at the same point. Kerodon alarm whistles
were usually given from an attend position.

Social Behavior
Contactual Patterns

Nursing postures. — Galea females generally sat
in an attend-iike posture while nursing; occasionally
adult females would lie down and dose their eyes.
Juveniles would generally suckle from the side; at
times they would suckle from in between the fore-
limbs of the female. When two juveniles were being

24

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

Fig. 15. — Adult female Kerodon grooming a suckling juvenile —
licking fur (a) and anogenital region (b).

nursed, they would suckle on the posterior pair of
teats from different sides or from the same side of
the body. When three juveniles were suckling, they
would utilize all four mammae. Kerodon females
sat in a posture similar to Galea while nursing
(Figs. 13, 14). Juveniles suckled both from the side
and from the front. A juvenile would sometimes
crawl completely under the belly of the female
while suckling. When there were two juveniles,
they suckled from the same side, or from different
sides, and the teat order was not specific. Juveniles
would occasionally pause while suckling and groom
the face. The female would also groom herself while
nursing, and at times would groom the juveniles
(Fig. 15). Before juveniles began suckling, they
often performed a series of climb-overs on the back
of the female. After a number of climb-overs, the

Fig. 16. — Two face-face contact positions in Kerodon — chin-on-
nose (a) and sniffing facial region (h).

female then allowed the juvenile to suckle. Females
also occasionally rejected juveniles by shaking
them from their backs, especially when the young
animals were nearly weaned.

Social grooming. — This behavior was rare in
Galea, and common in Kerodon. The only instance
of allogrooming observed in Galea was an adult fe-
male grooming her juvenile. In Kerodon, however,
there was a fairly complex network of allogrooming
associations. The most common association was
the grooming of juveniles by adult females, par-
ticularly while the young were suckling. Allo-
grooming between adult females, however, was
never observed. This was in sharp contrast to the
situation between adult males, where allogrooming
was quite frequent. The animal most often groomed
was the dominant male. He was generally groomed
by subordinate males (numbers three and four in
the hierarchy) who were his progeny or had been
introduced to the eolony as juveniles. The number
two male, introdueed as an adult together with the
dominant male, never allogroomed another animal.

1981

LACHER— CAVIID SOCIAL BEHAVIOR

25

Fig. 17. — Two adult Kerodon exchanging the nose-nose recognition gesture.

although he was himself allogroomed by the number
four male. The number three and four males once
allogroomed each other before their respective or-
ders in the hierarchy had been clearly established.
Intersexual ailogrooming was infrequent. An adult
female once groomed the dominant male and at-
tempted to groom the number three male, who
lunged at her. The number three male allogroomed
the dominant female once, and on a second occa-
sion allogroomed an estrous female after an at-
tempted mount. The areas allogroomed were pre-
dominantly the face, head, and neck, and
occasionally the forelimbs and shoulders. The ani-
mal performing the grooming w'ould either nibble or
nose the pelage of the other. In the observed case
of reciprocal ailogrooming, the two animals placed
their faces cheek to cheek, and nibbled at each oth-
ers neck and ear regions.

Nose -nose and kiss. — Face-face contact was
used as an apparent recognition gesture in both gen-
era (Figs. 16, 17). Initial nose-nose contact in Ker-
odon was sometimes followed by the cheek to
cheek gesture described in social grooming, as well
as a general sniffing of the body (Figs. 18, 19). In
addition, one animal would at times place its chin
on the back of the other following a nose-nose ges-
ture (Fig. 19a). A gesture resembling the kiss de-
scribed by Rood (1972) for Microcavia was also

Fig. 18. — Juvenile Kerodon nuzzling face (a) and side (b) of adult
female.

26

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

c

Fig. 19. — Contact-promoting behavior in Kerodon: chin on back (a), nose-nose (b), and mutual naso-anal (c).

observed in Keroclon, a mutual nuzzling of the
mouth region.

Crawl-over. — This behavioral component was
observed in both genera, though it was far more
common in Keroclon. One adult female Galea had

a habit of following another adult female around the
colony, and was observed crawling over the back
of the female on several occasions. The following
female also placed its chin on the back of the other.
Adult females would also occasionally place the

1981

LACHER— CAVIID SOCIAL BEHAVIOR

27

Fig. 20. — Young juvenile Kerodon exhibiting crawl-over behavior on resting adult female.

chin on the back of a juvenile. In Kerodon, crawl-
overs were observed in association with a variety
of other gestures. Adult males were often observed
either crawling-over, or resting their forelimbs, on
the backs of females (Fig. 19a). Female-female
crawl-overs were also observed, particularly in as-
sociation with huddling. The chin-on-back gesture
was also observed in Kerodon. Submissive males
placed their chins on the back of the dominant male
while allogrooming. Crawl-overs by males on fe-
males may actually be a component of sexual be-
havior, as males were twice observed exhibiting
crawl-overs in association with other sexual behav-
ior, like following and mounting. The most common
situation in which crawling-over was observed was
in adult-juvenile interactions. Juveniles would com-
monly crawl about on the backs of adults, particu-
larly adult females (Fig. 20). A variety of positions
was exhibited, including the chin on back. Juveniles
would occasionally approach the adult from the rear
and climb into a mounting-like position. Juveniles
would also climb completely onto the back of a rest-
ing female. Adult females in return would at times
place their chin on the back or rump of juveniles.
Juvenile crawl-overs may have a function in stim-
ulating nursing behavior in the adults (see Nursing
Postures).

Grappling. — Grappling behavior was observed

only in Kerodon juveniles. Animals would rear up
on their hind legs, grasp each other by the upper
body and wrestle back and forth. While wrestling,
the animals would often jump up and down, and
when one animal would drop down on all fours, the
other would climb or jump onto its back. Grappling
was similar both to frisky hops and adult agonistic
behavior, and is probably an important play behav-
ior in the development of certain aggressive ges-
tures. The actual aggressive counterpart in adults
is jousting and jump-turns, which have the same
basic structure, but are of a much higher intensity.
A juvenile on one occasion attempted to grapple
with an adult, who responded with an aggressive
lunge. Grappling, in this case, elicited a true ag-
gressive response from the adult. On another oc-
casion an adult male was sexually following a ju-
venile, whose response at the male’s approaches
was to turn and grapple. The male gesture of at-
tempting to mount served only to stimulate play
behavior in the juvenile.

Huddling. — Both Kerodon and Galea were ob-
served to rest in contactual positions while kept in
cages. Animals rested in a variety of positions —
side-side, head-over-side, rear-sit, and side-sit
(Rood, 1972). However, in the large colony rooms
animals generally rested in contact only when
frightened or disturbed.

28

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

Fig. 21. — Agonistic behavior in Kerodon: Male on left in partial submit while male on right assumes threat posture (a). Males begin
to joust (b and 0 and jump-turn (c through e). The animal on the left in sketch (e) has been caught flat-footed during a series of jump-
turns and has assumed a defensive posture.

Agonistic Patterns

Head thrust. — This was the lowest intensity ag-
gressive gesture in both genera. The dominant an-
imal would either jab the head forward or flick the
head sideways in the direction of an opponent. In
its most exaggerated form the aggressor would
lunge its entire body forward, or in the case of a
sideways thrust, would also throw its shoulders and
forelimbs towards the submissive animal. Head
thrusts did not appear to be associated with any
specific type of encounter, but rather seemed to in-

dicate a fairly high superiority of the aggressor over
the subordinate. It was the most common means
employed by an adult to rid itself of a pestering or
intruding juvenile. Dominant animals would often
approach a subordinate individual in a fairly slow,
deliberate walk with the head and shoulders held
low and directed forward. The appearance of the
body was thus much like the position assumed by
the aggressive animal after a head thrust. The low
body approach would often elicit a flee in a subor-
dinate, the functional similarity likely being related
to the structural similarity.

1981

LACKER— CAVIID SOCIAL BEHAVIOR

29

A

B

C

Fig. 22. — Agonistic behavior in Kerodon: aggressive chase (a), and
(b through d). The biting animal usually closes its eyes.

Attack lunge. — This act was also observed in
both genera, and was essentially an exaggerated
head thrust. The aggressor ran or jumped towards
another animal, terminating in the head-thrust po-
sition. The final position varied slightly between the
two genera. Kerodon attack lunges ended with the
body held low and directed forward, and the head
also directed forward and down. Galea terminated
an attack lunge with the body directed forward;
however, the forelimbs were usually erect and
spread, and the head was angled forward and up,
not down.

Chase. — If one of the above two gestures was not
sufficient to instigate a flee from the subordinate
animal, the dominant would then pursue the other
animal in a running chase. The end result was gen-
erally that the submissive animal fled, and the dom-
inant animal desisted. At times the chasing animal
would actually pursue the subordinate until he
caught it, and would attempt to bite the rump of the
fleeing animal. Kerodon chases in the field were
especially interesting, as the animals wove in and
out of rock crevasses, and even climbed trees. The
fleeing animal would often avoid an aggressor sim-
ply by crawling out onto a very thin branch.

Stand-threats. — Galea would frequently exhibit

three views of a dominant individual biting a submissive animal

curved-body postures in situations of unsettled
dominance hierarchies. The two animals would
characteristically slowly approach one another,
then stop and stiffen their limbs, hunch the back,
pull in the neck, and lower their heads. They would
frequently exhibit piloerection. One or both animals
would then lunge, and if neither fled, the animals
would again assume a curved-body posture and
continue the bout. When neither animal emerged
victorious, they would maintain the curved-body
for some time, then would mutually depart. Al-
though Kerodon did not exhibit curved-body pos-
tures, they would, in between bouts of jousting and
jump-turns (see below), space themselves parallel
to one another, but facing in opposite directions,
and slowly walk in large opposing semicircles.
When walking, the animals would take peculiar de-
liberate, bobbing steps. An animal would also oc-
casionally pause and rock the hindquarters to and
fro. This particular gesture, which I termed spacing
and pacing, was very uncommon. The act of align-
ing the bodies in parallel, but facing in opposite di-
rections, seems to have its origin in the curved-
body postures of Galea and Cavia.

Jump-turns. — Rood (1972) indicated that this trait
was observed only in Microcavia ; however, both

30

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

Fig. 23. — Sexual behavior in Kerodon: animal sniffs at sides (a)
and ano-genital region (b and c) of female.

Kerodon and Galea exhibited jump-turns during
colony and field observations. In Galea, jump-turns
were most commonly observed during bouts of
standthreats. One animal would lunge, and the oth-
er would jump into the air, both to avoid the lunge,
and to attempt to grab onto its opponent’s back.
This animal would in turn jump-turn, also attempt-
ing to get “on top’’ of the other. Such an aggressive
bout would generally consist of a rapid sequence of
jump-turns, both animals attempting to gain the top
position, and thus the advantage. The animal which
gained the advantage would bite the rump and back
of the animal on the bottom. The “loser’’ would

Fig. 24. — Sexual behavior in Kerodon: chin-rump follow (a),
naso-anal by male while female depresses hindquarters (b), fe-
male tail-up (c), and adult female in full submit position in re-
sponse to male advances (d).

then attempt to wrestle its way free, and would gen-
erally flee. At times neither animal would success-
fully dominate the other, and after a flurry of jump-
turns the pair would segretate into curved-body
postures, then begin the entire sequence again.
Jump-turns in Kerodon also consisted of a series of
jumps and twists during which one animal would
attempt to get on top of the other (Fig. 21). They
occurred most often in encounters of two
“strangers,” or of animals of approximately equal
status in the hierarchy. They tended to be far more
spectacular than Galea jump-turns, because of the
great jumping ability of Kerodon. Kerodon jump-
turns generally did not originate from a curved-
body posture, but rather from an upright position.
Animals would approach one another and rear up
on their hindlegs. The two animals would then
weave and bob on their hindlimbs, attempting to
grasp one another by the upper body. This would
then develop into an upright wrestling bout, termed
jousting (Fig. 21). As Galea would alternate be-
tween curved-body postures and jump-turns, Ker-

1981

LACHER— CAVIID SOCIAL BEHAVIOR

31

Fig. 25. — Four views of circling behavior in Keroilon. Adult male in front of adult female (a and b) stopping her forward advance. Male
then passes under the chin of female (c) and swings around, securing female by the nape of the neck while attempting to mount (d).

odon would typically alternate jousting and jump-
turns. In both cases, the winner was the animal
which would pin the other on the ground, with the
winner biting the back and rump of the loser (Eig.

22).

Retreat or flee. — A defeated animal would gen-
erally quickly run from the scene of an aggressive
encounter (Eig. 22a). On subsequent encounters
between the same two animals, the loser, or sub-
missive individual, would often exhibit a rapid re-
treat at the mere approach of the dominant individ-
ual.

Submit. — Both genera exhibited a submissive
crouch that was occasionally given by a subordinate
animal in response to excessive harrassment by a
dominant animal. In both genera, the posture was
quite similar; the animal would flatten itself on the
ground, belly pressed to the substrate, chin down,
ears flattened, eyes partially closed. Galea tended
to be somewhat more hunched than Kerodon. In
colony situations (large rooms) a submit can stop

a fight; however, in caged encounters it generally
was not effective.

Other aggressive traits described by Rood do not
apply here. In addition, I consider the tail-up an
aspect of sexual behavior, and I will discuss it in
the following section.

Se.xiial Patterns

Naso-anal. — This was the most common sexual
gesture in the two genera. One animal would ap-
proach a second, and sniff the anogenital region. In
Galea, the naso-anal was given by males to fe-
males, and frequently was the initiating gesture in
a mating chase. Kerodon males also sniffed the
vaginal area of females, with the frequency of naso-
anals being highest as the female approached es-
trous (Figs. 23, 24). Males would also sniff the an-
ogenital area of juveniles, both male and female,
often lifting the rump of the juvenile completely off
the ground. Shortly after the colony was estab-
lished, an adult female was observed giving the

32

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

naso-anal gesture to the dominant male on a number
of occasions. The female would also nose the fur
on the back, neck, and belly of the male. The re-
sponse of the male to these acts was typically either
to ignore the female, or to actively move away.

Chin -rump follow. — This also was a common be-
havior in both genera (Fig. 24a). A male would ap-
proach a female, and place its chin on the female’s
rump. The female would then attempt to move
away, and the male would follow. Each time the
male was close enough to the female, he would
again attempt to place the chin on the rump. This
would often deteriorate into an all-out chase, the
female running from the male, and the male follow-
ing in a hopping, or “slotting” gait, always attempt-
ing to place the chin on the rump. In the field, these
chases would weave in and out of vegetation and
rock crevices. The male would occasionally attempt
to bite the rump of the female, apparently to slow
or stop her. Often, a male Kerodon would place the
chin on the neck of the female, also an attempt to
impede her flight. Galea females generally vocal-
ized (peepy squeaks) while being followed.

Circling. — This was a complex gesture observed
only in Kerodon (Fig. 25). A male would approach
a female either from the front or from the rear.
When approaching from the front, the male and fe-
male would exchange a nose-nose; from the rear
the male would give a naso-anal, then at times place
the chin on the back. The male would then begin to
circle the female, passing close along her side. He
would pass in front of her face, occasionally giving
a nose-nose, or pass under the chin of the female.
He would then move along the other side to the rear
of the female. The male would then repeat the pass-
ing, but in the opposite direction. The male would
often give another naso-anal and then repeat the
process, passing a semicircle around the female,
then returning in the opposite direction. The female
would remain still, either standing, sitting, or in a
submit. When the male returned to the rear of the
female, she would often begin to move away, and
the male would follow, at times attempting to circle
again, and at times initiating a chin-rump follow.
The function of circling appeared to be to maintain
the female in place, so that the male could mount.
Although there was no behavior in Galea which
resembled circling, the "prowl” as described for
Cavia by Rood (1972) and the "figure-eight” dis-
play in DoUchotis (Dubost and Genest, 1974) are
both similar.

Foot tapping. — This also was observed only in

Kerodon. A sexually excited male would often rap-
idly tap the forefeet up and down on the ground.
This often occurred during circling, or between
mounts. This behavior seems related in structure
and function to the rump-tapping described for Mi-
crocavia (Rood, 1972). A variety of other hystri-
comorphs (Dasyprocta, Myoprocta, and Dinomys)
show foot-tapping motions (Kleiman, 1974), and a
Microcavia male will often tap the rump of a female
with his forepaws (Rood, 1972).

Copulations. — The basic pattern of copulation,
based on the criteria proposed by Dewsbury (1972),
was the same for both genera — no lock; intromis-
sion of short duration; multiple intromissions;
thrusting during intromission; single ejaculation.
The number of ejaculations present in Kerodon,
however, was based on the observation of only a
few copulations. As both Cavia and Microcavia
exhibited multiple ejaculations, further observa-
tions are necessary to confirm this aspect of the
pattern in Kerodon. The copulatory position dif-
fered between the two genera. Galea generally
mounted the female in an almost upright position,
resting the forepaws on the female’s rump. Kero-
don tended to rest the chest and chin on the fe-
male’s back and grasped the female behind the
shoulders with the forelimbs (Fig. 26). The upright
position, however, was also observed in Kerodon.
The position assumed seems to vary with the degree
of sexual excitement of the male, the upright as-
sociated with higher levels of excitement. A number
of gestures in the female seem to stimulate copu-
latory behavior in the male. Estrous females would
expose the perineum to males. In addition, the sim-
ple movement of lowering the head and depressing
the body would often stimulate males to mount.
Non-reproductive juveniles of both genera would
often exhibit mounting behavior, indiscriminantly
attempting to mate with adults and juveniles of both
sexes. On a single occasion, a juvenile male Ker-
odon achieved intromission on an adult female, but
did not exhibit thrusting.

Riding. — This behavior was described by Rood
for Microcavia and Cavia, and was also observed
in Kerodon. When the female attempted to pull
away from the male during mounting, the male
would often maintain his grasp, and hang on to the
female as she moved away.

Tad -up. — Rood described this act as being a de-
fensive aspect of agonistic behavior; however, here
it is considered as sexual in context. In both genera,
non-receptive females would respond to male fol-

1981

LACHER— CAVIID SOCIAL BEHAVIOR

33

lows by flattening the back, raising the perineum,
and squirting urine backwards onto the male (Fig.
24c). The male would then abruptly halt, shake his
head to and fro, and wipe the facial area. Females
would often present the tail-up without squirting
urine, which was also effective in halting males. An
additional gesture employed by females to halt ap-
proaching males was the tail-down, which is basi-
cally the same as the submissive crouch associated
with aggressive behavior (Fig. 24d). Females being
followed would suddenly stop and press the body
to the ground. This gesture was also surprisingly
effective in terminating a mating chase.

Vocalizations

Galea

Peepy squeaks. — These consisted of a series of
bubbly, high-pitched peeps and gurgling noises.
This was the most common vocalization. Peepy
squeaks seemed to indicate both arousal and/or
mild anxiety. Females would often give peepy
squeaks while being followed. Subordinate animals
would in addition give peepy squeaks while being
chased by a dominant animal. Peepy squeaks were
also emitted by animals when exploring a strange
area, and were highly contagious. They were very
common in associations of juveniles, and in fact,
any group of Galea would occasionally burst into
a chorus of peepy squeaks.

Stutter. — The stutter vocalization resembled a
nasal snort, and was often given in association with
the bark. Both these vocalizations were associated
with the aggressor in agonistic encounters, and in-
dicated a high level of aggression.

Squeak. — This was a single, high-pitched note,
and, as indicated by Rood, indicates pain. It was
commonly given by submissive individuals upon
being bitten in an agonistic encounter.

Drumming. — Animals in a high state of anxiety
would often drum the elongated feet up and down
on the substrate in rapid succession. It was most
commonly observed when animals were introduced
into a strange enclosure. An adult female was also
observed drumming when temporarily separated
from her juveniles.

Tooth chatter. — This was another vocalization
that appeared to indicate a high level of anxiety. It
was observed after aggressive encounters, and was
given by the loser.

Bark. — The winner of an aggressive encounter
would often bark at the fleeing animal. The domi-

b).

nant animal was by far the most frequent barker,
and at times would climb up on top of a pile of
rocks (the highest point in the colony) and spend 10
or 15 min barking. Barking, in conjunction with the
nasal snort, seemed to indicate both aggression and
dominance in Galea.

Kerodon

Churr. — This vocalization was very similar to
peepy squeaks and was, in essence, a low-pitched
version of the same. It was observed in animals
which were introduced into the extant colony, and
indicated mild anxiety.

Peepy squeaks. — Essentially the same vocaliza-
tion as in Galea, Kerodon peepy squeaks tended to

34

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

be more restricted in their use. Again, they were
common in animals released into strange surround-
ings, and also were observed in a variety of adult
male-female associations, and female-female asso-
ciations. They were extremely common among ju-
veniles, especially when they were with the adult
female. Kerodon adults, however, did not give
peepy squeaks in aggressive or sexual contexts.
Male-female peepy squeaks were given by animals
in contact or in close proximity, but were never
associated with sexual behavior.

Squeal. — This vocalization was analogous to the
Galea squeak, but was much louder and much more
protracted. Animals would emit a series of high-
pitched squeals in response to being captured by
hand or being bitten by an aggressor. This vocali-
zation obviously indicated fear and/or pain.

Slow whistle. — An additional vocalization asso-
ciated with anxiety, the slow whistle consisted of
a series of loud whistles, evenly spaced. They tend-
ed to be somewhat more muffled that the alarm
whistle. They were observed in a number of cir-
cumstances. Newborn juveniles would give slow
whistles when the female was out foraging. Adults
leaving the rocks to forage would also give the slow
whistle vocalization. In the field, this almost
seemed to function as a communication web, with
the animals whistling between one another as they
moved out into the trees to forage. More likely, the
slow whistle is an anxiety response to leaving the
rocks.

Alarm whistle. — Whenever a potential predator
moved into rocks occupied by Kerodon, the ani-
mals began to give an alarm whistle. The vocali-
zation started as a low clucking sound and in-
creased in both pitch and frequency as the source
of stimulus approached. The call was given both
while the animals were in the open, and when they
were down in the rocks. All animals, including Gal-
ea, would react to the alarm, but the whistle prob-
ably reflected high levels of fear and anxiety, as a
cornered animal would continue to whistle. Its
function as a true alarm call is uncertain (see Dis-
cussion).

Nasal hiss. — A rarely observed vocalization of
unclear function. The sound resembled a human
blowing his nose, and was observed in a number of
seemingly dissimilar contexts.

Tooth chatter. — As in Galea, an act which indi-
cated nervousness or anxiety. Given by a newly
introduced Kerodon juvenile while moving about
the colony.

Social Organization and Reproduction
Reproductive Behavior

Galea mating behavior is dominated by the chin-
rump follow, whereas Kerodon is more character-
ized by circling. Galea nuisteloides mating behav-
ior is described in detail by Rood (1972) and only
the differences observed for Galea spixii will be
noted here.

All males were observed chin-rump following fe-
males, with the frequency being highest as females
approached estrous. The churr vocalization was ab-
sent in males. Rearing, an extremely common be-
havior in G. musteloides, was also absent in G.
spixii. Galea spixii males exhibit a low tolerance to
other males when a female is near estrous, and are
extremely aggressive towards males which attempt
to follow females. A mating chase, presented in the
form of a raw interaction sequence, illustrates quite
well the basic format of reproductive behavior in
Galea spixii (Appendix II).

The chin-rump follow clearly dominates the mat-
ing behavior of Galea spixii. The practice of placing
the chin on the rump has the function of halting the
female, so that the male can mount. This was well
illustrated during the follow of B13 by MR. When
the female does not respond to the male’s chin
placements, the male often becomes frustrated, and
attempts to bite the rump of the female in order to
stop her, as was the case when MR followed B3. In
general, the female’s response largely determines
the sequence of events in a Galea spixii mating
chase (Fig. 27).

Both the gestures involved and the sequencing of
these gestures differ in Kerodon mating chases. The
typical patterns can be observed in the raw inter-
action sequences (Appendix II). The chase of 3 Oc-
tober 1977 was especially interesting. After the
chase, female B was captured, and possessed a cop-
ulatory plug. Throughout the chase, the only males
to follow were the dominant male, FR; subadult J2,
born in the colony; and subadult BR, who was in-
troduced to the colony as a very young juvenile.
The number two male, R, introduced as an adult
with FR, remained withdrawn throughout the du-
ration of the mating chase.

Although the chin-rump follow forms an integral
part of a Kerodon mating chase, circling is the most
effective behavior in stopping the withdrawal of the
female. The placing of the chin most likely serves
to inhibit the female tail-up as well as to aid in slow-
ing her retreat. Mating chases are most frequently

1981

LACHER— CAVIID SOCIAL BEHAVIOR

35

MALE AP

PROACH

FEMALE TAIL-UP MALE RETREATS

NASO-ANAL » CHIN RUMP FOLLOW

COPULATION

t

MALE ATTEMPTS MALE

MOUNT SUCCESSFULLY MOUNTS

FEMALE STOPS

MALE PLACES CHIN

FEMALE TAIL-UP

MALE RETREATS

Fig. 27. — Sequence of gestures in Galea spixii reproductive behavior. The flow of behavior can be either one-way (single arrows) or
two-way (double arrows). The chin-rump follow is the central behavior.

initiated by the naso-anal, although reproductive
behavior was also observed to initiate with a nose-
nose (Fig. 28).

One point of interest is the behavior of the dom-
inant male (FR) towards other males during the
mating chase. The dominant Kerodon male was tol-
erant of his male progeny but not of adult male R.
The two adult males R and FR held the initial dis-
pute over possession of the rock pile shortly after
introduction to the colony room with the four initial
females (B, F, M, J). Male FR, which won access
to the rocks and assumed the dominant position,
continued to be very aggressive towards subordi-
nate male R and actively excluded him from the
rock pile. All adult females of the colony inhabited
the rocks which were defended by dominant male
FR. The dominant male was not aggressive towards
other males subsequently bom in the colony or in-
troduced as young juveniles, even when these ani-
mals were sexually mature and participated in the
mating chase. Although the dominant male paid no
active attention to young bom in the rocks, this
tolerance could, in fact, be interpreted as a form on
paternal investiment. Even though FR dominated
the mating chases, and probably inseminated the
estrous females, the participation of his progeny in

the mating chase helped perfect their reproductive
behavior and would probably aid in successful re-
production by these juveniles later in life.

Reproduction and Growth

A periodicity in reproduction in the field is not
substantiated by the data collected. Very young
Kerodon were observed, and presumably born, in
all months except April, May, and June. Galea re-
productive data suffers from a large gap from March
to September due to a crash in the population. As
a result, reproduction was not observed in the
months of April through June (typically the rainy
season) for either of the two genera. More data are
required to determine if there is any regularity to
this observation.

In the colony, Kerodon and Galea exhibited a post-
partum estrous. There were no seasonal variations
in reproduction in the colonies. The vaginal mem-
brane remains closed until parturition, remaining
open for 1 or 2 days afterwards. Mating usually oc-
curred within a few hours after the female gave birth.
Postpartum mating chases were observed in both
genera, and postpartum copulation was observed in
Kerodon.

Galea females were more susceptible to trauma

36

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

Fig. 28. — Sequence of gestures in Kerodon mpestris reproductive behavior. As in Galea behavior, the flow can be either one-way or
two-way. The back and forth flow between the chin-rump follow and circling forms the center of the sequence.

than were Kerodon, thus complicating pre- and
postpartum observations. Galea would frequently
abort after being handled, and often stopped lactat-
ing if disturbed too soon after giving birth.

A gestation period of 49 to 52 days was observed
in Galea spLxii. This coincides closely with the 53-
day period reported by Rood ( 1972) for Galea mus-
telokles. The gestation period in Kerodon (75.0 ±
1 .42 SD days) is much longer than that reported for
other members of the subfamily Caviinae. Kerodon
litter size averages 1.41 ± .51 SD young (N = 17).
This is somewhat smaller than the average litter size
calculated for Galea spixii (2.20 ± .87 SD; N = 10).
The sex ratio at birth, however, is not significantly
different from 1 to 1 for either species (Lacher,
1980).

A few days before giving birth, a Kerodon female
would enter a period of low activity. She would
rarely leave the rocks and, when emerging, would
sit quietly. All births apparently occurred at night
and in the rocks. Both female and the newborn

would spend a large part of their time for the first
few days in the rocks, emerging only to feed and
nurse.

Both Kerodon and Galea juveniles are relatively
precocial (Fig. 29). Of 10 Kerodon weighed and
measured on their day of birth, the mean weight was
76 g ± 11.99 SD and the mean length 145.2 mm ±
10.50 SD. Galea newborn weighed 33.25 g ± 3.40
SD and had a mean body length of 103 mm ±3.16
SD. Newborn Kerodon are capable of running and
climbing rocks, but newborn Galea have relatively
weak hindlimbs. Although they are capable of scur-
rying for shelter, they do not possess the agility of
young Kerodon.

Growth curves for juvenile Kerodon and Galea
follow similar patterns though maximum adult
weights differ greatly (Walker, 1974; Lacher, 1980).
Both genera increase in weight until about 200 days
of age, when the rate of increase begins to level off.
Galea juveniles reach a plateau in total length much
sooner than Kerodon — 30 versus 80 days. The

1981

LACHER— CAVIID SOCIAL BEHAVIOR

37

Table 5. — Dates of the first observation of selected traits and
gestures in Galea spixii Juveniles. All values in days.

Individuals

Trait

BF

FMR

B13

B1

07

Maintenance and

locomotion

Grooming

22

Sandbathing

101

Foraging

11

30

Aggression

Chase

101

Flee

38

38

103

57

35

Reproduction

Follow

80

22

60

Followed

22

60

Mounting

30

Thrusting

38

Tail-up

121

Vagina open

80

Testes down

135

135

Contactual behavior

Crawl-overs

38

38

Vocalizations

Peepy squeaks

15

15

Bark

119

Weaning

49

49

43

35

42

levels at which the plateaus begin obviously reflect
the differences in adult size of the two genera.

Behavioral Development

Kerodon and Galea females, like other caviids,
give birth to precocial young. Juveniles of both gen-
era were observed moving freely about and foraging
at 2 to 3 days of age. Special attention was directed
towards juveniles during colony observations, and
^he first appearance of a variety of traits for both
Kerodon (Table 4) and Galea (Table 5) were noted.
Important events concerning changes in reproduc-
tive condition were also noted for both genera (Ta-
ble 6).

Certain aspects of these tables merit closer atten-
tion. Kerodon exhibit most basic maintenance be-
havior within a few days after birth. Juveniles run
with full coordination, forage on solid food, and ex-
hibit the complete adult grooming repertoire during
the first week. The first observed aggressive behav-
ior consists of grappling, which is essentially a type
of play-fighting, and fleeing from the aggressive ac-
tions of other animals. If the first five aggressive
actions against each of the colony-born animals is

Fig. 29. — A one-day old Kerodon rapes tris female. Note the
well-developed hindlimbs.

examined, it is found that 76.6% of the aggressors
were either adult females other than the mother, or
other colony-bom juveniles. These results will be
treated in more detail in the section on aggressive
behavior.

Kerodon juveniles do not exhibit more active ag-
gressive behavior until a much later date. Chasing
is observed first, well before lunging behavior.
Lunges tend to be given by animals of high domi-
nance, or by aggressors who hold a decided advan-
tage over their opponent, thus would be expected
to appear later than simple chases. Play-fighting, or
grappling, is last observed shortly after the appear-
ance of lunging behavior. The submissive posture
also first appears at approximately this time.

Table 6. — A comparison of the time of onset of selected repro-
ductive characteristics in Kerodon and Galea. Times with a sam-
ple size of N ^ 2 are presented ± one standard deviation.

Reproductive

characteristic

Kerodon

Galea

Weaning

35.17 ± 10.1 days
N = 6

42.25 ± 5.7 days
N = 4

Males testes descended

115 ± 2.7 days
N = 4

135 ± 0 days
N = 2

Female vagina open
(earliest observation)

60 days

80 days

Earliest pregnancy
observed

81 days

102 days

First partum

156 days

no data

38

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

JUMP-TURNS-

CHASE

-CURVED- BODY POSTURE

."DOMINANT" BITES

SUBMISSIVE FLEES/
GIVES SUBMIT

" DOMINANT" LUNGES

U, H. STAND OFF

APPROXIMATION

OF

TWO INDIVIDUALS

E.H

•- DOMINANT LUNGES ^ SUBMISSIVE FLEES

I »- DOMINANT BITES 1

i I » i ’

DOMINANT CHASES ^ DOMINANT DESISTS

f I t' t

SUBMISSIVE GIVES SUBMIT

I •- SUBMISSIVE FLEES

DOMINANT APPROACHES

I ►SUBMISSIVE AVOIDS / GIVES SUBMIT

Fig. 30. — The sequence of gestures in Galea spixii aggressive behavior differ between unestablished hierarchies (U.H.) and established
hierarchies (E.H.). The flow between various gestures often occurs in a continuous rapid sequence.

Juveniles are followed while only 2 or 3 weeks
old; a phenomenon also observed in other caviids
(Rood, 1972). Male juveniles do not begin following
females until much later, somewhat before the
testes descend. Mounting behavior is observed
much earlier than following behavior, generally
when juveniles mount their mothers during crawl-
overs.

Contact promoting behaviors differ in their fre-
quency of expression between younger and older
juveniles. Crawl-overs are one of the earliest ob-
served behavioral traits in Kerodon ; however, they
are restricted to fairly young individuals. The stan-
dard adult contactual gesture, the nose-nose, first
appears at approximately 2.5 months of age.

In summary, juvenile behaviors such as crawl-
overs and grappling, as well as the more passive
aspects of adult behavior (flee, followed) appear
first, before the animal is weaned. Juvenile behav-
iors begin to disappear and the remainder of the
adult repertoire appears after weaning.

Agonistic Behavior

Stand-threats in Galea spixii occurred at a much
lower frequency than that reported for Galea mus-
teloides (Rood, 1972). Rood stated that “their fre-
quency depends both on the population density and
the rigidity of the individual dominance relation-
ships,” with the frequency declining as density and
rigidity increased. This, in large part, explains the
infrequent occurrence of this trait in G. spixii. At
peak densities. Rood’s colony contained one animal
per 0.54 m-. Even after a number of juveniles were
removed, the density was only decreased to one
animal per m“. The G. spixii colony maintained in
this study had, at its highest density, one animal
per 5.33 m^, about five times lower than Rood’s
densities. The only occasion on which stand-threats
were observed was shortly after the formation of
the colonies, when previously isolated animals es-
tablished hierarchies. Once the hierarchies were
formed, the frequency of stand-threats diminished.

1981

LACKER— CAVIID SOCIAL BEHAVIOR

39

U. H.

APPROXIMATION

OF

TWO INDIVIDUALS

E.H.

SPACING AND PACING

JOUSTING

STAND OFF

DOMINANT LUNGES -

JUMP-TURNS-

•dominant" bites

"DOMINANT" LUNGES '

SUBMISSIVE FLEES

DOMINANT CHASES

SUBMISSIVE FLEES ■

DOMINANT APPROACHES

' SUBMISSIVE' FLEES /
GIVES SUBMIT

DOMINANT BITES

DOMINANT DESISTS

SUBMISSIVE Gl VES SUBM IT

SUBMISSIVE AVOIDS/GIVES SUBMIT

Fig. 31. — Sequence of aggressive gestures in Kerodon. Ke radon, as with Galea, presents two different sequences of aggressive
behavior — one for unestablished hierarchies (U.H.) and one for established hierarchies (E.H.). The major difference from the Galea
sequence is the substitution of the curved-body posture by jousting, and occasionally spacing and pacing.

The sequence of gestures in a given Galea ag-
gressive encounter therefore will differ, depending
upon the density of the population or the degree of
familiarity of the animals (Fig. 30). After the estab-
lishment of the hierarchy in the permanent Galea
colony, curved body postures were observed only
twice in 307 aggressive encounters (.65%), and thus
are not included as part of the normal behavioral
sequence for established hierarchies. The submit
posture was rarely observed in the colony, and oc-
curred primarily in situations of unestablished hier-
archies. Curved-body postures occasionally termi-
nated in a stand-off, with both animals withdrawing
simultaneously.

As with Galea, Kerodon aggressive behavior var-
ies with the degree of familiarity and rigidity of the
relationship between the animals (Fig. 31). The se-
quence of aggressive acts for established hierar-
chies is the same for both Kerodon and Galea.
There are a few minor differences in the sequence

of events for non-established hierarchies. The
curved-body posture is conspicuously absent from
Kerodon aggressive behavior. The uncommon ges-
ture spacing and pacing, in combination with joust-
ing, seems to occupy the functional position of the
curved-body posture.

This similarity in gestures and sequences is es-
pecially obvious in interspecific encounters. When
an individual Galea presents the curved-body pos-
ture to a Kerodon, the latter will align its body in
parallel to the former; Kerodon, however, does not
curve its body. In response to a Galea lunge, a
Kerodon exhibits jump-turns exactly as if it were
a Kerodon -Kerodon encounter. Kerodon tend to
simply avoid aggressive moves by Galea by jump-
turning away; however, if provoked, they easily
dominate Galea because of their larger size, supe-
rior agility, and jumping ability.

Interestingly, the flow of aggressive gestures for
well-established hierarchies is undirectional, as the

40

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

Table 7. — Summary of aggressive behavior observed in the Ga-
lea colony between September 1977 and March 1978. A = ap-
proach; L = lunge: C = chase. Losses mean the animal either
fled or gave a submissive display in response to the aggressor's
action.

Animals

Wins

Losses

A

L

c

A

L

c

Males

MR

5

19

13

0

0

0

BM

0

10

46

2

4

27

FR

0

1

6

0

5

26

FMR

0

1

4

1

9

57

BF

0

1

2

4

3

45

B2R

0

0

0

0

0

6

B2M

0

0

0

0

1

1

Total

5

32

171

7

22

162

Females

F

5

9

22

0

4

3

B3

5

2

15

2

8

10

R

0

5

25

2

0

17

M

0

0

3

3

5

26

B2

0

0

4

1

5

7

BI3

0

0

4

0

0

11

B1

0

0

0

0

2

4

BIJ

0

0

0

0

1

2

40

0

0

0

0

1

2

Total

10

16

73

8

26

82

outcome is predictable before the encounter. There
is a built-in cycle in the flow of aggressive behavior
for non-established hierarchies. This cycle will
often be repeated five or six times in a single ag-
gressive encounter. When the relationship between
the two animals is not established, exactly which
animal takes the initiative (that is, the “dominant”
individual) will often change each trip through the
cycle until one animal emerges as the convincing
victor. Being the winner of a single bout, however,
is often not sufficient to definitively set the hierar-
chical relationship. The “submissive” individual
(after a loss) will often initiate another bout with
the victor by assuming a curved-body posture or by
jousting. The order is often reversed in this manner,
the final established hierarchy being determined
through the action of numerous encounters between
the newly introduced members of the colony.

Galea females were overtly aggressive towards
one another beginning on the day of introduction
into the colony. Kerodon females, on the other
hand, exhibited amicable, or neutral, relations to-
wards one another throughout the first few months.
All initial aggression was male-male. If each female

Table 8. — Summary of aggressive behavior observed in the Ker-
odon colony between September 1977 and March 1978. See Ta-
ble 7 for explanation. A = approach; L = lunge; C = chase.

Wins

Losses

Animals

A

L

c

A

L

c

FR

23

26

Males

52

0

0

0

R

3

6

3

17

22

49

BR

3

2

16

2

17

19

J2

1

1

6

4

7

21

J5

0

2

3

1

3

2

B2

0

0

0

0

1

0

B3

0

0

0

0

1

2

Total

30

37

80

24

51

93

B

1

9

Females

31

0

0

1

F

0

8

5

1

2

10

J

0

5

23

1

4

10

JR

0

3

4

2

3

21

B1

0

0

0

0

1

6

JF

0

0

0

3

1

0

M

0

0

0

0

0

2

Total

1

25

63

7

11

50

is examined separately, an interesting pattern
emerges. The onset of aggressive behavior for each
colony female coincides with the first pregnancy for
that female. Thus the complete five-female hier-
archy was not established until the first pregnancy
of JR. Although difficult to demonstrate quantita-
tively, adult females were qualitatively far more ag-
gressive while pregnant. In addition, of six juveniles
born after 17 December (that is, the date on which
all four adult females were pregnant), four died
within a week to 10 days after birth. At least one
of these deaths was due to a failure of lactation in
the female. This high level of juvenile mortality is
probably related both to female-female aggression,
causing, for example, failure of lactation due to
stress, and to adult female-juvenile aggression,
which is especially common between females other
than the mother, and young juveniles. This high
level of female aggression is probably common in
the field, as the initial time period after the estab-
lishment of a colony represented a rather unnatural
situation. Field observations on Kerodon indicated
an absence of periodicity of reproduction, and a
reasonable proportion of the population is probably
pregnant at any given time.

Although the sequence of gestures in an aggres-
sive encounter is almost identical for Kerodon and
Galea, if the overall patterns of aggression within

1981

LACHER— CAVIID SOCIAL BEHAVIOR

41

a) MALES

t

)M

t

FR

t

FMR

f

BF

t

1

1

1 B2R

j t

B2

t

•

t

t

t

1 I

T

1

t

1

1

L_

_J

1

1

1

1

i

h = 1.00

b) FEMALES

2

t

1

1

B2

BI3

Bl

1

1

i

Bl J

40

]

]

h|= 1.00

h2=.548

Fig. 32. — Galea male (a) and female (b) hierarchies. The dominant animal in a given sequence is indicated by a dot and the animals it
dominates by arrows. Only animals located left of the dotted lines were used in calculating h. The calculation h for females includes
only sexually mature adults. The value hj includes adults and juveniles. The male h value is for sexually mature adults.

each colony are examined, some notable differ-
ences emerge. All aggressive encounters recorded
in both the Galea and Kerodon colonies between
September 1977 and March 1978 were tabulated
(Tables 7-8). As hierarchies were already well es-
tablished at this time, aggressive actions for domi-
nant individuals could be expressed as either ap-
proaches in an aggressive posture, lunges, or
chases. The submissive response was classified as

a “flee” from one of the above three dominant ac-
tions. In addition to the summary tables presented,
an individual record for each animal was compiled
in order to examine hierarchical relations, differ-
ences in use of aggressive gestures and intersexual
aggression.

Rank in the hierarchies was determined through
the balance of aggressive encounters between all
possible pairs of individuals of the same sex. The

42

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

a) MALES

h = .400

b) FEMALES

t

h=.900

Fig. 33. — Kerodon male (a) and female (b) hierarchies. The organization is the same as Fig. 32. Animals left of the dotted line are
introduced adults or juveniles which became reproductive adults during the course of the study.

rankings obtained were used to calculate an index 12 n / ^t - 1 \ ^

of linearity (h), using Landau’s method (Bekoff, ^ ~ ” 2 / ’

1977). The index of linearity varies between 0 and
1, and is calculated by the formula

where n = the number of individuals in the hier-

1981

LACHER— CAVIID SOCIAL BEHAVIOR

43

Table 9. — Summary of the comparisons of various aspects of aggressive behavior for the Kerodon and Galea colonies. Although no
direct comparisons are made between the two colonies, there are some marked differences present in the organization of agonistic

behavior.

Characteristic

Kerodon

Galea

Comparison of mean aggressive encounters
between sexes (excluding juveniles)

Males: mean = 61.4

Females: mean = 32.75

= 1.62, df = 4, 3; ns)

Males: mean = 76.4
Females; mean = 32.0
(t = 2.79, df = 9; F < 0.05)

Comparison of variation in aggressive
encounters between males and females
(excluding juveniles)

Maks: S" = 1,489.3
Females: = 54.25

F„ax = 27.45, df = 3; P < 0.05

Males: S" = 1,286.3
Females: S* = 207.4

Fmax = 6.20, df = 4, ns

Linearity of hierarchy

Males: h = 0.400

Females: h = 0.900

Males: h = 1.000
Females: h = 1.000

Both sexes use approaches plus lunges
and chases in the same proportions.

X\dj = 5.55; P < 0.025

X^<u = 2.457; P > 0.10

Intrasexual aggression occurs with the
same frequency for both sexes.

X\dj = 28.47; P < 0.005

X\di = 0.90; P > 0.10

archy and = the number of individuals domi-
nated by animal a. Hierarchies with a rating of 0.9
or higher are considered strongly linear.

Galea hierarchies {Fig. 32) are linear for both
males and females. Kerodon females also possess
linear hierarchies; however, males show a very
weakly linear relation (Fig. 33). The Kerodon male
hierarchy is even weaker when the two juveniles
excluded from the calculations are considered.
Both B2 and B3 were present in the colony for the
same amount of time as was juvenile female Bl.
Female Bl was aggressively attacked by three dif-
ferent adult females in the colony. Only her mother,
F, was nonaggressive. Males B2 and B3, on the
other hand, were involved in a total of only four
aggressive encounters, all with females. They were
involved in no aggressive encounters with males,
thus by definition could not be included in the cal-
culations of hierarchy linearity. The hierarchy, in
reality, should be considered even less linear than
is indicated by the h value.

Aggression between adults and juveniles is at its
minimum among Kerodon males (Fig. 33). The
dominant male, FR, directed only 12 of 101 aggres-
sive actions against juveniles. Eight of the 12 were
against BR, a male introduced into the colony as a
juvenile. Only four aggressive gestures were di-
rected against the 12 juveniles actually born in the
colony. Kerodon females are aggressive towards
juveniles other than their own progeny. Female F,
in addition, was aggressive towards her daughter,
JR, when JR was in late pregnancy.

There is no obvious pattern in Galea adult-ju-
venile aggression. Adults are aggressive towards

other juveniles, as well as their own progeny (Fig.
32).

Galea males were involved in 382 aggressive en-
counters during the observations, with a mean of

76.4 ± 35.9 SD per animal. Galea females, for the
same number of hours of observation, were in-
volved in 192 encounters with a mean of 32.0 ±

14.4 SD encounters per female. These mean values
are significantly different (t — 2.79, df = 9, P <
0.05). The variance of aggressive encounters among
males (S^ = 1,286.3) was much larger than among
females (S^ = 207.4); however, the difference is not
significant. The above calculations include only en-
counters involving adults. All encounters involving
juveniles were eliminated, as the potential number
of encounters varies with the amount of time that
each juvenile was present.

The Kerodon data were analyzed in the same
manner. Males were involved in 307 encounters,
giving a mean of 61.4 ± 38.6 SD encounters per
individual. As with Galea, the females were in-
volved in fewer encounters, 131, with a mean of
32.75 ± 7.4 SD per animal. The variances of ag-
gressive encounters for males (S^ = 1,489.3) and
females (S^ = 54.25) were significantly different
(Fmax = 27.45, df = 3, F < .05). Means were thus
compared using a t-test for unequal variances.
There was no significant difference (tg' = 1.62, df =
3, F > 0.1).

All agonistic actions were recorded as either
“approaches,” “lunges,” or “chases.” The fre-
quency of use of each of these various gestures was
compared between males and females of both gen-
era. The data used were the same as in the previous

ENCOUNTERS PER HOURS OBSERVATION

44

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

Fig. 34. — Relationships between the number of encounters ob-
served per hour and the number of days that the colony was in
existence for (a) a simultaneous introduction of both species

analysis. The number of “approaches” was ex-
tremely small for both Galea and Kerodon. Given
the functional similarities between “approaches”
and “lunges” (high superiority), the two gestures
were combined for the analysis. Galea males used
high and low superiority gestures in the same pro-
portions as females (x\dj = 2.457, P > 0.10), while
Kerodon males utilized a higher proportion of “su-
perior dominance” gestures than females (x^auj =
5.55, P < 0.025).

Finally, inter- and intrasexual aggression were
compared for the two genera. Data for both adults
and juveniles were used for these analyses. The hy-
pothesis that neither sex interacts disproportionate-
ly with the other was tested for Galea and Kero-
don. The hypothesis was accepted for Galea
Oi^acij = 0-9, P > 0.10) and rejected for Kerodon
(X^adi = 28.47, P < 0.005). Kerodon males interact
primarily with males, whereas females interact
equally with both sexes. These data are related to
the male FR-R conflict. There was no aggression
between FR and his harem females, and little
aggression towards females by the juveniles. The
actual number of aggressive encounters between
FR and R was large. These encounters generally
involved FR keeping R away from the rocks or the
females, a process involving an already established
hierarchy. High superiority gestures dominated in
these situations. There was aggression between R
and the progeny of FR, but levels were low because
of the vigilance of FR in guarding the rocks. There
is a straight-line hierarchy within juveniles (Fig. 33),
a factor which may be important in juvenile emi-
gration, but again, levels are low. The result is the
non-linear hierarchy and the high variance. The var-
ious comparisons of patterns of aggression within
each genus are summarized in Table 9.

Once the colonies were established, it was nearly
impossible to introduce an additional animal. An
adult female introduced into the Kerodon colony
was killed in two days, and a subadult male was
antagonized to such a degree that it was necessary
to remove him after a few hours. Only juvenile BR
was successfully introduced, and the colony con-

into an arena; (b) Galea introduced into a Kerodon colony; and
(c) Kerodon introduced into a Galea colony. Both intraspecific
and interspecific encounters are represented. The thin solid line
represents Kerodon wins over Galea, and the dotted line illus-
trates Galea wins over Kerodon. Intraspecific aggression is rep-
resented by the heavy solid line {Galea) and the dot and dash
line {Kerodon).

1981

LACHER— CAVIID SOCIAL BEHAVIOR

45

tained only seven animals at the time. Any animals
introduced into the Galea colony were subjected to
such violent aggressive actions by the dominant an-
imals that they would probably have been killed had
they not been removed. A large adult male intro-
duced one evening was scarred and bloody the fol-
lowing morning and had lost 10% of its body weight.

In order to evaluate the level of interspecific
aggression between the two genera, data were col-
lected on three mixed colonies — one in which both
genera were introduced simultaneously; a second
in which Kerodon were added to an existing Galea
colony; a third in which Galea were added to a
Kerodon colony. An hourly encounter rate for both
intraspecific and interspecific aggresion was calcu-
lated, and the daily trends in aggression compared
for all three colony situations (Fig. 34). The mean
encounters per hour for intraspecific aggression
fluctuate from day to day for both genera. The trend
in interspecific aggression varies with the colony
condition. When both genera were released simul-
taneously, the smaller Galea initially dominated the
larger Kerodon. After 2 days the trend reversed,
however, and Kerodon assumed an aggressive su-
periority. When Kerodon were introduced into a
Galea colony, interspecific aggressive actions were
equal for the first 2 days, after which Kerodon again
assumed a superiority. Galea were completely
dominated when placed in a established Kerodon
colony. Apparently, Kerodon can aggressively ex-
clude Galea. In the case in which Galea were in-
troduced, the 15 animals of the Galea colony used
throughout the study were placed in the large Ker-
odon room with the 14 resident animals. Kerodon
were extremely aggressive towards Galea in and
near the rocks, and by 6 days had effectively forced
the Galea to utilize the field, brush pile, and forest
areas. When startled. Galea fled into the rocks, but
after a delay of 1 to 5 min Kerodon would no longer
tolerate the Galea and would aggressively force
them out.

Galea probably use rock piles in the wild for tem-
porary escape from predators. The aggressive pres-
sure of Kerodon on Galea is probably less in the
wild than in the mixed colony, as the natural habi-
tat is structurally more diverse, offering potential
refugia to Galea. Microhabitat analyses, however,
indicated that Galea were significantly more abun-
dant in the forest thickets than in the other micro-
habitats, and these data, coupled with the above
observations, strongly imply that Galea is actively
excluded from the rocks by Kerodon aggression.

Interspecific aggression was also examined by
colony condition and type of gesture. The relative
proportions of Kerodon and Galea wins were
tested for all three colony conditions. Kerodon won
a significant porportion of interspecific encounters
only when Galea were introduced into the Kerodon
colony (x“ = 78. 19, df = 1, P < .005). Kerodon
superiority in the other two colonies was counter-
balanced by the high level of Galea dominance dur-
ing the first 2 days.

The use of '‘approaches,” “lunges,” and “chas-
es” was compared between encounters in which
Kerodon won and in which Galea won. There was
no significant difference when all three colony sit-
uations were summed (x“ = 0.428, df = 2, P >

0. 05). When each colony condition was examined
separately for both Kerodon wins and Galea wins,
Kerodon utilized a significantly greater proportion
of “chases” when attacking Galea introduced into
the Kerodon colony (x“ = 45.21, df = 2, P <
.005). Similarly, Galea used a significantly greater
proportion of “chases” when attacking Kerodon
introduced into the Galea colony (x“ = 1.86, df =
2, P < 0.005). In all other situations, the propor-
tions of the three gestures did not differ. These data
support the assumption made earlier that there ex-
ists a functional difference between the three ges-
tures. “Approaches” and “lunges” reflect a high
degree of superiority of the aggressor. “Chases”
are used in situations in which superiority has not
yet been established, as when strange animals are
introduced into an established colony.

The "‘'Galea introduced” mixed colony was ex-
amined in more detail than the other two, as two
established colonies were combined. The animals
used were those that had been maintained together
for the duration of the major part of this study. It
was thus possible to observe interspecific aggres-
sion between two populations with known estab-
lished hierarchies.

Aggression by Kerodon against Galea was ex-
amined for each animal in relation to its position in
the hierarchies. Kerodon males and females were
equally aggressive towards Galea (x^ == -59, df =

1, P > 0.05); however. Galea males were attacked
more frequently than females (x^ = 13.76, df = 1,
P < 0.005). If Kerodon interspecific aggression is
examined case by case, both females and males
were disproportionately aggressive towards Galea
males (x^ = 27.72, df = \, P 0.005). It was pre-
viously shown that Galea males are involved in sig-
nificantly more intraspecific encounters than fe-

46

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

Table 10. — Observed and expected capture frequencies for the five major species on the grid. The underlined pair of numbers for each
species indicate the point of greatest divergence from the theoretical distribution. The differences, D, were tested at P = .05 (Siegal,
1956). C.F.D. = Cumulative frecpiency distribution of traps per microhabitat type, beginning with the rarest and successively adding

each more common microhabitat.

Galea

Kerodon

Trichomys

Monodelphis

Didelphis

Total

Microhabital

C.F.D.

Ob-

served

Ex-

pected

Ob-

served

Ex-

pected

Ob-

served

Ex-

pected

Ob-

served

Ex-

pected

Ob-

served

Ex-

pected

Ob-

served

Ex-

pected

Cnidoscolus

flat

.007

3

1.61

0

.04

0

.20

0

.28

0

.32

3

2.45

Cultivated

area

.027

7

6.21

0

.16

2

.76

1

1.08

3

1.24

13

9.45

Scrub flat

.054

14

12.42

0

.32

2

1.52

1

2.16

4

2.48

21

18.90

Grassy area

.107

27

24.61

0

.64

5

3.00

2

4.28

6

4.92

40

37.45

Thorn thicket

.414

117

95.22

0

2.48

9

11.60

16

16.56

21

19.04

163

144.90

Boulder area

1.000

230

230

6

6.00

28

28.00

40

40.00

46

46.00

350

350.00

males. This higher level of aggressiveness in males
may have provoked the response of Kerodon males
and females. There are few gestural differences in
the repertoires of aggressive behavior between the
two genera, and the sequences are also quite simi-
lar.

Within Kerodon males, the frequency of aggres-
sive acts was not equally distributed among all in-
dividuals, even when young juvenile J13 was elim-
inated (x“ = 63.69, df = 6, f* < 0.005). The number
two male, R, exhibited 57.9% of the aggressive ac-
tivity against Galea. Male R was the only nonres-
ident of the ‘‘harem" of dominant male FR. Ker-
odon females, with juvenile J14 eliminated, were
also aggressive in unequal frequencies (x~ = 53.78,
df = 4, f < 0.005). The number four and five fe-
males in the hierarchy, JR and Bl, exhibited 91.1%
of the aggression. On the other hand. Galea males
(B2M eliminated) were all attacked in equal pro-
portions (x“ = 4.93, df = 4, f > 0.10). The small
sample size for Galea females prohibits a statistical
analysis, but aggression is approximately evenly
distributed throughout the hierarchy.

Use of Space

Only Galea were five-trapped in sufficient num-
bers to be able to calculate home range sizes. Male
home ranges averaged 872 ± 497 SD m^ and females
averaged 632 ± 978 SD m’; however, there was no
significant difference between sexes (t-test adjusted
for unequal variances; ts* = 619, P > .5).

Live-trap data were also used to calculate micro-
habitat preferences, not only for Kerodon and Ga-
lea, but for other mammals present on the grid as
well. The objective of the microhabitat analysis was

to determine if the two genera differed in their mi-
crohabitat distributions. The other species, none of
which was a likely competitor with the caviids,
were included in the analysis to provide material
for comparison.

A cumulative frequency distribution of traps by
microhabitat type was constructed based upon the
habitat analysis results. Deviations from this ex-
pected cumulative frequency distribution of cap-
tures were tested by the Kolmogorov-Smirnov
one-sample test (Table 10). Only Galea showed a
deviation from the theoretical distribution, being
significantly overabundant in the Croton thickets,
the microhabitat type which is physically adjacent
to the rockpiles. Kerodon was captured only in the
rocks, and although the sample size was too small
to indicate a significant preference for this habitat,
visual observations of 25 marked individuals indi-
cated that this species is a habitat specialist. All
other species are randomly distributed throughout
the six microhabitats. These data indicate either a
Galea preference for Croton thickets or an exclu-
sion of Galea from the rocks, leading to an over-
abundance in the habitat bordering the rockpiles.
Galea were constantly observed fleeing or retreat-
ing into the rocks at my approach, and showed no
obvious aversion to the rock areas. Also, Croton
thickets have little low vegetation available for for-
age by Galea to support the observed population
densities. The thickets offer little protection from
predators. These observations, coupled with the
data on aggression in mixed colonies, suggest that
Galea utilize the rocks as temporary refugia from
danger but are continually forced out by aggression
from Kerodon.

1981

LACHER— CAVIID SOCIAL BEHAVIOR

47

Table II. — Sumnuiry of the five time budget tables in relation
to four broad categories of behavior. All five groups spend the
majority of their time in various aspects of maintenance behav-
ior. All values except the activity ratio are percentages. The
activity ratio is equal to time inactive plus time sitting divided
by total time.

Behavior

Kerodon

Galea

Males

Females

Juveniles

Males

Females

Maintenance

98.34

98.77

99.01

99. n

99.73

Aggression

0.26

0.15

0.33

0.48

0.21

Reproduction

0.73

0.45

0.34

0.06

0.06

Contact

0.06

0.50

0.32

0.29

0.00

Activity-ratio

0.089

0.108

0.145

0.595

0.584

Time Budgets and Analysis

Time budgets were compiled for Kerodon adult
males, adult females, and juveniles (sexes com-
bined) as well as Galea adult males and females
(Lacher, 1980). Each time budget presents the av-
erage percent time an individual of a given group
was observed to perform a given behavioral act. An
activity ratio (time active divided by total time) is
also presented for all five groups.

Both genera spend the great majority of their time
involved in various aspects of maintenance behav-
ior (Tables 1, 11). The sum of aggressive, repro-
ductive, and contact-promoting behaviors repre-
sents approximately 1% of the time budget for each
group. Activity ratios were quite different between
Kerodon and Galea, with Galea active a much
greater proportion of the time. This difference was
associated with the large amount of time that Galea
spends foraging as compared to Kerodon.

In comparisons of the percent time spent in a giv-
en activity, some variables (for example, grooming)
were quite consistent among groups. Other vari-
ables indicate basic divergences between groups in
the allocation of time. Kerodon males, females, and
juveniles spent 1.29, 0.09, and 3.81%, respectively,
of their time foraging, whereas Galea males spent
55.07% and females 52.12%. These data were col-
lected for all groups approximately 12 h after food
was presented. Both genera are herbivorous, Ker-
odon feeding primarily on leaves and Galea on
grass. There were no major differences in caloric or
ash content in the plant species presented to the
colony animals (G. Eiten, personal communication)
and there is no likely physiological reason for the
difference in time allocation. One possible expla-
nation is that individual leaves of grass, particularly
younger shoots, are much lighter than, for example,
an individual tree leaf of Ziziphiis joazeiro or Cro-

ton jacohinensis. Galea spend more time foraging
simply because it takes more time to collect a suf-
ficient mass of food.

Kerodon females and juveniles spend much more
time “inactive” (for example, down in the rocks),
than the other three test groups. Although females
frequently emerge to nurse young, they undoubt-
edly nurse while in the rocks as well. Also, females
spend more time in the rocks even without young,
as the function of territorial defense is assumed by
the dominant male.

Mean population-durations, standard deviations,
and coefficients of variation were calculated for all
five groups for which time-budget data were col-
lected (Table 12). Within group variances for a num-
ber of behavioral acts were obviously quite differ-
ent from group to group. In order to attempt to
determine which acts were important in separating
the groups, a Kruskal-Wallis one-way analysis of
variance by ranks was performed to compare the
five populations (Siegel, 1956). Eourteen variables
were statistically different among the groups (Table
13). A distribution-free multiple comparisons test
based on Kruskal-Wallis rank sums (Hollander and
Wolfe, 1973) was used to compare the average
ranks of the above variables for the five groups
tested. This comparison allows for unequal sample
sizes, and is a more conservative procedure than
the test used for equal sample sizes. The experi-
ment-wise error rate was set at 0.05. Four variables
are effective in separating Galea adults from the
Kerodon groups; sitting rocks, sitting-ground, for-
aging, and activity ratio (Fig. 35). Kerodon juve-
niles cannot be separated from adults females, but
are statistically different from adult males in rela-
tion to two variables — inactive (that is, in the rocks)
and sitting in the trees. Based on Kruskal-Wallis
rank sums, no variables were effective in separating
adult males and females from one another for either
Kerodon or Galea.

Sexes were compared within each genus by a
Mann- Whitney U-test for two samples. Seven vari-
ables were effective in separating male and female
populations in Kerodon (inactive, sitting-trees, nurs-
ing, lunging, following, followed, crawl-overs),
whereas only one variable, foraging, differed be-
tween male and female Galea.

Although a variable by variable analysis is useful
in indicating which traits differed between groups,
such analyses do not permit a classification of in-
dividual samples in populations. In addition, a vari-
able by variable analysis will be less effective in

48

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

INACTIVE

KM

GF

GM

KF

KJ

SIT-ROCKS

GF

GM

KJ

KM

KF

SIT-GROUND

KJ

KF

KM

GF

GM

SIT-TREES

GM

GF

KJ

KF

KM

SIT-SAND

KM

KF

KJ

GF

GM

CLIMBING

GM

GF

KJ

KF

KM

WALKING

KF

KJ

KM

GF

GM

SUCKLE

GM

GF

KM

KF

KJ

NURSING

KJ

KM

KJ

GM

GF

FORAGING

KF

KM

KJ

GM

GF

GRAPPLING

GM

GF

KF

KM

KJ

FOLLOWING

KJ

KF

GF

GM

KM

CRAWL- OVER

GF

KM

GM

KF

KJ

ACTIVITY-RATIO

KF

KM

KJ

GM

GF

Fig. 35. — Results of the distribution free multiple comparisons test based on Kruskal-Wallis rank sums (Hollander and Wolfe, 1973)
for 14 significant variables. Groups are listed in order from lowest mean rank (left) to highest mean rank (right). All non-significantly
different groups are connected by the same line. GM = Galea males; GF = Galea females; KM = Kerodon males; KF = Kerodon
females; KJ = Kerodon juveniles. Experiment-wise error rate = 0.05.

1981

LACHER— CAVIID SOCIAL BEHAVIOR

49

Table 12. — Mean population-durations ± one standard deviation for 28 variables for two Galea groups and three Kerodon groups.

Coefficients of variation are given in parentheses.

Group

Variable

Galea males

Galea females

Kerodon males

Kerodon females

Kerodon juveniles

Total

Inactive

32.79 ± 27.7

(0.8)

34.01 ± 53.2
(1.6)

8.94 ± 15.7
(1.8)

174.93 ± 484.6
(2.8)

132.25 ± 212.6
( 1 .6)

76.34 ± 235.8
(3.1)

Sit rocks

10.04 ± 29.4
(2.9)

0.00

35.28 ± 42.8
(1.2)

42.64 ± 36.1
(0.8)

31.05 ± 22.4
(0.7)

25.30 ± 33.7
(1.3)

Sit ground

30.41 ± 22.4
(0.7)

48.41 ± 58.5
(1.2)

14.82 ± 30.8
(2.1)

9.74 ± 15.5
(1.6)

4.87 ± 5.6
(1.2)

20.03 ± 33.7
(1.7)

Sit trees

0.00

0.00

125.69 ± 144.0
(1.1)

186.35 ± 496.0
(2.7)

6.54 ± 10.7
(1.6)

66.07 ±231.1
(3.5)

Sit sand

7.38 ± 12.6
(1.7)

1.22 ± 2.9
(2.3)

0.00

0.00

0.00

1.47 ± 5.8
(4.0)

Running

1.59 ± 0.7
(0.5)

2.05 ± 0.9
(0.4)

2.06 ± 1.1
(0.5)

2.06 ± 0.9
(0.4)

2.09 ± 1.6
(0.8)

1 .98 ± 1.1
(0.6)

Climbing

0.00

0.00

1.80 ± 1.5
(0.9)

1.45 ± 2.3
(1.6)

0.59 ± 1.0
(1.8)

0.84 ± 1.5
(1.8)

Walking

2.96 ± 1.9
(0.7)

1.94 ± 1.6
(0.8)

1.48 ± 2.3
(1.6)

0.57 ±1.1
(2.0)

1.11 ± 1.4
(1.3)

1.56 ± 1.9
(1.2)

Sandbathing

0.33 ± 1.2
(3.5)

0.33 ± 1.2
(3.5)

0.00

0.00

0.00

0.11 ± 0.7
(5.9)

Grooming

4.51 ± 3.2

(0.7)

4.56 ± 1.8
(0.4)

5.8! ± 4.0
(0.7)

3.48 ± 2.5
(0.7)

4.75 ± 3.6
(0.8)

4.70 ± 3.2
(0.7)

Suckling

0.00

0.00

0.00

0.00

38.11 ± 64.9
(1.7)

8.7! ± 34.3
(3.9)

Nursing

0.00

8.19 ± 28.4
(3.5)

0.00

18.15 ± 33.7
(1.9)

0.00

4.78 ± 19.4
(4.6)

Foraging

60.24 ± 22.8
(0.4)

93.28 ± 41.6
(0.4)

6.46 ± 21.7
(3.4)

0.54 ± 1.9
(3.6)

17.27 ± 40.5
(2.3)

31.93 ± 44.9
(1.4)

Lunging

0.33 ± 0.5
(1.5)

0.17 ± 0.4
(2.3)

0.29 ± 0.5
(1.6)

0.00

0.13 ± 0.3
(2.7)

0.19 ± 0.4
(2.1)

Chasing

1.00 ± 1.4
(1.4)

0.27 ± 0.5
(1.8)

0.29 ± 0.7
(2.3)

0.65 ± 0.9
(1.4)

0.13 ± 0.5
(4.0)

0.44 ± 0.9
(2.1)

Fleeing

0.86 ± 0.9
(1.1)

0.87 ±1.1
(1.3)

1.08 ± 1.6
(1.5)

0.46 ± 0.8
(1.7)

0.60 ±1.1
(8.4)

0.78 ± 1.2
(1.5)

Grappling

0.00

0.00

0.21 ± 0.5
(2.3)

0.15 ± 0.4
(2.4)

1.71 ± 2.5
(1.5)

0.47 ± 1.4
(2.9)

Curve-body

0.00

0.75 ± 2.6
(3.5)

0.00

0.00

0.00

0.13 ± 1.1
(8.4)

Submit

0.00

0.00

0.00

0.15 ± 0.6
(3.6)

0.13 ± 0.5
(4.0)

0.06 ± 0.3
(5.9)

Follow

0.63 ± 1.2
(1.9)

0.00

1.97 ± 3.1
(1.6)

0.00

0.00

0.59 ± 1.8
(3.0)

Followed

0.00

0.29 ± 1.0
(3.5)

0.00

1.67 ± 4.1
(2.5)

0.93 ± 2.0
(2.1)

0.57 ± 2.1
(3.6)

Tail-up

0.00

0.29 ± 0.8
(2.6)

0.00

0.92 ± 3.3
(3.6)

0.00

0.22 ± 1.5
(6.6)

Repulsed

0.00

0.00

0.00

0.00

0.00

0.00

Mounting

0.08 ± 0.3
(3.5)

0.00

0.31 ± 0.8
(2.5)

0.00

0.22 ± 0.6
(2.8)

0.14 ± 0.5
(3.6)

50

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

Table 12. — Continued.

Group

Variable

Galea males

Galea females

Kerodon males

Kerodon females

Kerodon juveniles

Total

Crawl-over

2.63 ± 9.1
(3.5)

0.00

0.00

0.69 ± 1.6
(2.3)

1.82 ± 3.3
(1.8)

0.99 ± 4.1
(4.1)

Allogrooming

0.00

0.00

0.00

1.43 ± 5.2
(3.6)

0.00

0.27 ± 2.2
(8.4)

Scentmarking

0.00

0.00

0.56 ± 2.3
(4.1)

0.00

0.00

0.14 ± 1.1
(8.4)

Activity-ratio

0.59 ± 0.2
(0.4)

0.58 ± 0.3
(0.5)

0.08 ± 0.1
(1.2)

0.11 ± 0.2
(1.5)

0.14 ± 0.1
(1.0)

0.28 ± 0.3
(1.1)

separating groups than a simultaneous analysis of
all variables. The five populations therefore were
subjected to a multiple stepwise discriminant anal-
ysis (Table 14) to determine which linear combi-
nations of variables were most effective in separat-
ing the groups. In addition, each sample was
assigned a probability of classification to each of
the five populations.

The only variable not included in the multiple
stepwise D.A. was “repulsed.” All others were in-
corporated in the construction of the discriminant
functions. Eor the analysis of the five-group data,
there were three significant orthogonal axes (Table
15). The standardized discriminant function coeffi-
cients (Table 16) for these three axes indicate the
relative position and importance of all the variables
used in forming the discriminant functions. A two
dimensional scattergram illustrates the relative po-

sitions of the group centroids and all individual sam-
ples. As there were three significant axes (explain-
ing 96.3% of the total dispersion), the relative
positions of the five-group centroids can be visual-
ized much more clearly in three dimensional space
(Eig. 36). Discriminant Eunction I separates the
Galea groups from Keroclon. Four variables are
heavily weighted on the Galea side of the axis —
activity-ratio; sit ground; foraging; sandbathing.
The variables most heavily weighted on the Kero-
don side of the axis include nursing, suckling,
climbing, grappling, scentmarking, fleeing, submit,
mounting, and sit rocks.

The variables, which were important in separat-
ing the genera, corresponded fairly closely to the
univariate results (Fig. 35); however, discriminant
analysis was more sensitive in detecting other bio-
logically important differences between groups.

Table 13. — Results of a Kruskal-Wallis one-way analysis of variance by ranks for all five groups. Keroclon and Galea adults, and
Kerodon adults and juveniles. The Kruskal-Wallis test statistic (H) and the level of significance (P) are presented. All variables with

P < 0.05 are considered significant.

Variable

Groups compared

All five

groups

Four adult groups

Three .

Kerodon groups

H

F

H

p

H

P

Inactive

22.59

.0002

8.12

.0436

19.73

.0001

Sit Rocks

35.29

.0000

29.20

.0000

—

—

Sit Ground

26.25

.0000

17.91

.0005

—

—

Sit Trees

35.79

.0000

30.65

.0000

15.23

.0005

Sit Sand

23.82

.0001

16.96

.0007

—

—

Climb

25.67

.0000

23.95

.0000

6.68

.0355

Walk

14.30

.0064

13.17

.0043

—

—

Suckle

25.78

.0000

—

—

15.01

.0006

Nurse

14.01

.0073

9.72

.0211

10.85

.0044

Forage

46.06

.0000

41.44

.0000

—

—

Grapple

14.67

.0054

—

—

6.10

.0473

Follow

15.33

.0041

9.97

.0189

11.44

.0033

Crawl-over

12.51

.0139

—

—

7.52

.0233

Activity-ratio

39.25

.0000

33.58

.0000

—

—

1981

LACHER— CAVIID SOCIAL BEHAVIOR

51

Table 14. — Wilk's lambda (U-statistic), univariate F-ra!io, and
level of significance {P) for all 28 variables used in the discrim-
inant analysis. P-value is for the F-statistic with 4 and 65 degrees
of freedom. There were 10 significant variables for the five-group
comparison.

Variable

Wilks’ lambda

F

P

Inactive

0.92

1.36

>.25

Sit rocks

0.78

4.35

<.005

Sit ground

0.78

4.33

<.005

Sit trees

0.89

1.98

>.10

Sit sand

0.77

4.62

<.005

Running

0.97

0.48

>.50

Climbing

0.75

5.26

<.005

Walking

0.82

3.34

<.025

Sandbathing

0.94

0.97

>.25

Grooming

0.94

0.98

>.25

Suckling

0.77

4.61

<.005

Nursing

0.86

2.52

= .05

Foraging

0.40

23.44

<.001

Lunging

0.90

1.64

>.10

Chasing

0.87

2.28

>.05

Fleeing

0.96

0.62

>.50

Grappling

0.75

5.34

<.001

Curve-body

0.93

1.22

>.25

Submit

0.95

0.71

>.50

Follow

0.78

4.42

<.005

Followed

0.90

1.68

>.10

Tail-up

0.94

1.01

>.25

Repulsed

1.00

0.00

>.75

Mounting

0.93

1.12

>.25

Crawl-over

0.93

1.07

>.25

Allogrooming

0.93

1.10

>.25

Scentmarking

0.95

0.76

>.50

Activity-ratio

0.38

26.01

<.001

Sandbathing was indicated as an important variable
in separating the Galea population from Kerodon ;
qualitative observations had also indicated this to
be true. Grappling was indicated as an important
gesture in characterizing the Kerodon ethogram.
This is another difference which was not detected
by the non-parametric multiple comparisons test.
Eleeing was also heavily weighted for Kerodon,
primarily because of the relatively long extended

Table 15. — The three significant discriminant functions extract-
ed for the five-group comparison and associated statistics.

Discriminant function

Statistics

I

II

111

Percent variance

83.45

6.69

6.14

Wilk's lambda

0.0043

0.0878

0.2231

Chi-square

288.269

128.916

79.511

DF

108

78

50

Significance level

0.001

0.001

0.005

Table 16. — Standardized discriminant function coefficients for
the 27 experimental variables on the first three discriminant
axes. The sign and numerical value of the coefficient indicate
it’s relative position and relative importance on the axis. Vari-
able number 23, repulsed, was not included in the formation of
the axes.

Variables

Standardized discriminant function coefficients

Function 1

Function 11

Function 111

Inactive

-0.037

-0.246

0.170

Sit rocks

-0.091

-0.255

0.256

Sit ground

0.291

-0.187

0.107

Sit trees

0.051

0.017

0.263

Sit sand

0.040

-0.000

0.366

Running

0.024

0.011

-0.094

Climbing

-0.181

-0.181

0.201

Walking

0.087

-0.223

-0.004

Sandbathing

0.141

0.087

-0.492

Grooming

0.071

0.510

-0.129

Suckling

-0.181

-0.292

-0.379

Nursing

-0.217

-0.845

0.704

Foraging

0.098

-0.627

-0.433

Lunging

0.021

0.337

0.081

Chasing

0.022

-0.128

0.332

Fleeing

-0.139

0.078

-0.108

Grappling

-0.174

-0.200

-0.302

Curve-body

-0.007

0.296

-0.213

Submit

-0.133

0.079

0.179

Follow

0.036

0.541

0.173

Followed

-0.068

-0.514

0.034

Tail-up

0.024

0.277

0.156

Mounting

-0.127

0.087

-0.150

Crawl-over

-0.031

0.030

-0.020

Allogrooming

-0.027

0.268

-0.320

Scentmarking

-0.152

-0.041

0.035

Activity-ratio

0.590

0.211

0.468

chases when the fleeing animal weaved in and out
of the rocks.

The small sample size did generate some spurious
results, however. Submit and mounting were both
heavily weighted for Kerodon. Nonquantified ob-
servations indicated an approximately equal fre-
quency of occurrence of these traits for the two
genera. Scentmarking was observed only once in 70
trials, and was one of the three observations col-
lected for Kerodon in 2 years of study. Scentmark-
ing was actually far more common in Galea.

Discriminant Function II separates the males
from the female-juvenile complex. The males are
weighted most heavily by grooming, lunging, and
following, whereas the most important variables on
the female side include nursing, foraging, and fol-
lowed. The third axis is somewhat confused, but
serves primarily to separate Kerodon and Galea
females from Kerodon juveniles. Nursing and ac-

52

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

Fig. 36. — Five-group centroids in three dimensional space. Dotted lines represent the three significant discriminant functions, which
intersect at zero.

tivity ratio are heavily weighted for the adults’s
side. Suckling is one of the more important vari-
ables on the juvenile side; however, sandbathing
and foraging are more heavily weighted. A great
deal of this is due to the intermediate position on
this axis of Galea females, between Kerodon fe-
males and juveniles.

The results of the classification procedure indi-
cate that all samples were correctly classified by
genus (Table 17). There were some misclassifica-
tions by sex or age; however, the percent of cases
correctly classified was 84.3. One of the misclassi-
fied juveniles was a ’’borderline” case, a juvenile
male already presenting a few adult characteristics.
This particular juvenile was given a 56.9% proba-
bility of being an adult male and a 42.5% proba-
bility of being a juvenile . . . exactly what one might
expect for a maturing animal.

As juvenile behavior was extremely variable, an
additional discriminant analysis was run using only
the four adult groups to determine if a more efficient
classification could be obtained. As expected, there
were only two significant axes and the important
discriminant function coefficients remained basical-
ly the same. The scattergram and relative group
centroid distances were also very similar to the five
group analysis. The percent of correctly predicted
group memberships, however, was much higher
(90.7) when juveniles were eliminated from the
analysis.

All five groups analyzed represent relatively ho-
mogeneous populations based on mean durations.
The four adult groups were especially consistent.
A compound variable, like the mean population
duration, seems to facilitate group separation and
classification of individual samples. It is also more

1981

LACHER— CAVIID SOCIAL BEHAVIOR

53

Table 17. — Results of the SPSS Discriminant classification procedure. No samples were misclassified by genus, however there were

15.71% misclassifications by age or sex.

Actual group

No. of
cases

Predicted group membership

Group 1

Group 2

Group 3

Group 4

Group 5

Group 1

12

11

1

0

0

0

Galea males

91. Wo

8.3%

0.0%

0.0%

0.0%

Group 2

12

2

10

0

0

0

Galea females

16.7%

83.3%

0.0%

0.0%

0.0%

Group 3

17

0

0

16

1

0

Kerodon males

0.0%

0.0%

94.1%

5.9%

0.0%

Group 4

13

0

0

2

10

1

Kerodon females

0.0%

0.0%

15.4%

76.9%

7.7%

Group 5

16

0

0

2

2

12

Kerodon juveniles

0.0%

0.0%

12.5%

12.5%

15.0%

difficult to interpret. A variable which is signifi-
cantly different between populations may be larger
or smaller, it may be more common or more infre-
quent or it may differ because of combinations of
the above. The ideal situation would be to analyze
each variable in a variety of ways. For the purposes
of characterizing group differences and classifying
individual samples, however, frequency weighted
mean durations proved very effective. When uni-
variate methods were used to analyze the data, the
groups were difficult to distinguish. When all vari-
ables were analyzed simultaneously all five groups
were separated, and 85% of the individual trials
were correctly classified. The use of quantified as-
pects of a species’ behavioral repertoire (frequency
weighted mean durations in this case) as a system
of measurement was extremely effective in clari-
fying group differences. This was especially effec-
tive considering the sample sizes.

Principal components analysis was run on three
groups — Kerodon males; Kerodon females; and
Galea adults. Initial factors were extracted by the
principal components method of the BMDP pack-
age (Nie et al., 1975) from a correlation matrix with
all diagonal elements set at one. Only factors that
explained 10% or more of the total variance were
used in the interpretation of the results. In order to
facilitate interpretation, the factor matrix was ro-
tated to obtain a simpler structure via a Kaiser Nor-
mal Varimax rotation. The rotated factor loadings
were then sorted. The zero factor loading was set
at 0.2500.

Five factors in the Galea population accounted
for 10% (or more) of the total variance. None of the
factors, however, are especially strong; in fact, the

total variance accounted for by all five factors is
only 58%. The most interesting factor, II, tends to
separate submissive animals from dominant animals.
Factors 1 and V separate out two specific single
samples. Factor III separates the dominant female
from other animals and Factor IV distinguishes
males following estrous females from the rest of the
population.

The results obtained for Kerodon are more en-
lightening. There were five factors selected for Ker-

Table 18. — Factor loading matrix for Kerodon males. The ma-
trix has been rearranged so that columns are in order of percent
variance explained. The ron s also were rearranged, .so that for
each successive factor, loadings greater than 0.500 appear first.

The zero factor loading has been set equal to 0.250.

Variable

No.

Factors

I

II

III

IV

V

Foraging

9

0.98

0.00

0.00

0.00

0.00

Scentmarking

16

0.97

0.00

0.00

0.00

0.00

Activity-ratio

17

0.96

0.00

0.00

0.00

0.00

Walking

7

0.92

0.28

0.00

0.00

0.00

Fleeing

12

0.30

0.88

0.00

0.00

0.00

Sit ground

3

0.00

0.82

0.00

-0.31

0.00

Grappling

13

0.00

0.80

0.00

0.46

0.00

Follow

14

0.56

0.65

0.00

0.36

0.00

Running

5

0.00

0.00

0.89

0.00

0.00

Lunging

10

0.00

0.00

0.83

0.00

0.00

Inactive

1

0.00

0.00

0.00

0.70

0.00

Climbing

6

0.40

0.00

0.00

-0.67

0.00

Chasing

11

0.00

0.00

0.42

0.55

0.00

Sit rocks

2

0.00

0.00

0.00

0.00

0.90

Grooming

8

0.46

0.00

-0.28

-0.27

0.71

Sit trees

4

-0.25

0.00

-0.50

0.00

-0.59

Mounting

15

0.00

0.00

0.00

0.00

0.00

Eigenvalue

4.69

2.88

2.13

1.89

1.75

Percent variance

27.6

17.0

12.5

11.1

10.3

54

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

3 2 1- 0 1 —2 3

Fig. 37. — Graph of factor scores for Kerodon males in relation to factors II and III. Small graph in upper left hand corner indicates the
percent variance accounted for by each factor, as well as the behavioral component of the factor. Five of the seven trial runs on the
number one and two males are enclosed in area one. Area two represents subadult males; individuals which were submissive and
maintained a juvenile characteristic (grappling). Area three, near the origin, represents submissive adult males. Area four contains
three trials in which animals did nothing but sit, either in the rocks or in the trees, for nearly the duration of the trials.

odon males (Table 18), accounting for 78.5% of the
total variance. Factors I, IV and V separated single
trials from the rest of the population. Factor II dis-
tinguished obviously submissive males from other
males, and Factor III separated actively dominant
males from animals which spent little, if any, time
active. A plot of the factor scores for each trial on
these two axes give a good separation of dominant
and submissive individuals (Fig. 37).

Results for Kerodon females indicated six factors
which accounted for 10% or more of the total vari-
ance. Factor I distinguished behaviors associated
with nursing females from other animals. Factor II
separated estrous females from non-estrous fe-

males. Factors II through VI again separated indi-
vidual trials from the rest of the population. The
relatively small sample size used (13 to 24) is the
reason for the frequent occurrence of this phenom-
enon. Factor analyses are most effective with large
sample sizes, although the minimum allowable sam-
ple size is unknown (Frey and Pimentel, 1978).
Nevertheless there was some consistency in the re-
sults. Factors of biological coherence which ex-
plained 10% or more of the variance always keyed
on one of four things: gestures associated with
dominance; gestures associated with submissive-
ness; estrous females (and associated behaviors);
and nursing females.

1981

LACHER— CAVIID SOCIAL BEHAVIOR

55

DISCUSSION

Evolution of Behavior in the Caviidae

The evolution of morphological and behavioral
differences in the family Caviidae is strongly linked
to the different habitat requirements of the six gen-
era. In order to attempt a reconstruction of the evo-
lutionary trends within this family, it is necessary
to envision a caviid prototype, one from which the
living genera could have been derived. Rood (1972)
considered Microcavia to be most similar to the
caviine ancestor. This genus has a relatively high
social tolerance, amicable relationship among fe-
males, simplified sexual and reproductive reper-
toire, and occupies open thorn-bush formations.
This description is similar to what might be ex-
pected for the caviid prototype, an animal that pos-
sessed what was termed by Eisenberg (1963), a
“loose social system.”

This caviid prototype first appeared in South
America in the mid-Miocene, and by late Miocene
both subfamilies were present (Pascual, 1962). All
caviids show a reduction in the number of toes on
the hindfoot, indicating that at one point in their
evolutionary history, they were moderately digiti-
grade and cursorial. Although the actual foot pos-
ture and mode of locomotion now varies greatly
among the genera, this basic foot form remains con-
stant within the family. It is possible that changing
selective pressures have caused the caviids to be-
come secondarily small and less cursorial than their
precursors.

There are some major morphological and behav-
ioral differences between the subfamilies. The Dol-
ichotinae evolved in high plains, desert, and open
grasslands; their extensive morphological adapta-
tions for a cursorial life were described in detail by
Dubost and Genest (1974). In occupying this open,
structurally homogeneous habitat, it became in-
creasingly difficult for the female to protect and
care for the young alone, and selection subsequent-
ly favored a more active participation by the male
in protecting the female and young from predators.
Males would already have been tolerated by the
female during and after parturition in the “loose
social system” ancestor, thus no extensive modi-
fications in the social system were necessary. Sex-
ual patterns remained relatively simple. Basic hys-
tricomorph gestures like enurination and the tail-up
are present, and males display a figure-eight passing

behavior to the female during courtship. This ges-
ture may well have been part of the repertoire of
the caviid ancestor as it is also present in Cavia
(prowl) and in a modified form in Kerodon (cir-
cling). Aggressive behavior in Dolichotis is also
fairly simple, consisting of approaches, lunges,
chases with rump biting, and a mouth-to-mouth
threat posture (Dubost and Genest, 1974).

In the caviines, Microcavia maintains many sim-
ilarities to the hypothetical caviine ancestor. The
presence of limited shelter sites may have been an
important factor in favoring the maintenance of low
aggressive levels in Microcavia, particularly among
females.

Galea and Cavia, respectively, occupy increas-
ingly more productive and diverse habitats (scrub
forest, cerrado, pampas) offering an abundance of
shelter, food, and other resources. Although more
productive than the thorn-bush associations occu-
pied by Microcavia, these habitats remain struc-
turally relatively homogeneous. Where such habi-
tats occur, single individuals would not be able to
monopolize clumps of resources without large ex-
penditures of time and energy. Individuals within
populations would thus tend to be dispersed rather
than clumped. Selection would favor the develop-
ment of territoriality, effective aggressive gestures,
and complex displays for the attraction of females.
More effective means of visual, vocal, or olfactory
communication would also be favored. The end re-
sult would be a decrease in social tolerance within
the population and an increase in the complexity of
behavioral repertoires. Galea, and to a greater ex-
tent Cavia, seem to have evolved under these types
of selective pressures.

Kerodon, however, deviates markedly from this
caviine trend. It does occupy a relatively produc-
tive habitat; however, resources (rock piles) are
highly clumped. Thus it becomes quite feasible, in
terms of time and energy, for a single animal (most
likely a male) to monopolize a large amount of re-
sources critical to other animals (most likely fe-
males). This is a situation with a high potential for
the evolution of a harem-based mating system (Em-
len and Oring, 1977) requiring social tolerances
within resource clumps and complex aggressive
gestures to protect the clumps from intruders.

Kerodon was probably derived from a Kerodon-

56

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

Galea prototype which inhabited northeastern and
central Brazil, the area which is now covered by
the Cerrado-Caatinga complex. The Quaternary in
South America was dominated by alternations be-
tween wetter and drier periods (Vanzolini, 1970;
Vuilleumier, 1971) and the importance of these cli-
matic conditions in speciation in the Amazonian
fauna has previously been reported by various au-
thors (Haffer, 1969; Vanzolini and Williams, 1970;
Brown et al., 1974). It is quite possible that they
were also of importance in northeastern Brazil.
Geomorphological data indicate that the Caatinga
and Cerrado vegetation types expanded into the
Amazon Basin during dry periods (Vanzolini, 1973),
and populations of the Kerodoii -Galea prototype
could have been isolated in the rocky chains of hills
and plateaus in northeastern Brazil during a pro-
nounced drought. These hills and plateaus are more
humid that the Caatinga lowlands, and many re-
ceive orographic rainfall (Markham, 1972). The pla-
teaus currently serve as refugia in the dry Caatinga
for a number of more mesic-adapted species (for
example, Tamandua tetradactyla) and may have
been important refugia for mammals during the
Quaternary. Populations restricted to the more me-
sic plateaus (Kerodon precursors) would have di-
verged from the main body of the populations (Ga-
lea precursors) which was pushed southward or
eastward; Brown et al. (1974) have postulated a
large mesic refuge on the southeastern coast of Bra-
zil. As the plateau chains are fairly rocky, animals
isolated there would be expected to have evolved
the complex morphological and behavioral features
necessary for the successful exploitation of this
habitat. The Galea precursors, which migrated
southward with the refugia, would have changed
little morphologically or behaviorally, as they
would have experienced a minimal change in habi-
tat. A return to more mesic conditions allowed for
the recolonization of the lowland areas by the main
population (Galea), which probably maintained a
strong morphological resemblance to the initial pro-
totype. This would result in the current situation;
two closely related sympatric genera (Pascual,
1962), one of which shows morphological and be-
havioral adaptations to exploit rocky habitats (Ker-
odon ) and the other occupies the lowland areas and
shows morphological and behavioral similarities to
the basic caviine form (Galea). A discussion of the
major differences between these two genera and the
ecological factors which may have shaped these
differences follows.

Trends in Behavioral Evolution

Eisenberg’s (1963) study on the comparative be-
havior of heteromyid rodents provided valuable in-
sights on the evolution of social behavior within the
family. The data presented here allow a preliminary
evaluation of the trend in evolution of social behav-
ior within the family Caviidae. Rood (1972) fur-
nished both field and colony observations on Cavia,
Microcavia, and Galea miisteloides. Dolichotis was
studied under seminatural field conditions in Erance
(Dubost and Genest, 1974) and behavioral data on
Pediolagus was reported by Eisenberg ( 1974), Klei-
man (1974), and Wilson and Kleiman (1974). Behav-
ioral information available on the dolichotines is
somewhat sketchy and primarily descriptive. Be-
cause of the fine study on the Argentine cavies,
however, the patterns of ecology and behavior in
the subfamily Caviinae can be examined in some
detail (Table 19). Galea spixii and G. musteloides
are, in general, fairly similar. A few traits observed
by Rood (1972) were not observed in G. spixii.
Rearing, a common gesture in G. musteloides, was
completely absent in G. spixii. Climbing and dig-
ging were not observed for G. spixii, but both ges-
tures were probably associated with the chicken
wire outdoor enclosures used by Rood. The kiss
was observed for G. spixii', however, it was ex-
tremely rare. Two other aggressive gestures (jump-
turns and the submissive crouch) were very com-
mon in G. spixii but were not observed in G.
musteloides. There are thus three important gestur-
al differences at the species level — rearing, jump-
turns, and the submissive crouch.

Kerodon has by far the most complex overall rep-
ertoire among the caviine genera, especially in re-
lation to contactual, aggressive, and sexual pos-
tures. Kerodon vocalizations are also quite complex.
Rood suggested a trend in complexity among the
Argentine cavies, with Microcavia being the least
complex in terms of behavioral interactions, and
Cavia the most complex. If Kerodon is considered
on the same basis, it is far more complex than Ca-
via. Kerodon possesses nearly all of the gestures
present in the other genera, and in addition has a
complex contact-promoting play behavior (grap-
pling), a stereotyped aggressive gesture (jousting),
and two specialized reproductive displays (circling
and foot tapping). Also, the alarm whistle is a com-
plex vocal display with an apparent social function.

Rood (1972) also concluded that the three genera
of Argentine cavies differ in their degree of social

1981

LACKER— CAVIID SOCIAL BEHAVIOR

57

tolerance. Cavia was considered a distance animal,
with a diverse aggressive repertoire and a sparse
repertoire of contact-promoting gestures. Micro-
cavia was considered the most social of the two
genera, and this sociality is reflected in the complex
array of contactual gestures, and a correspondingly
simple aggressive repertoire.

Kerodon has both a complex aggressive behavior
and diverse contactual postures. This contradictory
situation is related to the maintenance of harems.
A great diversity of contactual gestures exists with-
in the harem to reduce aggression and maintain a
high level of communication among harem mem-
bers. There also exists a complex aggressive rep-
ertoire to ritualize aggressive interactions between
females and juveniles within the harem, thus reduc-
ing potentially crippling encounters, and to protect
the rocks from invasion by outsiders. Kerodon rep-
resents a major divergence from the typical cavy
trend of straight line dominance relationships. This
divergence is related to the distinctly different hab-
itat type occupied by this genus.

Vocalizations of the two genera maintained the
relative stereotypy of hystricomorph sounds noted
by Eisenberg (1974). In his study, the vocalizations
of a number of hystricomorph rodents were exam-
ined and classified according to their functional
context. Using Eisenberg's data and classifications,

I compared Kerodon and Galea spixii vocalizations
with those of a number of other caviid species
(Table 20).

Galea spixii possesses almost the same vocal rep-
ertoire as G. musteloides, one exception being the
ambiguous function of the tooth chatter. Tooth
chatters were given by G. spixii individuals in a
variety of aggressive contexts and were always giv-
en by the more subordinate individual in a given
encounter.

The slow whistle and alarm whistle are unique to
Kerodon, Dasyprocta punctata gives an alarm bark
(Smythe, 1978), and Dolichotis patagonum emits a
sharp “wheet” (Eisenberg, 1974). Octodon degas
uses a sharp squeak, at least part of which is re-
peated (Eisenberg, 1974). Both Spalacopus and
Lagidium give what appear to be warning calls, of
a very pure harmonic quality, although little is
known of their function in the wild. Lagidium in-
habits rock piles similar to Kerodon habitat, and
the call in this genus may have the same function
as is described here for Kerodon. No other hystri-
comorph has been reported to give a repeated, high-
pitched whistle like Kerodon' s.

Table 19. — Comparison of the relative frequencies of occurrence
of selected behavioral traits among five species of caviines. 0 =
absent, / = rare, 2 — common. Based on Rood {1972}.

Behavior

Micro-

cavia

Cavia

Galea

mustel-

oides

Galea

spi.xii

Kero-

don

Maintenance behavior

Climbing

2

1

1

0

2

Nosing

2

2

0

2

2

Combing

2

2

0

0

2

Rolling (Sandbathing)

2

1

2

2

0

Digging

2

1

1

0

0

Upright attend

2

1

1

1

2

Swimming

0

1

0

0

0

Frisky hops

2

2

2

1

2

Scentmarking

2

2

2

2

1

Contactual behavior

Climb-over (Crawl-over)

2

1

1

1

2

Side-sit

2

0

2

2

1

Rear-sit

2

0

2

1

1

Kiss

2

0

0

1

2

Social grooming

2

2

2

1

2

Grappling

0

0

0

0

Agonistic behavior

Jump turn

1

0

0

2

2

Stand threat

0

2

2

1

1

Tail-up

0

2

2

2

2

Facing

2

2

0

0

1*

Kick-back

1

1

0

0

0

Head-up

0

2

0

0

0

Submit

2

0

0

2

2

Jousting

0

0

0

0

Reproductive behavior

Rearing

0

0

2

0

0

Riding

2

2

0

0

2

Rumba

0

2

0

0

0

Rumping

0

2

0

0

0

Circling

0

p

0

0

2

Foot tapping

u

0

0

0

1

*Noles: Microcaviu males tap the rump of females. Cavia exhibit the prowl, a gesture
which superficially resembles circling. Kerodon juveniles exhibit a gesture similar to
facing while grappling.

The alarm whistle is interesting in that it resem-
bles a true alarm call. At the approach of a predator,
an animal will sit upright and begin to whistle. The
predator is probably detected by sound or smell, as
the whistling starts while it is still some distance
away. Other animals will respond to the whistle by
assuming the upright attend position. The animals
remain upright until the intruder is first seen by one
of the animals, at which time they all flee.

My impression is that the first individual which
detects the approach of a predator gives the alarm
whistle, alerting the other animals. All animals then
assume the upright position, and begin to watch for
the predator. The initial “whistler” would probably

58

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

Table 20. — Contextual classifications of various vocalizations present in the Caviidae. Data and classifications are from Eisenherg
(1974) and this study. Code: Roman numerals = syllable types (see reference above for detailed descriptions of sonogram analysis),
a = ascending frequency , d = descending frequency, ad = complex frequency modulation, s = short syllable, I = long syllable, r =

repeated emission, and u = no appropriate frequency modulation.

Warning sounds

Species

When

startled

When threatening
(to a conspecific
or slow predator)

Before or
after attack

Avoiding
while being
approached

When

injured

When defeated

Dolichotis patagonum
(Eisenberg, 1974)

sharp

wheet

tooth chatter
III; grunt

long wheet
with grunt

Pediolagus salinicola
(Eisenberg, 1974)

whine

Cavia porcellus
(Eisenberg, 1974)

tutt-tutt
II s, r

tooth chatter
III r

grunt, snort

low wheet I a
d, s, r

sharp

squeak

squeal 11 1

Cavia aperea
(Rood, 1972)

tooth chatter
111

grunt

bubbly squeak

squeal

Microcavia australis
(Rood, 1972)

tsit

tooth chatter

111

Galea musteloides
(Rood, 1972)

tooth chatter
III;

drumming

stutter

Galea spixii
(this study)

drumming;

tooth

chatter

stutter; bark;
tooth
chatter

peepy

squeaks

squeak

squeak; tooth
chatter

Kerodon rupestris
(this study)

alarm
whistle
I r s

squeal II

a d

Table 20. — Continued.

Species

When calling
young
(9)

When being
groomed

When

following

When seeking
contact

When courting

Special

contexts

Dolichotis patagonum

short grunts

cluck II a d.

wheet I a d, 1

(Eisenberg, 1974)

s

Pediolagus salinicola

cluck

wheet

(Eisenberg, 1974)

Cavia porcellus

cluck II a

gurgle; short

clucks II a d.

wheet lad,

purr II u s; r

inflected

(Eisenberg, 1974)

d

Had

s

s

wheets

Cavia aperea

rumble

(Rood, 1972)

Microcavia australis

rumble

(Rood, 1972)

Galea musteloides

churr

(Rood, 1972)

Galea spixii

peepy

(this study)

squeaks

(juv.)

Kerodon rupestris

peepy

slow whistle;

(this study)

squeaks

churr; tooth
chatter (all
anxiety)

1981

LACHER— CAVIID SOCIAL BEHAVIOR

59

Fig. 38. — A hypothetical representation of the trend in evolution of social behavior in the family Caviidae based on Eisenberg's (1963)
model for the classification of steps in the evolution of rodent social systems. The thin line illustrates the trend in behavioral evolution
for the six genera in the family, and does not indicate phylogenetic affinities. Microcavki approximates the transitional social system,
that is, showing trends towards dispersion as adults while maintaining a reasonable level of social tolerance. Environmental variables
were more important than phylogenetic constraints (for example, subfamily) in determining trends in the Caviidae.

not place himself in immediate danger, because at
the time of detection the predator is still distant. A
greater danger would be to flee without any idea of
the direction from which the predator is approach-
ing. This is especially true in the rugged, boulder-
strewn areas which Kerodon inhabit, because vi-
sual contact may not be established until the
predator is quite close. By alerting other animals in
the population, the “whistler” would gain the ad-
vantage of having numerous other eyes scanning
the pits and depressions for an approaching preda-
tor. Once an individual sees the predator and flees,
the rest of the population does likewise. All indi-
viduals gain the same advantage, and are exposed
to little individual risk. Indeed, indiscriminate
fleeing may well be far more risky than maintaining
an alert position on top of a boulder.

The resemblance of the alarm whistle to the
squeal (distress, pain) and the slow whistle (anxi-
ety, stress) aids in the development of a model for
the evolution of such an alarm call followed by the
“alert” reaction. Animals which initially gave a dis-
tress squeal upon hearing an approaching predator
would have gained the benefit of observing the re-
sponse of other animals to the squeal. Individuals
which did not respond to the squeal, or which fled
indiscriminately, would more likely have been lost
to predators; thus individuals which assume an alert
position would have been favored. Animals which
gave a “better” squeal, therefore alerting (and con-
sequently observing) more individuals, would also
have been favored, the end result being the modi-
fications of the distress squeal into a penetrating
whistle capable of alerting the maximum number of

60

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

individuals. If this were the case, the alarm whistle
should still maintain a relation to its initial, basic
function; that of a distress call. This is apparently
true, for as an animal is cornered it will give the
whistle.

Various authors have suggested that alarm calls
may actually benefit the caller (Trivers, 1971; Char-
nov and Krebs, 1975; Wilson, 1975; Sherman, 1977;
Staton, 1978) either directly, or indirectly by in-
creasing the caller’s inclusive fitness (Hamilton,
1964). By explaining the Kerodon whistle in the
most parsimonious way possible, I have concluded
that the call could have evolved through natural se-
lection acting on the individual by conferring a di-
rect benefit on the caller. Emlen (1973) has stated
that one of the important functions of kin selection,
however, may be in enhancing selection at the in-
dividual level. As Kerodon occur in semi-isolated
rock piles, living in groups of related individuals
(see below), the rate of selection of the alarm call
might have been greatly accelerated by a kin effect.

A proposed model for the evolution of caviid so-
cial behavior can be graphically represented in
terms of Eisenberg’s hypothetical model (Fig. 38).
The differing social organizations present for each
genus, considered in relation to its habitat require-
ments, offer an example of the importance of social
behavior in ecological adaptation.

Mating Systems

Orians (1969), in a discussion of the evolution of
mating systems, indicated a number of character-
istics of the biology of birds and mammals which
would favor a polygynous breeding system. Among
these were precocious young, limited nesting sites,
and juvenile survivorship, which is little influenced
by paternal care.

Emlen and Oring (1977) classified monogamous
and polygamous mating systems on the basis of cer-
tain ecological criteria. The environmental potential
for polygamy (EPP) was viewed as dependent upon
three factors — the spatial distribution of resources;
the temporal distribution of sexually receptive
mates; and the operational sex ratio, defined as
“the average ratio of fertilizable females to sexually
active males at any given time.”

When resources are clumped, they can be mo-
nopolized more easily by a single individual. A
small proportion of the population can thus control
a large proportion of the available resources. If
these resources are necessary for successful repro-
duction, the potential for accumulating multiple

mates increases with the potential for monopolizing
resources. Clumped resources should therefore ex-
hibit a high EPP.

When the temporal distribution of mates is
clumped (that is, sexual receptivity in the popula-
tion is highly synchronized), it becomes difficult for
one individual to control numerous individuals of
the other sex. This is especially true for species
which exhibit a prolonged courtship. The EPP
would be highest in those situations where one in-
dividual at a time is sexually receptive. A dominant
male could thus utilize his energies to monopolize
a sexually receptive female with little probability of
losing another mating opportunity. The potential for
males to accumulate multiple females should be
highest in populations which exhibit asynchronous
estrous periods.

The operational sex ratio provides a measure of
the ease with which one sex can monopolize the
limiting sex. When the proportion of sexually active
males is higher than sexually receptive females, it
becomes easier for a single male to gain control
over this small population of females, either directly
by herding or indirectly by holding and defending
necessary resources. Polygyny would be favored in
this case. The potential to shift towards polygyny,
however, will also be affected by spatial and tem-
poral clumping of females. When the ratio is
skewed in the other direction, the tendency towards
polyandry will vary with the degree of temporal and
spatial clumping of males.

Evaluations of Kerodon and Galea populations
by the above criteria (Table 21) indicate that both
genera exhibit a high potential for polygyny. When
the mating system of each genus is examined in
more detail, the means by which males control fe-
males appear to differ.

Kerodon exhibits resource defense polygyny, de-
fined by Emlen and Oring (1977) as “males control
access to females indirectly , by monopolizing crit-
ical resources.” The critical resource in this case
is the boulder-strewn rock face, the exclusive hab-
itat of Kerodon. Kerodon males actively defend the
rock piles, and accumulate multiple mates indirect-
ly through female choice of these limited sites.

Field and colony data strongly support this model
for Kerodon. Field observations indicated that Ker-
odon is an extreme habitat specialist, and marked
individuals maintained a fidelity to specific rock
piles. Kerodon born in captivity had a sex ratio of
unity. The sex ratio in field captured juveniles also
did not differ from unity. The adult sex ratio in the

1981

LACHER— CAVIID SOCIAL BEHAVIOR

61

Table 21. — Characteristics of Kerodon and Galea populations used to evaluate the ecological potential for polygyny.

Criteria

Kerodon

Galea

Spatial distribution of
resources

Highly clumped; animals occupy boulder
piles which have a very patchy
distribution.

Slightly clumped; more of a habitat generalist;
however, always found in areas with a
reasonable amount of ground cover.

Temporal distribution
of mates

High to total asynchrony; females exhibit a
post partum estrous. No periodicity in the
colony, no evidence of periodicity in the
wild.

Moderate to high asynchrony; females exhibit a
post-partum estrous. No periodicity in the
colony, some evidence of periodicity in the
wild.

Operational sex ratio

Highly skewed towards males; females
exhibit a post-partum estrous. Males
always receptive.

Highly skewed towards males; females exhibit a
post-partum estrous. Males always receptive.

Precocious young

Yes.

Yes.

Limited nesting sites

Yes.

Not likely.

Paternal care

Male tolerance of juveniles in harems. No
direct paternal care.

No paternal care observed.

field, however, is significantly skewed towards fe-
males (three males; 15 females, P < 0.01), a situ-
ation to be expected when single males monopolize
a resource about which females aggregate. This re-
sult implies that the imbalance present in adults oc-
curs only in the sexually active cohorts of the pop-
ulations, in this case through intense intrasexual
competition among males.

Additional behavioral observations for Kerodon
taken in the colony are supportive. Allogrooming
was observed in two contexts — adult females
grooming their progeny; and juvenile males groom-
ing the dominant male. The latter context seems to
be a contact promoting gesture used by the juve-
niles to appease the dominant male. The male does
not solicit the grooming; the juveniles cautiously
approach and the dominant male remains tempo-
rarily motionless. There are no such amicable re-
lations between the dominant male and the number
two male, nor among females. It has been shown
that females maintained a straight line hierarchy
(Fig. 33), and relations between adult females and
younger animals are predominantly agonistic. Al-
logrooming by the juvenile males may be of great
importance in their maintaining a position in the
hierarchy.

The relatively frequent use by Kerodon of crawl-
overs, another contact-promoting behavior, may be
important in maintaining social cohesiveness among
harem members. Although Kerodon scentmarked
infrequently, crawl-overs may serve to maintain an

olfactory cohesiveness among harem residents, par-
ticularly to mitigate aggressive behavior among ju-
veniles and between adult females.

Numerous data were presented in the section on
agonistic behavior which were supportive of re-
source defense polygyny. Hierarchy structures,
rates of inter- and intrasexual aggression and use of
gestures all supported the patterns expected for the
maintenance of harems.

Kerodon females only became aggressive to-
wards other females during their first pregnancy.
The hierarchy that formed within the harem corre-
sponded exactly to the order in which the female
residents became pregnant. A pregnant female has
much more to lose if expelled from the protection
of the harem male than a non-pregnant female.
Also, there was little aggression between juveniles
and their mothers or the dominant male. Most ag-
gressive interactions involving juveniles were with
other juveniles and females other than their moth-
ers. In a harem situation, a newborn would be a
direct threat to these two groups. There would be
direct competition with other juveniles for re-
sources and indirect competition with other adult
females in that competition with a female’s progeny
is a potential reduction in the fitness of that female.
Nevertheless, aggression among juveniles is not as
intense as might be expected. Juveniles are all re-
lated to some degree by the fact that they have a
common father. If the dominant male successfully
impregnates all the females, and all females have
the same number of progeny, the average coeffi-

62

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

dent of relationship among juveniles will be equal to

CR

l/2(p - 1) + */2(N - Dp
Np - 1

where p is the number of progeny per female and
N is the total number of females in the harem. If
females have unequal numbers of progeny, the av-
erage individual coefficient of relationship (I.C.R.)
between any given individual and all other juveniles
can be calculated as

l/2(b - 1) + 2 '/4Pi

where b is the number of sibs of the juvenile in
question. The sum of p, in the numerator thus in-
cludes the progeny of all females except the mother
of the juvenile for which the calculation is made.
Both equations assume that none of the females
were progeny of the dominant male. If this were to
occur the relationship among juveniles would be
higher. The average coefficient of relationship may
be important in limiting the number of females
which can form a harem. Selection would favor fe-
males and juveniles which became increasingly
more aggressive towards other females and juve-
niles as the total number of females in the harem
increased, and the average coefficient of relation-
ship declined.

The tolerance displayed by the dominant male
towards juveniles during reproductive behavior
may also be linked to relatedness. By allowing his
progeny to participate in the mating chase, the dom-
inant male is essentially “grooming an heir to the
throne.” He allows his progeny to gain experience
in developing their reproductive repertoire, and the
juveniles establish a relationship with the adult fe-
male harem residents. When the dominant male
dies, it is therefore more likely that one of his prog-
eny, rather than an outside male, will assume the
dominant position.

Wilson (1975) has proposed that, in harem birds,
communally-breeding groups of females may be
sibs, if they are genuinely cooperative. Emlen
(1978) presents empirical information implying that
communal females are competitive. Data collected
for Kerodon were supportive of Emlen.

There were relatively high levels of intraharem
aggression, especially among females, and between
females and juveniles. Emlen (1978) predicted that,
in birds, females may exploit the close proximity of

nesting areas to manipulate eggs and parasitize
nests of other females. Kerodon females seem to
manipulate progeny of other females, both by overt
aggression towards juveniles as well as aggression
against pregnant and lactating females. At least one
juvenile death was due to a failure to lactate in the
number three female. Possibly as a result of this
hierarchical aggression, the dominant female pro-
duced more progeny than any of the other females.
Dunbar and Dunbar ( 1977) reported a similar situ-
ation occurring among harem females in gelada ba-
boons, and Downhower and Armitage (1971) have
given indirect evidence that reproductive success
in yellow-bellied marmots is correlated with domi-
nance rank. Although the data presented here are
hardly conclusive, it leads me to speculate that this
may be a general phenomenon in harems. The ques-
tion merits further, more quantitative, investiga-
tion. These observations are based upon a single
colony; however, they indicate that, at least in this
situation, the harem is an internally competitive
group.

Galea is more difficult to classify within Emlen
and Oring’s system, but most closely approximates
the lek type of male dominance polygyny. Males
and females have overlapping territories and both
sexes establish linear dominance hierarchies. Sexes
do not differ either in their use of gestures or in the
proportions of inter- and intrasexual aggression.
Both males and females (in particular the alpha
male) are aggressive towards juveniles. Dominant
males apparently obtain access to estrous females
by excluding other competing males through overt
aggression. This is the probable reason for the sig-
nificantly higher number of mean encounters in
Galea males (Table 9).

Observations presented in the section on repro-
ductive behavior illustrate the high level of aggres-
sion present between the dominant male and other
colony males during the mating chase. The domi-
nant male was aggressive towards alt other males,
including animals introduced as juveniles as well as
colony-born males.

Paternal care in Galea is essentially nonexistent.
Adult males ignore juveniles, except for occasional
misdirected attempts at copulation. There is no
male-male allogrooming. Other contact-promoting
behaviors, especially crawl-overs, are rare. The
male contribution to parental investiment is also
minimal. Females apparently defend their own ter-
ritories and raise the young without assistance from
the male. As the survival of the young is little in-

1981

LACHER— CAVIID SOCIAL BEHAVIOR

63

Table 22. — Selected adaptations to an open scrub habitat in
Galea spixii.

Trait

Special function

Clawed feet

Improve traction on dusty, soft
substrate

Ground forager

—

Cursorial

—

Sandbathing

Mark trails and territories on
dusty substrate

Scentmarking

Mark trails and territories on
dusty substrate

Large litters

Predation adaptation (see text)

Less precocious
young

T rade-off for large litters

Male-subadult

aggression

Maintain a dispersed distribution
in the homogeneous thorn
scrub habitat

Male-female linear
hierarchies

Related to mating system (see
text)

Large temporal
allocation to
foraging

Related to grass-eating habitats
(see text)

fluenced by paternal care, females must choose
mates either on the basis of phenotype or territory
quality (Orians, 1969). Males stage aggressive con-
tests to obtain access to estrous females, but fe-
males still make the ultimate choice, selecting the
male directly, rather than indirectly, on the basis of
his aggressive phenotype.

This type of polygyny requires a skewed opera-
tional sex ratio, but does not result in an unbalanced
adult sex ratio. Mark-recapture data on Galea gave
an adult sex ratio of 34 males to 26 females, which
is not significantly different from unity. The sex ra-
tio at birth is also unity.

Variance in reproductive success among males in
a polygynous species is pronounced. In the specific
case of harems, one male maintains multiple mates,
whereas others do not mate at all. The resultant
difference in reproductive success among males
leads to intense intrasexual competition and per-
haps to the evolution of secondary sexual charac-
teristics (Trivers, 1972). Adult Kerodon males and
females captured on the study area do not differ in
mean weight, nor are there any obvious physical
differences between males and females. The only
differences observed concern the aggressive behav-
ior described above. Sexual selection, in addition,
may be mitigated in this situation by the fact that

Table 23. — Special adaptation to boulder pile habitat in Kero-
don rupestris.

Trail

Special function

Padded feet

Improve traction on rock
surface

Arboreal forager

Exploit abundant food source
available in rocks (ground
vegetation is minimal)

Climber

Exploit abundant food source
available in rocks (ground
vegetation is minimal)

Circling

Important in stopping female
during mating chase (female
can easily avoid male by
entering rocks)

Alarm whistle

Anti-predator adaptation (see
text)

Small litter size

"K-strategy” reproduction (see
text)

Highly precocious young

“K-strategy” reproduction (see
text)

Male subadult tolerance

Adaptation to mating system
(see text)

Female linear hierarchy

Adaptation to mating system
(see text)

Sally forager

Related to ability to collect a
relatively large mass of food
in a very short time

females choose mates indirectly, by selecting harem
sites on the basis of territory quality, and not on
the basis of some external secondary sexual char-
acteristic in males. The observed differences in
aggression between the sexes, however, may in
some way be related to the capability of males to
hold and defend rock piles.

Morphological and Behavioral
Adaptations to Microhabitat

I have emphasized the relationship which exists
between behavioral gestures and environmental
conditions. Galea spixii has numerous morpholog-
ical and behavioral adaptations to the more open,
relatively homogeneous areas it inhabits (Table 22,
Fig. 39), whereas Kerodon rupestris is extremely
well adapted for life in the boulder-strewn rock
faces (Table 23, Fig. 40).

The difference in litter size and gestation period
between Kerodon and other caviines (Table 24) in-
dicates a possible change in reproductive strategy
within the subfamily. Kerodon has the longest ges-

64

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

Fig. 39. — Hind foot of Galea spixii. Note the claws and stiff,
bristly hairs; two adaptations which improve traction on sandy
substrates.

Fig. 40. — Hind foot of Kerodon nipestris. Claws and hairs are
absent, and the sole is covered with a leathery epidermal padding
to improve traction on rock surfaces.

tation and smallest litter size of the caviines. Ker-
odon young are born larger and heavier than Galea,
develop faster behaviorally (Tables 4, 5), grow fast-
er, and, although sample sizes are small, apparently
mature faster sexually (Table 6). The length of ges-
tation period is probably related to the larger adult
size of Kerodon in comparison with the other gen-
era. The difference in litter size appears to be a
special adaptation to the habitat.

Caviines in general occupy open formations.
Births either occur in the open or in a shallow
depression (Rood, 1972). There are no data indi-

cating any nest construction for Rood’s Argentine
cavies; however, a Galea spixii nest was found on
the Eazenda Batente site containing an aborted fe-
tus. The nest was constructed with dry grasses and
consisted of a shallow depression with low walls.
In any event, newborns and juveniles have little
protection from predators.

Kerodon births occur deep in rock crevices
where the danger of predation is minimal. Eemales
nurse juveniles both in the rocks and on top of boul-
ders; however, they are always near a crack of fis-

Table 24. — Gestation period and litter size data for various caviine species.

Kerodon
(this study)

Galea spixii
(this study)

Galea nwsteloides
(Rood. 1972)

Cavia apcrea
(Rood, 1972)

Microcavia australis
(Rood. 1972)

Gestation (days)

75.0 ± 1.42

49-52

53.2 ± 0.2

60.9 ± 0.4

54.2 ± 0.4

Litter size

1.41 ± 0.5

2.2 ± 0.9

2.7 ± 0.0

2.1 ±0.1

2.8 ± 0.3

1981

LACHER— CAVIID SOCIAL BEHAVIOR

65

sure when doing so. Juvenile mortality is probably
much lower in Kerodon than in Galea.

All caviines give birth to precocious young. In
relation to Kerodon, however. Galea, Cavia, and
Microeavia have maintained relatively large litters
of less precocious young. In the more open, unpro-
tected areas inhabited by Galea, there is probably
little advantage in having more precocious young.
A juvenile, once discovered by a predator, would
have little chance of escape.

Kerodon have small litters of highly precocious
young which subsequently grow and develop more
rapidly. Because of this rapid development and the
protection offered by the habitat, juveniles mortal-
ity is lower. Females expend relatively less energy
on reproduction, and would be expected to live
longer and have more litters. It should be noted that
litter size estimates in this study (Table 24) are
probably overestimated, as nutritional levels in the
colony were probably higher than in the field. No
female was seen with more than one juvenile in the
field.

Behavior associated with olfactory communica-
tion differed between genera. Of interest was the
absence of sandbathing behavior in Kerodon, even
though an area suitable for sandbathing was provid-
ed. Scentmarking was observed only three times.
Galea frequently sandbathed and marked, both
with urine and by dragging the perineum. These
differences may be related to habitat requirements.
Galea typically occupy open formations where the
soil is granular or sandy. Runways were abundant
on the study area, with numerous circular bare
areas present. These areas served as foci for Ga-
lea's sandbathing behavior in the wild and often
contained accumulations of feces. Although suit-
able sandbathing areas utilized by Galea were at
times interspersed among Kerodon' % rocky habitat,
Kerodon apparently does not need to sandbathe.

The same differences of substrate may also ac-
count for the depressed frequency of scentmarking
in Kerodon. The smooth granite surfaces would not
hold a scent like the shallow, dusty depressions
used by Galea. This would be especially true during
the rainy season. However, the extensive accu-
mulations of feces which are found in Kerodon pop-
ulated boulder piles may have an important olfac-
tory function. Certain piles could, in this manner,
contain olfactory information about the individual
residents and could provide information about ter-
ritorial boundaries.

Previous attempts at relating social organization

to environmental variables have been largely un-
successful (Clutton-Brock, 1974), primarily because
comparisons have been made across widely differ-
ing taxonomic groups. Exactly how a given species
adapts to a given environment will depend largely
on its phylogenetic background. Such constraints
have led to a variety of different responses to sim-
ilar environmental problems. Basic differences in
felid and canid social systems, for example, are par-
tially related to phylogenetically old behavioral
traits and morphological adaptations which influ-
ence hunting strategies (Kleiman and Eisenberg,
1973).

In order to evaluate the importance of environ-
mental factors on social organization it is far more
profitable to examine closely related species which
occupy markedly different habitats. Barash (1974)
was successful in correlating differences in social
organization with environmental factors for three
species of Marmot a.

Hoeck ( 1975) studied two sympatric species of
rock hyraxes, P roc avia johnstoni and Heterohyrax
hrucci, in Tanzania. Both species have similar mor-
phology and social structure, although Procavia is
somewhat larger. Both species are highly modified
for life in the kopjes (rock outcroppings) of the Af-
rican plains, and the males of both species maintain
harems. The hyraxes represent a situation where
the extensive morphological adaptations they have
undergone limit them to a certain habitat type,
much like Kerodon. Existence in the rock piles, a
clumped resource, has favored the evolution of a
harem-based mating system in both genera. In the
case of the hyraxes, the only behavioral modifica-
tion to allow for coexistence has been the devel-
opment of differential feeding behavior in the two
genera. Heterohyrax browses predominantly on
trees and shrubs, whereas Procavia is predomi-
nantly a grazer. The hyraxes illustrate an example
where both morphology and habitat place restric-
tions on the degree in which social behavior is mod-
ified. Although both genera are physically capable
of feeding in shrubs (Procavia in fact does browse
during the dry season), the modified foraging be-
havior in Procavia may be important in facilitating
coexistence between the two genera.

The strong morphological and behavioral similar-
ities between Kerodon and the hyraxes demon-
strate an interesting example in convergence. Both
the hyraxes and Kerodon are tail-less, guinea pig-
like animals with padded feet. Both have four claw-
less toes on the forepaws and three on the hindfeet.

66

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

the innermost toe being equipped with a small
grooming claw. The hyraxes, subungulates, have
rodent-like skulls with elongated rostrae, very
much like the skull of Kerodon. Both groups oc-
cupy rock piles, and both have harem-based mating
systems. Both even demonstrate the peculiar trait
of defecating only in cetain areas of the rock piles.

A number of other rock-dwelling mammals of
widely differing phylogenetic backgrounds have
also evolved the same specialized morphology — pi-
kas (Ochotonidae), gundis (Ctenodactylidae), and
yellow-bellied marmots (Sciuridae). The marmots
(Mannota flaviventris) have also been studied be-
haviorally and possess a harem-based mating sys-
tem (Armitage and Downhower, 1974; Armitage,
1975). Detailed behavioral studies on pikas and gun-
dis would provide valuable information on the im-
portance of social behavior in ecological adapta-
tion.

I have argued that two closely related species
show quite distinct patterns of social organization
in response to differing environmental pressures.
The paleontological work of Pascual (1962) indicat-

ed that the genera Galea and Kerodon were more
closely related to each other than either was to Ca-
via or Microcavia. If behavioral differences closely
reflected taxonomic affinities, Kerodon and Galea
would have exhibited more similar behavioral rep-
ertoires and social organizations than Galea and
either Cavia or Microcavia. The fact that Galea
exhibited more similarity to the other open forma-
tion caviines (Cavia and Microcavia) implies that
behavioral repertoires are responsive to ecological
conditions. The rather strict habitat requirements
of Kerodon were reflected in associated behavioral
traits.

Social organization thus appears to be strongly
associated with ecological adaptation. This implies
a far greater flexibility in the genetic control of be-
havior (or a far greater genetic variability) than, for
example, with morphology. Whereas morphology
provides the “basic hardware” of ecological adap-
tation, the lability of the behavioral response allows
an organism to fine tune its interrelationship with
the environment.

ACKNOWLEDGMENTS

I am extremely grateful to my major advisor, M. A. Mares,
who obtained the grant, pointed me in the direction of this proj-
ect, and offered his continual moral and intellectual support
throughout. Second, I must thank M. Willig, who suggested the
format for the equation of sib-relatedness in harems, K. Strei-
lein, and L. Vitt; their input throughout the course of the re-
search stimulated its development beyond the limit of my own
resources. Drs. W. P. Coffman, S. Gaulin, R. T. Hartman, and
R. J. Raikow offered valuable advice and criticism. The manu-
script was greatly improved by the reviews of Drs. J. F. Eisen-
berg, H. H. Genoways, and P. T. Handford. Drs. J. F. Eisenberg
and D. Kleiman kindly allowed me to use facilities at the Na-
tional Zoological Park. Additional thanks go to my family for
their patience during my long years of graduate study.

In Brazil, I wish to extend special thanks to Drs. A. P. Leao
and P. E. Vanzolini. Drs. J. J. Cruz and F. Cunha were ex-
tremely helpful with logistic problems, and Drs. M. de Ataide
Silva and D. de Andrade Lima of the Institute de Pesquisas

Agronomicas in Recife, Brazil, identified my plant collections.
J. Belino, A. Lemos da Silva, R. Lopes da Silva, J. Luna de
Carvalho, and F. Canute assisted with various aspects of the
field work in Exu, Pernambuco. Z. Saraiva, E. C. Teixeira, and
my wife, Susana, provided friendship, hospitality, and a shoul-
der to lean on during difficult times. E. Ventura graciously al-
lowed me unrestricted use of his land for my field work and he
and his family provided invaluable assistence throughout the
study. I. Sa provided indispensable counsel and good Montilla,
enabling me to overcome my frequent bouts of intellectual de-
spair.

The research was funded by the Academia Brasileira de Cien-
cias, project number 85 (Ecology, Evolution, and Zoogeogrpahy
of Mammals), as part of the larger program, “Ecological studies
of the semi-arid region of northeastern Brazil.” The untiring
efforts of Dr. A. P. Leao in obtaining the financing are deeply
appreciated. Additional funds were provided by the Carnegie
Museum of Natural History.

LITERATURE CITED

Ab'Saber, a. N. 1970. Provincias geologicas e dominios mor-
foclimaticos no Brasil. Geomorfologia Univ. Sao Paulo,
Inst, de Geografia, 20:1-26.

Armitage, K. B. 1975. Social behavior and population dynam-
ics of marmots. Oikos, 26:341-354.

Armitage, K. B., and J. E. Downhower. 1974. Demography
of yellow-bellied marmot populations. Ecology, 55:1233-
1245.

Aspey, W. P., and J. E. Blankenship. 1977. Spiders and snails
and statistical tails: application of multivariate analyses to

1981

LACHER— CAVIID SOCIAL BEHAVIOR

67

diverse ethological data. Pp. 75-120, in Quantitative methods
in the study of animal behavior {B. A. Hazlett, ed.), Aca-
demic Press, New York.

Barash, D. P. 1974. The evolution of marmot societies: a gen-
eral theory. Science, 185:415^20.

Bekoff, M. 1977. Quantitative studies of three areas of clas-
sical ethology: social dominance, behavioral taxonomy and
behavioral variability. Pp. 1^6, in Quantitative methods
in the study of animal behavior (B. A. Hazlett, ed.). Aca-
demic Press, New York.

Bekoff, M., H. Hill, and J. Milton. 1975. Behavioral tax-
onomy in canids by discriminant function analyses. Science,
190: 1223-1225.

Blackith, R. E., and R. A. Reyment. 1971. Multivariate mor-
phometries. Academic Press, London.

Brown, K. S., Jr., P. M. Sheppard, and J. R. G. Turner.
1974. Quaternary refugia in tropical America: evidence
from race formation in Heliconius butterflies. Proc. Roy.
Soc., Ser. B, 187:369-378.

Cabrera, A. 1953. Los roedores Argentines de la familia Ca-
viidae. Escuela de Veterinaria, Facultad de Agronomia y
Veterinaria, Buenos Aires.

Cabrera, A., and J. Yepes. 1960. Mamiferos Sudamericanos.
Historia Natural Ediar, Buenos Aires.

Carvalho, J. O. 1973. Plano integrado para o combate pre-
ventivo aos efeitos das secas no Nordeste. Ministerio do
Interior, Brasilia, Brasil.

Charnov, E. L., and J. R. Krebs. 1975. The evolution of alarm
calls: altruism or manipulation? Amer. Nat., 109:107-112.

Clutton-Brock, T. H. 1974. Primate social organisation and
ecology. Nature, 250:539-542.

Cody, M. L. 1978. Habitat selection and interspecific territo-
riality among the sylviid warblers of England and Sweden.
Ecol. Monogr., 48:351-396.

Colgan, P. W. (ed.). 1978. Quantitative ethology. John Wiley
and Sons, New York.

Contreras, J. R. 1965. Un caso de simpatria entre los generos
de roedores de la subfamilia Caviinae. Neotropica, 11:81-
83.

Cooley, W. W. and P. R. Lohnes. 1971. Multivariate data
analysis. John Wiley and Sons, New York.

Davies, W. G. 1978. Cluster analysis applied to the classifi-
cation of postures in the Chilean Flamingo {Phoenicopteriis
chilensis). Anim. Behav., 26:381-388.

Dewsbury, D. A. 1972. Patterns of copulatory behavior in
male mammals. Quart. Rev. Biol., 47:1-33.

Dixon, W. J. (ed.). 1975. BMDP: Biomedical Computer Pro-
grams. Univ. California Press, Berkeley.

Downhower, j. F., and K. B. Armitage. 1971. The yellow-
bellied marmot and the evolution of polygamy. Amer. Nat.,
105:355-370.

Dubost, G., and H. Genest. 1974. Le comportement social
d'une colonie de Maras Dolichotis patagonum Z. dans le
Parc de Branfere. Z. Tierpsychol., 35:225-302.

Dunbar, R. I. M., and E. P. Dunbar. 1977. Dominance and
reproductive success among female gelada baboons. Na-
ture, 266:351-352.

Eidt, R. C. 1968. The climatology of South America. Pp. 54-
81, in Biogeography and ecology in South America (E. J.
Fittkau et al., eds.). Junk N. V., The Hague.

Eisenberg, J. F. 1963. The behavior of heteromyid rodents.
Univ. California Publ. Zool., 69:1-100.

. 1967. A comparative study in rodent ethology with em-
phasis on evolution of social behavior. I. Proc. U.S. Nat.
Mus., 122:1-51.

. 1974. The function and motivational basis of hystrico-

morph vocalizations. Pp. 211-247, in The biology of hys-
tricomorph rodents (1. W. Rowlands and B. J. Weir, eds.),
Symp. Zool. Soc. Lond. No. 34.

Emlen, J. M. 1973. Ecology: an evolutionary approach. Ad-
dison-Wesley, Reading, Massachusetts.

Emlen, S. T. 1978. The evolution of cooperative breeding in
birds. Pp. 245-281, in Behavioral ecology: an evolutionary
approach (J. R. Krebs and N. B. Davies, eds.), Sinauer
Associates, Inc., Junderland, Massachusetts.

Emlen, S. T., and L. W. Oring. 1977. Ecology, sexual selec-
tion and the evolution of mating systems. Science 197:215-
223.

Fagan, R. M., and D. Y. Young. 1978. Temporal patterns of
behavior: durations, intervals, latencies, and sequences. Pp.
79-114, in Quantitative ethology (P. W. Colgan, ed.), John
Wiley and Sons, New York.

Frey, D. F., and R. A. Pimentel. 1978. Principal component
analysis and factor analysis. Pp. 219-246, in Quantitative
ethology (P. W. Colgan, ed.), John Wiley and Sons, New
York.

Frota-Pessoa, O., a. B. Coutinho, D. A. Lima, A. F. Fur-
TADO, M. J. A. Lima, S. M. Pereira, and E. A. Mansur.
1971. Biologia Nordeste I. Ecologia e taxinomia. Univ.
Federal de Pernambuco, Recife.

Guimaraes, L. R. 1972. Contribui^ao a epidemiologia da peste
endemica no Nordeste do Brasil e estado da Bahia. Estudo
das pulgas encontradas nessa regiao. Revista Bras, de Ma-
lariologia e Doengas Tropicais, 24:95-164.

Haffer, j. 1969. Speciation in Amazon forest birds. Science,
165:131-137.

Hamilton, W. D. 1964. The genetical theory of social behav-
ior, 1, II. J. Theoret. Biol. 7:1-52.

Hazlett, B. A. (ed.). 1977. Quantitative methods in the study
of animal behavior. Academic Press, New York.

Hoeck, H. N. 1975. Differential feeding behaviour of the sym-
patric hyrax Procavia Johnstoni and Heterohyrax hnicei.
Oecologia, 22:15-47.

Hollander, M., and D. A. Wolfe. 1973. Nonparametric sta-
tistical methods. John Wiley and Sons, New York.

James, P. E. 1942. Latin America. Odyssey Press, New York.

Karr, J. R., and F. C. James. 1975. Ecomorphological config-
urations and convergent evolution. Pp. 258-291, in Ecology
and evolution of communities (M. L. Cody and J. M. Dia-
mond, eds.), Belknap Press, Cambridge.

King, J. A. 1956. Social relations of the domestic guinea pig
living under semi-natural conditions. Ecology, 37:221-228.

Kleiman, D. G. 1974. Patterns of behavior in hystricomorph
rodents. Pp. 171-209, in The biology of hystricomorph ro-
dents (I. W. Rolands and B. J. Weir, eds.), Symp. Zool.
Soc. Lond. No. 34.

Kleiman, D. G., and J. F. Eisenberg. 1973. Comparisons of
canid and felid social systems from an evolutionary per-
spective. Anim. Behav., 21:637-659.

Kunkel, P., and I. Kunkel. 1964. Beitrage zur Ethologie des
Hausmeerschweinchens Cavia aperea f. porcelliis (L.). Z.
Tierpsychol., 21:603-641.

Lachenbruch, P. a. 1975. Discriminant analysis. Hafner,
New York.

68

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 17

Lacher, T. E., Jr. 1979. Rates of growth in Kerodon mpestris
and an assessment of its potential as a domesticated food
source. Papeis Avulsos Zool., Sao Paulo, 33:67-76.

. 1980. The comparative social behavior of Kerodon m-

pestris and Galea spi.xii in the xeric Caatinga of northeastern
Brazil. Unpublished Ph.D. thesis, Univ. Pittsburgh.

Landry, S. O. 1957. The interrelationships of the new and old
world hystricomorph rodents. Univ. California Publ. Zool.,
56:1-118.

Mares, M. A. 1975. South American mammal zoogeography:
evidence from convergent evolution in desert rodents. Proc.
Nat. Acad. Sci. U.S.A., 72:1702-1706.

. 1976. Convergent evolution of desert rodents: multi-
variate analysis and zoogeographical implications. Paleo-
biology, 2:39-64.

. 1980. Convergent evolution among desert rodents: a

global perspective. Bull. Carnegie Mus. Nat. Hist., 16:1-
51.

Markham, C. G. 1972. Aspectos Climatoldgicos da Seca no
Brasil-Nordeste. Ministerio do Interior, Recife, Brasil.

Moojen, J. 1952. Os roedores do Brasil. Rio de Janeiro.

Neff, N. A., and G. R. Smith. 1979. Multivariate analysis of
hybrid fishes. Syst. Zool., 28:176-196.

Nie, N. et al. (ed.). 1975. SPSS: Statistical Package for the
Social Sciences. 2nd ed., McGraw-Hill, New York.

Orians, G. H. 1969. On the evolution of mating systems in
birds and mammals. Amer. Nat., 103:589-603.

Osgood, W. H. 1943. The Mammals of Chile. Pubs. Field Mus.
Nat. Hist., Zool. Ser., 30:1-268.

Pascual, R. 1962. Un nuevo Caviinae (Rodentia, Caviidae) de
la formacion arroyo Chasico (Plioceno Inferior) de la Pro-
vincia de Buenos Aires. Ameghiniana, 2:169-174.

Patterson, B., and R. Pascual. 1968. The fossil mammal
fauna of South America. Quart. Rev. Biol., 43:409^51.

Pimentel, R. A., and D. F. Frey. 1978. Multivariate analysis
of variance and discriminant analysis. Pp. 247-274, in Quan-
titative ethology (P. W. Colgan, ed.), John Wiley and Sons,
New York.

Ringo, J. M., and R. H. Hodosh. 1978. A multivariate analysis
of behavioral divergence among closely related species of
endemic Hawaiian Drosophila. Evolution, 32:389-397.

Rood, J. P. 1970. Ecology and social behavior of the desert
cavy {Microcavia australis). Am. Midland Nat., 83:415-
454.

. 1972. Ecological and behavioral comparisons of three

genera of Argentine cavies. Anim. Behav. Monogr., 5: 1 -83.

Sherman, P. W. 1977. Nepotism and the evolution of alarm
calls. Science, 197:1246-1253.

Sick, H. 1965. A fauna do cerrado. Arq. Zool., Sao Paulo,
12:71-93.

Siegel, S. 1956. Nonparametric statistics for the behavioral
sciences. McGraw-Hill, New York.

Simpson, G. G. 1945. The principles of classification and a
classification of mammals. Bull. Amer. Mus. Natl. Hist.,
85:1-350.

. 1958. Behavior and evolution. Pp. 507-535, in Behavior

and evolution (A. Roe and G. G. Simpson, eds.), Yale Univ.
Press, New Haven.

Slater, P. J. B, 1978. Data collection. Pp. 7-24, in Quan-
titative ethology (P. W. Colgan, ed.), John Wiley and Sons,
New York.

Smythe, N. 1970. Ecology and behavior of the agouti (Dasy-
procta punctata) and related species on Barro Colorado Is-
land, Panama. Unpublished Ph.D. thesis, Univ. Maryland.

. 1978. The natural history of the Central American agou-
ti (Dasyprocta punctata). Smithsonian Contrib. Zool.,
257:1-52.

SoKAL, R. R., and F. J. Rohlf. 1969. Biometry. W. H. Free-
man Co., San Francisco.

Souza Reis, A. C. de. 1976. Clima da Caatinga. An. Acad.
Brasil. Cienc., 48:325-335.

Sportorno, a. E. 1979. Contrastacion de la macrosistematica
de roedores caviomorfos por analisis comparative de la
morfologia reprodutiva masculina. Arch. Biol. Med. Exper.,
12:97-106.

Stacey, P. B., and C. E. Bock. 1978. Social plasticity in the
Acorn Woodpecker. Science, 202:1298-1300.

Staton, M. A. 1978. “Distress calls" of crocodilians-whom
do they benefit? Amer. Nat., 112:327-332.

Svendsen, G. E., and K. B. Armitage. 1973. Mirror-image
stimulation applied to field behavior studies. Ecology,
54:623-627.

Trivers, R. L. 1971. The evolution of reciprocal altruism.
Quart. Rev. Biol., 46:35-37.

. 1972. Parental investment and sexual selection. Pp.

136-179, in Sexual Selection and the Descent of Man, 1871-
1971 (B. Campbell, ed.), Aldine, Chicago.

Vanzolini, P. E. 1970. Zoologia, sistematica, geografia e a
origem das especies. Univ. Sao Paulo Inst. Geogr., Ser.
Monografias e Teses 3.

. 1973. Paleoclimates, relief and species multiplication in

equatorial forests. Pp. 255-258, in Tropical forest ecosys-
tems in Africa and South America: a comparative review
(B. J. Meggers, E. S. Ayensu, and W. D. Duckworth, eds.),
Smithsonian Institution Press, Washington, D.C.

. 1974. Ecological and geographical distribution of lizards

in Pernambuco, northeastern Brazil (Sauria). Papeis Avul-
sos Zool., Sao Paulo, 28:61-90.

. 1976. On the lizards of a cerrado-caatinga contact: evo-
lutionary and zoogeographical implications (Sauria). Papeis
Avulsos Zool., Sao Paulo, 29:111-119.

Vanzolini, P. E., and E. E. Williams. 1970. South American
anoles: geographic differentiation and evolution of Anolis
chrysolepis species group (Sauria, Iguanidae). Arq. Zool.,
Sao Paulo, 19:1-298.

Vasconcelos SoBRiNHO, J. 1971. As regioes naturals do Nord-
este, o meio e a civilizagao. Condepe, Recife, Brazil.

Vaughan, T. A. 1972. Mammalogy. Saunders, Philadelphia.

Vuilleumier, B. S. 1971. Pleistocene changes in the fauna and
flora of South America. Science, 173:771-780.

Walker, E. 1974. Mammals of the World, Vol. III. Johns Hop-
kins Press, Baltimore.

Wilson, E. O. 1975. Sociobiology, the new synthesis. Belknap,
Cambridge, Massachusetts.

Wilson, S. C., and D. G. Kleiman. 1974. Eliciting play: a
comparative study. Amer. Zool., 14:341-370.

1981

LACHER— CAVIID SOCIAL BEHAVIOR

69

APPENDIX I

Botanical survey of the Fazenda Batente study site. Plant identifications done by Drs. Dardano de Andrade Lima and Marcelo de
Ataide Silva of the Instituto de Pesquisas Agrondmicas in Recife, Pernambuco, Brazil-

Family

Scientific name

Common name

Acanthaceae

Jacobina sp.

Canela de nambu

Ruellia asperula

Camara branco

Ruellia paniculata
Ruellia bahiensis

Camara

Ruellia sp.

Camara de boi

Amaranthaceae

Amaranthus spinosus

Bredo de porco

Anacardiaceae

Astronium immdeuva

Aroeira

Boraginaceae

Cordia globosa

Maria preta

Cordia insignis

Orelha de Leao

Bromeliaceae

Encholyrium. spectabile

Croata

Byttneriaceae

Wakheria sp.

Malvinha

Cactaceae

Cereus jamacaru

Mandacaru

Pilosocereus goimellei

Xique-xique

Pilosocereus piauhyensis

Facheiro

Arrojadoa rodantha

Rabo de raposa

Opimtia palmadora

Palma

Capparaceae

Crataeva tapia

Trapia

Commelinaceae

Commelina sp.

Cipd

Compositae

Centratherum punctatum

Peripeta

Blainvillea rhoinboidea

Bamburra

Convolvulaceae

Ipomoea sobrevoluta

Ipomoea sp.

Gitirana

Cucurbitaceae

Cayaponia tayuya

Tajuja

Mamordica charantia

Melao de Sao Caetano

Cucumis anguria

Maxixe

Cyperaceae

Cy penis sp.

Capim barba de bode

Erythoroxylaceae

Erythroxyhun sp.

Pau vidro

Euphorbiaceae

Croton jacobinensis

Marmeleiro

Croton campestris

Velame

Croton argyrphylloides

Casatinga

C nidoscolus urens

Cansacao

Flacoutiaceae

Prockia crucis

Gramineae

Rhyncheiytrmn repens

Capim rosado

Panicum sp.

Capim touceira

Aristida sp.

Capim mimoso

Cenchrus echinatus

Capim carrapicho

Brachiaris mutica

Capim de planta

Guttiferae

Vismia guianensis

Lacre

Labiatae

Leonotis nepetaefolia

Cordao de Sao Francisco

Hyptis pectinata

Bamburra

Hyptis sp.

Camara branco

Leguminosae

Piptadenia zehntneri

Angico brabo

Piptadenia sp.

Espinheiro Preto

Cassia excelsa

Canafistula

Pterogyne nitens

Madeira Nova

Erythrina velutina

Mulungu

indigofera suffruticosa

Anil de bode

Microptilium longepedunculatum

Orelha do Prea

70

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 17

APPENDIX I — Continued

Family

Scientific name

Common name

Phaseohis semierectus

Orelha do Prea

Phaseolus peduncidaris

Feijao de rolinha

Malvaceae

Gaya sp.

Melosa

Bogenhardia tiubae

Melosa

Sida panicidata

Malva preta

Sida galheirensis

Malvinha

Meliaceae

Cedrela sp.

Cedro

Portulacaceae

Portulaca elatior

Bedoegua de ovelha

Rhamnaceae

Rhamnidium sp.

Ziziphus joazeiro

Jua

Rubiaceae

Mitracarpus sp.

Peripeta

Sapindaceae

Sapindus saponaria

Sabonete

Talisia esculenta

Pitombeira

Cardiospenniun halicacabiun

Chocalho de vaqueiro

Serjania caracasana

Cipo saltocora

Serjania sp.

Cipo folha de came

Scrophulariaceae

Scoparia dulcis

Vassourinha

Solanaceae

Solanum panicidatiim

Jurubeba

Solanum aniehcanum

Erva moura

Lycopersicon escidentum

Tomate brabo

Vitaceae

Cissus simsiana

Parreira

Cissus sicyoides

Cipo mole

APPENDIX II

Protocols of Reproductive Behavior for Kerodon and Gtlea

Galea Protocol

24 March 1978-10:05 pm: The dominant male (MR) began fol-
lowing female B3, attempting to place his chin on the female’s
rump. B3 gave a tail-up, spraying urine on MR. MR shook his
head back and forth, and continued following B3. The number
two male, BM, twice tried to join the chase, once taking the
lead, but both times was aggressively chased by MR.

MR then directed aggressive lunges toward three males; BM,
BF. and FMR. MR resumed the chase, following B3 through the
rock piles, always attempting to place his chin on her rump. MR
would occasionally attempt to bite the rump of B3, in an appar-
ent attempt to halt her flight. During the chase B3 passed female
M, and as MR approached M, she gave a tail-up and in quick
succession twice urinated on MR. MR retreated and groomed.

MR resumed the chin-rump follow on B3, and as he passed close
to M, she gave a tail-up and MR avoided her. MR then walked
towards the middle of the room using short, deliberate strides,
pausing as he strode, then lunged at FMR. A contagious chorus
of peepy squeaks had overtaken the colony. MR lunged again
at FMR, then sat down in the sandbathing arena. MR rose and
lunged at BF, then at BM; then began chin-rump following B3
again. As before, at every opportunity MR attempted to chin the
rump of B3. Other males were obviously excited, but MR re-

stricted them from entering the chase. MR again attempted to
bite the rump of B3.

Male BF, excited by the mating chase, began to follow female
B13. MR immediately stopped following B3 and lunged at and
bit BF. MR then lunged at B13. MR directed a series of espe-
cially violent lunges at BF. BF then lunged at juvenile male
B2M.

MR began a chin-rump follow, this time on B13, attempting to
mount after placing his chin on her rump. Each time MR placed
his chin on the rump of B13, she paused, and MR attempted to
mount. B13 then moved away, and MR reinitiated the follow.

The other colony males were obviously excited, but could not
approach B13. MR lunged at BM, then at BF. MR relinquished
the chase, went up on the rock pile, and began to bark. The
activity level rapidly declined. MR groomed himself, and con-
tinued barking. B13 began to answer the low bark of MR with
a higher-pitched, peepy bark. Male FMR attempted to approach
B13, and was attacked by MR.

Kerodon Protocols

A) 28 September 1977-10:35 am: Female F came out of the rear
rocks, and the dominant male FR ran back to her and sniffed

1981

LACHER— CAVIID SOCIAL BEHAVIOR

71

the vaginal area. They separated. At 10:46, FR entered the
rocks where F was sitting; a few seconds later she emerged
and FR followed with his chin on her rump. She re-entered
the rear rocks, and FR sat on top, face-wiped twice, and
then returned to the central rock pile.

At 10:55, the number two male, R, descended from his perch
in a tree and entered female F's rocks. After a delay of a few
seconds, she emerged and circled back into the rocks, with
R following. As they entered the rocks, R attempted to
mount F. FR almost immediately ran back and chased R
from the presence of F.

B) 28 September 1977-3:40 pm: Male R approached female F
and sniffed the vagina. He then attempted to mount and she
resisted. FR immediately arrived and chased R. R climbed
one of the trees in the colony room, and FR climbed another.
FR descended, ran over to the tree in which R was perched,
and chased R from the tree. R then ran over to female F and
attempted to mount a number of times. The dominant male
then aggressively chased female F from the area. FR then
began to follow' F, always attempting to mount. He then
mounted successfully, with intromission, thrusting five or six
times. The number two male, R, then approached and lunged
at FR, who was still mounting female F. FR turned on R,
who fled.

FR then sat behind F, who exposed her perineum to FR,
apparently receptive. FR then began to follow F, attempting
to mount, but the female continually moved away.

R again approached the female, and gave a series of twisting
lunges, attempting to separate the male from the female. FR,
however, would not leave F.

At 3:50 PM, the activity level declined. R climbed a colony
tree, and after a short delay, FR climbed the same tree and
began a series of three prolonged aggressive chases on R.
After the last chase, R fled at the mere approach of FR.

C) 3 October 1977-3:58 pm: Dominant male FR sat at female
B’s left side and placed his chin on her back. FR then circled,
again placed his chin on the female’s back, circled again and
gave a naso-anal. Female B began to move away and FR
followed. Subadult male J2 immediately joined in, and male
BR also began to follow after about 30 seconds. The follow
deteriorated into an all out chase, with B fleeing and all three
males pursuing in an odd hopping, “stotting” gait. FR was
generally in front, although J2 at times took the lead. When-
ever the leader w'as sufficiently close to B, he would place
his chin on the rump. The leading male would often try to

hop on the fleeing female and mount. The chase then dimin-
ished in intensity and ended at 4:00 pm. at 4:02 FR ap-
proached female B, gave a naso-anal, and attempted to
mount. B then moved away from FR and J2 attempted to
mount. B lunged at J2, who fled. Subadult male BR then
approached the dominant male and gave a naso-anal and at-
tempted to mount. FR merely moved away.

At 4:07, an all out follow-chase began again, initiated by FR
sniffing the back and rump of B. Was chased by FR, J2, and
BR; with the latter attempting to mount FR. B had little
trouble in avoiding all three males. The follow-chase stopped
briefly, then began again at 4:18; same participants and same
behavior. BR again attempted to mount FR.

FR continued to chase B, who ran a bit, then stopped and
presented a tail-up with urine-squirting. FR stopped, shook
his head to and fro, and face-wiped. After a delay of about
10 minutes, FR and J2 resumed the chase on B. B again
responded with a series of tail-ups. J2, however, managed to
mount, and achieved intromission, but there was no thrust-
ing.

Short chase on B by all three males again took place at 4:36,
4:43, and 4:45. During the third chase, FR seemed confused
and began to follow a non-estrous female, F. As soon as he
got sufficiently close, however, he turned away. At one point
while being followed, B turned and lunged at BR. FR and
BR continued the pursuit. The female actually had no diffi-
culty in avoiding the pursuing males. B climbed a tree and
the males temporarily desisted. At 4:50 B descended, and
FR again initiated the chase. BR attempted to mount FR,
who turned on BR in an upright aggressive posture. J2 then
began to follow B, and BR attempted to mount him also. At
4:53 this last chase deteriorated, and the activity level quick-
ly decreased.

D) 2 December 1977-7:30 am: J2 repeatedly attempted to mount
subadult female JR. At J2's approach, JR would assume a
submit posture; J2 then giving a naso-anal. J2 would place
his chin on the rump of JR and attempt to mount. JR would
pull away, J2 riding the female for a short distance. After a
series of attempted mounts by J2, BR approached and chased
J2, then began to follow JR. The two males then followed
alternately, with a high frequency of circling. The male
would approach the female and give either a naso-anal or a
nose-nose, then would begin circling. While being followed,
JR would avoid the males, but she remained motionless while
being circled; either standing, sitting or in the submit pos-
ture.

Copies of the following of Carnegie Museum of Natural History may be obtained at the prices

listed from the Publications Secretary, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pitts-
burgh, Pennsylvania 15213.

1. Krishtalka, L. 1976. Early Tertiary Adapisoricidae and Erinaceidae (Mammalia, Insectivora) of North

America. 40 pp., 13 figs $2.50

2. Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the

late Pleistocene paleoecology of the central Appalachian Mountains of Virginia. 88 pp., 21
figs $12.00

3. Wetzel, R. M. 1977. The Chacoan peccary, Catagonus wagneri (Rusconi). 36 pp., 10 figs. . . $6.00

4. Coombs, M. C. 1978. Reevaluation of early Miocene North American Moropus (Perissodactyla,

Chalicotheriidae, Schizotheriinae). 62 pp., 28 figs $5.00

5. Clench, M. H., and R. C. Leberman. 1978. Weights of 151 species of Pennsylvania birds analyzed

by month, age, and sex. 87 pp $5.00

6. Schlitter, D. A. (ed.). 1978. Ecology and taxonomy of African small mammals. 214 pp., 48 figs.

$15.00

7. Raikow, R.J. 1978. Appendicular myology and relationships of the New World nine-primaried oscines

(AvesiPasseriformes). 43 pp., 10 figs $3.50

8. Berman, D. S, and J. S. McIntosh. 1978. Skull and relationships of the Upper Jurassic sauropod

Apatosaurus (Reptilia, Saurischia). 35 pp., 11 figs $3.00

9. Setoguchi, T. 1978. Paleontology and geology of the Badwater Creek area, central Wyoming. Part

16. The Cedar Ridge local fauna (Late Oligocene). 61 pp., 30 figs $4.50

10. Williams, D. E. 1978. Systematics and ecogeographic variation of the Apache pocket mouse (Roden-

tia: Heteromyidae). 57 pp., 23 figs $4.00

11. Guilday, J. E., H. W. Hamilton, E. Anderson, and P. W. Parmalee. 1978. The Baker Bluff Cave

deposit, Tennessee, and the Late Pleistocene faunal gradient. 67 pp., 16 figs $5.00

12. Swanepoel, P., and H. H. Genoways. 1978. Revision of the Antillean bats of the genus Brachyphylla

(Mammalia; Phyllostomatidae). 53 pp., 17 figs $4.00

13. Schwartz, J. H., and H. B. Rollins, eds. 1979. Models and methodologies in evolutionary theory.

105 pp., 36 figs $6.00

14. Engesser, B. 1979. Relationships of some insectivores and rodents from the Miocene of North Amer-
ica and Europe. 68 pp., 12 figs., 20 plates $5.00

15. Bentz, G. D. 1979. The appendicular myology and phylogenetic relationships of the Ploceidae

and Estrildidae (Aves: Passeriformes). 25 pp., 5 figs $2.00

16. Mares, M. A. 1980. Convergent evolution among desert rodents: a global perspective. 51 pp., 25

figs $3.50

BULLETIN

0/ CARNEGIE MUSEUM OF NATURAL HISTORY

I ■ ■ ^U. -

I ANNOTATED CATALOGUE OF THE DINOSAURS

(REPTILIA, ARCHOSAURIA) IN THE COLLECTIONS OF
CARNEGIE MUSEUM OF NATURAL HISTORY

JOHN s. McIntosh

PITTSBURGH, 1981

& NUMBER 18

s

BULLETIN

of CARNEGIE MUSEUM OF NATURAL HISTORY

ANNOTATED CATALOGUE OF THE DINOSAURS (REPTILIA,
ARCHOSAURIA) IN THE COLLECTIONS OF
CARNEGIE MUSEUM OF NATURAL HISTORY

JOHN s. McIntosh

Research Associate, Section of Vertebrate Fossils;
Department of Physics,

Wesleyan University, Middletown, Connecticut 06457

NUMBER 18

PITTSBURGH, 1981

BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY
Number 18, pages 1-67, 22 figures

Issued 16 October 1981

Price: $6.00 a copy

Craig C. Black, Director

Editorial Staff; Hugh H. Genoways, Editor; Duane A. Schlitter,
Associated Editor: Stephen L. Williams, Associate Editor;
Barbara A. McCabe, Technical Assistant.

© 1981 by the Trustees of Carnegie Institute, all rights reserved.

CARNEGIE MUSEUM OF NATURAL HISTORY, 4400 FORBES AVENUE
PITTSBURGH, PENNSYLVANIA 15213

CONTENTS

Foreword 5

Order Saurischia

Suborder Theropoda 7

Suborder Prosauropoda II

Suborder Sauropoda II

Order Ornithischia

Suborder Ornithopoda 30

Suborder Stegosauria 35

Suborder Ankylosauria 38

Suborder Ceratopsia 40

Literature Cited 43

Appendix 1 45

Appendix 2 65

mm

a

,.’i'

■=

I tJ

J

I

I

4

FOREWORD

This catalogue of the dinosaur specimens of the
Carnegie Museum of Natural History employs a
format similar to those used by Woodward (1889,
1891, 1895, 1901) and Lydekker (1890) for the fish,
amphibians, and reptiles of the British Museum
(Natural History). The specimens are arranged sys-
tematically from order down to species. Under each
species the type specimen is listed first if it occurs
in the Carnegie Museum of Natural History collec-
tion, followed by the more important articulated
specimens, and then by the individual elements,
with skull elements first, followed by vertebrae,
forelimb bones, and hind limb bones. With one ex-
ception accession and field numbers are omitted
from the text but are given in the appendix, where
all the specimens are listed serially by their cata-
logue numbers. Only field numbers for specimens
from Dinosaur National Monument are given and
are indicated in the text by the abbreviation DNM,
as, for example, DNM 130. These field numbers
were assigned in serial order by Earl Douglass, who
directed the excavation of the quarry, to what he
considered single individuals. Individual bones of
a specimen or blocks containing bones of a single
specimen were assigned subnumbers, for example
DNM 130/2 or DNM 130/B. Inevitably some series
of bones thought to belong to a single individual
proved upon preparation to belong to several ani-
mals, or even several genera. Less often prepara-
tion has shown that specimens assigned two field
numbers, for example DNM 240 and part of DNM
270, belong to a single individual. Despite these
drawbacks, Douglass’ field numbers have proved
useful. Citations within the text that refer to a par-
ticular specimen are expanded to include the pages
and figures pertaining directly to that specimen.

Specimens originally assigned Carnegie Museum
of Natural History catalogue numbers but subse-
quently transferred to other institutions, are also
listed, because many were described and figured as
specimens of the former and because of the impor-
tance of knowing their present location. Catalogued
casts are also included, because some of these may
be useful for study; for example, the Struthiosaurus
casts are perhaps the only representative materials
of this genus in America. Unprepared specimens
and those transferred to other institutions before
preparation and cataloguing are not included. In a
few cases the same catalogue number had been

used for two specimens, usually for a dinosaur and
for a non-dinosaur. In these instances the dinosaur
specimens have been recatalogued, but the original
numbers are also given.

It is beyond the scope of this catalogue to provide
a revision of species or a novel classification. In
general, the classification used here is a slightly sim-
plified version of that found in Romer’s (1966) third
edition of Vertebrate Paleontology, it is updated by
recent studies and by employing a modified classi-
fication of the Sauropoda. Troublesome is the ques-
tion of validity of many of the congeneric species
from the Morrison Formation which have been
based on very incomplete skeletons. There are cer-
tainly two or more valid species of Stegosaurus,
probably at least two of Camarasaurus, and, al-
though the evidence is not yet unequivocal, prob-
ably several of Diplodocus. Determination of the
validity of other congeneric species from the Mor-
rison, however, must await discovery and prepa-
ration of much more material. Where species are
clearly identical, as in the case of Haplocanthosau-
rus priscus and H. utterhacki, they are synony-
mized. Where it is not known whether minor dif-
ferences between congeneric species warrant spe-
cific separation, as with Apatosaurus excelsus and
A. louisae, the conservative position of retaining
both species is taken. Isolated sauropod elements
can seldom be identified as to species. As not more
than one species per genus has been recognized in
the material from Dinosaur National Monument the
isolated specimens belonging to each have been list-
ed after those which are identifiable, but in Appen-
dix 1 they appear as, for example, Camarasaurus
sp. In the case of Camptosaurus, where the type
specimens of four of the six described species have
come from a single quarry — a number that is almost
certainly too high — Gilmore’s ( 19256) assignment of
the Carnegie Museum of Natural History specimens
of this genus to C. medius is accepted, pending a
revision of the species by Dodson. Where possible
individual bones are given a generic assignment.
This has undoubtedly led to some errors, particu-
larly with poorly preserved specimens, but the ad-
vantages of attempting an identification appear to
outweigh the disadvantages. In some cases it was
not possible to provide more than a family desig-
nation for isolated elements, particularly vertebrae,
limb, or foot bones of hadrosaurids.

5

6

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

-Carnegie Museum dinosaur quarries. Sheep Creek, Wyoming.

1981

McIntosh— DINOSAURS of carnegie museum

7

Thanks are due to C. Black, M. Dawson, and D. Berman for
making this work possible and overall encouragement. J. Os-
trom, M. Brett-Surman, J. Horner, and P. Dodson were con-
sulted concerning several hadrosaur specimens and D. Weis-
hampe! reviewed critically the entire section on the ornithopods.
L. Plowman retouched photos in Figs. 3, 10, 16, 17, and 20.
Their help is gratefully acknowledged. Most of all my thanks are
offered to Lee Schiffer who painstakingly cross-checked the
main text and appendix for inconsistencies and provided invalu-
able aid in delving into the records; lastly to Elizabeth Hill, v/ho
typed the manuscript, aided with the illustrations, and in count-
less other matters.

Abbreviations used to refer to collections or localities:
AMNH, American Museum of Natural History; BM, British
Museum (Natural History); CM, Carnegie Museum of Natural
History; CMNH, Cleveland Museum of Natural History;
DMNH, Denver Museum of Natural History; DNM, Dinosaur
National Monument; DU, Duquesne University; GSC, Geolog-
ical Survey of Canada; ROM, Royal Ontario Museum; UM,
University of Michigan; UNHSM, Utah Natural History State
Museum; USNM, National Museum of Natural History; UU,
University of Utah; YPM, Yale Peabody Museum.

ORDER SAURISCHIA
SUBORDER THEROPODA

Family Teratosauridae

Teratosaurus and its allies are known from only
very fragmentary material, and their position has
been the subject of a great deal of controversy.
From its traditional assignment as a primitive the-
ropod some have argued that it is really a carnivo-
rous prosauropod, while others have questioned
whether it is a saurischian at all. For purposes of
this catalogue the conservative position is accepted.

Genus Teratosaurus Meyer, 1861
Teratosaurus suevicus Meyer, 1861

358 A tooth from the Keuper of Aixheim, Ger-
many; obtained by exchange in 1901.
Collected by E. Fraas

Fa.mily Megalosauridae
Genus Allosaurus Marsh, 1877

Syn.: (?) Antrodemus Leidy, 1870.

Although it is not unlikely that the half caudal
centrum upon which Leidy founded Antrodemus
belongs to this animal, the realization in recent
years that a number of large theropods were present
in the Morrison Formation makes positive verifi-
cation of the synonymy difficult. Most writers today
have returned to the use of the name Allosaurus —
the position adopted here.

Allosaurus fragilis Marsh, 1877

1 1844 Cranium, mandible, and the greater portion
of the skeleton lacking the forelimbs and a
few other bones; from the Morrison Forma-
tion of Carnegie Museum Quarry at Dinosaur
National Monument, north of Jensen, Uintah
County, Utah (DNM 202). The partial skull,
cervicals, scapula-coracoid, and some ribs

were originally catalogued CM 1 1868 but
they are part of this skeleton. The specimen
was mounted and placed on exhibition in
1938. The right fibula was supplied from
DNM 171 found near the skeleton and likely
belonging to it. The incomplete skull was re-
placed in the mount by a cast of one from
the same quarry, UU 6000. The forelimbs
were cast from USNM 4734. Pictured by
Kay — Carnegie Mag. (1940c), 13:303-304;
mentioned by Stovall and Langston — Amer.
Midland Nat. (1950), 43:713.

Collected by Douglass et al., 1913-
1915

The following specimens all came from the Mor-
rison Formation at the Carnegie Museum Quarry at

Dinosaur National Monument near Jensen, Utah,

and were collected by Douglass et al., 1909-1923.

21703 Cranium, presacrals, caudals, ilium, ischium
(DNM 22).

1 1843 Cranium (only partially prepared), several
centra, ribs, 1 coracoid, and other parts
perhaps belonging to a young individual of
this genus (DNM 366).

3387 Teeth and fragments (DNM 2).

3382 A tooth (DNM 14).

3383 Small vertebrae etc., perhaps belonging to
this form (DNM 187).

33965 Two anterior caudal centra, four spines,
three arches (DNM 120/C).

38341 Caudal and claw (marked DNM 130, proba-
bly really 102).

21705 Caudal centrum (DNM 193/A).

33957 Two caudals (with DNM 232).

21757 Three caudals.

33901 Several vertebrae.

8

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

33903 Gastralia perhaps belonging to this form (as-
sociated with DNM 210).

21713 Left ischium and a metatarsal and other ma-
terial not yet prepared (DNM 302, part).

21726 Left femur (DNM 39/60L).

21769 Distal end of right femur (DNM 120/N).

37004 Distal end of metatarsal (DNM 270/26).

38349 Several incomplete metatarsals (?DNM 197,
part).

10002 Proximal end right tibia and fibula from the
Morrison Formation, east of Nielsen Gulch,
northeast of Carnegie Museum Quarry at
Dinosaur National Monument and not far
from the latter (DNM 165). The horizon is
the same as that of the main quarry. The fib-
ula was incorporated into the mounted skel-
eton CM 11844.

Collected by Douglass et al., 1913

82 Anterior caudal centrum from the Morrison
Formation of Carbon County, Wyoming. It
was brought back to Pittsburgh in 1899 by
W. J. Holland from his first trip to the West
to obtain dinosaurs for the museum. It is of
historical importance as the first dinosaur
specimen (aside from a cast) to be entered
into the museum catalogue. A small section
of a sauropod rib found with the centrum
bears the same catalogue number.

Collected by Holland, 1899

21736 A large right scapula-coracoid probably be-
longing to this species. The data concerning
it has been lost, but it may have come from
the Morrison Formation at Dinosaur Nation-
al Monument, despite its somewhat unusual
coloration.

1254 Both ischia to which has been added a right
premaxilla, two teeth, two sacral vertebrae,
left humerus, four metatarsals, and several
phalanges. These are part of a skeleton of a
small individual, some bones of which were
formerly catalogued as part of the “catchall”
no. 1255 (p. 15) from Quarry N, Freezeout
Hills, Carbon County, Wyoming. Many ver-
tebrae and other parts of the skeleton were
poorly preserved and were discarded.

Collected by Gilmore, 1902

2045 A femur from the Morrison Formation of
Wilson Creek, Colorado.

Collected by Utterback, 1901

36037 Caudal. Quarry B, Red Fork of the Powder
River, Johnson County, Wyoming.

Collected by Utterback, 1903

Genus Torvosaurus Galton and Jensen, 1979
Torvosaurus tanneri Galton and Jensen, 1979

14955 Cast of a very large claw (original BYU 2020)
from the Morrison Formation at Lily Park,
Moffat County, Colorado. Figured by Galton
and Jensen — Brigham Young Univ. Geol.
Studies (1979), 26(1): Fig. IM.

Presented by James Jensen, 1963

Genus Ceratosaurus Marsh, 1884
Ceratosaurus nasicornis Marsh, 1884

21706 Incomplete dentary from the Morrison For-
mation at the Marsh-Felch Quarry, Garden
Park, north of Canon City, Fremont County,
Colorado.

Collected by Utterback, 1901

Small Theropod Genus

Under study by J. Madsen

21704 Cranium, cervicals 1-6, five articulated dor-
sals. The remainder of this specimen was
transferred to the Royal Ontario Museum. It
came from the Morrison Formation on the
west side of Nielsen Gulch northeast of the
Carnegie Museum Quarry at Dinosaur Na-
tional Monument, Uintah County, Utah.

Collected by Douglass et al.

Family Tyrannosauridae
Genus Tyrannosaurus Osborn, 1905
Tyrannosaurus rex Osborn, 1905

9380 Skeleton consisting of partial skull and man-
dible, one cervical, seven dorsals, five sa-
crals, three abdominal ribs, right scapula,
left humerus, ilia, pubes, ischia, left femur
and part of right, tibia, right metatarsals II
and III, left metatarsal IV. Type specimen.
It was collected in the Lance Formation at
Hell Creek, Garfield County, Montana, and
was described and figured by Osborn — Bull.
Amer. Mus. Nat. Hist. ( 1905), 21:262-263,
Fig. 1; Bull. Amer. Mus. Nat. Hist. (1906),
22:281-296, Figs. 1-12; Bull. Amer. Mus.
Nat. Hist. (1917), 35:762-771, Figs. 19-21;
Mem. Amer. Mus. Nat. Hist., n.s. (1912),
LFigs. 20-24. This skeleton (formerly
AMNH 973), obtained from the American
Museum of Natural History, is mounted and
was placed on exhibition in 1942.

Collected by B. Brown, 1902-1903

9379 Brain cast of AMNH 5029 collected in the

1981

McIntosh— DINOSAURS of carnegie museum

9

Fig. 2. — Skeleton of Tyrannosaurus rex, CM 9380 (formerly AMNH 973): missing parts restored in outline except neck and first dorsal
vertebra which are drawn from AMNH 5866, now mounted and on exhibition in the British Museum (Nat. Hist.), after Osborn.

Lance Formation (Hell Creek beds) on the
west side of Big Dry Creek, 44 mi south of
Glasgow, Garfield County, Montana. This
cast was also obtained from the American
Museum of Natural History and was figured
by Osborn — Mem. Amer. Mus. Nat. Hist.,
n.s. (1912), 1:13-24, Pis. 3-4 and Figs. 16-
17.

Original collected by Brown and Ka-
isen, 1908

1400 Part of the skull including the maxilla, cra-
nium and both lower jaws, two dorsals, sev-
en caudals, ribs, chevrons, part of pubis, il-
ium, femur, and other bones. It was obtained
from the Lance Formation on Snyder Creek,
Niobrara County, Wyoming. The upper jaw
is on exhibition.

Collected by Peterson, 1902

244 A phalanx from the Lance Formation on
Lance Creek, Niobrara County, Wyoming.

Collected by Hatcher, 1900

Tyrannosauridae, indeterminate

30749 Thirteen teeth of varying size belonging to
more than one individual from the Lance
Formation, on Sheep Mountain, Carter

County, Montana. The smaller teeth may
belong to a different genus.

Collected by Kay, 1938

12102 Left femur. Same data as above.

Collected by Kay, 1937

9401 Right lachrymal bone from the Judith River
beds on Willow Creek, three miles east of
the Nolan Archer Ranch, Fergus County,
Montana. It was originally catalogued CM
963 but altered because this number had also
been assigned to a specimen of Deinosu-
chus.

Collected by Hatcher, 1903

Family Compsognathidae
Genus Compsognathus Wagner, 186!
Compsognathus longipes Wagner, 1861

53 Cast of an essentially complete skull and
skeleton from the Kimeridge Clay (Solnho-
fen Shale) of Solnhofen, Bavaria. The origi-
nal is the type specimen and is in the Bavar-
ian Museum in Munich. It was described and
figured by Wagner — Abh. Bayer Akad.
Wiss. (1861), 9:94-102, PI. 3.

Original collected bv Oberndorfer

Fig. 3. — Pelvis of Haplocanihoscumis prisciis, CM 572; a) anterior view; b) lateral view; c) posterior view.

Family Coeluridae
Genus Stenonychosaurus Sternberg, 1932
Stenonychosaurus inequalis Sternberg, 1932

30748 Left femur, ?right femur, tibia, half a hu-
merus, ulna perhaps not belonging to a single

individual and doubtfully referred to this
form. It came from the Lance Formation on
Sheep Mountain, Carter County, Montana.
Collected by Kay, 1938

1981

McIntosh— DINOSAURS of carnegie museum

1 1

Family Ornithomimidae
Genus Ornithomimus Marsh, 1890
Ornithomimus sp.

593 Fragmentary metapodiai from the “Belly
River” Formation near Havre, Montana.
Collected by Douglass, 1902
38322 Four claws of the manus from the Lance
Formation of Sheep Mountain, Carter Coun-
ty, Montana.

Theropoda, insertae sedis

38326 Right femur from the Lance Formation at

Sheep Mountain, Carter County, Montana.
Collected by Kay, 1938
38323 Two claws. Same data as above.

Collected by Kay, 1938

SUBORDER PROSAUROPODA

Family Plateosauridae
Gtxm% Plateosaurus H. von Meyer, 1837
Piateosaurus sp.

11908 Cast of a complete right pes from the Upper
Trias of Germany.

Purchased in 1933

SUBORDER SAUROPODA

Family Cetiosauridae

Genus Haplocanthosaurus Hatcher, 1903

Syn.: Haplocanthus Hatcher (preoccupied), 1903.

Haplocanthosaurus priscus Hatcher, 1903

Syn.: Haplocanthus priscus Hatcher, 1903.

Haplocanthosaurus utterbacki Hatcher, 1903.

572 Partial skeleton consisting of the last two
cervicais, 10 dorsals, five sacrals, caudals 1-
19, many ribs, two chevrons, ilia, pubes, is-
chia, left femur. Type specimen. This skel-
eton is from the Morrison Formation of the
Marsh-Felch Quarry No. 1 at Garden Park,
north of Canon City, Fremont County, Col-
orado. It was described briefly by Hatcher
as Haplocanthus priscus — Proc. Biol. Soc.
Washington (1903&), 16:1-2 — but altered to
Haplocanthosaurus — Proc. Biol. Soc.
Washington (1903o), 16: 100. He fully de-
scribed and figured all parts of it in Mem.
Carnegie Mus. ( 1903J), 2: 1-27, Figs. 3-4, 7-
14, Pis. 1, 2, 4, 5.

Collected by Utterback, 1901

879 Partial skeleton consisting of 10 cervicais, 13
dorsals, five sacrals, caudals 1-7, five or six
ribs, left scapula, right coracoid. Type spec-
imen of Haplocanthosaurus utterbacki .

Found a few feet from the above skeleton.
It was described and completely figured by
Hatcher — Mem. Carnegie Mus. (1903J),
2:27-43, Figs. 15-20, PI. 2.

Collected by Utterback, 1901
33995 Left scapula-coracoid . These well preserved
bones are marked 94, the catalogue number
of the paratype of Diplodocus carnegii from
Sheep Creek Quarry D. This is an obvious
error because ( 1) both scapulae of 94 are ac-
counted for and this one is much smaller,
and field records indicate that there was no
“extra” scapula found, and (2) this bone is
clearly not that of Diplodocus. It resembles
that of Haplocanthosaurus CM 879 closely.
Not unlikely is it the “lost” scapula-cora-
coid which field records indicate was found
with CM 572.

2043 Right tibia, fibula, and astragalus. Same data
as the above. This specimen probably be-
longs to this form.

Collected by Utterback, 1901
2046 Left tibia and fibula. Same data as the above.
It probably belongs to this form.

Collected by Utterback, 1901
36034 A median caudal exhibiting the large chevron
facets of this genus from Quarry B, Red Fork

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

Fig. 4. — Skull and mandible of Canuinisaurus lentiis, CM 1 1338,
right lateral view.

of the Powder River, Johnson County, Wy-
oming (originally part of CM 1256). May be-
long to Haplocanthosamus .

Collected by Utterback, 1903

Eamily Camarasauridae
Genus Camarasaurus Cope, 1877

Syn.: Morosaurus Marsh, 1878.

Uintasaiirns Holland, 1919.

Camarasaurus lentus (Marsh) 1889

Syn.: Morosaurus lentus Marsh, 1889.

Uintasaiirns douglassi Holland, 1919.

Caniarasaiirus annae Ellinger, 1950.

This species has never been satisfactorily sepa-
rated from the gigantic C. siiprenms Cope, type
species of the genus.

11338 Skull and skeleton, articulated and complete
except for caudal centra 9, 10, 11, left ilium,
left ischium, and some of the ribs on the left
side; it is the most perfect sauropod skeleton
ever found. This juvenile animal is from the
Morrison Eormation of the Carnegie Mu-
seum Quarry at Dinosaur National Monu-
ment, north of Jensen, Uintah County, Utah
(DNM 333). It was described by Gilmore —
Mem. Carnegie Mus. (1925u), 10:347-384,
and figured as follows: skull — Eigs. 1-4, PI.
16, skeleton — Pis. 14, 15, 17, sternal plate —
Pig. 5. Pictured by Kay — Carnegie Mag.
(1951), 25:91. It is mounted and was placed
on exhibition in 1924.

Collected by Douglass et al., 1919-
1920

11373 Skull and skeleton, articulated and complete
except for the tail (DNM 300 and 301). Same
data as the above. Although much larger
than the above specimen, this one is far from
a full sized animal. It was transferred to the
National Museum of Natural History in
Washington in 1935 (USNM 13786) where it
is mounted and on exhibition. It is pictured
in Glut — The Dinosaur Dictionary (1972), p.
41.

Collected by Douglass et al., 1918-
1919

30743 Cast of the skull and mandible of the above
specimen (DNM 300).

Made in the museum

3379 Articulated tail of 47 caudals complete al-
most to the tip. Same data as the above
(DNM 130, part). It was displayed at the
Dallas Exposition in 1936 and transferred to
the National Museum of Natural History
(USNM 15492) where it was used to com-
plete the mounted Camarasaurus skeleton
(USNM 13786) (see above). Douglass con-
sidered it likely that this specimen was the
detached tail of 210 (see CM 8942 and CM
33916 below).

Collected by Douglass et al., 1912

11393 Disarticulated skeleton of an adult animal
consisting of the partial skull and mandible,
five-i- cervicals, 10+ dorsals, sacrum, 25
caudals, many ribs, several chevrons, scap-
ulae, coracoids, right humerus, right radius,
right ulna, part of manus, ilia, pubes, ischia,
femora, left fibula, calcaneum, two metatar-
sals, three phalanges, two claws (DNM 240
and 270, part). The skull was later recata-
logued CM 12020 (see below). Same data as
the above. Two articulated cervicals have
been transferred to the University of Michi-
gan (V 16995) where they are on display.

Collected by Douglass et al., 1915-
1916

12020 Partial skull and mandible found among the
bones of the above skeleton (DNM 240/L).
It was initially given the catalogue number
of the remainder of the skeleton, 11393, but
was later recatalogued 12020, and a cast of
it was used until recently to complete the
headless mounted skeleton of Apatosaurus
louisae, CM 3018. It is referred to by Gil-

198!

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

13

more — Mem. Carnegie Mus. (193fo), 11:188-
190, Fig. 4.

21751 Right premaxilla, both rami of the mandible,
many teeth of a large individual found in the
same quarry as the above and quite near it,
but in a distinctly higher stratum than the
usual bone layer in which the above speci-
mens were found (DNM 201/A and B).

Collected by Douglass et a!., 1914

11069 Five articulated posterior cervicals (or per-
haps four cervicals and the first dorsal) of a
medium-sized animal (DNM 150B/M). Same
data as the above. Originally thought to be
two large cervicals of a nearby Barosaurus
skeleton, these proved to be five smaller ver-
tebrae, which were assigned by Holland as
the type specimen of a new genus and
species, Uintasaurus douglassi', Holland —
Carnegie Mus. Ann. Rept. (1919), p. 38;
Ann. Carnegie Mus. (1924fl), 15:119-138,
Figs. 1-7, Pis. 10-14. This species was
shown by White to be synonymous with
Camarasaurus ientus — J. Paleo. (1958),
32:482.

Collected by Douglass et al., 1914-
1915

8942 A complete anterior dorsal vertebra from the
same locality as all of the above (DNM 130/2)
described and figured by Ellinger — Amer.
Nat. (1950), 84:225-228, Figs. 1-2. It was
originally cited in a Master’s Thesis of L. L.
W'hite at Duquesne University, 1950 (unpub-
lished). Ellinger made it the type specimen
of a new species, Camarasaurus annae. The
specimen had been transferred to Duquesne
University (DU 1), but was later returned to
the Carnegie Museum of Natural History.
The characters cited by Ellinger in founding
the species are here considered to be due to
individual variation, and the animal is as-
signed to C. ientus. An associated dorsal of
the same size and also complete (DNM 130/
1) no doubt belongs to the same individual
as do several other presacrais. Furthermore,
these dorsals were found only 4 ft from the
above cervicals, CM 11069, and in all prob-
ability belong to the same individual. Finally,
these Camarasaurus vertebrae lay across an
articulated series of Apatosaurus presacral
vertebrae, CM 33916. Many different limb
and girdle elements belonging to several gen-
era were found nearby and assigned the same

Fig. 5. — Two anterior dorsal vertebrae of Camarasaurus an-
nae," CM 8942, anterior view.

field number as the latter, DNM 210, but
some may belong with this animal (see under
CM 33916 below).

Collected by Douglass et al., 1912
The following specimens were all collected by
Douglass et al. in the Morrison Formation at the
Carnegie Museum Quarry at Dinosaur National
Monument, north of Jensen, Uintah County, Utah,
between 1909 and 1922. They probably belong to
this species.

11969 Cranium and part of the neck (the latter un-
prepared) of a large individual (DNM 325).
21732 A maxilla, dentary, quadratojugal and frag-
ments of a poorly preserved skull (DNM
201/C, D, E, and F) associated with, but dis-
tinct from CM 21751.

3381 A large tooth (DNM 3).

21702 Right maxilla without teeth (DNM 39/60a),
juvenile.

36701 Cervical (from DNM 39/65 Ca).

30760 Two and one-half dorsals (found with DNM
40).

33955 Dorsal (DNM 101).

!4

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

21733 Several posterior dorsals articulated with the
sacrum and right ilium. A few centra were
transferred to the Everart Museum in Scran-
ton, Pennsylvania, where they are on display
(DNM 344).

21712 Thirteen articulated caudal vertebrae, the
arches as yet unprepared (DNM 45).

37001 Two neural arches of anterior caudals or sa-
crals (DNM 302, part).

21716 Five caudal centra (marked, apparently in-
correctly, DNM 60).

30769 Two caudals (DNM 70/4).

30771 Caudal arch and spine (with DNM 210), ju-
venile.

33986 A caudal (DNM 39, part).

33909 A caudal (DNM 232, part).

38342 Caudal (DNM 102, part).

33925 Right scapula and coracoid (DNM 210/2 and

6).

33929 Right scapula, three cervicals (DNM 270/1).

38320 Large left humerus, more robust than usual,
but probably too slender for Apatosaurus
(DNM 39/13).

21781 Left humerus (DNM 45/21), perhaps same
individual as tail CM 21712.

33972 Right forelimb and foot (DNM 120, part).

33963 Right humerus (DNM 210/0), very young an-
imal.

33971 Left humerus (DNM 120/M, part).

33988 Head of right humerus (DNM 210/16). The
remainder of the bone is on exhibit at the
California Academy of Science, San Fran-
cisco, California.

33979 Right humerus (two ends) (DNM 102, part).

38334 Distal end of right humerus (DNM ?39).

33948 Left ulna of a smaller individual (DNM
232/23).

33954 Both ends of a left ulna (in DNM 39/65H).

33949 Right radius (DNM 232/17).

37003 Metacarpal (DNM 232, part) — might belong
to skeleton CM 11393 but found 20 ft away
from it.

21735 Left ilium (DNM 147), juvenile, perhaps be-
long to this form.

33981 Right ilium (DNM 130/10).

21750 Left femur, lacking the head. A part of the
head received the separate catalogue number
11971 (DNM 155/Fe2).

21759 Left femur (small) (DNM 167) (originally cat-
alogued as part of CM 9000).

21772 Left femur of a very small individual (DNM

350/3“), found associated with Stegosaurus
skeleton CM 11341. It perhaps belongs to
this species. An equally small dorsal centrum
and right coracoid found associated may be-
long to the same animal.

33973 Right femur (DNM 210/20) perhaps belong-
ing to CM 1 1069 above. The proximal end of
the bone is at the California Academy of Sci-
ence, San Francisco, California.

21777 Left tibia, astragalus, metatarsal I, claw
(DNM 270, part).

21783 Right tibia (DNM 150/45), found with a skel-
eton of Diplodocus but not belonging to it.
The right fibula of the same limb is in the
Denver Museum of Natural History where
it was transferred with skeleton DNM 150.

33956 Lower half of a small right tibia (with DNM
130/D).

33927 Right tibia (DNM 232/2).

33958 Two claws (DNM 232, part).

Camarasaurus grandis (Marsh, 1877)

Syn.: Apatosaurus grandis Marsh, 1877.
Morosaurus grandis Marsh, 1878.
Morosaurus iinpar Marsh, 1878.

Pleurocoelus inontanus Marsh, 1896.

21760 Casts of bones belonging to the type speci-
men of M. grandis (YPM 1901), paratype
(YPM 1905, 1903), and type specimen of M.
impar (YPM 1900). These were found in the
Morrison Formation intermingled at Quarry
1, Como Bluff, Wyoming. As the separation
of the individuals is speculative, it was con-
sidered better to catalogue them under one
number. The individuals were of about the
same size. Sacrum (YPM 1900), left scapula,
coracoid, humerus, radius, ulna, femur
(YPM 1901), right ilium and ischium, left tib-
ia, fibula, and pes (YPM 1905), left pubis
(YPM 1903). Described by Marsh originally
as Apatosaurus grandis — Amer. J. Sci.
(1877), 14:515, and referred to Morosaurus,
Amer. J. Sci. (1878), 16:412-413, Pis. 6-7.
These figures were reproduced by Marsh in
later articles and monographs. The individ-
ual elements are figured in all views in Os-
trom and McIntosh — Marsh’s Dinosaurs
(1966), Pis. 43, 49, 51, 53, 65, 68, 72, 73, 75,
76, 78, 79, 81-89.

Originals collected by Reed, Carlin,
and Williston, 1877-1878

1981

McIntosh— DINOSAURS of carnegie museum

15

Camarasaurus sp.

113 Left maxilla and postorbital, right dentary,
other skull fragments from the Morrison For-
mation at Quarry C, Sheep Creek, Albany
County, Wyoming.

Collected by Peterson and Gilmore,
1902

584 Two cervicals, eight dorsals, 31 caudals,
many ribs, chevrons, right ilium, pubis, is-
chium, and scapula-coracoid. These beauti-
fully preserved and uncrushed bones came
from the Morrison Formation at Quarry D,
Sheep Creek, Albany County, Wyoming. It
was thought at one time that a nearly com-
plete neck and also bones of both the fore
and hind limbs which had been catalogued
as CM 555 and 556 were probably part of this
skeleton and plans were made to mount it.
However, further preparation of the neck
showed it to belong to Apatosaurus as do
the limb bones 555 and 556 (see below), Hol-
land— Carnegie Mus. Ann. Rept. (1909), p.
31.

Collected by Peterson and Gilmore,
1900

The following material was all collected at Quarry

N, Freezeout Hills, Carbon County, Wyoming, by

C. W. Gilmore, 1902-1903.

1255 A general catalogue number applied to the
bulk of the material from the Morrison For-
mation of Quarry N, Freezeout Hills, Car-
bon County, Wyoming. At the time of cata-
loguing a small Allosaimis skeleton was
separated as CM 1254, and a Camarasaurus
tail (below) as CM 1252. The remainder of
the collection was given this number. It in-
cluded the bones of two intermingled medi-
um large Camarasaurus skeletons, a small
number of Apatosaurus bones, and a dozen
or so Stegosaurus bones. Recently the latter
have been recatalogued as CM 21737 (below)
and an associated Camarasaurus fibula and
pes as CM 21730 (below). The remainder of
the material (largely Camarasaurus) consists
of a large spatulate tooth, 22 caudals, two
ribs and fragments of two others, 1 1 chev-
rons, four (or five) scapulae, three coracoids,
two humeri, two radii, one ulna, two carpal
bones, five metacarpals, one claw, one ilium,
four pubes, two ischia, two femora, four tib-
iae, two fibulae, three astragali, one calca-

neum, five metatarsals, 19 metapodials, 17
phalanges. Now these have also been reca-
talogued with the aid of the original quarry
diagram. The number 1255 is now applied to
a few elements, principally ribs, which can-
not be referred to one of the two Camara-
saurus skeletons with any degree of certain-
ty. There are also a large spatulate tooth,
several chevrons, a fragmentary pubis, and
some fragments.

1252 Forty largely articulated caudals reaching to
the tip of the tail with a number of accom-
panying chevrons.

21730 Left fibula, astragalus, calcaneum, and com-
plete pes found semi-articulated. This spec-
imen was found near the anterior end of tail
1252 and not improbably belongs to the same
individual.

36663 Both scapulae and coracoids, right humerus,
right ilium, pubes, ischia, left femur-tibia-
fibula-astragalus-calcaneum.

36664 Both scapulae and coracoids, left humerus
and radius, left femur, both tibiae, right fib-
ula. This skeleton is a little larger than the
last.

36671 Four caudals.

36689 Six caudals.

36693 Four caudals.

36694 Caudal.

36695 Caudal.

36678 Carpal bone perhaps belonging to this genus.

36686 Three metacarpals.

36685 Two phalanges of the left manus.

36666 Incomplete pubis found near tail CM 1252.

36668 Left tibia-metatarsal I-metatarsal IV of a
half grown individual.

36669 Fibula lacking both ends.

36675 Astragalus, three metatarsals, phalanx of
pes.

36677 Phalanx.

36679 Two phalanges.

36684 Seven phalanges perhaps belonging to sev-
eral individuals.

36680 Ungual phalanx of the pes.

1256 This number was formerly assigned to the
bulk of a large amount of material collected
from the Morrison Formation at Quarry B,
Red Fork of the Powder River, Johnson
County, Wyoming. Initially an articulated
Diplodocus tail, CM 307, was separated out.

16

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

At the time that the latter was prepared in
connection with the mounting of the Diplod-
ocus skeleton, several other boxes of mate-
rial were also worked up and catalogued CM
312 (see below). Three bones of the left fore-
limb were also prepared, but it is unclear
whether they were assigned to CM 312 or
not. They have recently been recatalogued
CM 21775. Several years later the remainder
of the material was prepared and found to
represent a number of individuals. A series
of chevrons received the number CM 1253
and all the remainder were catalogued CM
1256. Recently the bones of the several in-
dividuals have been separated and recata-
logued, a task complicated by the fact that
the quarry diagram has been misplaced. A
small number of bones, largely ribs and frag-
mentary specimens retain the number 1256
until recovery of the diagram allows their
correct assignment to one of the new cata-
logue numbers.

Collected by Utterback, 1903
The following specimens were all collected in the
Morrison Formation at Quarry B, Red Fork of the
Powder River, Johnson County, Wyoming, by Ut-
terback in 1903. Unless otherwise stated, all but the
first four (312, 1253, 36670, 21775) were originally
catalogued CM 1256.

312 A series of 22 median and postmedian caudal
vertebrae with a few attached chevrons and
the distal end of a possibly associated right
ischium. The vertebrae were for the most
part articulated. A left dentary, left scapu-
la, both coracoids, series of ribs, two meta-
tarsals and a phalanx, originally assigned to
312, almost certainly belong to other individ-
uals and have been recatalogued.

1253 A well preserved series of 12 chevrons pos-
sibly belonging to CM 36031 below.

36670 Left dentary, formerly referred to CM 312.
It is very doubtful that it belongs to the same
individual as the tail to which that number is
now restricted.

21775 Left humerus, radius, and ulna of Canuira-
saurus (perhaps originally assigned to CM
312) which were used in the mounted skele-
ton of Diplodocus carnegii, but not in the
casts of the skeleton sent to other museums
around the world (blown up models of CM
662 replaced these bones in the cast skele-

ton). To this individual has also been as-
signed a left scapula, both coracoids and a
series of ribs which were originally part of
CM 312; also a right scapula and sternal plate
originally CM 1256.

36039 Four anterior cervicals.

36040 A large anterior dorsal.

36042 Incomplete dorsal centrum.

36032 First caudal and another anterior caudal.

36036 Posterior caudal centrum.

36031 Series of 16 articulated anterior caudals.

36019 Both scapulae, right coracoid, both humeri,
left radius, left ulna.

36028 Upper end of large right scapula.

36043 Left scapula.

36030 Left scapula.

36029 Small right scapula.

36027 Right coracoid.

36025 Left coracoid.

28846 Both ischia, right pubis.

36024 Right ischium and pubis.

28847 Right ischium.

36021 Both femora.

36020 Right tibia and fibula of a smaller individual.

36023 Right tibia.

36022 Right fibula.

21742 Caudal centrum from the Morrison Forma-
tion at Quarry L, Freezeout Hills, Carbon
County, Wyoming.

Collected by Gilmore, 1902

21743 Anterior caudal arch. Same data as the
above.

Collected by Gilmore, 1902

1200 Six caudal centra from the Morrison For-
mation of Joe Wittecombe’s Ranch, Sweet
Creek County, Montana.

Collected by Silberling, 1903

2789 Caudal from the ?Morrison Formation of the
Texas panhandle.

Collected by Hammon, 1926

574 Right pubis. This number was intended for
the greater portion of a Camarasaurus skel-
eton from the Morrison Formation of Quarry
C, Sheep Creek, Albany County, Wyoming.
The bulk of the material proved to be poorly
preserved and was discarded. All that re-
mains (in addition to the pubis) are two dor-
sals, a half dozen caudals, right scapula,
right coracoid and a left ilium, of which four
of the caudals (DU 2 to 5), scapula (DU 9)
and coracoid (DU 10) were transferred to

1981

McIntosh— DINOSAURS of carnegie museum

17

Duquesne University, but recently returned
to the Carnegie Museum of Natural History.

Collected by Peterson and Gilmore,
1900-1901

21734 Dorsal, right pubis and left ischium from the
Morrison Formation of Quarry K, Sheep
Creek, Albany County, Wyoming. The right
femur has been transferred to the University
of California, Berkeley.

Collected by Gilmore, 1902
21720 Cast of right metacarpals I-V. The original,
AMNH 965, was found in the Morrison For-
mation at Bone Cabin Quarry, north of Med-
icine Bow, Wyoming. It was figured by Os-
born— Bull. Amer. Mus. Nat. Hist. (1904),
20:182, Fig. 1, correctly as Morosaiirus sp.
(a junior synonym of Camanisaiiriis). Later
he reconsidered and assigned the foot to Di-
plodociis, sending a cast to Pittsburgh for use
in the Diplodocus mount, where scaled
down models were employed. The manus
was figured by Abel — Abh. Zool.-Bot. Ges.
Wien (1910), 5:27, Pigs. 3-5.

Original collected by Kaisen, 1903
The following centra referred by Hatcher to As-
trodon (Pleurocoelus) all belong to very young in-
dividuals. They do bear a resemblance to that Low-
er Cretaceous Maryland genus in their enlarged
pleurocoels. On similar grounds Marsh (Dinosaurs
of North America, 16th Ann. Rept. U.S. Geol.
Surv., 1896, p. 184, Figs. 35-41) named some ju-
venile remains from Quarry 1, Como Bluff, Wyo-
ming, Pleurocoelus montanus. The latter are al-
most certainly a juvenile Camarasaurus gnindis,
and it appears likely that Hatcher’s vertebrae are
likewise referrable to Camarasaurus, the enlarged
pleurocoels being a juvenile character.

578 A cervical and a dorsal centrum from the
Morrison Formation at Quarry C, Sheep
Creek, Albany County, Wyoming. They are
described and figured by Hatcher — Ann.
Carnegie Mus. (1903c), 2:9-10, Figs. 1-4, as
Astrodon johnstoni.

Collected by Gilmore, 1901
585 Distal caudal centrum found in the Morrison
Formation at Quarry E, Sheep Creek, Al-
bany County, Wyoming. It was figured by
Hatcher — Ann. Carnegie Mus. (1903c), 2:11,
Pigs. 5-6, as Astrodon johnstoni.

Collected by Gilmore, 1901

Family Diplodocidae
Genus Diplodocus Marsh, 1878
Diplodocus longus Marsh, 1878

The following specimens from Dinosaur National
Monument have been referred to D. longus in the
literature, although it is not unlikely that a definitive
study will show that they should be transferred to
D. carnegii, if as is likely the specific differentiation
of these two forms is verified.

3452 Skull, mandible, cervicals 1-6 in articula-
tion. This is the only specimen of Diplodo-
cus in which a reasonably complete skull has
been found in articulation with postcranial
elements. It was found in the Morrison For-
mation at the Carnegie Museum Quarry at
Dinosaur National Monument, north of Jen-
sen, Uintah County, Utah (DNM 220). It was
mentioned by Holland — Mem. Carnegie
Mus. ( 19246), 9:385-386, 403, and figured PI.
15, Fig. 2. Described by McIntosh and Ber-
man— J. Paleo. (1975), 49:187, Fig. 2 — and
by Berman and McIntosh — Bull. Carnegie
Mus. Nat. Hist. (1978), 8:14-16, Figs. 2-3.
The specimen is mounted and was placed on
exhibition in 1915.

Collected by Douglass et al., 1915
37006 Half a cervical bearing the number DNM
220/3, thus possibly it might pertain to the
last specimen. However, it appears not to.
11161 Skull and mandible, complete and uncrushed
(DNM 160/10). This skull was found beneath
the tail of Apatosaurus, CM 3378, and was
otherwise relatively isolated at the same
quarry as the above. It was described and
figured by Holland — Mem. Carnegie Mus.
(19246), 9:385-402, Figs. 1-3 and 8-11, Pis.
40-42 — by Haas — Ann. Carnegie Mus.
(1960), 36: Figs. 7-8 — by McIntosh and Ber-
man—J. Paleo. (1975), 49:187-195, Figs. 1,
3, 5 — and by Berman and McIntosh — Bull.
Carnegie Mus. Nat. Hist. ( 1978), 8: 14, Figs.
2, 9. This superbly preserved specimen,
which has furnished more detailed informa-
tion about the sauropod skull than probably
any other was discovered by Douglass ap-
propriately on Thanksgiving Day, 1912, after
three relatively frustrating years of searching
for skulls, at the great dinosaur quarry.

Collected by Douglass, 1912
11255 Skull and mandible (juvenile) (DNM 351).
Same data as the above. Mentioned by Hoi-

Fig. 6. — Skull of Diplotlociis longus, CM 11161; a) dorsal view; b) right lateral view; c) palatal view; d) posterior view.

land — Mem. Carnegie Mus. (1924/7), 9:386,
403, and figured PI. 43. CM 1 1394 is a second
number applied to this specimen.

Collected by Douglass et al., 1921
21763 Right fibula and astragalus (DNM 150/11).
These bones are part of an articulated skel-
eton from the same locality as the above, the
remainder of which has been transferred to
the Denver Museum of Natural History
(DMNH 1494) where it is mounted and on
display. It consists of the vertebral column
complete from cervical 8 to caudal 20, right
scapula-coracoid, complete pelvis, and both
hind limbs without feet. The right femur was

originally catalogued CM 11970. Pictured in
Colorado Mus. Nat. Hist, (now Denver Mus.
Nat. Hist.) (1947), popular series No. 1:64-
65, and in subsequent guides of the Denver
Museum.

Collected by Douglass et al., 1912-
1915

21738 Left radius and ulna found with the above
skeleton (DNM 205/C) and not unlikely be-
longing to it.

Collected by Douglass et al., 1913
26552 Braincase (formerly catalogued CM 1201,
which had been assigned to two specimens)
(DNM 175/A). Same data as the above. De-

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

19

scribed and figured by Berman and Mc-
Intosh— Bull. Carnegie Mus. Nat. Hist.
(1978), 8:19-21, Eigs. 3-6.

Collected by Douglass et al., 1913
The following specimens were all collected in the
Morrison Formation at the Carnegie Museum Quar-
ry at Dinosaur National Monument, north of Jen-
sen, Uintah County, Utah, by Douglass et al.,
1909-1922, and may belong to the same species as
the above.

21718 Teeth preserved in place, the jaws having
decayed away (DNM 130/CA).

33984 Several cervicals (DNM 102, part).

30772 Three cervicals, questionably referred to this
genus (DNM 66).

21745 Articulated series of dorsals, sacrals and
caudals, left ilium (incomplete), and right fe-
mur (DNM 145).

3388 Two dorsal centra with bases of arches
(DNM 15).

30768 Dorsal centrum (DNM 285/53).

21728 Sacrum, incomplete left ilium, ischia (DNM
89).

33977 Articulated series of antero-medial caudals
(DNM 148).

11975 Four articulated postmedian caudals (DNM
168/A). A right femur originally catalogued
with these bones has been removed to 21745.
21714 Postmedian caudal (DNM 39/65H).

33904 Posterior caudal (DNM 99).

30763 Three caudals, right tibia-fibula-astragalus,
ribs, and an associated right scapula-cora-
coid which may belong to the same individ-
ual (DNM 285, part). (Part of what was for-
merly catalogued CM 2969.)

38339 Rib (found with 202/F) perhaps belonging
here.

30764 Right scapula (DNM 285/H). (Formerly part
of CM 2969.)

10004 Left scapula-coracoid (DNM 210/9). (Two
scapulae, a coracoid and a fibula of other in-
dividuals formerly referred to this number
have been recatalogued; see 33924, 33925,
33926.)

33926 Right scapula (DNM 210/3). (Formerly part
of 10004.)

33961 Left scapula (DNM 45/18?).

38335 Small right scapula (DNM 285/H).

21721 Right humerus, radius and ulna (DNM 228).

30773 Left humerus, left pubis, left ischium (DNM
350 and 345/A). A large part of the vertebral
column of this animal remains to be pre-

Fig. 7. — Mandible of Dipiodocus longiis, CM 11161; a) dorsal
view; b) left lateral view; c) ventral view.

pared. (Pelvic bones originally catalogued as
part of CM 9000.)

33959 Left radius (DNM 238/1).

33962 Incomplete right ilium and pubis (DNM
146/B).

21747 Pair of pubes (juvenile) (DNM 231).

33947 Incomplete right pubis (DNM 228/7) proba-
bly belonging to this form.

3389 Right ilium (juvenile) (DNM 4).

33987 Left ischium (DNM 270/14).

33991 Left femur and tibia (DNM 160/G and H).
These belong to a young individual found
associated with Apatosaurus CM 3378 and
formerly catalogued as part of it.

3377 Left femur (juvenile) (DNM 172).

21710 Left femur (DNM 334). This femur is on ex-
hibition as a “touch” bone. It was pictured
by Berman — Carnegie Mag. (1971), 45:95.

21753 Left femur and head of right tibia, perhaps
belonging to two individuals (DNM 151/A).

21754 Left femur (small) (DNM 219).

21771 Right femur (DNM 60/14), transferred to the
South African Museum, 1977.

21788 Left femur (small), transferred to the USNM
with Barosaurus material and returned.

30762 Right femur (small) (DNM 120/P).

33951 Right tibia (DNM 228/6).

20

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

33953 Right tibia (DNM 285/38).

33967 Left tibia (DNM 160/K), found associated
with Apatosaurus skeleton CM 3378 and
originally catalogued as part of it.

33924 Right fibula (DNM 210/2) (originally part of
CM 10004).

33950 Left fibula (DNM 285/10).

30767 Right fibula, astragalus and pes (DNM
175/G) originally catalogued as part of CM
2905.

33928 Right fibula (DNM 232/22).

21741 Right astragalus (DNM 60/16). The associ-
ated right femur was transferred to the North
Carolina Museum in Durham, North Caroli-
na, where it is on display. The associated
right tibia and fibula have been transferred
to the Royal Ontario Museum in Toronto.

33941 Left metatarsal III (DNM 39/15).

37013 Chevrons 11, 12, 13 being casts of the “tran-
sition region" chevrons of AMNH 223 from
the Morrison Formation of Como Bluff, Wy-
oming. Figured by Osborn — Mem. Amer.
Mus. Nat. Hist. (1899), 1: Figs. 10 and 13.

Original collected by Osborn and
Brown

Diplodocus carnegii Hatcher, 1901

84 Cervicals 2-15, dorsals 1-10, sacrals 1-5,
caudals 1-12, 18 ribs, left scapula (not right
as stated by Hatcher), left coracoid, right il-
ium and a fragment of the left, pubes, ischia,
right femur, both sternal plates, supposed
clavicle. Type specimen. It was found in the
Morrison Formation of Quarry D (not C as
stated by Hatcher) also referred to as Quarry
3 of 1899 of Sheep Creek, Albany County,
Wyoming. It was discovered by A. Cogges-
hall on 4 July, 1899. It was described by
Hatcher — Mem. Carnegie Mus. (19016), 1:1-
57 and figured as follows: cervicals. Pis. 2-
6 and Fig. 6; dorsals. Pis. 7-8; caudals, PI.
9; sternal plates. Fig. 12; “clavicle,” Fig. 13.
Preliminary reports by Holland — Science,
n.s. (1900), 11:816-818 — and Hatcher — Sci-
ence, n.s. (19006), 12:828-830. This speci-
men forms the core of the skeleton which
was mounted and put on display in 1907. The
latter was completed by additions from sev-
eral other individuals as follows: CM 94
(median caudals, right scapula-coracoid.

right tibia-fibula-pes), CM 307 (distal cau-
dals). The skull was modelled from the brain-
case of CM 662 and skull USNM 2673. The
right forelimb (and also the left forelimb of
the eleven casts of the skeleton sent to mu-
seums throughout the world) was accurately
modelled from the smaller individual CM
662. The forefeet were modelled from the
larger manus AMNH 965 now known to be-
long to Camarasaurus, and too many pha-
langes were assigned to the manus. In the
Carnegie Museum of Natural History origi-
nal only, the left forelimb CM 21775 now as-
signed to Camarasaurus was used, as were
left fibula and partial pes CM 33985. Casts of
the skeleton were sent to London, Berlin,
Paris, Vienna, Madrid, St. Petersburg (now
Leningrad), Bologna, La Plata, Mexico City,
Munich, and another was made in Vernal,
Utah.

Collected by Wortman, W. H. Reed,
A. S. Coggeshall, and W. C. Reed,
1899

94 Nine cervicals, nine dorsals, sacrum, 39 cau-
dals, ribs, five chevrons, scapulae-cora-
coids, sternal plates, ilia, pubes, ischia, left
femur, right tibia, right fibula, right astraga-
lus, complete right pes. Para type. A second
individual is indicated by a second pair of
ischia, a fragment of a pubis, and an incom-
plete second left femur. In addition eight of
the caudals clearly belong to a larger individ-
ual, which might possibly be CM 84 above.
It came from the same quarry as CM 84. As
stated above the right scapula-coracoid, tib-
ia, fibula, pes and some caudals and chev-
rons were used to complete the mount of CM
84. The left femur (as well as the other frag-
mentary one) and six caudals were trans-
ferred to the Cleveland Museum of Natural
History to supplement skeleton CM 662
which was transferred there. Later these
were again transferred to Houston where the
femur was incorporated into the mount, but
the caudals were returned to Pittsburgh. The
skeleton was described by Hatcher — Mem.
Carnegie Mus. (19016), 1:1-57, and figured
as follows: co-ossified caudals. Fig. 11; scap-
ula-coracoid, Fig. 14; sacrum and pelvis, PI.
10, Figs. 1-2; femur, Figs. 15-17, 23; tibia-
fibula, Fig. 18-19, PI. 11 1-2; pes. Figs. 20-
21, PI. 11 1-2. The pelvis was figured by

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

21

Hatcher — Mem. Carnegie Mus. (1903J), 2:
PI. 4, Fig. 2 — and the pubis was figured by
Gilmore — Mem. Carnegie Mus. (1936u),
11:263, Fig. 37, where it was erroneously
stated to be that of Apatosaurus CM 563.

Collected by Peterson and Gilmore,
1900

33985 Left fibula and probably left metatarsals III,
IV, V from the same quarry as the type and
paratype but not belonging to either. These
were used to complete the mounted skele-
ton.

Collected by Peterson and Gilmore,
1900

30765 Two metatarsals from the Morrison Eorma-
tion of Wyoming, quarry not known.

Collected by Wortman et al., 1899

Diplodocus hayi Holland, 1924

662 Only the “clavicle” and perhaps two dorsal
arches remain at the Carnegie Museum. The
greater part of this skeleton, consisting of the
cranium, 10 cervicals, five dorsals, 33 cau-
dals, numerous ribs and chevrons, scapulae,
coracoids, sternal plates, humeri, radii, left
ulna and part of manus, pubes, ischia, left
tibia, fibulae, astragali, part of pes, was
transferred to the Cleveland Museum of Nat-
ural History (CMNH 10670) and later to the
Houston Museum of Natural Science, where
the skeleton has been mounted and is on ex-
hibition. The sacrum and ilia were not trans-
ferred to Cleveland but were sent to Houston
at a later date. The left femur was not trans-
ferred to Cleveland either, and its current
whereabouts are unknown. Type specimen.
The skeleton from the Morrison Formation
was found isolated at Quarry A, Red Fork of
the Powder River, Johnson County, Wyo-
ming. The cranium was described and fig-
ured by Holland — Mem. Carnegie Mus.
(1906), 2:225-246, Figs. 4-10. Some of Hol-
land’s interpretations were criticized by
Hay — Science, n.s. (1908), 28:517-519 — and
answered by Holland — Science, n.s. (1908),
28:644-645. Other figures and description
was provided by von Huene — Neues Jahrb.
Min. Geol. Pal. (Beil.-Bd.) (1914), 37:579-
580, PI. 9, Fig. 1. Still further discussion ap-
pears in Holland — Mem. Carnegie Mus.
(1924/?), 9:397-401 — where it is established
as the type of the new species D. hayi.

Hatcher described and figured the radius and
ulna — Mem. Carnegie Mus. (1903c), 2:72-
73, Figs. 1-2.

Collected by Utterback, 1902-1903
36041 Spine of anterior caudal marked “662” on
one side and “1256” on the other. It is from
the Morrison Formation on the Red Fork of
the Powder River, Johnson County, Wyo-
ming. The former specimen is from Quarry
A, the latter from Quarry B. Two posterior
dorsal arches and spines similarly marked
both with “662” and “1256” also belong to
Diplodocus, but have not yet been recata-
logued.

Collected by Utterback, 1903
Diplodocus sp.

86 Right femur from the Morrison Formation of
Quarry 2, Dyer’s Ranch, Carbon County,
Wyoming. It was figured by Hatcher — Mem.
Carnegie Mus. (1901/?), 1: PI. 11, Figs. 3-4 —
and has been transferred to the Utah Natural
History State Museum, Vernal, Utah
(UNHSM V 16B).

Collected by Wortman et al., 1899
1201 A dorsal and a caudal centrum from the
?Morrison Formation on Joe Wittecombe’s
Ranch, Sweet Creek County, Montana.

Collected by Silberling, 1903
3395 Caudal from the Morrison Formation 50 mi
south of Grand Junction, Colorado.

Collected by McMillen, 1914
307 Articulated series of 38 posterior caudals in-
cluding “whiplash.” This tail was found in
the Morrison Formation at Quarry B, Red
Fork of the Powder River, Johnson County,
Wyoming. It was described and figured by
Holland — Mem. Carnegie Mus. (1906), 2:252-
255, PI. 29 — and was used to complete the
tail of the mounted Diplodocus.

Collected by Utterback, 1903

2098 Right femur (distal half), right tibia, right fib-
ula, right astragalus, right pes from the Mor-
rison Formation near Brown’s Ranch, Elk
Mountains, Johnson County, Wyoming.

Collected by Utterback, 1906

2099 Left tibia and part of pes. Same data as
above.

Collected by Utterback, 1906
21749 An astragalus probably belonging to Diplod-
ocus, which was collected with astragalus
CM 21748 of lApatosaurus at the American

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

T")

Fig. 8. — Cervical vertebra of Barosaiinis leiitiis, CM 1198.

Museum Quarry at Bone Cabin Quarry,
north of Medicine Bow, Wyoming.

Collected by Gilmore, 1902

887 Seven postmedian caudals, a phalanx, and a
chevron taken from the Morrison Formation
at the Marsh-Felch Quarry at Garden Park,
north of Canon City, Colorado.

Collected by Utterback, 1901

Genus Barosaurus Marsh, 1890
Barosaurus lentus Marsh, 1890

11984 Articulated series of presacral vertebrae ex-
tending from the anterior part of the neck to
the mid-dorsal region, several ribs. At the
time of this writing less than half of the cer-
vicals and dorsals have been freed from the
matrix. Associated and perhaps belonging to
the same individual is the greater part of the
right manus (see however 21744 below). The
specimen comes from the Morrison Forma-
tion at the Carnegie Museum Quarry at Di-
nosaur National Monument, north of Jensen,
Uintah County, Utah (DNM 310).

Collected by Douglass et al., 1918-
1919

1198 An associated series of cervical vertebrae
(DNM 150/B) from the same horizon and lo-
cality as the above. Ten vertebrae were col-
lected but only four were sufficiently well
preserved to be saved. The trunk, sacrum,
and tail found in association with this neck
were originally thought to belong to Diplod-
ociis and were given a separate field number
(DNM 155). Same data as the above. The

latter specimen was transferred to the Royal
Ontario Museum in Toronto (ROM 3670).
Some of the limb bones numbered 155 may
belong to this specimen, but many do not.
One caudal each transferred to North Caro-
lina Museum, University of Texas, and Uni-
versity of Cincinnati.

Collected by Douglass et al., 1913-
1915

21744 Distal half of humerus, radius, ulna, and
most of manus of an articulated right fore-
limb, not improbably belonging to this genus
(DNM 312). Only the humerus among fore-
limb elements has been found with remains
of this animal before. The metacarpals are
more elongate and slender than in Diplodo-
cits. If this specimen does indeed belong to
Barosaurus, then the manus mentioned un-
der 1 1984 does not and most likely is that of
Apatosaurus. Same data as the above.

Collected by Douglass et al.

33978 Anterior half of left ilium and head of left
pubis, perhaps belonging to same individual
as CM 1198 (DNM 155/C, incorrectly la-
belled 150/C).

Collected by Douglass et al., 1915

21719 Right humerus which lay close to the neck
CM 1 198 and was assigned the same number
as the Toronto specimen (DNM 155/Hum 2),
but would appear to be too long to go with
it. It is very long and slender, more so than
the only known Barosaurus lent us humerus
AMNH 6341. Although crushed, it is ques-
tionable that the crushing could account for

1981

MclNTOSH— DINOSAURS OF CARNEGIE MUSEUM

23

the differences. It may belong to the Denver
DiplodocHs (DNM 150) found nearby, or
may even belong to the genus Bracluosau-
rns .

Collected by Douglass et al.

11878 Postmedian caudal vertebra and upper ends
of tibia and femur (DNM 347). Same data as
above. The caudal has been transferred to
the University of Michigan (V 16778).

Genus Apatosaurus Marsh, 1877

Syn.: Brontosaurus Marsh, 1879.

Elosaurus Peterson and Gilmore, 1902.

Apatosaurus excelsus (Marsh), 1877

Syn.: Brontosaurus excelsus Marsh, 1877.
Brontosaurus ampins Marsh, 1881.

Elosaurus parvus Peterson and Gilmore,
1902.

563 Skeleton consisting of nine cervicals, nine
dorsals, five sacrals, 18 caudals, left scapu-
la-coracoid, humeri, radii, ulnae, right ma-
nus, left ilium, pubes, ischia, right femur, tib-
iae, right fibula, right metatarsals I, III, IV,
V, a phalanx and a claw, collected from the
Morrison Formation at Quarry E, Sheep
Creek, Albany County, Wyoming. At pres-
ent only the pubes, ischia, and a caudal spine
remain in Pittsburgh, the balance having
been transferred to the University of Wyo-
ming W. H. Reed Museum in Laramie,
where it is mounted and on exhibit. The fore-
limb and foot were described by Hatcher —
Science, n. s. (1901c), 14:1015-1017 and in
more detail in Ann. Carnegie Mus. (190E/),
1:356-376, Pis. 19-20. Hatcher figured the
pelvis and sacrum in Mem. Carnegie Mus.
(1903t/), 2: PI. 4, Pig. 1. Gilmore described
the entire skeleton — Mem. Carnegie Mus.
(1936</), 11:245-264, and figured the follow-
ing elements: cervicals, PI. 31; dorsals, PI.
32; sacrum and pelvis. Pigs. 31 and 36; cau-
dals, PI. 33; scapula-coracoid. Fig. 32; hu-
merus, Fig. 33; radius and ulna. Fig. 34; ma-
nus. Fig. 35.

Collected by Gilmore, 1901

37010 Left astragalus, claw of pes found in the
wash of Quarry E, Sheep Creek, Albany
County, Wyoming. They are placed here as
they might belong to the above skeleton, in
which these elements were missing.

Collected by Peterson and Gilmore,
1900

21785 Left femur, metacarpal, and 10 caudals be-
longing to a skeleton largely discarded, from
the same locality as the above. A caudal and
the right femur have been transferred to Du-
quesne University (DU 11).

Collected by Peterson and Gilmore,
1900

21740 Seven anterior caudals, left pubis, right is-
chium, right femur, and right tibia from the
Morrison Formation at Quarry A, Sheep
Creek, Albany County, Wyoming. The fe-
mur (erroneously marked CM 85) has been
transferred to the Royal Ontario Museum
where it is on exhibit.

Collected by Peterson and Gilmore,
1900

1 188 Sacrum, right ilium, five anterior caudal cen-
tra, collected from the Morrison Formation
at Quarry F, Sheep Creek, Albany County,
Wyoming. The ilium and sacrum were fig-
ured by Hatcher — Mem. Carnegie Mus.

( 19016), 1: PI. 10, Fig. 3 — but they have been
transferred elsewhere and their current
whereabouts are unknown.

Collected by Peterson and Gilmore.
1900

33994 Proximal extremity of a truly colossal femur,
rivaling in size that of the gigantic ^'Atlan-
(osaurus imtnanis" {=Apatosaurus ajax)
YPM 1840. It was collected in the Morrison
Formation at Quarry B, Sheep Creek, Al-
bany County, Wyoming.

Collected by Peterson and Gilmore,
1900

87 Two dorsals, three caudals, left pubis, is-
chia, left femur, fragment of limb, from the
Morrison Formation of Quarry 4, Sheep
Creek, Albany County, Wyoming. One dor-
sal has been transferred to the University of
Michigan (V 16777) where it is on exhibition.

Collected by Wortman et al., 1899

595 Caudal centrum of a very young animal from
the Morrison Formation at Sheep Creek,
Albany County, Wyoming, probably belong-
ing to this animal.

Collected by Gilmore, 1901

89 Complete right pes with astragalus from the
Morrison Formation at Quarry 5, Sheep
Creek, Albany County, Wyoming. It was fig-
ured by Hatcher — Mem. Carnegie Mus.

24

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

M viiiA

n

\

V%

^ .

a

Fig. 9. — Limb bones of “Elosaurus parvus” (=Apatosaurus excelsus), CM 566; a) right scapula, lateral view; b, c) left humerus,
posterior and anterior views; d, e) right humerus, anterior and posterior views; f, g) right ulna, anterior and lateral views; h, i) right
femur, anterior and posterior views; j, k) left fibula, lateral and medial views.

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

25

(1901^), 1:52, Fig. 22 — and described by Gil-
more— Mem. Carnegie Mus. (1936r/), 11:234-

242.

Collected by Wortman et al., 1899

555 A right radius and ulna (CM 555) and a left

556 tibia-fibula-astragalus (CM 556) together
with II cervicals (field nos. fl-f8, y, y',
z) found together in the Morrison Forma-
tion at Quarry D, Sheep Creek, Albany
County, Wyoming, and of proper propor-
tions to belong to a single individual. The
vertebrae originally and the limb bones later
were referred to a Camarasaurus skeleton
CM 584 found nearby, but this specimen is
clearly referrable to Apatosaurus. In all like-
lihood the articulated vertebral series fl-f8
are those originally catalogued as CM 663.

Collected by Peterson and Gilmore,
1900

577 Two distal caudals from the Morrison For-
mation of Quarry J, Sheep Creek, Albany
County, Wyoming.

Collected by Axtell, 1901

21748 Left astragalus collected with astragalus
21749 of IDiplodocus in the Morrison For-
mation at Bone Cabin Quarry, north of Med-
icine Bow, Wyoming.

Collected by Gilmore, 1902

""Elosaurus parvus Peterson and Gilmore, 1902”

566 Cervical arch, dorsal arch, right scapula, hu-
meri, right ulna, right femur, left fibula, frag-
ment of ilium, fragment of pubis, and per-
haps two sacral centra. This specimen of a
very immature individual, forming the type
specimen of Elosaurus parvus is clearly that
of a young apatosaur as demonstrated con-
clusively by the forelimb bones. As it was
found intermingled with CM 563, it probably
belongs to A. excelsus. It was collected in
the Morrison Formation at Quarry E and was
described and figured by Peterson and Gil-
more— Ann. Carnegie Mus. (1902), 1:490-
499, Figs. 1-10, PI. 25.

Collected by Gilmore, 1901

Apatosaurus louisae Holland, 1916

3018 Cervicals 1-15, dorsals 1-10, sacrals 1-5,
caudals 1-64, 18 ribs, chevrons 1-3, scapu-
lae, coracoids, humeri, left radius, ulna and
manus, ilia, pubes, ischia, right femur, tibia
and fibula, left pes. Type specimen. This

skeleton, the most perfect of Apatosaurus
ever collected, was discovered 17 August,
1909, marking the discovery of by far the
greatest Jurassic dinosaur quarry so far dis-
covered anywhere in the world. It was in the
Morrison Formation in what later was to be-
come Dinosaur National Monument, north
of the little town of Jensen, Uintah County,
Utah (DNM 1 and 25). It was described brief-
ly by Holland — Ann. Carnegie Mus. (19157>),
10:143-145, and cervical 11 was figured by
him in Ann. Carnegie Mus. ( 1924r/), 15:124,
Fig. 2. The entire skeleton was described in
detail and all parts figured by Gilmore —
Mem. Carnegie Mus. (1936ri), 1 1: 175-300,
Figs. 5-30, Pis. 24-27, 29-30, 34. The follow-
ing specimen, a skull, may belong to this
skeleton (see below).

Collected by Douglass, 1909-1911
11162 Skull without mandible found in association
with two nearly complete skeletons of this
species, CM 3018 above and the former CM
11990 below (DNM 60/10). Holland— Ann.
Carnegie Mus. ( 1915u), 9:274-277 — argued
that it probably belongs to CM 3018, a sug-
gestion opposed by Osborn, who followed
Marsh in believing that Apatosaurus did not
possess a Diplodocus-\\kQ" skull. Holland
continued the discussion in Mem. Carnegie
Mus. (1924/?), 9:391, Fig. 5, but his argu-
ments were rejected by Gilmore — Mem. Car-
negie Mus. (1936^/), 11:188 — because of a
confusion of this skull CM 11162 with
another, CM 11161, of Diplodocus. Field let-
ters from Douglass to Holland clearly dem-
onstrate that Holland’s association of CM
11162 with CM 3018 and CM 11990, and of
CM 11161 with CM 3378 is the correct one,
and that Gilmore’s is incorrect, see also
McIntosh and Berman — J. Paleo. (1975),
49: 187. This skull is described and figured in
Berman and McIntosh — Bull. Carnegie Mus.
Nat. Hist. (1978), 8:21-30, Figs. 3D, 7, 8,
9 — where it is argued that Holland was prob-
ably correct and that this skull is probably
that of Apatosaurus.

Collected by Douglass et al., 1910
1 1990 This number was formerly employed for the
prepared parts (a tibia and some caudals) of
an almost complete, but slightly smaller, ar-
ticulated skeleton found partly overlapping
CM 3018. It lacked only the skull, a femur.

26

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

Fig. 10. — Right femur of Apatosaurus louisae, CM 3018, anterior view.

one of the forelimbs, the pedes, and probably
the tip of the tail. The skeleton was trans-
ferred to the Los Angeles County Museum
in 1939 (LACM 52844) and the left femur has
been placed on exhibit. The skeleton is pic-
tured as it lay in the rock in Gilmore — Mem.
Carnegie Mus. (1936a), 11: PI. 21, Eig. 6,
where it is erroneously stated to be that of
Diplodociis 150. This specimen is DNM 40.
3378 Vertebral column complete and articulated
from the mid-cervical region to the tip of the
tail (caudal 82). This specimen (DNM 160)
from the same quarry as all of the above,
was found associated with several disartic-
ulated limb bones and a skull (DNM 160/0)
CM 11161 referred to Diplodocus. It is
doubtful if any of the limb bones really be-
long to CM 3378. The specimen is described
and figured by Holland — Ann. Carnegie
Mus. (1915«), 9:277-278, PI. 59— and by Gil-
more— Mem. Carnegie Mus. (1936r/), 11:204-
205, PI. 28. The specimen shows minor dif-
ferences from CM 3018, but these may be
due to age.

Collected by Douglass et al., 1913-
1915

The following specimens were all collected in the
Morrison Eormation at the Carnegie Museum Quar-
ry at Dinosaur National Monument, north of Jen-
sen, Uintah County, Utah, between 1909 and 1922.
They show no characters to distinguish them from
the above specimens of Apatosaurus louisae.

33916 An articulated series of the last two cervi-
cals, all 10 dorsals, the first sacral, and three

disarticulated sacral centra with ribs (DNM
210). The left ribs were found articulated in
place, but only a few remain. A large number
of limb and girdle bones belonging to more
than one genus were found associated with
the vertebrae and assigned the same field
number 210. A number of these have been
transferred to the Junior Museum and the
California Academy of Science in San Fran-
cisco, others to the University of Delaware
in Newark, Delaware, and some to the Lake-
side Museum and Art Center in Peoria, Illi-
nois. Future preparation and study will be
needed to determine which belong to CM
33916.

Collected by Douglass et al., 1914

3390 Articulated vertebral column consisting of

3391 most of the dorsals, the sacrum, and 12 cau-
dals, many ribs, left ilium, left pubis, left is-
chium of a very immature individual (DNM
24). It is likely that a series of six articulated
cervicals (DNM 37) found 20 ft east of 24
belong to the same skeleton, according to
Douglass. These were given the separate cat-
alogue number 3391. Douglass also felt that
a small jaw (DNM 35, now lost) found with
neck 37 also belonged to the same animal.
The latter contained slender peg-like teeth of
the Diplodocus kind.

Collected by Douglass et al., 1909-
1910

11339 A series of seven dorsals of a very young
animal, perhaps belonging to Apatosaurus,

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

27

Fig. 11. — Skull of Apatosaurus louisae, CM 11162 (not 11161 as painted on bone); a) lateral view; b) dorsal view; c) posterior view.

28

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

found with a Stegosaurus skeleton CM
11341.

33968 Two dorsals (DNM 130/B).

30766 Eleven anterior caudals, several chevrons
(formerly part of CM 2905); left femur, right
tibia and probably associated right radius,
ulna, and partial manus (formerly CM 10000)
(DNM 175/G). Other bones belonging to this
skeleton have not yet been prepared.

33912 Centrum (with DNM 45).

21752 Two dorsals, sacrum, ilia, pubes, ischia, fore
and hind limbs and feet of a juvenile (DNM
120/M, N).

3384 Four caudals (DNM 7, part).

21731 Anterior caudal (field number in doubt).

33918 Four caudals (DNM 130/P).

33921 Caudal, transferred to USNM and returned.

36687 Left scapula-coracoid (DNM 55).

21708 Complete left forelimb and manus (DNM
12), found somewhat to the east of the quarry
in the same stratum.

21715 Left humerus, radius and ulna found with
some Stegosaurus remains (DNM 39/9, 10,
11).

21717 Right ulna (DNM 39/53).

33996 Right radius and ulna (DNM 345/A).

38338 Metacarpal I (found with DNM 130).

36699 Left metacarpals II, III, IV (found with
DNM 160) (juvenile).

21780 Two metacarpals (DNM 197/66 and 71), and
?radius (DNM 197/38).

38352 Metacarpal (DNM 232, part).

21746 Left ilium (DNM 39/50), juvenile, used to re-
place the missing bone in the mounted Cam-
arasaurus skeleton CM 11338.

33980 Left pubis (DNM 130/11).

33902 Left pubis (DNM 210/10), incomplete and
referred to this genus with doubt.

11998 Left femur (DNM 155/FA) transferred to the
Newark Museum; a second (right) femur also
assigned this catalogue number (DNM 155/S)
is part of the "'Diplodocus" skeleton trans-
ferred to the Royal Ontario Museum (ROM
3670).

21756 Left femur (small) (DNM 238/2) perhaps be-
longing to this form.

21784 Right femur, perhaps belonging to Apato-
saurus.

33976 Left femur, perhaps belonging to Apatosau-
rus (DNM 232/12).

33997 Left femur formerly part of CM 3378 (DNM
146/B).

21729 Right hind limb (DNM 171).

33964 Left tibia and fibula (DNM 210/0).

33952 Right fibula (DNM 175/2).

30761 Left astragalus (DNM 103).

11253 Right pes (DNM 60/2). Associated right fe-
mur transferred to the University of Cincin-
nati; right tibia and fibula transferred to Ju-
nior Museum, San Francisco, California,
where the tibia is on exhibition.

30770 Right metatarsal I (found with DNM 210).

33989 Left metatarsal III (DNM 173).

33990 Right metatarsal IV (DNM 243).

Apatosaurus sp.

85 Five ribs, right pubis, right ischium and frag-
ment of the left, right femur, right tibia, right
fibula from the Morrison Formation at Quar-
ry 1, Dyer Ranch, Freezeout Hills, Carbon
County, Wyoming.

Collected by Wortman et al., 1899

83 Anterior third of a very large femur. It is of
historic importance as the “130 foot dino-
saur” discovered by W. H. Reed, reported
in the New York Journal. This story got Mr.
Carnegie’s attention and led him to commis-

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

29

sion the original Carnegie Museum dinosaur
expedition, see Coggeshall — Carnegie Mag.
(1951), 25:238-239.

Collected by W. H. Reed and brought
to Pittsburgh by Holland

36033 Caudal from Quarry B, Red Fork of the Pow-
der River, Johnson County, Wyoming (for-
merly part of CM 1256).

Collected by Utterback, 1903
36035 Two caudals. Same data as above.

2044 Dorsal (exchanged to University of Cincin-
nati, 1941), Morrison Formation, Cope Quar-
ry, Garden Park, Colorado.

Collected by Utterback, 1901
37007 Right scapula from the Morrison Formation
at Quarry C, Sheep Creek, Albany County,
Wyoming.

Collected by Peterson and Gilmore,
1900

28849 Right humerus from the Morrison Formation
of Quarry B, Red Fork of the Powder River,
Johnson County, Wyoming.

Collected by Utterback, 1903
36026 Right humerus. Same data as above.

21782 Left ulna and metacarpal V, and perhaps as-
sociated metatarsal I from the Morrison For-
mation at Quarry O, Freezeout Hills, Carbon
County, Wyoming.

Collected by Gilmore, 1903
28848 Right ischium from Quarry B, Red Fork of
the Powder River, Johnson County, Wyo-
ming (formerly part of CM 1256).

Collected by Utterback, 1903
936 Right metatarsal I. It was collected at the
Marsh-Felch Quarry in Garden Park north of
Canon City, Colorado in the Morrison For-
mation.

Collected by Utterback, 1901
36038 Right metatarsal II from Quarry B, Red Fork
of the Powder River, Johnson County, Wy-
oming (formerly part of CM 1256).

Collected by Utterback, 1903

The following specimens were collected by Gil-
more in 1902-1903 in Quarry N, Freezeout Hills,

Carbon County, Wyoming. Most of them were orig-
inally catalogued under the “catchall” number CM
1255 (p. 15).

36691 Anterior caudal.

36692 Eleven median and posterior caudals.

36690 Caudal.

36665 Right radius and ulna, juvenile.

36672 Left manus with one carpal, five metacarpals
and two phalanges.

36676 Two metacarpals.

36682 Two metacarpals.

36667 Greater portion of the right pes, juvenile.
36674 Four left metatarsals.

36683 Metatarsal I (crushed).

38347 Left phalanx II- 1 of pes.

Sauropoda, indeterminate

All of these specimens are from the Carnegie
Museum Quarry at Dinosaur National Monument,
north of Jensen, Utah. They were collected by
Douglass et al., 1909-1922.

37008 Cervical and rib (DNM 302, part).

33960 Dorsal centrum (DNM 39/65, part).

33900 Two very small dorsals (DNM 84).

37012 Sacral centrum (DNM ? 130/3).

33969 Caudal (DNM 312, part).

38351 Caudal and metacarpal (DNM 320, part).
37002 Rib (DNM 338).

37009 Rib (DNM 328 and 329).

38336 Proximal part of left scapula (DNM 146).
33944 Head of large right humerus (DNM 39/83).

33970 Incomplete humerus, proximal half of ra-
dius.

38348 Small metacarpal.

38346 Left metatarsal IV, left phalanx I-l, fragment
(with DNM 285/C).

33907 Chevron (DNM 228, part).

33914 Left radiale, probably sauropod.

33905 Incomplete right astragalus (DNM 39/65,
part).

38350 Claw of manus (DNM ? 45)

33982 Huge claw of digit I of pes (found with
Stegosaurus skeleton CM 11341, DNM 350).

30

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

ORDER ORNITHISCHIA
SUBORDER ORNITHOPODA

Family Hypsilophodontidae
Genus Dryosaurus Marsh, 1894
Dryosaurus altus Marsh, 1878

Syn.: Laosaurus altus Marsh, 1878.

3392 Skull, mandible, and greater part of skeleton
from the Morrison Formation at the Carnegie
Museum Quarry in Dinosaur National Mon-
ument, north of Jensen, Uintah County,
Utah (DNM 26). It was described and figured
by Gilmore — Mem. Carnegie Mus. (19256),
10:394-403, Figs. 3-6 — and the skull was fig-
ured by Gabon — Nature (1977c/), 268:231,
Fig. 2a, b; in Milwaukee Public Mus., Spec.
Publ. Biol. Geol. (19776), 2:44-46, Figs. 2a,
b; and in Mem. Soc. Geol. France (1980),
59:104, Fig. I04a, b. The mounted skeleton,
placed on exhibition in 1940, was pictured by
Kay — Carnegie Mag. (1940b), 14:215.

Collected by Douglass et al., 1910
11340 Skull and large part of skeleton of a very
young individual; same data as above (DNM
360). It was described and figured by Gil-
more— Mem. Carnegie Mus. (19256), 10:403-
409, Figs. 7-8, as Laosaurus gracilis but is
referred to this species by Gabon.

Collected by Douglass et al., 1922
1949 Three dorsals, about 28 caudals, right ilium,
right femur, right tibia, right fibula, metatar-
sals, from the Morrison Formation of Elk
Mountains, Johnson County, Wyoming. It
was described and figured by Shepherd et
al. — Brigham Young Univ. Geol. Studies
(1978), 24:2, 11-15, Figs. 3-5— and by Gal-
ton — Mem. Soc. Geol. France (1980), 59:104,
Fig. lU, IX.

Collected by Utterback, 1906

21786 About 10 dorsals, sacrum, 30 caudals, hu-
meri, both hind limbs with the shafts of many
of the elements incomplete, both pedes, oth-
er bones, found by Jess Lombard at Lily
Park, Moffat County, Colorado, in the Mor-
rison Formation. Figured by Gabon — Nature
(1977c/), 268:231, Figs. 2u and 2x, and also
Milwaukee Public Mus., Spec. Publ. Biol.
Geol. (19776), 2:46, Figs. 2U and 2X. It was
described and figured by Shepherd et al. —
Brigham Young Univ. Geol. Studies (1978),
24:2, 11-15, Figs. 1-5.

Collected by Kay and Lloyd, 1955

Genus Thescelosaurus Gilmore, 1913
Thescelosaurus neglectus Gilmore, 1913

9900 Two left humeri about the same size, locality
data unknown.

246 Dorsal centrum from the Lance Formation
on Lance Creek, Niobrara County, Wyo-
ming.

Collected by Hatcher, 1900

21787 Two humeri, two ulnae, two radii, fragments
of ilium, metatarsal, six phalanges of several
individuals from the Lance Formation of
Sheep Mountain, Carter County, Montana.

Collected by Kay, 1938

38327 Three dorsals, 10 phalanges, four claws
found with above.

Collected by Kay, 1938

Family Iguanodontidae
Genus Camptosaurus Marsh, 1885
Camptosaurus medius Marsh, 1894

This species has not been separated satisfactorily
from the type species, C. dispar, the type speci-
mens of both having come from the same quarry,
YPM Quarry 13 east of Como Bluff, Wyoming. Its
smaller size might be due to age or sex.

1 1337 The most complete skeleton of this genus yet
found lacking only the skull, mandible, atlas,
caudals beyond no. 14, right coracoid fore
arm and manus, pedes. It came from the
Morrison Formation at the Carnegie Mu-
seum Quarry at Dinosaur National Monu-
ment, north of Jensen, Uintah County, Utah
(DNM 370 and 353). It was described and
figured by Gilmore — Mem. Carnegie Mus.
(19256), 10:385-393, Figs. 1-2, PI. 18— and

1981

MclNTOSH— DINOSAURS OF CARNEGIE MUSEUM

31

Fig. 14. — Skeleton of Cumptosaiirus medius, CM 1 1337.

the skeleton, mounted and placed on exhi-
bition in 1940, is pictured by Kay — Carnegie
Mag. (1940^), 14:214.

Collected by Douglass et al., 1922
The following specimens are all from the Carne-
gie Museum Quarry at Dinosaur National Monu-
ment in the Morrison Formation. They were col-
lected 1909-1922.

15780 Right femur-tibia-fibula-astragalus-meta-
tarsal (DNM 315).

21778 Distal end of ischium (with DNM 130/D).

21722 Right femur (small) (DNM 217a).

21723 Right femur (small) (DNM 102, part).

21724 Upper end of left femur (DNM 210/12).
21707 Left tibia (DNM 197/11).

38337 Incomplete tibia and fibula (DNM 130).

21725 Three metatarsals (DNM 358).

Gqwus Rhabdodon Matheron, 1869
Syn.: Mochlodon Seeley, 1889.

Rhabdodon suessii (Bunzel, 1871)

Syn.: Iguanodon suessii Bunzel, 1871, referred to

Mochlodon by Seeley, 1881.

965 Cast of two teeth, from the Gosau Formation
of Neue Welt, Wiener Neustadt, Austria.
The originals in the University of Vienna
Museum were figured by Seeley — Quart. J.

Geol. Soc. London (1881), 37: PI. 27, Figs.
2-4 as Mochlodon suessii.

Originals collected by Pawlowitsch

Family FIadrosauridae
Subfamily Hadrosaurinae
Genus Hadrosaurus Leidy, 1858

Syn.: Kritosaurus Brown, 1910 (see Horner, J. Pa-
leo., 1979, 53:568).

Hadrosaurus sp.

1077 Cranium, part of left maxilla, left jugal and
part of right from the Judith River Formation
on Mud Creek, Fergus County, Montana. It
was mentioned by Lull and Wright — Spec.
Papers, Geol. Soc. Amer. (1942), 40:22.

Collected by Utterback, 1903

Genus Edmontosaurus Lambe, 1917

Syn.: Anatosaurus Lull and Wright, 1942 (see
Brett-Surman, Nature, 1979, 277:560).

Edmontosaurus regalis Lambe, 1917

26258 Skull and greater part of skeleton from the
Edmonton Formation (member A) of the Red
Deer River Valley, Alberta. It was obtained
from the Royal Ontario Museum in 1973
(ROM 1931-7) and was mentioned by Rus-

32

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

Fig. 15. — Skull and mandible of Edmontosaurus regalis, CM 26258.

sell and Chamney — Nat. Hist. Papers, Nat.
Mus. Canada ( 1967), 35:11.

Collected by L. Sternberg, 1931

Edmontosaurus annectens (Marsh), 1892
Syn.: Claosourus annectens Marsh, 1892.

105 Pelvis, many caudals, epidermis, etc. in poor
state of preservation from the Lance Eor-
mation 2 or 3 mi north of Doegie Creek, Nio-
brara County, Wyoming. The epidermis was
figured by Hatcher — Science, n.s. (1900c;),
12:719-720 — see also Hatcher — Ann. Car-
negie Mus. (190U), 1:386. Mentioned by
Lull and Wright (as CM 106) — Spec. Papers,
Geol. Soc. Amer. (1942), 40:23.

Collected by Hatcher, 1900
The following specimens were collected in the
type locality for this species, that is, the Lance Eor-
mation on Lance Creek, Niobrara County, Wyo-
ming, and likely belong to it. They were collected
by Hatcher in 1900. They are mentioned by Lull
and Wright — Spec. Papers, Geol. Soc. Amer.
(1942), 40:23.

243 Pour caudals and two phalanges.

245 Two caudals and one phalanx.

254 Left scapula.

252 Left astragalus and distal end of tibia.

Edmontosaurus sp.

30745 Rami of mandible, three dorsal centra, two
humeri, right ulna, three pubes, metatarsal,
ungual of more than one individual from the
Lance Pormation on Sheep Mountain, Car-
ter County, Montana.

Collected by Kay, 1938

38321 Small right dentary. Same data as above.

Collected by Kay, 1938

9970 Right ramus of the mandible. Same data as
above.

Collected by Kay, 1938

38324 Astragalus, three phalanges, two unguals.
Same data as above.

Collected by Kay, 1938

38325 Tooth. Same data as above.

Collected by Kay, 1938

38328 Metatarsal. Same data as above.

Collected by Kay, 1938

38333 Pubis. Same data as above.

Collected by Kay, 1938

38329 Caudal. Same data as above.

Collected by Kay, 1938

11745 Five dorsals, five sacrals, 27 caudals, pelvis,
hind limbs, and feet from the Lance For-
mation 15 mi south of Camp Crook, Harding
County, South Dakota.

Collected by Kay, 1938

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

33

1652 Left ilium, pubis, ischium, and fibula from
the Lance Formation, Hell Creek, Garfield
County, Montana. Mentioned by Lull and
Wright — Spec. Papers, Geol. Soc. Amer.
(1942), 40:23.

Collected by Utterback, 1903

38355 Right tibia from the Lance Formation of
Billy Creek, Garfield County, Montana.

Collected by McCrady, 1978

Subfamily Lambeosaurinae
Genus Corythosaurus Brown, 1914
Corythosaurus casuarius Brown, 1914

Syn.: Corythosaurus brevicristatus Parks, 1935.

9461 Skeleton lacking only skull, mandible, tip of
tail, left radius and both manus from the Old-
man Formation, 11 miles southeast of Steve-
ville, Alberta, south of the Red Deer River.
It was obtained from the Royal Ontario Mu-
seum in 1940 and was mounted and placed
on exhibition in 1941. The missing skull was
replaced by a cast of the type of Corytho-
saurus brevicristatus ROM 5856, a species
which has recently been shown to be syn-
onymous with C. casuarius, Dodson — Syst.
Biol. (1975), 24:37-52. It is referred to by C.
M. Sternberg in “Notes on the Dinosaur
Quarries” accompanying Map 969A Steve-
ville, Geol. Sur. Canada (1950), as ITetra-
gonosaurus, map no. 86.

Collected by L. Sternberg, 1920
The following specimens were obtained from the

Geological Survey of Canada in 1925. They are

from the Oldman Formation near Steveville, Red

Deer River, Alberta, and were collected by C. M.

Sternberg in 1919.

11375 Top of skull.

11376 Jaw.

Hadrosauridae, indeterminate

12101 Part of jaw, four caudals, foot bones from
the Oldman Formation of Sheep Mountain,
Long Pine Hills, Montana, about 15 mi
southwest of Camp Crook, South Dakota.

Collected by Kay, 1937

12100 Thirty-one caudals, six ribs or so, three
chevrons, pubes, left ischium, epidermis
from the Oldman Formation on Fred Town-
send’s Ranch, Carter County, Montana.

Collected by Kay, 1937

The following specimens are all from the Judith

River Formation on Mud Creek, Fergus County,
Montana, and were collected by Utterback, Silber-
ling, and Douglass in 1903.

1074 Right dentary and part of right maxilla, a
caudal and fragments mentioned by Lull and
Wright — Spec. Papers, Geol. Soc. Amer.
(1942), 40:22. Collector — Utterback.

38344 Tooth. Collector — Silberling.

1072 Sacrum. Collector — Utterback.

1073 Part of sacrum and fragments. Collector —
Silberling.

1071 Two dorsals, part of sacrum, two ribs, seven
chevrons, scapula, ulna, radius, ilia, pubes,
ischium, left tibia, fibula, three metatarsals,
part of pes. Collector — Utterback.

1064 Twelve caudals, part of ilium, fragments.
Collector — Silberling.

38356 Large section of the tail. Collector — Utter-
back (the following year, 1904).

1066 Left humerus. Collector — Utterback.

324 Incomplete right femur. Collector — Silber-
ling.

344 Both femora. Collector — Utterback.

345 Both femora. Collector — Utterback.

325 A tibia. Collector — Silberling.

326 Right tibia. This and the next specimens
were mentioned by Lull and Wright — Spec.
Papers, Geol. Soc. Amer. ( 1942), 40:22. Col-
lector— Silberling.

327 Small tibia and part of fibula. Collector — Sil-
berling.

328 Right tibia (large). Collector — Silberling.

1068 Incomplete tibia. Collector — Silberling.

38353 Large metatarsal, three phalanges, section of

limb bone. Collector — Douglass.

1069 Foot bone. Collector — Silberling.

38343 Ungual. Collector — Silberling.

1076 Nine caudals, six phalanges, two unguals.
Collector — Utterback.

The following specimens were all collected at
various sites in the Judith River Formation of Mon-
tana in 1903 by Douglass, Silberling, and Utterback.
3319 Right maxilla collected by Douglass between
Fish Creek and the Musselshell River. This
and the following six specimens were men-
tioned by Lull and Wright — Spec. Papers,
Geol. Soc. Amer. (1942), 40:22. Also men-
tioned by Horner — J. Paleo. ( 1979), 53:572.

329 Dorsal, metapodial and phalanx collected by
Silberling on Mud Creek, Fergus County.
Mentioned by Lull and Wright — Spec. Pa-
pers, Geol. Soc. Amer. (1942), 40:22.

34

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

b

Fig. 16. — Corylhosaurus casuarius\ a) partial skull, CM 11375; b) mandible, CM 11376.

281 Sacrum. Same data as above.

1065 Right humerus collected by Utterback on
Willow Creek, Fergus County.

1067 Part of right femur. Same data as above.
Mentioned by Lull and Wright — Spec. Pa-
pers, Geol. Soc. Amer. (1942), 40:22.

282 Partial limb and foot bone. Same data as
above but collected by Silberling.

305 Part of foot and fragments collected by
Douglass on Mud Creek, Sweetgrass Coun-
ty. Mentioned by Horner — J. Paleo. (1979),
53:572.

306 Fragments. Same data as above.

1202 Skull parts Gugal) and part of lower jaw from
the Judith River Formation on Crawford’s
Ranch, Sweetgrass County, Montana. Men-
tioned by Lull and Wright — Spec. Papers,
Geol. Soc. Amer. (1942), 40:22.

Collected by Silberling, 1904
594 Various remains including a caudal from the
Upper Cretaceous near Havre, Montana.

Collected by Douglass, 1902
3363 Incomplete right tibia and astragalus from
the Fish Creek Beds on Fish Creek, Mon-
tana.

Collected by Douglass, 1902

1981

McIntosh— DINOSAURS of carnegie museum

35

Family Pachycephalosauridae
Genus Pachycephalosaurus Brown and Schlaikjer,
1943

Pachycephalosaurus wyomingensis (Gilmore), 1931
Syn.: Troddon wyomingensis Gilmore, 1931.

3180 Fronto-parietal cap from the Lance Forma-
tion probably of Wyoming. Described and
figured by Gilmore — Ann. Carnegie Mus.
(1936/7), 25:109-112, Figs. 1-3 as Troddon

wyomingensis . Mentioned by Brown and
Schlaikjer — Bull. Amer. Mus. Nat. Hist.
(1943), 82:143.

Collected by Utterback

Ornithopoda(?), incertae sedis
33974 Presacral vertebra from the Morrison For-
mation of the Carnegie Museum Quarry at
Dinosaur National Monument.

Collected by Douglass et al.

SUBORDER STEGOSAURIA

Family Stegosauridae
G^nus, Stegosaurus Marsh, 1877
Stegosaurus ungulatus Marsh, 1879

There are two common and clearly distinct
species of Stegosaurus, S. ungulatus and S. sten-
ops. Both would appear to be represented in the
fauna at Dinosaur National Monument, based on
the characteristic radius and ulna. However, the
three mounted skeletons compiled from material
collected by Carnegie Museum of Natural History
parties at that site are all to a certain extent com-
posites, and it is not clear that bones from both
species have not been used in a single skeleton. No
attempt will be made here to separate individual
bones as to species, pending a detailed study of this
material sometime in the future.

11341 Partial skeleton consisting of the posterior
dorsals, ribs, complete sacrum, pelvis and
tail with tail spikes in place, and left femur
(not used in mount) lacking distal end, from
Carnegie Museum Quarry at Dinosaur Na-
tional Monument, Uintah County, Utah
(DNM 350). It forms the core of the mounted
skeleton placed on exhibit in 1940. Supple-
menting 350 in the mount is the contents of
a large block DNM 39/60AA consisting of the
complete left pectoral girdle, limb, and ma-
nus, left tibia, fibula and pes, seven cervi-
cals, as well as six dorsals and a caudal not
used in the mount as they duplicated material
from 350. The right radius, ulna and manus
and the right tibia, fibula and pes were re-
stored in plaster. One femur was taken from
197. The source of the other elements is un-
known. Pictured by Kay — Carnegie Maga-
zine ( 1940/H, 14:211 .

Collected by Douglass et al., 1920-
1922

11372 Composite partial skeleton from same local-
ity as above (DNM 39/60 and other material
not recorded), transferred to and mounted at
the Nebraska State Museum, University of
Nebraska at Lincoln, in 1946.

Collected by Douglass et al.

The following specimens were all collected in the
Carnegie Museum Quarry at Dinosaur National
Monument, north of Jensen, Uintah County, Utah,
in the Morrison Formation. No attempt is made to
differentiate them as to species. They were collect-
ed by Douglass, 1909-1922.

12000 Cranium and other skull fragments (DNM
365).

21727 Dorsal, right scapula, found associated with
Camarasaurus CM 11393.

21774 Dorsals, ? caudals, right scapula-coracoid,
originally part of CM 2969 (DNM 285/C and
H), and one right fibula. One dorsal was
transferred to the University of Cincinnati in
1941.

33917 Dorsal, ribs and scapula (DNM 39/65a).

33934 Centrum and two arches (DNM 270/A).
21711 Sacrum and two caudals (DNM 39/17).

33931 Sacral rib (DNM 39/76).

21766 Caudal (formerly catalogued as part of Apa-
tosaurus CM 3378).

21776 Caudal (DNM 231/A).

33935 Caudal (DNM 244).

33937 Four caudals (with DNM 130).

33938 Caudal centrum (DNM 120/Ra).

37005 Caudal and rib (DNM 39/60). Part of former
CM 11372 (see above).

33983 Two caudals (DNM 39/78).

36

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

Fig. 17. — Skeleton of Stegosiiiiriis "i{ngi/latus," CM 11341.

33939 Caudal centrum (with DNM 130).

33966 Distal caudal (with DNM 146).

33940 Caudal spine (with DNM 45).

33930 Three caudals and a rib (DNM 102, part).

33908 Chevrons (DNM 197/31) perhaps belonging
to 21770 below.

21768 Right scapula-coracoid (with DNM 120/Ra).

21770 Scapula-coracoid, right femur, three cau-
dals, rib and dermal spike (DNM 197/33).
Another femur associated with these remains
was used in the Stegosaurus mount with CM
11341.

38354 Right coracoid (DNM 226/10).

21765 Left humerus, radius and ulna (found with
Barosaurus skeleton DNM 310 and formerly
catalogued CM 11984 and CM 11985). This
limb belongs to Stegosaurus stenops.

21739 Right humerus (with DNM 240), perhaps
same individual as CM 21727 above.

21761 Right humerus (?DNM 39), formerly cata-
logued as part of Apatosaurus CM 3378.

33946 Right humerus (DNM 270), perhaps belong-
ing with CM 33934 above.

21767 Distal end of left humerus (DNM 348/B), for-
merly catalogued as part of Barosaurus CM
11878.

33945 Distal half of radius (DNM 39/60, part).

33920 Lower half of right ulna, carpal, metacarpal
I, metacarpal, transferred to Washington
with Barosaurus (DNM 340) neck and re-
turned.

33922 Carpal.

33923 Carpal.

33936 Ulna (DNM 39/60Q).

33932 Metacarpal I.

33933 Metacarpal.

33992 Incomplete left ilium (DNM 102/6).

38340 Left femur (damaged), tibia, fibula, astraga-
lus (DNM 320/A, B, C).

10001 Right femur, coracoid (DNM 197), trans-
ferred— present location unknown; probably
used in one of the three mounts in Pitts-
burgh, Lincoln, and Toronto.

21709 Left femur (DNM 319).

21755 Left femur (DNM 59).

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

37

c d

Fig. 18. — Cranium of Stegosaurus sp., CM 106 (not 105 as chalked on paroccipital process); a) dorsal view; b) right lateral view; c)
palatal view; d) posterior view.

21758 Femora, right tibia and astragalus (DNM
232).

33919 Right femur (cast).

21773 Calcaneum (DNM 120/G).

10003 Dermal plate (DNM 39/60), transferred.
33942 Dermal spike, three caudals, and end of ilium
(DNM 39/60Ca).

Stegosaurus sp.

106 Cranium and right ramus of mandible from

the Morrison Formation at Quarry D, Sheep
Creek, Albany County, Wyoming. (The

bones are incorrectly labelled CM 105, which
is that of a Cretaceous hadrosaurid.)

Collected by Peterson and Gilmore,
1900

88 Partial skeleton consisting of three dorsals,
sacrum, one caudal, scapulae, one coracoid,
ilia, pubes, ischia, femora, dermal plate, and
?ungual. Same data as the above. The fem-
ora are the only parts which remain in Pitts-
burgh, the pelvic bones etc. were transferred
to the Royal Ontario Museum where they
form part of a composite mount; a caudal

38

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

went to the University of Michigan (V
16779).

Collected by Wortman et al., 1899 and
by Peterson and Gilmore, 1900

579 Three vertebrae. Same data as above.

Collected by Peterson and Gilmore,
1900

580 Two vertebrae. Same data as above.

Collected by Peterson and Gilmore,
1900

598 Spine of vertebra. Same data as above.

Collected by Peterson and Gilmore,
1900

599 Dorsal centrum. Same data as above.

Collected by Peterson and Gilmore,
1900

600 Left radius (juvenile). Same data as above.

Collected by Peterson and Gilmore,
1900

36700 Right radius from the Morrison Eormation at
Quarry O, Ereezeout Hills, Carbon County,
Wyoming.

Collected by Gilmore, 1902

597 Several vertebrae from the Morrison Eor-
mation at Quarry E, Sheep Creek, Albany
County, Wyoming.

Collected by Gilmore, 1901

557 Eemora, left scapula from the Morrison Eor-
mation at one of the Sheep Creek quarries,
Albany County, Wyoming.

Collected by Peterson and Gilmore,
1900

21762 Scapula found at Quarry G, Sheep Creek,
Albany County, Wyoming.

Collected by Peterson and Gilmore,
1900

21737 One dorsal, three caudals, right scapula-cor-
acoid-humerus. These bones were formerly
catalogued as part of 1255 and are from the
Morrison Eormation at Quarry N, Ereezeout
Hills, Carbon County, Wyoming.

Collected by Gilmore, 1902

36696 Eemora, right scapula. Same data as above.

Collected by Gilmore, 1902

36697 Two vertebrae. Same data as above.

Collected by Gilmore, 1902

36698 Vertebra. Same data as above.

Collected by Gilmore, J903

581 Small dermal plate collected in the Morrison
Eormation at either Quarry C or E, Sheep
Creek, Albany County, Wyoming.

Collected by Gilmore, 1901

596 Incomplete right humerus and fragment of
limb bone perhaps belonging to Stegosaurus
from Quarry J, Sheep Creek, Albany Coun-
ty, Wyoming.

Collected by Gilmore, 1901

SUBORDER ANKYLOSAURIA

Eamily Nodosauridae
Genus Struthiosaurus Bunzel, 1871

Syn.; Cratueomus Seeley, 1881.

This genus is customarily included in the ill-de-
fined family Acanthopholidae. It is here referred to
the Nodosauridae on the authority of a recent study
of the group by Coombs (Columbia University
Ph.D. thesis, 1971).

Struthiosaurus austriacus Bunzel, 1871

Syn.: Cratueomus pawlowitschii Seeley, 1881.
Cratueomus lepidophorus Seeley, 1881.

The following casts are all from the Gosau Eor-
mation of Neue Welt, near Wiener Neustadt, Aus-
tria, and were collected by Pawlowitsch for Suess
over a period from the late 1860s and 1870s. The

originals are in the Geological Museum of the Uni-
versity of Vienna. Nopcsa referred them all to this
species — Geol. Hungarica Ser. Palaeont. (1929),
4:32-34.

972 Cast of cranium. The original, which is the
type specimen, was described by Bunzel —
Abh. Geol. Reichsanst. Vienna (1871), 5:11-
12, PI. 5, Pigs. 1-6; and Seeley — Quart. J.
Geol. Soc. London (1881), 37:628, PI. 17,
Pigs. 5-6.

970 Cast of right dentary, described by Seeley —
Quart. J. Geol. Soc. London (1881), 37:638,
PI. 27, Pigs. 9-10 as Cratueomus — and by
Nopcsa — Geol. Hungarica Ser. Palaeont.
(1929), 4:35, Pigs. 4, 6.

974 Cast of dorsal vertebra, described and fig-
ured by Bunzel — Abh. Geol. Reichsanst.

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

39

Fig. 19. — Casts of Stnithiosaunis aiistriaciis', a, b) anterior and posterior views of right femur, CM 971; c) dermal spine, CM 967.

Vienna (1871), 5:2-3, PI. 6, Figs. 1-3 and PI.
7, Fig. 24 — and by Seeley — Quart. J. Geol.
Soc. London (1881), 37:644, PI. 30, Fig. 3
where it is made a cotype of C. pawlowit-
schii.

976 Cast of a dorsal vertebra, described and fig-
ured by Bunzel as a crocodilian — Abh. Geol.
Reichsanst. Vienna (1871), 5:3-4, PI. 1, Figs.
24-26 — and by Seeley — Quart. J. Geol. Soc.
London (1881), 37:669-670, PI. 30, Fig. 5, as
a cotype of C. pawlowitschii.

968 Cast of a caudal, described and figured by
Bunzel — Abh. Geol. Reichsanst. Vienna
(1871), 5:4, PI. 2, Figs. 4-6 — and by Seeley —
Quart. J. Geol. Soc. London (1881), 37:646-
647, PI. 30, Fig. 4, as a cotype of C. paw-
lowitschii.

971 Cast of a right femur, described and figured
by Seeley as a cotype of C. lepidophorus —
Quart. J. Geol. Soc. London (1881), 37:664-
666, PI. 31, Fig. 5.

967 Cast of a dermal spine superficially resem-
bling a ceratopsian horn-core referred by
Seeley to Crataeomus — Quart. J. Geol. Soc.
London (1881), 37:650-651, PI. 28, Fig. 4.
Marsh, who first synonymized Crataeomus
with Stnithiosaunis, remarked on the resem-
blance of this specimen to a horn-core of
Ceratops, and later referred Stnithiosaunis
to the Ceratopsidae. However, Hatcher fol-
lowed by all subsequent authors has rejected
this suggestion.

964 Cast of a shorter, stouter scute, figured by
Seeley — Quart. J. Geol. Soc. London ( 1881),
37: PI. 30, Fig. 2, as Crataeomus.

966 Cast of a dermal scute.

975 Cast of a dermal shield, referred to Crataeo-
mus by Seeley — Quart. J. Geol. Soc. Lon-
don (1881), 37:651, PI. 28, Fig. 2.

969 Cast of a dermal bone, figured by Seeley —
Quart. J. Geol. Soc. London (1881), 37: PI.
28, Fig. 3.

40

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

Fig. 20. — Skeleton of Protacemtops iindrewsi, CM 9185.

973 Cast of an incomplete bone identified by
Nopcsa as an ilium of Struthioscmms\ Seeley
described it as a costal plate of Pleuropeltus
suessii — Quart. J. Geol. Soc. London (1881),
37:691, PI. 30, Eig. 15.

Ankylosauria, indeterminate

33975 Cast of a scute. The original in the British
Museum (BM R 1656) was collected in the
Wealden of Brook, Isle of Wight. It was
compared to Ceratops and Crataeomus by

Lydekker — Quart. J. Geol. Soc. London
(1890^), 46:185-186, and again in Catalogue
of the fossil Reptilia and Amphibia in the
British Museum (Natural History) (1890c/),
Part 4:256.

591 Vertebra from the Judith River Formation
near Havre, Montana.

Collected by Douglass, 1902
1609 ?Dermal plate from the Lance Formation,
Lance Creek, Niobrara County, Wyoming.

Collected by Utterback, 1905

SUBORDER CERATOPSIA

Family Protoceratopsidae
Genus Protoceratops Granger and Gregory, 1923
Protoceratops andrewsi Granger and Gregory, 1923

9185 Skull and skeleton, obtained from the Amer-
ican Museum of Natural History (AMNH
6471). It was collected in the Djadochta For-
mation at Shabarakh Usu (now known as
Bain Dzak), Mongolian People’s Republic.
Many parts were figured by Brown and
Schlaikjer — Ann. New York Acad. Sci.
(1940), 40; mandible, 198, Fig. I8c, 204, Fig.
20b, 209, Fig. 21a; scapula-coracoid, 229,

Fig. 26b; humerus, 232, Fig. 27b; radius, 233,
Fig. 28c; ulna, 233, Fig. 28a; ilium, 237, Fig.
30a; femur, 241, Fig. 32a. Pictured by Kay —
Carnegie Mag. (1946), 19:241.

Collected by American Expedition,
1925

9186 Hyoids from the above skeleton. It was de-
scribed and figured by Colbert — Amer. Mus.
Nov. (1945), 1301:5-9, Figs. 3-4.

11390 Cast of nest of six eggs. Same data as the
above.

23169 Egg shell fragments. Same data as the above.

1981

McIntosh— DINOSAURS of carnegie museum

41

Family Ceratopsidae
Genus Centrosaurus Lambe, 1904
Centrosaurus apertus Lambe, 1904

11374 Skull obtained from the Geological Survey
of Canada in 1925, and later transferred to
Museo de La Plata, where it is on exhibition.
It came from the Oldman Formation on the
south side of the west branch of Little Sand-
hill Creek, Red Deer River area, Alberta. It
was mentioned by Lull — Mem. Peabody
Mus. (1933), 3:3, 10, and is no. 42 in Stern-
berg's list of the Steveville Map.

Collected by C. M. Sternberg, 1919

Genus Triceratops Marsh, 1889
Triceratops brevicornus Hatcher, 1905

1219 Skull from the Lance Formation on Hell
Creek, Montana, which is mounted and was
placed on exhibition in 1905. It was pictured
in Holland — Carnegie Mus. Ann. Rept.
( 1905), p. 94 — and was mentioned by Lull in
Hatcher, Marsh, and Lull — Monogr. U.S.
Geol. Surv. ( 1907), 49:182; and in Mem. Pea-
body Mus. (1933), 3:3, 14, and described 1 19.

Collected by Utterback, 1904

Fig. 21. — Hyoid bones of Protoceraiops undrewsi, CM 9186; a)
first ceratobranchial, external view; b, c) second ceratobranchi-
als, dorsal view.

Triceratops sp.

30744 Both rami of the mandible from the Lance
Formation on Sheep Mountain, Carter Coun-
ty, Montana.

Collected by Kay, 1938

Fig. 22. — Skull and mandible of Triceratops brevicornus, CM 1219. right lateral view.

42

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

30746 Teeth from the same horizon and locality as
the above.

Collected by Kay, 1938

998 Partial skeleton consisting of fragments of
the skull, 14 vertebrae, ribs, right scapula,
right coracoid, right humerus, left ulna, part
of femur, etc., from the Lance Eormation on
Snyder Creek, Montana.

Collected by Peterson, 1902
1070 Nasal horn. Same data as above.

Collected by Peterson, 1902
249 Nasal horn from the Lance Eormation on
Lance Creek.

Collected by Hatcher, 1900
1619 Two horn cores from the Lance Eormation
on Lance Creek, Niobrara County, Wyo-
ming.

Collected by Utterback, 1906
1651 Two horn cores from the Lance Eormation
on Hell Creek, Montana.

Collected by Utterback, 1906
1618 Rostral bone, left squamosal, both rami of
mandible, half dozen vertebrae, sacrum,
right scapula, coracoids, left humerus, ra-
dius, right ilium, pubes, ischia, left femur,
right tibia, etc., from the Lance Formation

of the C. A. Sheldon Ranch, Niobrara Coun-
ty, Wyoming.

Collected by Utterback, 1906-1907
The following specimens were all collected by

Hatcher from the Lance Formation on Lance

Creek, Niobrara County, Wyoming, in 1900.

247 Cervical.

248 Cervical.

253 Vertebra.

241 Dorsal centrum.

242 Metapodial.

255 Phalanx.

257 Phalanx.

Ceratopsidae, indeterminate

11377 Left humerus from the Oldman Formation
near Steveville, Red Deer River area, Alber-
ta. It was obtained from the Geological Sur-
vey of Canada in 1925.

Collected by C. Sternberg

12103 Right tibia from the Lance Formation, Sheep
Mountain, Carter County, Montana.

Collected by Kay, 1937

38345 Tooth from the Lance Formation, Hell
Creek, Garfield County, Montana.

Collected by Silberling, 1904

DINOSAURS, INCERTAE SEDIS

The following specimens from Dinosaur National
Monument are so incomplete as to make their iden-
tifications difficult or impossible, but they are prob-
ably dinosaurian. They were collected by Douglass
and his assistants between 1909 and 1922.

33915 Small centrum associated with AUosaurus
CM 11843.

3393 Fragments (DNM 6).

33906 Chevron (DNM 47/2).

33911 Long slender bone (with DNM 130).

3385 Caudal (DNM 185).

33910 Phalanx (DNM 79).

33913 Small caudal (with DNM 39/65).

33943 Juvenile right ulna (DNM 39/60?).

36688 Foot bone from Quarry N, Freezeout Hills,
Carbon County, Wyoming, in the Morrison
Formation.

Collected by Gilmore, 1902
36681 Strange bone. Same data as above.

1075 Part of mandible, caudals, etc. from the Ju-
dith River Formation of Mud Creek, Fergus
County, Montana.

Collected by Utterback, 1903
38330 Metatarsal from the Lance Formation of
Sheep Mountain, Carter County, Montana.

Collected by Kay, 1938

— Egg shell fragments from Aix-en-Provence,
France. They may be dinosaurian (perhaps
belong to Hypselosaiints prisons) .

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

43

LITERATURE CITED

Abel, O. 1910. Die Rekonstruktion des Diplodociis. Abh.
Zool.-Bot. Ges. Wien, 5; 1-60.

Berman, D. S. 1971. Stone bone, a new touch exhibit in the
Dinosaur Hall. Carnegie Mag., 45:93-96.

Berman. D. S. and J. S. McIntosh. 1978. Skull and relation-
ships of the Upper Jurassic sauropod Apatosaurus (Reptilia,
Saurischia). Bull. Carnegie Mus. Nat. Hist., 8:1-35.

Brett-Surman, M. K. 1979. Phylogeny and palaeobiogeog-
raphy of hadrosaurian dinosaurs. Nature, 277:560-569.

Brown, B., and E. M. Schlaikjer. 1940. The structure and
relationships of Proioceratops. Ann. New York Acad. Sci.,
40:133-266.

. 1943. A study of the troodont dinosaurs with a descrip-
tion of a new genus and four new species. Bull. Amer. Mus.
Nat. Hist., 82:115-150.

Bunzel, E. 1871. Die Reptilfauna der Gosauformation in der
neuen Welt bei Wiener-Neustadt. Abh. Geol. Reichsanst.
Vienna, 5: 1-18.

CoGGESHALL, A. S. 1951. How “Dippie” came to Pittsburgh.
Carnegie Mag., 25:238-241.

Colbert, E. H. 1945. The hyoid bones in Protoceratops and
in Psittacosaurus. Amer. Mus. Nov., 1301 : 1-10.

Colorado Museum of Natural History. 1947. Popular Se-
ries no. 1 , 7th ed., 80 pp.

Coombs, W. 1971. The Ankylosauria. Unpublished Ph.D. dis-
sert., Columbia Univ., New York, 487 pp.

Cope, E. D. 1877. On a gigantic saurian from the Dakota epoch
of Colorado. Pal. Bull., 25:5-10.

Dodson, P. 1975. Taxonomic implications of relative growth
in lambeosaurine hadrosaurs. Syst. Zool., 24:37-54.

Ellinger, T. U. H. 1950. Camarasaurus annae, a new Amer-
ican sauropod dinosaur. Amer. Nat., 84:225-228.

Galton, P. M. \911a. The ornithopod dinosaur Dryosaiinis
and a Laurasia-Gondwanaland connection in the Upper Ju-
rassic. Nature, 268:230-232.

. 1911b. The Upper Jurassic ornithopod dinosaur Dry-

osaiirus and a Laurasia-Gondwanaland connection. Pp. 41-
54, in Paleontology and Plate Tectonics (Robert M. West,
ed.), Milwaukee Public Mus., Spec. Publ. Biol. Geol., 2: 1-
109.

. 1980. Dryosauriis and Camptosaurus, intercontinental

genera of Upper Jurassic ornithopod dinosaurs. Mem. Soc.
Geol. France, 59: 103-104.

Galton, P. M., and J. A. Jensen. 1979. A new large theropod
dinosaur from the Upper Jurassic of Colorado. Brigham
Young Univ. Geol. Studies, 26( 1): 1-12.

Gilmore, C. W. 1925«. A nearly complete articulated skeleton
of Camarasaurus, a saurischian dinosaur from the Dinosaur
National Monument, Utah. Mem. Carnegie Mus., 10:347-
384.

. 19256. Osteology of ornithopodous dinosaurs from the

Dinosaur National Monument, Utah. Camptosaurus med-
ius. Dryosauriis altus, Laosaurus gracilis. Mem. Carnegie
Mus., 10:385-409.

. 1936(1. Osteology of Apatosaurus, with special refer-
ences to specimens in the Carnegie Museum. Mem. Car-
negie Mus., 11:175-300.

. 19366. Remarks on a skull cap of the genus Troddon.

Ann. Carnegie Mus., 25: 109-1 12.

Glut, D. 1972. The dinosaur dictionary . Citadel Press, Secau-
cus. New Jersey, 218 pp.

Haas, G. I960. A proposed reconstruction of the jaw muscu-
lature of Diplodociis. Ann. Carnegie Mus., 36: 139-157.

Hatcher, J. B. 1900u. The Carnegie Museum paleontological
expedition of 1900. Science, n.s., 12:718-720.

. 19006. Vertebra! formula of Diplodociis (Marsh). Sci-
ence, n.s., 12:828-830.

. 1901«. Some new and little known fossil vertebrates.

Ann. Carnegie Mus., 1:128-144.

. 19016. Diplodociis (Marsh), its osteology, taxonomy,

and probable habits, with a restoration of the skeleton.
Mem. Carnegie Mus., 1:1-63.

. 1901c. On the structure of the manus in Brontosaurus.

Science, n.s., 14:1015-1017.

. I90l(/. Structure of the fore limb and manus of Bron-
tosaurus. Ann. Carnegie Mus., 1:356-376.

. 1901c. The dermal covering of Tracliodon. Ann. Car-
negie Mus., 1:386.

. 1902. Field work in vertebrate paleontology at the Car-
negie Museum for 1902. Science, n.s., 16:752.

. 1 903u . A new sauropod dinosaur from the Jurassic of

Colorado. Proc. Biol. Soc. Washington, 16:1-2.

. 19036. A new name for the dinosaur Haplocanthus

Hatcher. Proc. Biol. Soc. Washington, 16:100.

. 1903c. Discovery of remains of Astrodon {Pleiirocoe-

lus) in the Atlantosaiiriis beds of Wyoming. Ann. Carnegie
Mus., 2:9-14.

. 1903(/. Osteology of Huplocanthosaiirus, with descrip-
tion of a new species, and some remarks on the probable
habits of the Sauropoda and the age and origin of the At-
lantosaurus beds. Mem. Carnegie Mus., 2:1-72.

. 1903c. Additional remarks on Diplodociis. Mem. Car-
negie Mus., 2:72-75.

. 1 903/. Vertebrate paleontology at the Carnegie Mu-
seum. Science, n.s., 18:569-570.

Hatcher, J. B., O. C. Marsh, and R. S. Lull. 1907. The
Ceratopsia. Monogr. U.S. Geol. Surv., 49: 1-300.

Hay, O. P. 1908. Dr. W. J. Holland on the skull of Diplodociis.
Science, n.s., 28:517-519.

Holland, W. J. 1900. The vertebral formula in Diplodociis.
Science, n.s., 1 1:816-818.

. 1905. Annual Report of the Director for the year ending

March 31, 1905, Carnegie Museum. Carnegie Mus. Ann.
Rept., 70 pp.

. 1906. The osteology of Diplodociis Marsh. Mem. Car-
negie Mus., 2:225-264.

. 1908. Dr. O. P. Hay on the skull of Diplodociis. Sci-
ence, n.s., 28:644-645.

. 1909. Twelfth annual report of the Director, Section of

Paleontology, The Carnegie Museum. Carnegie Mus. Ann.
Rept., 93 pp,

. 191 1 . Note on the finding of remains of dinosaurs. Ann.

Carnegie Mus., 8:2,

. 1 9 1 5rt . Heads and tails: a few notes relating to the struc-
ture of the sauropod dinosaurs. Ann. Carnegie Mus., 9:273-
278.

. 19156. A new species of Apatosaurus. Ann. Carnegie

Mus., 10:143-145.

44

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

. 1916. [Note on the discovery of a skull of D//7/oJoc«.v.J

Ann. Carnegie Mus., 10:3.

. 1919. Twenty-second annual report of the Director,

Section of Paleontology, The Carnegie Museum. Carnegie
Mus. Ann. Rept., 87 pp.

. 1924;. Description of the type of Uintasaurus doug-

lassi Holland. Ann. Carnegie Mus., 15:119-138.

. 19246. The skull of Diplodocus. Mem. Carnegie Mus.,

9:379-403.

Horner, J. R. 1979. Upper Cretaceous dinosaurs from the
Bearpaw Shale (Marine) of southcentral Montana, etc. J.
Paleo., 53:566-577.

Huene, F. von. 1914. Uber die Zweistammigkeit der Dinosau-
rier, mit Beitriigen zur Kenntnis einiger Schiidel. Neues
Jahrb. Min. Geol. Pal. (Beil.-Bd.), 37:577-589.

Kay, j. L. 1940;;. The Antrodemiis. Carnegie Mag., 13:303-
304.

. 19406. “Behold the mighty dinosaur.” Carnegie Mag.,

14:211-215.

. 1942. The king of dinosaurs. Carnegie Mag., 16:200-

202.

. 1946. Protoceratops comes to town. Carnegie Mag.,

19:241.

. 1951. More dinosaurs. Carnegie Mag., 25:90-91, 102.

Leidy, j. 1870. Remarks on Poicilopleuron valens, CUdastes
intermedins, Leiodon proriger, Baptemys wyomingensis,
and Eniys stevensonianus. Proc. Acad. Nat. Sci. Philadel-
phia, pp. 3-5.

Lull, R. L. 1933. A revision of the Ceratopsia or horned di-
nosaurs. Mem. Peabody Mus., 3:1-175.

Lull, R. S., and N. E. Wrigh r. 1942. Hadrosaurian dinosaurs
of North America. Spec. Papers, Geol. Soc. Amer., 40:1-
242.

Lydekker, R. 189(V;. Catalogue of the fossil Reptilia and Am-
phibia in the British Museum (Natural History). Part 4. Con-
taining the orders Anomodontia, Ecaudata, Caudata, and
Labyrinthodontia; and supplement. British Mus. (Nat.
Hist.), London, 295 pp.

. 18906. On a peculiar horn-like dinosaurian bone from

the Wealden. Quart. J. Geol. Soc. London, 46:185-186.

Marsh, O. C. 1877. Notice of new dinosaurian reptiles from
the Jurassic formation. Amer. J. Sci., 14:514-516.

. 1878. Principal characters of American Jurassic dino-
saurs. Part 1. Amer. J. Sci., 16:411-416.

. 1879. Notice of new Jurassic reptiles. Amer. J. Sci.,

18:501-505.

. 1896. The dinosaurs of North America. Sixteenth Ann.

Rept., U.S. Geol. Surv., Part 1, pp. 133-244.

McIntosh. J. S., and D. S Berman. 1975. Description of the
palate and lower jaw of the sauropod dinosaur Diplodocus
(Reptilia: Saurischia) with remarks on the nature of the skull
of Apatosaurus. J. Paleo., 49:187-199.

Nopcsa, F. 1929. Dinosaurierreste aus Siebenburgen. Geol.
Hungarica Ser. Palaeont., 4:1-76.

Osborn, H. F. 1899. A skeleton of Diplodocus. Mem. Amer.
Mus. Nat. Hist., 1:189-214.

. 1904. Manus, sacrum, and caudals of Sauropoda. Bull.

Amer. Mus. Nat. Hist., 20:181-190.

. 1905. Ty/;;/;;!o,vnz;;7;i and other Cretaceous carnivorous

dinosaurs. Bull. Amer. Mus. Nat. Hist., 21:259-265.

. 1906. Tyrannosaurus, upper Cretaceous carnivorous

dinosaur (2d comm.). Bull. Amer. Mus. Nat. Hist., 22:281-
296.

. 1912. Crania of Tyrannosaurus and Allosaurus. Mem.

Amer. Mus. Nat. Hist., n.s., 1:1-30.

. 1917. Skeletal adaptations of Ornitholestes, Struthio-

niimus. Tyrannosaurus. Bull. Amer. Mus. Nat. Hist., 35:
733-771.

OsTROM, J. H., and J. S. McIntosh. 1966. Marsh's dinosaurs.
Yale Univ. Press, New Haven, 388 pp.

Peterson, O. A., and C. W. Gilmore. 1902. Elosaunis parvus,
a new genus and species of the Sauropoda. Ann. Carnegie
Mus., 1:490-499.

Romer, a. S. 1966. Vertebrate paleontology. Univ. Chicago
Press, Chicago, 3rd ed., 468 pp.

Russell, D. A., and T. P. Chamney. 1967. Notes on the bio-
stratigraphy of dinosaurian and microfossil faunas in the
Edmonton formation (Cretaceous), Alberta. Nat. Hist. Pa-
pers, Nat. Mus. Canada, 35:1-22.

Seeley, H. G. 1881. The reptile fauna of the Gosau formation
preserved in the geological museum of the University of
Vienna. With a note on the geological horizon of the fossils
at Neue Welt, west of Wiener Neustadt, by Edw. Suess.
Quart. J. Geol. Soc. London, 37:620-707.

Shepherd, J. D., P. M. Galton, and J. Jensen. 1978. Addi-
tional specimens of the hypsilophodontid dinosaur Dry-
osaurus altus from the Upper Jurassic of western North
America. Brigham Young Univ. Geol. Studies, 24(2): 1 1-15.

Sternberg, C. M. 1950. Notes on the dinosaur quarries [on
Map 969A Steveville west of fourth meridian. Alberta. Geol.
Surv. Canada].

Stovall, J. W., and W. Langston, Jr. 1950. Acrocanthosau-
rus atokensis, a new genus and species of Lower Cretaceous
Theropoda from Oklahoma. Amer. Midland Nat., 43:696-
728.

Wagner, J. A. 1861. Neue Beitrage zur Kenntniss der urwelt-
lichen Fauna des lithographischen Schiefers. 2. Schildkro-
ten und Saurier. Abh. Bayer. Akad. Wiss., 9:65-124.

White, L. L. 1950. Duquesne University collection of dinosaur
bones. Unpublished Master’s dissert., Duquesne Univ., 56

pp.

White, T. E. 1958. The braincase of Camarasaurus lentus
(Marsh). J. Paleontol., 32:477-494.

Woodward, A. S. 1889-1901. Catalogue of the fossil fishes in
the British Museum. Part 1, 474 pp.; Part 2, 567 pp.; Part
3, 544 pp.; Part 4, 636 pp. British Mus. (Nat. Hist.), London.

198!

McIntosh— DINOSAURS of carnegie museum

45

O 1 ^ S

£ -S

S 3

s I e

5 2 3

X'

tH

<L»

X)

£

z

D

3

CJ)

o

e

o
>,

^ (

'yi o o j. -

. U CJ o ^ .

a: 1 a: 1

c c

Lij u o a

w

6Jj 0£( 00

C C O- c

._ {J ._ 4^

E £

w u w 4;' O ^ O

^ X' ^ ^ ^ ^

\ a (A ^

■>^>>§>.§>. §

: o t o fc o t o

s o 3 u 3 u § u

^ O' O' O'

oc c c c c

o o o o

I 'E 'E f I

.5000 o
^ S S S S

2 I

^ 00 c

4^ o> O 4>

w O >. o

U ^ S u

^ ^ ^

00 ^

Q i
>>=>.=
t o t O

• = u = -s
E a E I
O u o Z
>, <u >, ,

^ ^ ^

u

4i OX) 4> 00

E U

^ 0)

> 5

£ U E U

E U
o

>< rv

^ E

E U
o ^

^ E

E U
o ..
>«

^ 5

OX) 4)

E u

00 4) 0X1 4> OX) 4)

E U
>' 2
3: 5

E U
° o

^ S

E O E U
>■ 5 >■ 2
^ E s S

E u £ '-' 2

S g S

_]

= S
=:

X

)-H

Q

Z

U

Oh

Oh

<

C3

U

>.

X

T3

<U

hJ

C/D

C

<U

s

"o

aj

a

in

u

3

3

y)

O

3

I I I I

I I - I

^

H- ^ S

I I

I I

I I I I I I I I I I I

E

> ju _ ^

! E I i

3 2 s S

£ a, a,

S _

"S o

5 5

■5 c -g _

is £-i i

o g o g
2- S
t o

_> i

D O

O

O .12 ,« ox>

ill

S i; E

3 -a 3 .2

S I S' S

(£ o

I I I I I I I I I I I I I I I

E s -a

u O

J Q

s£j \o

o

Cl 5 <3 o

—

46

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

— So

CJO CC Ot) BO

£ ^ ^ £

iy5

_ _ _ w Lu w
c?5 D D ui ai

S E
S O

±•^30

c c c c

O o O o

o o o o

U

II

U M

II

X> > TJ >

O ' O '
Q- ^ Q. ^

O t > t .>
!S § Qi g Oi
O' O

o o o o

^ ^ ^
iJ ij V <u
1» 4> D

u u u u

o o o o

UUUU UOUU

3 3 3 3 3 3

CiO BO BO 00 BO 00

U 0> U U <U U

U. U. U. U. U. U. I

4> <U D «J p'

U U U U c

T3T3T3T3 T3X>'UT3-C

llll

CO C

.= o

u E a

Q ^ Q

.£ U .£
E a £

U ^ ^ «
' ^ O

3 L_ 3

©to
O g U

o

a

U .£
a E

P- 2
jz 3
^ o
« O

Xj ±1 — —

.£ U .£ U .H O

E a £ a E a

o w o o> o w

.3 3
£ £
o <

^ ^ ^ ^ W ^

i u id id i

' © h' = ^ ^ ^

O C O i- O 1- O

O 2 O 5 u g U

o .£ O .£

a £ a £

<u o 4> o

« o> >.

, c/^ ^ ^

O E =->
_ o >.

S' ^ t

4> ^ g

oi o:

J3 £

0) <U 0) OJ

■© ■© T3 -3

>X — —

X ^
0>

Q c

Z -3

S §

^ u

&H

<

I I

I I

^ 00

I I I I I I I I

rr ^

9\ 0\ ^

82

I ^

I I I I I I I I I

TT ^ ^ 00

I I

I I

I I

® s

a 3;

^ E
•c .2

^ op £

•= -3 ' £ Q

■g ^ .2 BO E

o o o .2^

•- 3 3 i2

“ -s

,'i E '£ ~3 ■^S-S^

t»

— .Xu

•3 *3 *3 •©

■3 -3 *3 ■© S

0000

•3 -3 -3 -3

1 S X X

0000

X X X X ^ ^

1981 McINTOSH— DINOSAURS OF CARNEGIE MUSEUM 47

48

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

X -n

l-H TD

Q ^
Z .S

W c

Oh ^

<

X> X)

S S I s

D D c/5 D

c c c c

o o o o
S S S S

□ 3 3 3
O O O O

4/ 0) 4>

lU 0>

s s s s

ci:: dr; Q/ in::

jz ^ ^ x:

I I I I

I I I I I I

Os 0\

a r=
a 0)
5 ^

S E iS
i- ao

B O ~

O -3 -3

s s

O' D

E

E >; •

O '0> S '4. s

s o

oi o

I

■3 O

I I

D

-3 — 3

E S E .E

3 ^

cjj ca

3 C

■3 *3 T3 TJ
■C 'C 'C 'C

O O O O

X K X Dh

o o o o

^ 3

o — —

- S o a:

E .{i* 4> J= "P

= 2 = 'Efl 5 3

O 2 o g ;H f

O O J5 D 2

s (_) (_)

s =

N E

at o
<L» >.
[J-. ^

z >-
>> 5
t o

== ON

^ p a

6 J2 m 2 aC Ji

c 22 3. — 3 00

— '<T

sO — .

— ^

E S
-S CJ

5 c o -c

o .£ t C/ o

N E tu o z

4> o „ 9

U ^ = -

“ o

Z i’ m "I

>. E >. 4>

§ £ 4

g o g (2

o o

u

: o

o

z

ac o> 0&

C 4> C

1 6 1

° s °

O E

I I

?? _ = 5 -s

E 2

3 u 3

= ^

-J

I I

o — —

1

o .1

I i> c c
iS « iS 3
§ ^

X

c —

•3 4>

E ^

^ « u

I I

NO —

E -E

.1 13 3 «r

s-

It

1981

McIntosh— DINOSAURS of carnegie museum

49

z z

Q Q

t o
^ u
O'

O 3

g

Ml o

•E "u
E =o

u ^ u

s

i S

O' D O D o

I E i

= o 3 o

S u s o

OJ ^ flj *3 ^

'5L *- '5b —

w c w c w

§ 5 E 5 5

I o 5 o 5

I I ^ I ^

I <U C V c

3 o S o

: S u S CJ

E u

O D
E =;

O D

, e

C u c u C 1)

0 3 0 3 0

0 3 0 3

ij gj SZ gj

5 O' 5 o 3 O' 5

C u C u c 1> c

3 3 (/5 3 tzi □

0 3 0 3 0 3 0

o s o U S O

^ 4; -3 4> ^ 0-3

iH S) iS Si i3 OX) w

S O

t D
U

0 p

it

(/) 3
3 O

0> ^
'S 2

O) c

1 ^

U

£ S

X -3

5? (U

Q c

z -B
S §

Oh U
Oh
<

I I I I

I

o —

-o —

2 i ^ 3 ^

'o M 3 O —

O E

CJ> o. '

S £

—

o [J
C 4)

X- _ -o

3

■- 2

z £

E X X

c S g

•2 o

S A

Cl. ^

■5

X Q

X —

o —

50

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

X -a

HH <u

Q c

W §
Oh U
Oh
<

OS 5

1^0
§ -c

= 03

4> -5
'SI) -O

•U C

I 5

O' 5 O' D O' ;

3 0 3 0 3

O S o s o s

o W) -2 'Sii iS '5b

— >u c o c i>

u b o £ 5 £

c c ^

I o 5 o :S S

§ W 4>

f I ^

J o o

n o w 5

C

i ^

. I .E

O c«

S c

U

I I I

S ^ S =
I 5

^ ~ 3: ;:: g I

z ■» z -1 ^ 3

2 s !2 -

I I

-3 >, ® M O

O. 1>2

Q

— X) 4>

<; o Cj D 2

E £

§ -j

.C t:

I I

I I

_ CJj

— x:

Z Z

Q Q

O D
E >;

D O ^ O D

5 2 ^ SO

0 3 0 3 0

u

X X o

I w 'Sb i3 '51) iS

c u c <u c

5 I 5 I 5

u 0

a 3

e ^

s o

O' p

o D
E =;

o P

B

s O S u S P

s •§ s •§ s ■§

3 2 3 2 3 S

0 D O D O O

1 - I - I

§ j> = <u 5

3 0 3 0 3 0

S o s o 2 O

Ci£ X bfi —

O C 4> C

£ D E D £ :

o o o

^ 4»

£ ^

-TT = o ;£

'O ,4> -— 3:

3 O w

o iS

H-. O.

c —

’§1 — ”5) E ^

£ 5 (5

? E 3 ^

os —

£ D E b

O U u

o 3

£ D

— sD

c ^ Ji
3 -x u_. ^

I g ° I

~ -5 E

V C 4> "c

“ E ^ I

'§) *o 3

.0

Q

i981

McIntosh— DINOSAURS of carnegie museum

5!

X

Q
X
w

u
Pu

<

8 & 8 a

i I i

O' D a

G e

O' D
; G ^

« I
' 5 o
e

D 2

sugusosus

•U'5.2^

'5i 2 '5j 2 'ob w ‘5b ^

^ .E ^ .5 ^ .E ^ .E c

|p£D|D|Djg
O U O U ^

§ ^

-= 72

CO ■«

"S -o

Xi ^ Xi

E E E

o «;>

^ X

S 2 S I

0 0 0 <

8 8

8 8

8 8 8

A C

§ ^

c r c

r~i ^ rtj

I i

o

< <

s

z

Q

i 2

o a

&

O 3

G s;

a a a ^

ml

2 2 0 2 0

1-5 2
o g .E

s S 3

o

P 3 O 3 O

S § § I

e s e ■§

2 g g 3

O O o 'o'

MM

I MM

On On ^ Os Os f"4

t 3 =

<2^2

1 1

“ t

i

2 1

2 J O

C' C' 2 O

o o 4j £
O O u 0,

o o o ^

r-j ri o\ ON

SO NO NO On — ‘ — ‘

cS

S-o

4J = ^

s u s o s o

■a 2 M 2 M 2

« C (U C « C

I 5 £ 5 £ 5

u u o

3 2 ^ 3 2 3 2

O' D « O D o 3

I ^ g I =; I =;

^ § o ^ ^ 5

= o ' 3 O = O

S O ^ S O S O

4^ ^ 2 ^ ^

'ob 2 If '5b 2 M 2

w E w E ^ E

E P “ ^ D ^ D

U SO U

I I s

I I

ri

^ 2

I I S

■3 2

= E

2 2

s

' S £
o o

3 O'

E

c 2
3 lo
O 3

o S

■G

2 M

3 Ot

5 I

.a E
2 Ji

2- ■§ S ^
■g ^ f' 2

I I -c o

“ ° 1 f-

I G E 1

2 S;

■> NO 3

^ c3

P 2

lU

2 2*^

PP « 3
2 ^ ^
liill

■sa

1 I 2 i 2 i 2

3 O 3 O 3 O' 3
>^ £ E E P

^3^3^3 —
C 4J C 4^ C C

3 w 3 wa 3 w 3

0 3 0 3 0 3 0

o S o s o s o

2 'a 2 -a 2 ^a 2

C flj C 4; C (U C

D I 5 I D I P

U U U

S ^

p so ^
E ^ .3

= s ^

^ U -u
^ T3 E
— Q C

■ o

c "i I

3 ^ 2

■ g 3 ^ ^

•O ^ J3 S

Q H

a p
£ >;
1 1
I O

Cj -O

■5b 2

flj c

b 5

S

Z

Q

i I

O' P

E =;

i I

3 ©

S ^

O -3

2

§ ^

CJ

E ^ t S
i E .g

3 s2 ^ 3

l-I?
i i| I

o 3 2 os
o

g S

52

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

S O = . 3

: o c

3 S B

o ^ o ^ o

I (j 5i s Si s

■ JZ JZ

3 2 3 2 3 2

a D O' O' D

i ^ I ^ I i

u C 1> c lU c

I/' 3 c« 3 ^ 3

3 0 = 030

S u S u S u

u u

O' p

E

S o

u .c

M 3
u

o 5

E ^

s s

z z

Q Q

S S
z z
Q Q

D iS 3 3 2

O p O p O D

E E i ^

x: o

X 5

S E

O p
E ^

O p
E

O p
E

;E;iE^E:iE

S P

V J=

3030^0^0333030 = 030 = 030

SPSPSPSPSUSPSPSUSPSP20

(jj >C ^ JZ (U XI (]j ^ ^ XI

'oil 2 ’5b 2 'ob 2 '5b 2 '5b 2

UC4jCliC«^C4>C

£5|o£5£5|p

u O u u u

'5b 2 '5b

W C 4>
3 XT C

60 w

§ 5

& 5 O' O ^ ^

I ^ I i I s

4) C « C 3 O

c/5 3 t/) 3 O CD

= 0 = 0^^

S o s u .2 %

1— — ^

'5b '5b U S3

Ji .E y .S w w

I s =* § s

U O m

X

I— H

Q
Z

w

CSh u
Oh

I I I

I I

I I

^ —

= is

'O’ 4>

O ^

~ o —

00 ^

sO — —

0\ 0\
ri

Tf Tf hT

— NC- —

Ilf

1 = 1;^
on u

•7; *H CC

E <

S 8 S

o — —

— — 0\

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

53

ass

S S S

ftfi hfi Q

s s

s

•E

2 M fxi

ell

c © W
S E i2
>- rr oa

l°l

® Rfi ofi ori

s s s

§5

<§5

a D

E ^ 6 e ^ E

D O' D

=; g

il

£ t j-

5 O' 3

^ I ^

c S C

I g'O'

E g

S U

A

s o

« -5
■a 2

lU c

i 2

U

JZ m JZ JZ

•G) iS -Sa 5 M 2 •© 5
(U C w _c « ,c o .c

EdE5|5|d

u u u u

0 3 0 3 0

S u
o -5
•a 5

c

E 5

SOSO

■a 2 -a 2

45 C fli C

I 5 E 5
u u

o £ ^

. S o

3 2

a D

s o
•SA

j=

.£ t

"I

g> O' ;

11

« .

i “

u I ;

5

M

O

^

I2! ilii

'to § “aoao
Scgogcgc

3 C ^ 0^ 3 <5^ 3
-'-=> 3 0 3 0

2 U 5 U

.‘ 4> -C -5

i -a iS -a iS

I 4^ C 4^ C

' I p I 5

U D

s o £ ^

I (;j

' -a

5

z

3 '

a

2 I a 5

a, E S
9j O 3 1:i

< ^ 43

>. 5 ‘So 2

_ _ >.

as

3 ? 3 “OP

o .5 o ,E ^ . _ .

jjsseE5;|>,

g?,g?,i|ig

I o I u

0 i £? i ^ ^

S ® ® 3

2 O 2 u i-

3 D

5 £ P

4> -S

'a 3

I 5

u

E ’E

X

Q

Z

w

^ o

Oh
<

T3

SJ

3

•S

C

O

r"

£ 2

£

2 ?
z z

®N — — .

— — OS

— 3

m M

-O —

« <a

£ w
o

s e

3 o

1 1

2 I

r E s-

33 w

E S 3

^ 2
2 E

■ESS

e ° c

.2 s 2

I g S

C 3 ^ u BS

--||E

■s:

2 .3 ^

-I O «f

S -S 3

O 5 E

■c := 2

z -c S

is

E U

54

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

X

HH (L>

Q g

Z -a

W §

Oh (J
Oh

t; 'g t

O' D
E ^

ilf

s O

» O'
> E

a O

DOS
E ^

O D

i.. E ^ £

cz ^ c ^ c

D O D

O D
E ^

0 3 0 = 030

E ^ U

O 3 o >.

O D
E ^

E .2

O g

U ,S
o. E

SUSUSUSOSUSUSUSO t g>SU

S S S

I I

— —

D t O
U

do w DO L> d£ J: do

lU c V c

5 ? ^

.-r o ^
£ ^ -a

a

I ^

- "E i E

E

.5 «

0 D

1 i-

g D
u

s &

S u

O D
, E ^

3 .-
W 3
^ 3

3 O

S o

D O D O D O D

O

■1-3-3

0> C V c u

0303030=0

CXi W DA

S D 2 u S

ca — 00 — tfi

I § I

I 3^ I

s

c D

O I
E ,S

•3 s:",-

Q 5 3 :§•

Q I

,E £

O S U S O S O

•2 ’db 2 ’db 5 ‘db 5

_c <u c 4> _c w _c

5 £ 5 I D £ 5

U U CJ

-: z: ^

— Hj-

dO ‘C dX; <!

£ S

3

,w O

1981

McIntosh— DINOSAURS of carnegie museum

55

X

HH

Q

z

w

Oh (J
Oh
<

O' D O' D
E ;; E

u u

1

u- O

>
O '

a. ^

o 3

t >

o

a 5 o 5 a 5

I ^ e ^ I ^

C c C

C/) 3 un 3 (/I 3

3 0 3 0 3 0

s u S u s u

HJ JZ HJ fU

z - 2 S;

CO ^ CQ 7 <

, — , 00 —

o - 2: ri
’ ^ — 02

CO

03

^ i M

■3 O. 'C

H3 OS ^

's ^ «'

"ra 5 ^ 3

S S G ^

Q c/3

o -Tr, Tr,

£33
r3 o O

O Q Q

3 «

a D
E ^
S c

3 o
S U
1} .n

■a 2

aj c

£ 5

cy D

E ;;

MOD

1 I ^

O' D ” 2f
I >■ a E

r ■?S 4» O

cy o
E =;

E

1-. o

Oi >.

I ^

^ 1

' o

>■ Oh

1 E s
° S ° z

» 1 > Q
' O ' -c
>■ CL, >< £■

O 3
E

O ■£ O ■£ U -c U

O D
E

. S u s O

_ JZ

3 O i

5 u

a, JO O

g .£ t § g .E
E D g u I 3
u o 0

S U

.0 p

S u

— 3-3 H oa

o —

= -I i 5

u o

3

—

E 3

I I

I I

•C TS

E .E

□

OXJ — 1>

“ ^ t .>

3 Oi

O'

I

03

> fc

I'i

J

ON <

v5

s ::

o

tz

s u
•a B

lU c

I 5

u

S 2 2 2

&| &|

I I

•C “ —

o Z *

^ c/5 "3

a"l

- 'C

^ 'C
JO 4>

^ E

56

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

D g P g
0 U U

'E 'E

'E 'E

X

1-H

Q
X

W ^

(J

Dh

<

I I

I I I

r-. — .

Tj- — —

2 g;

sO —

— —

r-. —

On — —

-r -C

= .2 a.

Uintah County, Utah

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

57

X

Q
X

W ^

CLi ;j

pL,

-a

<u

3

_C

c

o

— 00

I I I I

I I I

2 S

I I

I I

Tf

g- I

I I

r

^ e-

e-

U

t

U

75 73

^ >

APPENDIX 1

Continued

58 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 18

E ^ i E

O' 3
. E

O D

e ^

O 3
E ^

O' D O' 3

=3 5 3

0 3 0 3
E =; E ^

O 3
E i

O 3
E ^

O 3
E

O 3
E >;

O 3
E

O 3
E

O 3
E ^

030303030303

£;;,E:;E^e^S;;E;;

O 3
E ;;

4>co>ca>c:a>c

^ g D t ^ D

u u u

3 0 3 0

“ CJ S u

3 0 3 0 3 0

t D t
u u

lU c o c

E D I D

U 6

x:

I

D

f a>-^

LJ o z:

— —

sD —

— sC

2 ^

3

— w

S” e -S,

^ =
E V «3
^ O E

C C/!!

A

c

o —

x> )P

o> ^

Uintah County, Utah

1981

McINTOSH— DINOSAURS OF CARNEGIE MUSEUM

59

E JS « iS

5 O' 5 O' 5 O'

5 O' 5

O D
E ^

O O
E ^

a p
, E C

” -g “

3 2

O' p £

O p O
E ^ E

O p
E

O p

e c

O' p
, E

O P
E

O P
, E ^

■ O p O p O p

SUSP

^ 3 (/I

3 0 3 0 3

S u S u S I

3 0 3 0 3 0

-S 4J W -5

SPSUSPSU

4.) C 4> C C 0>

._ __ aow ofl - _

C4>C4>C4JCU

' w w CJ)

ii S p s

<U "C w

‘5b B

U S O

030303030

w -5

S O

w -c

^ u ^ u ^ u

C-- C-- cijr Cur cur cur cur O

iucoc-^a«ciuc4^ciuci>

00 w >,

O D

^ 6

^ c V C 0> c

0 3 0 3 0

O' D

^30

X 00 — 00 i- 00«

3 o C O C O C

P c ur c ur c--

U ^ D D ^ D

0 O O

00 *- 00

Is t
<u o

=t s

X

HH

Q
Z

Oh (j
Oh

— sO

I 2 2 -

“2 CQ

—

sC —

— —

— Tj-

_ >

J=.

60

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

Q 3

z -B

^ u

Ph

i2 i2 x>

3 2 3 iS

O' D O' D

s 5 1 S

0*0^'

ISIS

O ' o *

a, ^ a. >

E

u. o u
(U >. W
73^-0
» > $
0-0
D-, ^ Ph

- o - o -

E

>« w

» 1

O

E E E

o L. o «- o

>' ^ >< >

^ (2 ^ (£

Oti

c

1^'

on

00

^ ^ a<

ISIS

o - o -

0- ^ Pk ^

E E E E

o u o >- O >- O

>, W ^ >* ii ><

^ j ^ 5 ^ j ^

- o - o - o

>^P« >-Q« >•&, >-Pn >>

E e

U .E

3 « 3 w ^ ^

5> = J» = W >,

3 0 3 0 ^^

S o s O

l; ^ lU ^ CQ

'5b 2 'ob -2 >. E

^ -H g .E t o

^ D |3 D 2 U

E o -£.

ace

O' =) O' p

Isis

u C U c
1/5 3 3

3 0 3 0

S u S ^

oj .3 (U

'5b w '5b 5
0) 3 u c

§ 5 E D

03030303030

k ^
® ^ ® -

tJ- S ti- E

Pi -S a:

030303030

j>Uj^Us/'-’v^PrffOvOv'Os/Ov

TUuCjuuifuuUuUuUTiuu

O^O^O^O^O^O^O^O^O

•3 </) 73 '^13 -O <y^ 73 <y: 73 (/5-0 <^73 i/5'3

i^3q^3^3i^3^3i^3^3^3i^
C^J=C^X:r^x:£^j:c^£C^J=/VJ=r^£c^

“ n““ i^‘-“ rv'*" i»s^

0 3 0 3 0

^0^0-

•n u i: u u

U. C 3 u,

■£ U u U

O r- O ^

^ E E

•> t •>

^ —

05 —

3 ^

i g; g; ?: S I o

CQ . CQ .

t > t >

^ Pi ^ Pi

O' o

P ^ r: -

o —

. CQ .CD

.CD .CD .

I 4J >.4) >^ 4)

•> t .> t ,>

Pi § Pi g Pi

' O O

I I

I I

U <sj U CJ

t .>

§ Pi
O

o —

1 ^
ca ffi

>« 4» >> 4)

t > t >

§ Pi I Pi

o o

3 4) 3 4> C 4»

•S os -S os -S

os os

’. CQ

>.4) >v4» >%4) >.4{

w -u

fc .>

3

o

t >

c«

O'

fc •> t •>

« Pi § Pi

o o

I I I

I I

1981

McIntosh— DINOSAURS of carnegie museum

61

T3 > T? > *U >

(J u U
C c

> ^
' o
^ cu

I

CD

I

CD

E ^
o 1=

Ss =

= £

o 3 oij

U o =

g N 1
O 4> O
^ JU >,

£ u. ^

N E
0) o
<u >.

u- ^

o .H o = o

E

> u.

oa 3 M

-E o E
E E

X X X X

OXj 3 6U 3 0£i 3

X K X

3 0XJ3 CJJ3 0£i3 0fl3 OO3 603 6*3 0ti3 003 C1O3 0^3

ocococo<=o^oco^o = o=^o = o^o

?iES5E?^ESE?^ES2E?iES5ESESSE?^£S5

Dooo'uo'uowoaiouoifo^soiuovo'y

0J>,«>.'U>.4>>,'U>,W>.4)>.W>-.<U>.t)>.4>>.flJ

r’r :>■ .'r ^ ,'r :>. ,'r :> ,'r ^ r’r :> ,'r :> .'r •>■ ,'r ^ ,'r •>■ .'r •>■ ,b-

^ ^ tu ^

u ^ ^

N E It

4» O _

tu ^ a:

u ^ u ^

^ U, ^ U, ^

w >,

£ ^

t > t > fc >

^ QC ^ DC g DC

a a o

0) >> 0> >> 4)

> fc

DC §

' O'

> t o
DC § U

o

^ z
E

O t

o t o

o § o

O

■CD Z ^ Z

1: > t

z ^ z ^ z

= 3

o c o
U 5 U

>1;^ >>z ^z ^

t ^
« o

a O' O' O' O' O'

0

1

■E

o

S

o

s

'E

o

S

'E

o

S

'E

o

S

o

S

X

Q

Z

u

0.

Dh

<

TD

(U

X

.S

c

o

U

2

0)

U,

I I

'C

It

o

u.

'a

<

E

E

c

&

"3

O

I I

Ci.

X

3

£.

5 3

3

so

SO

sO

SO

so

62

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 18

2 OA bO
_E 3 3

O Q Q

_ Of!

O o
Q

>- = = = = =:

0 = 0 = 0 =
% E % S ^ B
V O O V o

00 □ on

E o =

e S E

u- ^

z 6

=

t' o

E ^

ao 3

E o
E a

CO 3 00 3 60

E o
E “

E o
E S5

E o
E ^

0) >> W >, W >.

^ [i: ^

' O D

D a D O' D a D
^ E

• z ^

gogugu^cjgu

O' O' O' O' O'

-SOS

o s o s o 2 o

pC ^ pC ^ pC ^ pC

" ~ iS ’5) w 'Sb 3

O' D

I ^

II

O' o

E

s o

“ -s

<U C V C QJ C 0>

O ^ D t t D

U U U

00 -lo

V c
=

£! 00 O O
a E I

Si ° ^

S

(J ^ <u
C "w>
Cow
E o =

3 o _J5

o

•C

o

I

I

o

‘E

o

s

c

o

'E

o

'E

o

2

‘E

h

o

S

X

h— H

Q

Z

Ui

a

a

<

-o

0>

3

c

o

U

O

c

<u

iZ

?

z

00 --

I 2

— so

— — ON

U

1981

McIntosh— DINOSAURS of carnegie museum

63

S £ ob

2 5 o

S. O

s a

SS !S

a D
E

S U

H M a D

U .5 c - E

p- g i S’ s

O' 5

E >;

fT^

3 O

S u ^

S' c 'C
I 5

u a

E ®
<

S U

' 4^ ^

it! M 2

3 «J _C

D

S 2
^ o

8 I

8 S 8 2

2 2 S 2

■ o o

O c O c O c O

s ^

& §

8 5

2 2 S 2 S 2

2 2

2 8 2 8 2 8 I

Q, Q 0-0 ^ O

2 2
a. o
8 2

O' 5 O' 5

E ^ E ^

3 t; 3 t;

(U 3 C
3 </■' □

3 0 3 0

^ o s u ;

4^ "E .4^ -E .

'Bl « 'oa •« '

4J C (U C

E D E D

CJ o

O' 5 o 5

E E ^

t D t D

u u

I-- i'-

0 3 O' D

1 i I i

K i 8 i

3 0 3 0

S u s o

iu -E w -E
■a 2 M 2
c w c

E 5 I 5

o u

I' - |A t

O' 5 O' 5 O'
E ;; E E

g I S I §

3 0 3 0 3

^ U ^ U ^

5a i2 ’Si S 5a

I 5 I 5 I

j j o

X ^

S 0)

Q 2
X -5

? o

^ u

Ph

<

I rn 0^ ©V ©N

I i

I I I I I I I I I I

■'I? — —

— —

— — O

§ S

©\ —

E

r-i

u ^3

3 g
u I 15

^ O

^ g s

«= t:

— Cj

o —

Uintah County, Utah

38343 Hadrosaurid Ungual — — — Lance Hell Creek, Montana Silberling

38344 Hadrosaurid Tooth — — — Lance Hell Creek, Montana Silberling

64

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

1981

McIntosh— DINOSAURS of carnegie museum

65

1899

18 May
1899

27 Oct.
1899

3 Nov.
1900

APPENDIX 2

Accessions to the Carnegie Museum of Natural
History Containing Dinosaurian Fossils

-\cc. no. Description Date

414 Bavarian Museum in Munich —

original from Solnhofen, Bavaria
cast of the skeleton of the type of

Compsognathus longipes

954 W. J. Holland, collector
Carbon County, Wyoming
fragment of caudal and rib

1004 W. H. Reed, collector, obtained by W. J.

Holland

Carbon County, Wyoming
upper part of very large sauropod femur

1246 J. L. Wortman, W. H. Reed, A.

Coggeshall, collectors
Carbon and Albany counties, Wyoming,

Quarries 1, 2, 3(D), 4, 5
large part of Diplodocus skeleton, CM 84,
and partial skeletons of Apatosaurus and
Stegosaurus

1617 O. A. Peterson and C. W. Gilmore,
collectors

Sheep Creek, Albany County, Wyoming,

Quarries A, B, C, D, E, F, G
92 boxes containing a large part of second
Diplodocus skeleton, CM 94, and partial
skeletons of Camarasaurus,

Apatosaurus, etc.

1629 J. B. Hatcher and party, collectors
Lance Creek, etc.. Converse (now
Niobrara) County, Wyoming
disarticulated hadrosaur and ceratopsid
bones

1858 Exchange from E. Fraas
Triassic of Germany
various vertebrate fossils including a
theropod tooth

1885 C. W. Gilmore, collector

Sheep Creek, Albany County, Wyoming,

Quarries C, E, J

50 boxes containing two skeletons of
Apatosaurus from quarry E, one of them
CM 563, and other partial skeletons,
mainly sauropod

1934 W. H. Utterback, collector

Classic Marsh Quarry No. 1, Garden Park,
north of Canon City, Fremont County,

Colorado

26 cases containing two partial skeletons of
Haplocanthosaurus and other scattered
bones

2129 O. A. Peterson, collector

Snyder Creek, Converse (now Niobrara)

County, Wyoming

APPENDIX 2

Continued

Description

Date

6 Nov.
1900

1901

1901

1901

3 Nov.
1902

12 boxes containing partial skeletons of
Triceratops and Tyrannosaurus

2131 C. W. Gilmore, collector 3 Nov.

Freezeout Hills, Carbon County (Quarries 1902

L, M, N, O) and Sheep Creek, Albany
County (Quarry K), Wyoming
40 boxes containing several partial

skeletons and much scattered material of
many Morrison genera

2132 E. Douglass, collector Nov.

near Havre, Montana 1902

various scattered dinosaurian bones

2145 W. H. Utterback, collector 15 Nov.

Red Fork of the Powder River, Quarry A, 1902
Johnson County, Wyoming
22 boxes of a Diplodocus skeleton, CM 662

2312 C. W. Gilmore, collector 26 June

Freezeout Hills, Quarries N, O, Carbon 1903

County, Wyoming

7 boxes with the conclusion of the previous
season’s collections from Quarry N (and
a few bones from O)

2345 J. B. Hatcher, collector 3 Sept.

Judith River of Montana 1903

reptilian remains

2346 Exchange from Pal. Inst, of Vienna Univ. 8 Sept,

originals from near Vienna, Austria 1903

collection of casts mostly of

Struthiosaurus

2394 J. B. Hatcher, E. Douglass, W. H. 18 Nov.

Utterback, collectors 1903

32 boxes containing remains of various
Cretaceous dinosaurs, largely
hadrosaurid

2395 W. H. Utterback, collector 18 Nov.

Red Fork of the Powder River, Quarries A 1903

and B, Johnson County, Wyoming
29 boxes containing remainder of
Diplodocus skeleton CM 662

2417 A. C. Silberling, collector 5 Jan.

Morrison and Judith River beds of 1904

Montana

3 boxes containing Cretaceous and Jurassic
reptilian fossils

— American Museum of Natural History 1904

casts of Jurassic dinosaur bones from
Wyoming

casts of 3 chevrons of Diplodocus (AM
222) and manus of “Diplodocus”

( Camarasaurus) (AM 965)

2643 W. H. Utterback, collector 10 Oct.

Judith River and Hell Creek, Garfield 1904

County, Montana

66

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 18

APPENDIX 2

Continued

Acc. no. Description Date

10 boxes containing Triceratops skull and
other remains

3083 W. H. Utterback, collector 1906

Lance Creek, Converse (now Niobrara)

County, Wyoming

2 boxes with partial Triceratops skeleton
and other ceratopsian material

3085 W. H. Utterback, collector 1906

Elk Mountain, Johnson County, Wyoming
limbs of two Diplodocus and a partial
skeleton of Dryosaurits

3343 W. H, Utterback, collector 1907

Lance Creek, Converse (now Niobrara)

County, Wyoming

remainder of Triceratops partial skeleton

3723 Exchange from Yale University Peabody 1908
Museum

originals from YPM Quarry 1, Como Bluff,
Wyoming

4 boxes of casts of limb and girdle bones
and sacrum of cotypes of Camarasaurus
grandis

4256 E. Douglass and party, collectors 1910

(what is now) Dinosaur National
Monument north of Jensen, Uintah
County, Utah

First shipment, 70 boxes (nos. 1-70, 73, 75)
of vast dinosaur collection from DNM
including Apatosaums skeleton CM
3018, and Dryosaurus skeleton CM 3452

4791 E. Douglass and party, collectors 1913

(what is now) Dinosaur National
Monument north of Jensen, Uintah
County, Utah

Second shipment, 177 boxes (nos. 71, 76-
250) from DNM, containing Diplodocus
CM 11161 and Apatosaurus CM 11162
skulls and many partial skeletons

5246 McMillen, collector 3 Nov.

50 mi south of Grand Junction, Colorado 1914
sauropod caudal

5314 E. Douglass and party, collectors 1914

(what is now) Dinosaur National
Monument north of Jensen, Uintah
County, Utah

Third shipment, 3 boxes (nos. 262, 266,

267) containing Diplodocus skull and
Camarasaurus jaws

5973 E. Douglass and party, collectors 7 Dec.

mostly from Dinosaur National Monument, 1917
Uintah County, Utah

Fourth shipment, 119 boxes (nos. 251-261,

263-265, 268-341, 343-373) from DNM
containing Allosaurus skeleton CM

APPENDIX 2

Continued

Acc. no. Description Date

11844, Diplodocus skeleton (Denver
Museum of Natural History),

Camarasaurus skeleton CM 11393 and
many partial skeletons

6169 Donated by the American Museum of 4 Mar.

Natural History 1919

original from the Hell Creek beds of Big
Dry Creek, Montana
cast of skull and mandible of
Tyrannosaurus

7195 E. Douglass and party, collectors 1923

Dinosaur National Monument, Uintah
County, Utah

Fifth and final shipment, 70 boxes (373A-
374, 377, 379-446) from DNM including
two Camarasaurus skeletons (CM 11338,

USNM 13786), Stegosaurus skeleton
(CM 11341), Camptosaurus skeleton
(CM 11337), and much more material

7620 Exchange with Geological Survey of 14 May

Canada 1925

Red Deer River of Alberta, Canada
skulls of Corythosaurus and Centrosaurus

7877 F. H. Hammon, collector 1926

Panhandle of Texas
sauropod caudal

7974 Exchange with American Museum of 24 Jan.

Natural History 1927

Shabarakh Usu (now Bain Dzak), Mongolia
casts of Protoceratops eggs

8011 Return from U.S. National Museum 14 Mar.

Dinosaur National Monument, Uintah 1927

County, Utah

return of part of Stegosaurus CM 11341
and other material contained in blocks of
sauropod neck sent to USNM

8411 J. L. Kay, collector 8 Oct.

Harding County, South Dakota 1928

rear two-thirds of skeleton of
Edmontosaurus

9834 Purchased from Paul Mauersberger 1933

Upper Trias of Germany
cast of Plateosaurus pes

11348 J. L. Kay, collector 1937

Oldman and Lance formations at Sheep
Mountain, Carter County, Montana
various Cretaceous dinosaur bones
including a good hadrosaur

11767 J. L. Kay, collector 1938

Lance Formation at Sheep Mountain,

Carter County, Montana
scattered remains of numerous dinosaur
genera

1981

McIntosh— DINOSAURS of carnegie museum

67

APPENDIX 2

Continued

APPENDIX 2

Continued

Acc. no.

Description

Date

Acc. no.

Description

Date

12863

Acquisition from American Museum of

31 Dec.

20674

Donated by James Jensen

7 Nov.

Natural History

1941

Moffat County, Colorado

1963

Hell Creek, Montana

casi of gigantic Torvosaurus claw

type skeleton of Tyrannosaurus rex

26805

Acquired from Royal Ontario Museum

Dec.

13498

Acquisition from American Museum of

Feb.

Edmonton Formation, Red Deer River,

1973

Natural History

1945

Alberta, Canada

Shabarakh Usu (now Bain Dzak), Mongolia

skull and skeleton of Edmontosaurus

almost complete skull and skeleton of

regalis

Protoceratops

—

A. McCrady and party, collectors

June

—

Exchange with Royal Ontario Museum

1940

Hell Creek and Billy Creek, Garfield

1978

Red Deer River, Alberta, Canada

County, Montana

skeleton (without skull) of Corythosaurus

various bones of hadrosaurs and

17486

Acquired from Jess Lombard

Jura 3, Lily Park, Moffat County, Colorado

somewhat fragmentary skeleton of

1955

Triceratops

Drvosaurus

f

•y

-)

i

Copies of the foWowing Bulletins of Carnegie Museum of Natural History’ may be obtained at the prices
listed from the Publications Secretary, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pitts-
burgh, Pennsylvania 15213.

1 . Krishtalka, L. 1976. Early Tertiary Adapisoricidae and Erinaceidae (Mammalia, Insectivora) of North

America. 40 pp., 13 figs $2.50

2. Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the

late Pleistocene paleoecology of the central Appalachian Mountains of Virginia. 88 pp., 21
figs $12.00

3. Wetzel, R. M. 1977. The Chacoan peccary, Ca?ago/u/.v irug/jcn (Rusconi). 36 pp., 10 figs. .. $6.00

4. Coombs, M. C. 1978. Reevaluation of early Miocene North American Moropus (Perissodactyla,

Chalicotheriidae, Schizotheriinae). 62 pp., 28 figs $5.00

5. Clench, M. H., and R. C. Leberman. 1978. Weights of 151 species of Pennsylvania birds analyzed

by month, age, and sex. 87 pp $5.00

6. Schlitter, D. A. (ed.). 1978. Ecology and taxonomy of African small mammals. 214 pp., 48 figs.

$15.00

7. Raikow, R. J. 1978. Appendicular myology and relationships of the New World nine-primaried oscines

(Aves:Passeriformes). 43 pp., 10 figs $3.50

8. Berman, D. S, and J. S. McIntosh. 1978. Skull and relationships of the Upper Jurassic sauropod

Apatosaurus (Reptilia, Saurischia). 35 pp., 1 1 figs $3.00

9. Setoguchi, T. 1978. Paleontology and geology of the Badwater Creek area, central Wyoming. Part

16. The Cedar Ridge local fauna (Late Oligocene). 61pp., 30 figs $4.50

10. Williams, D. F. 1978. Systematics and ecogeographic variation of the Apache pocket mouse (Roden-

tia: Heteromyidae). 57 pp., 23 figs $4.00

11. Guilday, J. E., H. W. Hamilton, E. Anderson, and P. W. Parmalee. 1978. The Baker Bluff Cave

deposit, Tennessee, and the Late Pleistocene faunal gradient. 67 pp., 16 figs $5.00

12. Swanepoel, P., and H. H. Genoways. 1978. Revision of the Antillean bats of the genus Brachyphylla

(Mammalia; Phyllostomatidae). 53 pp., 17 figs $4.00

13. Schwartz, J. H., and H. B. Rollins, eds. 1979. Models and methodologies in evolutionary theory.

105 pp., 36 figs $6.00

14. Engesser, B. 1979. Relationships of some insectivores and rodents from the Miocene of North Amer-
ica and Europe. 68 pp., 12 figs., 20 plates $5.00

15. Bentz, G. D. 1979. The appendicular myology and phylogenetic relationships of the Ploceidae

and Estrildidae (Aves: Passeriformes). 25 pp., 5 figs $2.00

16. Mares, M. A. 1980. Convergent evolution among desert rodents: a global perspective. 51 pp., 25

figs $3.50

17. Lacher, T. E., Jr. 1981. The comparative social behavior of Kerodon rupestris and Galea spixii and

the evolution of behavior in the Caviidae. 71 pp., 40 figs $6.00

THE STRATIGRAPHICAL PALEONTOLOGY
^ OF THE TERTIARY NON-MARINE

SEDIMENTS OF ECUADOR

C. R. BRISTOW and JUAN J. PARODIZ

NUMBER 19 PITTSBURGH, 1982

BULLETIN

of CARNEGIE MUSEUM OF NATURAL HISTORY

THE STRATIGRAPHICAL PALEONTOLOGY
OF THE TERTIARY NON-MARINE
SEDIMENTS OF ECUADOR

GEOLOGY

C. R. BRISTOW

Institute of Geological Sciences, London, England

PALEONTOLOGY

JUAN J. PARODIZ

Curator Emeritus, Section of Invertebrates

NUMBER 19

PITTSBURGH, 1982

BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY
Number 19, pages 1-53, 24 figures, 1 map, 3 tables

Issued 5 March 1982

Price: $5.00 a copy

Craig C. Black, Director

Editorial Staff: Hugh H. Gennoways, Editor: Duane A. Schlitter,
Associate Editor: Stephen L. Williams, Associate Editor:
Barbara A. McCabe, Technical Assistant.

@ 1982 by the Trustees of Carnegie Institute, all rights reserved.

CARNEGIE MUSEUM OF NATURAL HISTORY, 4400 FORBES AVENUE
PITTSBURGH, PENNSYLVANIA 15213

CONTENTS

Abstract

Part 1. Geology

Introduction

Stratigraphy

Chota Basin

Cuena Basin

Biblian Formation

Loyola Formation

Azogues Formation

Guapan Formation

Mangan Formation

Santa Rosa Formation

Turi Formation

Nabon Basin

Loja and Malacatos Basins

Loma Blanca Formation

Trigal Formation

San Cayetano Formation

Quillollaco Formation

Flora, Fauna, and Age

Deposits of the Rio Pachitea and Vicinity in Eastern Peru

Magdalena Embayment, Colombia

Esmeraldas Formation

Mugrosa Formation

La Cira Formation

Caqueta Basin, Colombia

Canama, Peru

Iquitos and Pebas, Peru

Summary of Occurrences in Cuenca Basin

Part 2. Paleontology

Introduction

Systematic Accounts

Class Bivalvia

Superfamily Unionacea

Family Hiriidae

Superfamily Mutelacea

Family Mycetopodidae

Superfamily Sphaeriacea

Family Corbiculidae

Family Sphaeriidae

Superfamily Myacea

Family Corbulidae

Class Gastropoda

Order Prosobranchia

Superfamily Trochacea

Family ?Trochidae

Superfamily Neritacea

Family Neritidae

5

5

5

5

5

6

6

9

1 1

12

12

15

15

15

15

16

16

16

16

16

17

17

17

18

18

19

20

20

21

22

22

22

22

22

22

27

27

29

29

31

31

31

34

34

34

34

34

34

Superfamily Cyclophoracea 38

Family Aperostomatidae 38

Superfamily Viviparacea 38

Family Ampullariidae 38

Superfamily Rissoacea 38

Family Hydrobiidae 38

Superfamily Cerithiacea 43

Family Pleuroceridae 43

Family Thiaridae 47

Order Pulmonata 50

Superfamily Lymnaeacea 50

Family Planorbiidae 50

Superfamily Succineacea 50

Family Succineaidae 50

Incertae sedis 50

Literature Cited 51

ABSTRACT

The stratigraphy of all the Andean Tertiary non-marine basins
of Ecuador is described. The rich molluscan fauna (30 species
of molluscs have now been identified) has been thoroughly re-
vised and three new species described — Diplodon (Ecuudorea)
bristowi, Neritina loyolaensis , Paleoancidosa kennerleyi. The

stratigraphical position of the previously described Ecuadorian
faunas has been reinterpreted. Half of the species have been
recorded from localities outside Ecuador, and with this evidence
the chronostratigraphy of the non-Ecuadorian deposits has been
critically reviewed.

PART 1. GEOLOGY
C. R. Bristow
INTRODUCTION

At the present day, Ecuador is divided into three
physiographical regions — the relatively low-lying
coastal belt (La Costa), the Andes, and the Oriente,
which is the western part of the Amazon Basin
within Ecuador. Despite their present day promi-
nence the Andes are relatively young; fully marine,
deep-water deposition took place along the length
of the Ecuadorian Andes throughout Maestrichtian
times, and at least locally during Middle and Upper
Eocene times (Bristow and Hoffstetter, 1977:349-
51; Henderson, 1979). It is not possible to date pre-
cisely the first major uplift of the Andes but it ap-
pears to have taken place before the Upper Oligo-
cene Saraguro Eormation'. The Saraguro Eorma-
tion consists of lavas, pyroclastics, ignimbrites,
which are thought to have been deposited sub-
aerially, and, locally, conglomerates; it has been
dated by the potassium/argon method at 26 mil-
lion years. By Miocene times, a number of fresh-
water sedimentary basins were established through-
out the length of the Andes of Ecuador. These are
from north to south: Chota, Cuenca, Giron, Nabon,
Loja, and Malacatos, of which the Cuenca, Loja,

and Malacatos are the best known. In the Oriente
there is a thick cover of, at least in part, non-marine
sediments (Curaray, Chalcana, and Pastaza forma-
tions), but they have not been described in detail
and little is known of their fauna and flora.

In this account, details will be given of the stra-
tigraphy and paleontology of all the Ecuadorian
Tertiary non-marine Andean basins, together with
an assessment of their respective ages. The fauna
of the best known basin, Cuenca, is now known to
contain several species which have been recorded
from other South American localities. The ages as-
signed to the fauna and sediments of many of these
other localities are speculative as there are no ma-
rine marker horizons within the sequences, and also
because the true age ranges of the molluscs are un-
known. It is only at Cuenca, where the fossiliferous
sediments rest on a dated lava of Lower Miocene
age, that a reliable oldest date can be given to the
fauna. This, however, does not necessarily mean
that a similar fauna elsewhere cannot be older than
this date.

STRATIGRAPHY

Chota Basin

This basin is best exposed on the south side of
the Rio Chota in the Province of Imbabura.

The sequence commences with the Tumbatu For-
mation which rests unconformably on metamorphic
rocks. The formation is divisible into three units of

* The Ecuadorian names and their age ranges used in this account are those defined
in Bristow and Hoffstetter (i977).

which the lowest consists of some 90 m of conglom-
erates, sandstones, and red, green, and khaki
shales. The only fossils found are plant fragments.
The middle unit comprises about 1,000 m of pale-
colored shales, locally bentonitic, thin ( 1 to 2 m)
sandstones and beds of lignite; two thin shell beds
(10 to 20 cm) are composed almost entirely of the
gastropod Liris aff. miniiscuUi (Gabb). The upper
unit, about 230 m thick, consists principally of grey-

5

6

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 19

wackes, tuffaceous sandstones, and locally con-
glomerates, shales, and tuffs. Silicified wood is the
only fossiiiferoLis material. The Tumbatii Formation
has a gradational junction with the overlying Chota
Formation. The Chota Formation, some 144 m
thick, is divided into four units of breccias and con-
glomerates, and unfossiliferous sediments.

Hall {in: Bristow and Hoffstetter, 1977:268)
noted the similarity of the bentonitic shales and
lignites of the middle part of the Tumbatii Forma-
tion to the similar lithologies of the Mangan For-
mation (Upper? Miocene) of the Cuenca Basin,
and to the Arajuno Formation (Upper Miocene)
in the Oriente. This similarity can be extended
to the Curaray (Upper Miocene) and Upper Pastaza
(Upper? Miocene) formations, also of the Oriente,
and to the San Cayetano (Upper? Miocene) of Loja
and Malacatos, and to the Nabon (Miocene) For-
mation.

The only fossil, Liris minuscula (Gabb), was de-
scribed from Pebas, Peru. As discussed below, the
age of the Pebas deposits are thought to be of Upper
Miocene/Lower Pliocene age.

Cuenca Basin

The Cuenca Basin is the largest and best known
of the Andean Tertiary basins (Table 1). It has been
studied by various workers from the time of Hum-
boldt ( 1823). The most recent account is by Bristow
(1973) and the following stratigraphic notes are
based on that account. The paleontology is com-
pletely revised.

The early stratigraphy (Sheppard, 1934; Fiddle,
in Fiddle and Palmer, 1941) was somewhat con-
fused as it was not appreciated that the so-called
“Cuenca Shales” were a hybrid deposit, with the
similar lithologies of the Loyola, Guapan, and Man-
gan formations not stratigraphically distinguished.
The correct sequence was first recognized by Erazo
(1957) and this has been adopted, with minor no-
menclatural changes, by subsequent authors. The
evolution of the terminology, and the ages assigned
to the various formations by differing authors is
shown diagrammatically in Bristow (1973: fig. 2).

The sediments of the Cuenca Basin are strongly
folded and faulted. In the area between Azogues
and Cuenca where they are best known, the struc-
ture consists principally of a central north to south
or northeast to southwest trending Biblian Anti-
cline. On the east side of the anticline is the similar
trending Azogues Syncline. West of the anticline as
far as the Deleg Fault the sediments are steeply

dipping, or even overturned, but west of the Deleg
Fault the strata are virtually horizontal.

North of Azogues the Biblian Anticline and Azo-
gues Syncline cannot be recognized. The sediments
are commonly steeply dipping, but faults are the
dominant structural element.

In the center of the basin the contact with the
pre -Tertiary formations is not seen; in the east the
younger formations successively overlap the older
to rest with marked unconformity on the Maestrich-
tian Yunguilla Formation.

The relationship of the Tertiary sediments to the
older beds on the west side of the basin is obscure.
The strata are weathered, not well exposed, and
there is much superficial volcanic or glacial debris
mantling the surface. The boundary may be faulted,
or as in the east the younger sediments may have
overlapped on to the older.

The Cuenca Basin has been mapped at a scale of
1:50,000. The major part of the outcrop falls on the
two 1:50,000 Azogues (73 NW) and Gualaceo (73
SW) geological sheets (1974), published by the Di-
reccion General de Geologia y Minas, Quito; the
northern part of the basin is included on the
1:100,000 Canar (72) Sheet (1975); most of the re-
mainder of the southwestern portion falls on the
1:100,000 Giron (1974) Sheet^. All the localities
mentioned in the text can be found on these sheets.
The local grid references are taken from these same
maps.

Bihlidn Formation

The formation crops out over some 40 km in the
core of the main anticline, extending north-north-
east to south-southwest from north of Biblian to
near El Valle (740, 250) south-southeast of Cuenca.
In addition there is a large isolated outcrop of sim-
ilar trend to the east and this is seen fom Jadan
(360, 810) to the Quingeo area (300, 650) south of
Santa Ana.

It was only the Biblian Anticline outcrop which
was known to Sheppard (1934) and Fiddle and
Palmer (1941). However, the fossiliferous horizon
within the “Biblian Sandstone and Conglomerate”
locality of the above authors, exposed on the flanks
of the small anticline (377, 990) between Biblian and

2 The Giron Basin is separated from the Cuenca Basin by a blanket of Pleistocene
volcanic deposits. Little is known of the deposits of this area — originally they were
mapped as the undivided Ayancay Group ( = Mangan + Santa Rosa Formation), but
it is now thought that sediments of the Azogues and Mangan formations occupy the
basin. The only fossil found to date is a juvenile Neocorhicula sp.

1982

BRISTOW AND PARODIZ— ECU ADORIAN TERTIARY SEDIMENTS

7

Table 1. — Stratigraphic column showing the composition of the Cuenca Busin, Ecuador.

STRATIGRAPHIC COLUMN OF THE CUENCA BASIN

2

<

GENERALIZED

LLi

h-

CO

cc O

VERTICAL

SECTION

LITHOLOGY

FOSSILS

>

m

O f-
LL

scale 1:30,000

LU

White, buff and pink silts: fine-grained tuffs and
fuffaceous sandstone.

W Z

ul H

CC

D

Wood, mammoth

D. O

1-

Conglomerate.

UJ

z

UJ

O

^ o o o 0^0
<=» o ^ tr

Alternating sequence of pebble and boulder beds,
with red silty and sandy shales

H <
V CO

g

O ^ Cr o

^ O o C) C)

□

L

Oo° ° °0°o°

n o o Q

Passage from Mangan Formation

C> O o C^

o <? C>

o ® ^

Alternating sequence of coarse brown fuffaceous

sandstones, pebble beds and green silty shales

Cahari coal seams

Alternating sequence of coarse brown fuffaceous
sandstones, and red and green silty shales. More
shaly towards the base.

cc

UJ

Q.

Z

<

Fossula derbyi, Neocorbicula cojitamboensis,
Doryssa bibliana, Aylacostoma browni, Aylacostoma
sulcata. Paleoanculosa kennerleyi. Neritina
pacchiana. Vetustocytheridea bristovd, Hoplias sp.

o

z

<

Q.

Bentonite

Washington coal seams
Flard white silica rock

Paleoanculosa kennerleyi

LU

Alternating sequence of coarse brown fuffaceous

z

sandstones and buff and brown shales

LU

O

o

LU

Buff fissile shales with bentonite

Q

O « ^

Coarse, buff fuffaceous sandstone with thin shale
beds. Locally basal conglomerate

Q

Neocorbicula cojitamboensis

—

Buff fissile shale with occasional limestone lenses.
Much gypsum veining.

Diplodon guaranianus biblianus, D. bristowi, D.
liddlei. Monocondylaea azoguensis, M. pacchiana,
Erodona iquitensis, Pisidium sp.. Neocorbicula
cojitamboensis, Ostomya fluviatilis, Calliostoma sp.,
Neritina loyolaensis, N. pacchiana, Puperita

<

_j

o

>-

o

_J

—

sphaerica, Poteria bibliana, Pomacea manco,
Hydrobia ortoni, Lyrodes sp., Doryssa bibliana,

tr

UJ

—

Argillaceous sandstones, coarse sandstones and
conglomerates on east side of basin

Aylacostoma browni, Gyraulus sp.

§

— ■o-0~-^-c:> —

o

Diplodon guaranianus biblianus, Doryssa bibliana

o — ^ o

o O <3

Coarse white to buff fuffaceous sandstones
and pebble beds, alternating with red silty and
sandy shales

z

<

. C o

^ Oa o o o

CQ

A O ^ ^ o o,
^ " o r> o

CQ

g Q ^ •

£3 . c> o

0 p fpi _

Unconformity

8

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 19

Azogues, is in fact the overstepping basal bed of
the Loyola Formation.

The base of the formation is not exposed in the
core of the main anticline and it is possible there
that it rests on older Tertiary strata at depth. In the
isolated outcrop east of Cuenca the formation rests
with marked unconformity on the Maestrichtian
Yunguilla Formation. In this vicinity, where the
Biblian Formation is not overlain by younger Ter-
tiary sediments, the thickness exceeds 1,000 m. In
the center of the basin there is a gradual upward
transition over a few meters into the overlying Loy-
ola Formation.

The basal bed of the Biblian Formation is a fairly
coarse pebble bed often incorporating pebbles of
the underlying Yunguilla Formation. Near Jadan
(355, 813) an abundance of plum-red volcanic ag-
glomeratic material occurs close to the Biblian/Yun-
guilla contact, and although the field relationships
are not clear it appears to form a local base to the
formation. In this general area coarse clastic units
up to 30 m thick dominate the sequence, whereas
it is more argillaceous in the Biblian-Azogues area.
In this latter area the buff or light-brown, soft, tuff-
aceous sandstones, grits, and conglomerates occur
in thinner lenticular units within the red and pur-
plish-red silty and sandy shales which typify the
formation. Locally, greenish-gray colors occur.
There is much gypsum veining of the deposits.

As mentioned above, in the center of the anti-
cline there is an upward passage into the Loyola
Formation. In three localities at least, there are
however andesitic volcanic deposits at, or very
close to, the junction between the Biblian and Loy-
ola formations. Thin andesitic lavas occur in at least
two localities (352, 972) near Cojitambo in the up-
permost part of the Biblian Formation. Cojitambo,
which forms such a prominent landmark (see Plate
5 of Fiddle and Palmer, 1941) has long been regard-
ed as an intrusive andesite at the junction of the
Loyola and Biblian formations. However, the con-
tact with the overlying Loyola Formation shows lit-
tle or no evidence of metamorphism, although Fiddle
and Palmer ( 1941 : 18) remarked that the shales were
slightly baked. The general fine-grained aphanitic
texture and the presence of volcanic glass in the mat-
rix shows that it is in fact extrusive and probably con-
temporaneous with the El Shalal deposits described
below. At El Shalal (360, 990) there is an outcrop
of interbedded lavas, volcanic agglomerates, fine-
grained tuffs and coarse tuffaceous sandstones and
grits, which are well exposed along the Biblian-

Azogues road (see Plate 4, Fig. 2, of Fiddle and
Palmer, 1941) at the junction of the Biblian and
Loyola formations.

At Descanso (368, 867) on the east side of the
basin a fourth outcrop of andesite underlies the
overstepping basal Loyola Formation. The base of
the andesite is not seen, but the andesite does in-
corporate blocks of the Yunguilla Formation. The
relationship of the andesite to those described
above is not known, but it is thought to be a cor-
relative of them. The outcrop at Descanso is re-
garded as the residual remnant of the highest Bib-
lian Formation prior to its disappearance beneath
the overstepping Loyola Formation. The signifi-
cance of this andesite is that two Lower Miocene
K:Ar age determinations of 19 and 20 million years
ago have been obtained on samples from it (Snell-
ing, 1974, unpublished report of the Institute of
Geological Sciences, London). Regardless of its ex-
act stratigraphic position, it means that the fossil-
iferous basal Loyola Formation which rests on the
andesite cannot be older than uppermost Lower
Miocene.

Fauna and age. — Only a very sparse fauna has
been collected from this formation. The fossilifer-
ous Biblian Sandstone and Conglomerate locality of
Sheppard (1934) and Fiddle and Palmer (1941) is the
basal bed of the Loyola Formation and is fully dis-
cussed below. The author has found a thin shell bed
composed almost entirely of Doryssa bihliana
(Marshall and Bowles), but with rare Diplodon
giiaranianus biblianus (Marshall and Bowles) in the
uppermost part of the Biblian Formation at only
two localities near El Valle (250, 757; 278, 780) (loc.
CRB 8 and 5). This sparse fauna is also known in
the overlying Loyola Formation and since there is
a passage from one formation to the other in the
center of the basin accompanied only by a change
of sedimentation and not by an orogenic break,
there is thought to be no great time difference be-
tween the two formations. On this evidence the
Biblian is regarded as of Lower Miocene age.

Repetto ( 1977) found a tooth in the lower ( ?) Bib-
lian 8 km west of Azogues, of a notungulate toxo-
dont, close to but distinct from Prototoxodon rothi
Kraglievich of Middle Miocene age.

Sigal (1968) examined 64, and the author one,
samples from the formation, but all proved to be
barren of microfauna, other than reworked Maes-
trichtian foraminifera (Rugoglobigerina sp.) and
ostracods.

Pollen recorded by Savoyat et al. (1970:57-60,

1982

BRISTOW AND PARODiZ— ECUADORIAN TERTIARY SEDIMENTS

9

and Fig. 2) from the Biblian and Loyola formations,
gave a Paleocene-Lower Eocene date to the Bibli-
an, and a Lower-Middle Eocene date to the Loyola
Formation. Because the sparse data of the Loyola
Formation are so at variance with the radiometric
age determinations, the palynological dating of the
Biblian Formation is also thought to be grossly in
error. If the pollen identifications are correct the
palynology may be at fault either because the
spores are derived, or because their true age ranges
within Ecuador in particular, and South America in
general, are not known with certainty.

Loyola Formation

The type locality is the small village of Loyola,
also known as Chuquipata (373, 908), 7 km SSW of
Azogues.

The formation has an extensive outcrop on the
flanks of the Biblian Anticline. A thin, but impor-
tant, outcrop occurs on the east side of the Azogues
Syncline. North of Biblian the Loyola Formation is
either faulted out or disappears beneath a younger
deposit. A small fault-bounded outcrop is seen
(368,163) to the south of Ingapirca.

The Cuenca White Shales of Sheppard (1934)
were defined as occurring between the Biblian
Sandstones and Conglomerates, and the Azogues
Sandstone. By this definition they are synonymous
with the Loyola Formation, but, from an exami-
nation of Sheppard's Fig. 2, it is clear that the Gua-
pan Formation as now recognized was included
within his Cuenca White Shales. Liddle (in Liddle
and Palmer, 1941:22) also included the shales and
coal of what is now known as the Mangan Forma-
tion in the Cuenca Shales. Additionally, the fossil-
iferous “Biblian Sandstone and Conglomerate” lo-
cality of Sheppard (1934) and Liddle and Palmer
(1941) at the classic anticlinal section between Bib-
lian and Azogues, is in fact the overstepping basal
bed of the Loyola Formation. Erazo (1957:13-14)
was the first to recognize much of the confusion,
but no new names were given to the revised strata.
These were eventually introduced by the United
Nations (for example, UNDP, 1969).

The dominant lithology of the formation consists
of a monotonous sequence of fissile dark gray
shales and silty shales which weather buff or cream.
Locally, lenses of limestone occur together with
thin layers of soft sandstone. Gypsum veining and
coatings to joints and bedding surfaces is a common
feature of the weathered rocks. In the center of the

basin the Biblian passes upward, with no marked
break, into these fissile shales. Fossils, other than
fish remains and leaves, are uncommon but local
shell beds composed almost entirely of Doryssa
hihliana occur.

On the eastern side of the basin there is a well-
developed basal series of sandstones and conglom-
erates. It is possible to trace a gradual overstep of
the Loyola Formation across the Biblian Formation
and on to the Yunguilla Formation. This is well seen
between Descanso (368, 867) and a point (410, 953)
south of Azogues, and at the historically important
road and railway cuttings in the small-scale anti-
clines and synclines to the northwest of Azogues
(377, 990). At this latter locality the basal beds are
some 45 m thick. The pebbles of the conglomerates
consist of tuff, quartzite, quartz, and fragments of
the Yunguilla Formation. At Descanso, where these
beds rest on andesite, much weathered angular
andesite is incorporated in it. The basal beds are
locally richly fossiliferous and have yielded the
principal faunas. Fossils consist dominantly of mol-
luscs, but crustacean debris, fish teeth and scales,
ostracods and charophytes are common.

The maximum thickness of the formation is about
360 m in the center of the basin.

Flora, fauna, and age. — Until recently the fauna
of the Cuenca Basin was thought to be endemic
(Marshall and Bowles, 1932; Liddle and Palmer,
1941). Parodiz (1969) first demonstrated that certain
species did occur outside Ecuador. Further collect-
ing extended the number of non-endemic species
(Bristow, 1973). This new fauna, together with ad-
ditional molluscan material collected during 1974
has been completely revised by Parodiz and the re-
sults incoiporated here.

The known fossils from the Loyola Formation
include:

Plants

Gymnospermae

Trigonia varians Engelhardt (CRB 18)

(Berry, 1934, 1945)

Macrolohium tenuifoliiun (CRB 18)

Engelhardt (Berry, 1934, 1945)

Charophytes

Char a sp.

Pollen — various spores listed by Savoyat et al.

( 1970:57-60, fig. 2), but of doubtful value
isee below)

10

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Mollusca

Bivalvia

Diplodon {Ecuadorea)
guaranianus hiblianiis
(Marshall and Bowles)

Diplodon (Ecuadorea)
bristowi, new species

Diplodon (Ecuadorea) liddlei
(Palmer)

Monocondylaea azoguensis
(Palmer)

Monocondylaea pacchiana
(Palmer)

Monocondylaea sp.

Anodontites olssoni (Palmer)

Pisidium sp.

Neocorbicula cojitamboensis
(Palmer)

CRB 18, 28

CRB 18

CRB 7, 18

CRB 18

(CRB 18)

CRB 18
CRB 18
CRB 7, 11,
26, 46

CRB 6, 7, 9,
10, 11, 15,
18, 26, 34,
48, 49
CRB 7

Erodona icjuitensis (de
Greve)

Ostomy a cfr. fluviatilis (H.

Adams)

Gastropoda

ICalliostoma sp.

Neritina loyolaensis, new
species

Neritina pacchiana (Palmer)

Neritina sp.

Puperita aff. sphaerica
(Olsson and Harbison)

Poteria (Pseudoaperastoma)
bibliana (Marshall and
Bowles)

Pomacea (Limnopomus)
nianco Pilsbry
Hydrobia ortoni (Gabb)

Lyrodes sp.

Potaniolithoides biblianus
(Conrad)

Aylacostoma browni
(Etheridge)

Aglacostonia dickersoni

(Palmer in\ Liddle and Palmer, 1981)
Doryssa bibliana (Marshall CRB 2, 12,

and Bowles) 14, 17, 18,

28

Pulmonata

Gy raid us sp. CRB 7

Succinea sp. CRB 7

CRB 26

CRB 7
CRB 9

CRB 34
CRB 7
CRB 18, 46,
48

CRB 26, (18)

CRB 34

CRB 7, I 1
CRB 7

CRB 26b, (18)
CRB 18

Crustacea

Ostracoda

Vetustocytheridea bristowi
Bold

Brachyura

Necronectes proavitus
Rathbun (Bristow, 1973;
Collins and Morris, 1976)

Echinoidea

Unsubstantiated record
(Erazo, 1965:9)

Eish

Characoids

Hoplias sp., Leporiniis sp.
(Roberts, 1975)

CRB 11, 26,
27, 30, 35,
42, 51

CRB 26

CRB 26

CRB 7, 9,

10, 11, 18,
26, 30, 31,
51

Parentheses signify material described by pre-
vious authors in same localities (CRB) recollected
by the author.

Arenaceous foraminifera recorded from the basal
beds of the Loyola Eormation (Bristow, 1973) are
almost certainly derived from the underlying Yun-
guilla Formation (J. Whittaker, personal commu-
nication). The calcareous Miocene ISiphogenerina
senni also listed in Bristow (1973) is now regarded
as a contaminant. S. senni occurs in the marine
Miocene coastal deposits of Ecuador and samples
from the coastal provinces were being processed at
the same time as the author’s samples.

With one exception, there is little in this fauna to
provide an accurate independent date, but collec-
tively a Miocene age is indicated. The following in-
ferences can be made.

The leaves recorded by Berry (1934, 1945) also
occur in the Miocene Loja Basin where the flora
has been studied in greater detail (see below).

Chara sp. indicates a post-Middle Eocene, but
more probably an Oligocene age (Grambast, per-
sonal communication).

The spores listed by Savoyat et al. ( 1970) are sup-
posedly of Lower to Middle Eocene age. These
dates are wildly at variance with the other evidence
and are not accepted by Bristow (1973). (See re-
marks about the pollen of the Biblian Formation.)

Diplodon (Ecuadorea) guaranianus biblianus is
also known in the Miocene Monagas Series of Ven-
ezuela (Parodiz, 1969).

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

Ponuicea (Limnopomus) manco occurs in the
Upper Oligocene or Lower Miocene strata in Peru
(Piisbry, 1944).

Hydrobia ortoni is known from Iquitos and Pebas
(de Greve, 1938).

Aylacostoma browni, which is also known in the
Mangan Formation, is one of the more widely dis-
tributed species outside of the Cuenca Basin. Else-
where it is known in ?Pliocene beds at Canama,
Brazil (Etheridge, 1879), and the ?Pliocene La Ta-
gua beds, Colombia (C. P. Nuttall, personal com-
munication).

Puperita sphaerica was first described from the
Pliocene of Florida (Olsson and Harbison, 1953).

Although Vetiistocytheridea bristowi is endemic
to the Cuenca Basin, other members of the genus
and allied genera (cf. Cyprideis stephensoni in the
Malacatos Basin) are generally regarded as indica-
tive of the Miocene.

Necronectes proavitus is the species which prob-
ably provides the best independent date for the
Loyola Formation. The type specimen came from
the Middle Miocene Gatun Formation. It is also re-
corded from the Middle Miocene of Puerto Rico
(Gordon, 1966: 184) and the Middle Miocene Brasso
Formation of Trinidad (Collins and Morris,
1976: 125). This Middle Miocene date is in close
agreement with the Lower Miocene radiometric
dates of 19 and 20 million years ago obtained on the
andesite at Descanso which is immediately overlain
by the crustacean-rich (not specifically identified at
this locality) basal pebbly beds of the Loyola For-
mation.

Azogues Formation

There has been a gradual restriction in the appli-
cation of the name ‘Azogues’ to the Cuenca sedi-
ments. Wolf (1879, 1892) used the term “Azogues
Sandstone” for all the sediments of the Cuenca Ba-
sin. Sheppard’s (1934:361) use of Azogues Sand-
stone was for the post-Cuenca White Shales ( Loy-
ola Formation). Fiddle (in Fiddle and Palmer,
1941:23) used the modified term “Rio de Azogues
Sandstone,” as the outcrops on the east side of the
Rio Azogues ( = Rio Burgay on modern maps)
offered a better type locality. Erazo (1957) further
restricted the Azogues Sandstone to those beds be-
tween the underlying Cuenca Shale ( = Loyola For-
mation) and the (unnamed) overlying Guapan For-

1 1

mation. This restriction is logical, but it now means
that the town from which the formation takes its
name does not overlie it; Azogues is sited on the
Guapan Formation. Subsequent authors have
adopted Erazo’s usage, although Fiddle’s name
“Rio de Azogues Sandstone” would have been
more applicable.

The Azogues Formation is best developed on
either side of the Azogues Syncline where it ex-
tends from just north of Azogues in the north, to
near El Valle (266, 750) in the south. On the west
side of the Biblian Anticline the Azogues Formation
extends from a short distance north of Cojitambo,
southwards to Boqueron (230, 740) where it disap-
pears beneath the unconformable Turi Formation.
Southwestwards of Boqueron the Azogues Forma-
tion reappears but has been mistakenly grouped
with the Mangan Formation and the two deposits
have been mapped as the undivided Ayancay Series
(1:100,000 Giron geological sheet). The disappear-
ance of the formation north of Cojitambo is attrib-
uted to faulting, but it may be due to facies change.

The base of the formation is transitional over
some 10 to 20 m with the underlying Loyola For-
mation. The dominant lithology is medium to
coarse-grained, brown weathering tuffaceous sand-
stones, but beds of siltstone, clay, and shale occur,
generally not more than 1 m thick and principally
in the lower part. A characteristic feature of the
sandstones is their curved weathering surfaces. The
siltstones are generally off-white to pale yellow,
have fine regular ferruginous laminae in places, and
are of varying hardness. The shales are usually off-
white to pale gray or yellow, locally slightly silty,
usually poorly laminated and medium hard.

On the east side of the basin the Azogues For-
mation oversteps the older beds to rest directly on
the Yunguilla Formation. In the area to the north
and east of Azogues where the older Tertiary sed-
iments are absent there is a well-developed basal
conglomerate. At such points (for example, 402,
994), and along the road from Azogues church to
Uchupucum (416, 985), the conglomerate is well
exposed. Conglomerates, often with distinct cross
bedding, are also developed at other levels. Well-
rounded pebbles from I to 10 cm diameter, consist
mostly of weathered igneous rock, with pebbles of
quartz, and Yunguilla shales and limestone. The
higher beds of the Azogues Formation in the syn-
cline to the east of Paccha (304, 800) consist of vol-
canic agglomerate with pumice fragments.

Where the Guapan Formation is developed in the

12

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Azogues Syncline the junction with the Azogues
Eormation is interdigitational over a few meters.
The Guapan Eormation appears to be a lateral fa-
cies of the Azogues Formation, as a thickening of
either is accompanied by a thinning of the other.
Where the Guapan Formation is absent, and the
Mangan Formation directly succeeds the Azogues
Formation, the contact between the two formations
is also transitional, but over a somewhat greater
thickness of beds than with the Guapan.

The maximum thickness appears to be about 280
m in the El Tablon vicinity (265, 792), some 4 km
east of Cuenca.

Fauna and age. — Fossils have not previously
been found in the Azogues Formation. Sheppard’s
(1934:362-363) Paccha samples clearly came from
the Loyola Formation (see his Fig. 4) while his ma-
terial from the Biblian Sandstone and Conglomer-
ates (=basal Loyola Formation) he confusedly in-
cludes within the Azogues Sandstone. Fossils found
by Olsson (in Fiddle and Palmer, 1941:24) said to
be from this formation between Puente del Descan-
so and Cerro Tahual are almost certainly from the
basal Loyola Formation (=CRB 9). The somewhat
doubtfully identified Nucula cf. andersoni Clarke
recorded from the base of the formation (UNDP,
1969:15), and the marine fossils found by Erazo
(1965:9) also come from the condensed basal beds
of the Loyola Formation. Thirty-two samples col-
lected by Sigal (1968) were barren of microfauna.

The author has collected a limited, but signifi-
cant, fauna from the basal beds of the Azogues For-
mation at three localities (CRB 1, 305, 835; CRB 4,
258, 786, and CRB 13, 308, 832). Neocorhicula co-
Jitamboensis occurred abundantly at all three lo-
calities. Additionally, Aylacosloma cfr. dickersoni
and Diplodon sp. were present at CRB 1.

A. dickersoni is known from only one other lo-
cality— in the beds of the Loyola Formation on the
southwest side of Cojitambo (Fiddle and Palmer,
1941:36, map).

In view of the transitional contact with the Mid-
dle Miocene Loyola Formation, and as there is no
distinct faunal difference between the two forma-
tions, the Azogues Formation is also regarded as
Middle Miocene.

Guapdn Formation

Because of lithological similarity, earlier workers
had confused this formation with the older Loyola
Formation (for example, Sheppard, 1934: Fig. 2).
Erazo (1957:13-14, section 2) first recognized this

discrepancy but gave no name to the formation. The
name “Guapan,” taken from the outcrops near the
Guapan cement works (394, 996), was introduced
by the United Nations (for example, UNDP, 1969).

The formation is confined almost entirely to the
center of the Azogues Syncline; a small isolated
outcrop occurs on the west side of the Biblian An-
ticline in the area between Ayancay and Cojitambo
(339, 892-346, 930).

The junction between the Guapan and Azogues
formations is gradational and, as mentioned above,
although the Guapan Formation overlies the Azo-
gues Formation it also passes laterally into it.

The Guapan Formation consists characteristical-
ly of finely laminated dark brown to black shales,
weathering white or yellow and with much limonite
staining. The laminae tend to be thicker than in the
Loyola Formation, but otherwise the two forma-
tions are very similar. Tuff and tuffaceous sand-
stones were noted on the roadside (385, 925) just
south of Charasol and have been recorded in the
lower half of the Guapan Formation near Ayancay.
Sands, sandstones, and occasional conglomerates
occur. Bentonite is known at Charasol (Nunez del
Arco, 1971). Also in this locality a gypsum seam
was located some 60 m above the base of the for-
mation.

The maximum thickness appears to be about 100
m in the center of the syncline.

Flora, fauna, and age. — Excellently preserved
impressions of fossil leaves abound at many hori-
zons, but await a detailed study. Fish scales are
common; one incomplete characoid fish cf. Moen-
khausia has also been recorded (Bristow, 1973).
Seventeen samples taken by Savoyat et al. (1970)
for microfauna proved to be barren.

In view of its facies relationship with the Azogues
Formation, the Guapan Formation is thought to be
also of Middle Miocene age.

Mangan Formation

The formation takes its name from the several
localities to the west of Nazon (310, 010) which in-
corporate Mangan in their title. It is an unfortunate
choice of name because all the localities overlie the
Santa Rosa Formation.

The Mangan Formation has a wide and extensive
outcrop on the west side of the Cuenca Basin from
Ingaprica (365, 190) in the north, where it appears
from beneath the Tarqui Formation, to just south
of Cuenca where it disappears under the Turi For-
mation. Southwest of Turi where the Tertiary sed-

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

13

iments reappear in the area of the Giron Sheet
(1974) they have been mapped as the undivided
Ayancay Group ( Mangan and Santa Rosa forma-
tions). They extend southwards as far as a point
about 25 km NW of Saraguro. However, as men-
tioned above, the Ayancay Group as mapped in this
area is misnamed, as the sediments comprise, at
least in part, the Azogues and Mangan formations.

The coal-bearing strata of the formation have
been known for many years, and were included in
the upper part of Wolf s “Arenisca de Azogues.”
Liddle {in Liddle and Palmer, 1941) thought that
they were part of the Cuenca Shales. Erazo (1957)
was the first worker to separate off from the “Azo-
gues Formation,” the strata later to be named the
Ayancay Group, but to which he gave no name.

There has been much research into the coal of
the Cuenca Basin, of which the most recent and
complete is that by the United Nations (UNDP,
1969). For that report the coal field was mapped at
1:10,000 scale and a considerable amount of detail
exists for the quantity and quality of the coals, and
for the structure of the coal-bearing deposits (see
also Putzer, 1968).

For descriptive purposes it is convenient to di-
vide the Mangan Formation into three units — a low-
er, including all the strata up to, but excluding, the
lower (Washington) coal seams; a middle unit com-
prising everything from the Washington to the Ca-
nari coal seam; and an upper unit above the Canari
seams. These three divisions can only be recog-
nized in the area where the coal seams are devel-
oped.

There appears to be a transitional upward pas-
sage from the Azogues Formation, or from the Gua-
pan Formation where present, into the Mangan For-
mation. The upper part of the Azogues Formation
loses its massive sandstone beds, and red and green
blocky shales appear, alternating with thinner sand-
stones. This alternating sequence characterises
much of the Mangan Formation. The United Na-
tions (UNDP, 1969), however, claims that there is
a disconformity at the base of the Mangan, but there
appears to be little evidence of this unconformity
in the field. In the Cushumaute area (338, 954) the
beds consist predominantly of light-colored silt-
stone, shale, and fine-grained sandstones, interbed-
ded in layers generally less than 1 m thick.

The maximum thickness of this unit appears to
be 870 m in the San Nicolas area (340, 940).

The gastropod Paleoancnlosa kennerleyi, new
species, found in a 10 cm seam (338, 953) in the

lower third of the unit near Cushumaute, scarce fish
teeth from this same bed (Roberts, 1975:263), and
leaves found stratigraphically below (339, 954) have
been the only fossils found to date. The occurrence
of a shell bed also composed entirely of the above
gastropod near Ingapirca (365, 161), in an area
where the Mangan Formation cannot readily be di-
vided into three units, may indicate a potential
stratigraphical marker for the lower Mangan For-
mation. The latter locality, where the road to In-
gapirca crosses the Rio Canar, is probably the same
locality as that at which Reiss (see Wolf, 1892:257)
found fossils. The Mangan Formation at this local-
ity is in fault contact to the east with the Foyola
Formation.

The middle unit is characterized in the center of
the basin by the presence of several coal seams at
its top and bottom. The Washington seams do not
extend beyond the Rio Sidcay (290, 840) in the
south, and the Canar seams are here contaminated
with elastics. Neither is well developed north of
Nazon (330, 015) (UNDP, 1969:18).

Thick beds of shales, lithologically identical to
those of the Loyola and Guapan formations, are
well developed only in the lower part of the unit in
close association with the Washington coal seams.
Bentonite, in beds up to 15 m thick, has been noted
at several localities. Higher in the sequence there
is a well-bedded series of tuffaceous sandstones,
silts, thin conglomerates, and thin shales. A con-
spicuous, persistent bed some 2 m thick of hard
white, almost pure silica rock, first noted by Wolf
( 1892), occurs some 20 to 30 m below the Canari
seams and forms an excellent marker horizon. A
similar bed above the Canari seams was also noted
by the United Nations (UNDP, 1969). Dark green
tuffs have been recorded at the San Luis mine (334,
977). The coal seams, of sub-bituminous C grade,
are “lenticular, sheared, faulted and otherwise un-
predictably discontinuous.” Other unfavorable fac-
tors, such as the near vertical foot and hanging
walls, led to the closure of the last large-scale work-
ing mine, San Luis, in 1967 (UNDP, 1969:42). The
maximum thickness of the middle unit appears to
be 600 m.

All the fossils so far found within the formation,
with the one exception in the lower unit mentioned
above, are associated with the coal seams: directly
over one of the Washington coal seams at Cocha-
huaicu (335, 995), and in a similar position close to
the San Luis mine (334, 976). Crocodile teeth and
other vertebrate remains have been reported from

14

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

the coal seams (UNDP, 1969:18). The fossils are
listed below.

The upper unit consists principally of coarse
brown tuffaceous sandstones, commonly pebbly
and becoming increasingly conglomeratic upwards.
The arenaceous and conglomeratic beds alternate
with blocky weathering green and brick-red silty
clays. In the area (270, 865) north of Ricaurte, the
passage into the overlying Santa Rosa Eormation is
gradational, but represented by a coarsening up-
ward sequence and the incoming of red beds. In the
Nazon area the change is more abrupt and much
coarser conglomerates typify the Santa Rosa For-
mation.

The full thickness of the upper unit is unknown
as nowhere is there a complete unfaulted sequence;
it certainly exceeds 700 m. No fossils have been
found in this unit.

Flora, fauna, and age. — The flora consists of
well-preserved, but as yet unstudied, impressions of
leaves.

Spores from the coal seams have been examined
by Grebe {in Putzer, 1968:479-480, 486-487). They
indicate a correlation between the Washington
seams of the Cuenca Basin with the coal seams of
the Loja and Malacatos basins.

The fauna is dominated by molluscs and includes
the following:

Bivalvia

Fossula cf. derhyi Ihering

Neocorhicula cojitamhoensis
(Palmer)

Gastropoda

Doryssa hihliana (Marshall and
Bowles)

Aylacostoma browni
(Etheridge)

Aylacostoma sulcata (Conrad)

Ncritina pacchiana Palmer

Paleoanculosa kennerleyi, new
species

Additionally, the ostracod, Vetustocytlieridea bris-
towi Bold, has been found in sample CRB 42c, and
jaw teeth of the erythrinid fish, Hoplias sp., have
been found in sample CRB 36a (Roberts, 1975:263).

The fauna contains a mixture of species known
from the underlying Loyola Formation, and several
new species (Fossula cf. derbyi, Paleoanculosa ken-
nerleyi, Aylacostoma sulcata). Of the new species for

this fauna, F. cf. derbyi was described from speci-
mens from strata at Santa Maria do Boca do Monte,
Rio Grande do Sul, Brazil. The age of the strata is un-
certain. Ihering (1907) thought they were of Creta-
ceous age, but Parodiz (1969:83) was of the opinion
that they were Upper Tertiary. A. sulcata is known
at Pebas (Conrad, 1871), Pichua (Woodward, 1871),
and Iquitos (de Greve, 1938).

In view of the occurrence of species common to
both the Loyola and Mangan formations, including
Vetustocytlieridea bristowi thought to be indicative
of the Miocene, the Mangan Formation is also re-
garded as Miocene. However, the vast thickness of
sediments, approximately 1,800 m, separating the
two formations suggests a significant time interval
between them. The Mangan Formation is accord-
ingly regarded as Upper Miocene in age.

Santa Rosa Formation

The uppermost formation of the Tertiary sedi-
ments of the Cuenca Basin consists predominantly
of coarse clastic units alternating with red silty and
sandy shales. They crop out on the west side of the
basin. There appears to be an upward transition
from the Mangan Formation. The contact with
the overlying Turi Formation appears also to be
transitional, at least in the center of the basin.
Very coarse boulder beds in the Nazon area (362,
015) are probably indicative of proximity to their
source area. The maximum thickness at outcrop
appears to be about 500 m.

No fossils have been found in the formation and
only a speculative Pliocene age can be assigned to
it.

Turi Formation

Succeeding the Santa Rosa Formation in the cen-
ter of the basin is a series of well-bedded, generally
horizontal, conglomerates, volcanic breccias, ash-
es, shales, and sandstones. In the type area (214,
771) to the south of Cuenca, and in the central por-
tion of its outcrop, there is a well-developed basal
conglomerate. Around Turi the formation rests un-
conformably on the Loyola and Azogues forma-
tions, but further west there appears to be a tran-
sitional upward passage from the Santa Rosa
Formation.

North of Biblian there is much volcanic debris in
the formation which has expanded in thickness
from approximately 280 m at the type locality to
aproximately 1,200 m. In this northern sector of the
basin the sediments are steeply dipping.

CRB 36b
CRB 36a,
42c

CRB 36a

CRB 36b,
42c

CRB 42c
CRB 36b
CRB 20, 60

!982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

15

The only fossils found to date are unidentified
gastropods, silicified wood, and a possible mam-
moth bone (Erazo, 1957:28). The age is considered
to be Pleistocene.

The Turi Eormation is succeeded by a series of
volcanic deposits (Tarqui Formation and Llacao
Volcanics), till, terrace gravels, and alluvium. The
Tarqui Formation provides the only independent
date for the minimum age of the sediments of the
Cuenca Basin. At two localities, carbonized wood
fragments gave C*'* dates of 24,900 ± 1 ,200, and
34,300 ± 1,950 BP (UNDP, 1969:22, 1972:13).

Sample sites

(Cuenca

Local grid

Basin)

references

Formation

CRB 1

305, 834

basal Azogues

CRB 2

311, 832

high Loyola

CRB 4

258, 786

basal Azogues

CRB 5

287, 780

uppermost Biblian

CRB 6

356, 846

basal Loyola

CRB 7

357, 842

basal Loyola

CRB 8

250, 757

high Biblian

CRB 9

367, 867

basal Loyola — Descanso

section; Loyola
Formation rests on lava
radiometrically dated at
19 and 20 million years
ago

CRB 10

370, 870

basal Loyola

CRB 11a

361, 855

high Loyola

1 lb, c

361, 855

middle Loyola

CRB 12

311, 832

high Loyola

CRB 13

308, 882

basal Azogues

CRB 14

311, 832

high Loyola

CRB 15

400, 923

basal Loyola

CRB 17

378, 915

high Loyola

CRB 18

378, 987

basal Loyola — “Biblian

sandstone and
conglomerates” locality
of Marshall and Bowles
(1932); loc. 1038 of
Liddle and Palmer (1941)

CRB 20

337, 953

lower unit of Mangan

CRB 26

406, 947

basal Loyola — “echinoid’
locality of Erazo ( 1965)

CRB 28

373, 016

basal Loyola

CRB 30b

245, 773

high Loyola

CRB 34

398, 912

basal Loyola

CRB 35

400, 918

basal Loyola

CRB 36

333, 976

middle unit of Mangan

CRB 39

179, 663

“Ayancay Group” of Giron
Basin

CRB 42

332, 994

middle unit of Mangan

CRB 46

400, 929

basal Loyola

CRB 48

400, 929

basal Loyola

CRB 49

400, 929

basal Loyola

CRB 60

365, 161

Mangan

CRB 61

368, 159

Loyola

Nabon Basin

Nabon is a village about 50 km south of Cuenca
and 75 km north-northeast of Loja. The deposits of
the basin, the Nabon Formation, crop out over an
area 16 by 6 km, and rest unconformably on the
Oligo-Miocene Saraguro Formation. The general
strike of the formation is northeast to southwest.
The formation is divided into three units — a lower
one, some 40 m thick, of bedded tuffs; a middle
one, 130 m thick, of conglomerates, sandstones, silt-
stones, shales, diatomites, lignites, and minor tuffs;
and an upper unit, 160 m thick, of bedded tuffs and
agglomerates. The presence of lignite has been tak-
en by most authors to indicate an equivalence with
the coals of the Mangan Formation of the Cuenca
Basin, and the San Cayetano Formation of the Loja
and Malacatos basins. This is supported by the lim-
ited flora found in the middle unit which is also
known from Loja (Bristow, 1976: 107). Additionally,
the rodent Olenopsis aequatorialis of uncertain
stratigraphic provenance, was described from Na-
bon (Anthony, 1922). O. aequatorialis is common
in the Miocene of La Venta in Colombia (Fields,
1957). At the time of writing. Fields regarded the
La Venta sediments as Upper Miocene in age (Vin-
dobonian/Friasian) . These stages are now regarded
by modern authors (for example. Van Eysinga,
1975) as part of the Middle Miocene. To date, no
Mollusca have been found in the Nabon Formation.
A tentative Middle-Upper Miocene date is hereby
given to the Nabon sediments.

Loja and Malacatos Basins

These two separate basins have been studied by
numerous authors, of whom the most recent is Ken-
nerley ( 1973; 1 : 100,000 Loja, 1975, and Gonzanama,
1975 sheets). Initially Kennerley (1973) gave sepa-
rate formational names to the similar deposits in
each of the basins, but later, on the Loja and Gon-
zanama sheets, the stratigraphical nomenclatures of

16

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 19

the two basins were united and the following for-
mational names, in ascending order, are currently
in use: Loma Blanca, Trigal, San Cayetano, and
Quillollaco.

Loma Blanca Formation

The type locality is a hill (895, 338) 4 km west of
Malacatos. It is a volcanic formation comprising a
well-developed basal agglomerate, succeeded by
tuffs and agglomeratic tuffs. In the Loja Basin an-
desitic and basaltic lavas occur. It rests unconform-
ably on older formations.

No fossils have been found in the formation.
Kennerley (1973:29) thought that it was possibly
equivalent to the Saraguro Formation. At the time
Kennerley was writing the Saraguro Formation had
not been dated, but it is now known to be of Oligo-
Miocene age (26 and 28 million years ago, Snelling,
in Bristow, 1976:107).

Trigal Formation

The name is taken from the Rio Trigal (963, 613)
some 4.5 km northwest of Loja, where there are
excellent outcrops. It crops out as a narrow belt on
the west side of the basin where it rests unconform-
ably on the Loma Blanca Formation. The dominant
lithology is coffee-colored shales and clays; gypsum
is common as a coating to bedding surfaces and
joints.

In the Malacatos Basin the Trigal (formerly Al-
garobillo) Formation overlies the Loma Blanca For-
mation conformably. The lithology is similar to that
of Loja but with minor beds of sandstones, tuffs,
and in the upper part, thin seams of coal.

The presence of the ostracod Cyprideis stephen-
soni (Sandberg) in one sample from the Malacatos
Basin, dates the formation as Miocene.

San Cayetano Formation

The type locality is the village of the same name
(003, 596) 1 km north-northeast of Loja. The for-
mation crops out extensively in the center of the
Loja Basin, and as two belts in the Malacatos Ba-
sin. In both basins there is a conformable upwards
passage from the Trigal Formation.

The lithology consists of a well-bedded sequence
of sandstones, silicified shales, calcareous shales,
coarse conglomerates, diatomites, and low-grade
coals. The conglomerates are most common at the
top and bottom of the formation. The sub-bitumi-
nous coal or lignite occurs in five principal seams
in the Loja Basin, and in eight in the Malacatos

Basin. In the latter basin the coal seams are found
in the lower part of the formation. The maximum
thickness appears to 700 to 800 m.

Quillollaco Formation

The type locality is the stream (997, 502) 7 km
south of Loja. The formation, which appears to suc-
ceed the San Cayetano Formation conformably in
the Loja Basin, crops out on the west side of the
basin. In the Malacatos Basin the contact with the
San Cayetano is not seen, and the Quillollaco For-
mation rests unconformably on the older Tertiary
formations.

The formation consists predominantly of coarse
conglomerates with interbedded grits, greywackes,
and sandstones. No fossils have been found in the
formation. Kennerley (1973) thought that it was of
Pliocene age; it is probably equivalent to the Santa
Rosa Formation of the Cuenca Basin.

Flora, Fauna, and Age

Grebe (in Putzer, 1968:480), working on pollen,
suggested a correlation between the coals of the
Loja, Malacatos, and Cuenca basins.

The flora, unfortunately not stratigraphically lo-
cated but probably from the San Cayetano Forma-
tion, was first studied by Engelhardt (1895) and sub-
sequently by Berry (1918, 1929, 1934, 1945). Berry
initially (1918) placed this tropical flora in the Low-
er Miocene, but in later revisions (1929, 1934, 1945)
of this same flora, he regarded it as Mio- Pliocene,
or Pliocene age. Some of these plants are known in
the Nabon (Bristow, 1976) and Cuenca (Berry,
1934, 1945) basins.

Fish scales are common in the San Cayetano For-
mation, but the only fish so far identified, whose
exact provenance is uncertain, is Carrionellus du-
morterei White (1927).

The molluscs consist of Dyris cf. gracilis Conrad
“form” tricarinata (Boettger) occurring on an un-
located slab sent to the British Museum of Natural
History by Prof. Carrion, and in shales collected by
the author and B. Kennerley from the San Cayetano
Formation (JW424 [008, 582]). The latter are found
as scattered impressions on the bedding surfaces of
some of the shales.

D. gracilis tricarinata was described from Pebas,
Peru (Boettger, 1878). It is also known from Iquitos,
Peru (de Greve, 1938), Pichua near Cochiquinas,
Peru (Hauxwell, Colin. BMNH), Canama on Rio Ja-
vari, Peru (Etheridge, 1879 as Melania hicarinata
nov. sp.), and in ?Pliocene beds at La Tagua, Ca-

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

17

queta River, Colombia (C. P. Nuttall, personal
communication).

The Miocene ostracod Cy pride is stephensoni
(Sandberg) has already been mentioned.

It is thus evident that the oldest non-volcanic de-
posits (Trigal Formation) is of Miocene age, as evi-
denced by C. stephensoni. The San Cayetano For-
mation, based on the similarity of its lithology and
pollen to the Mangan Formation, is probably Upper
Miocene.

Deposits of the Rio Pachitea and
Vicinity in Eastern Peru

The red beds of the Rio Pachitea area were
mapped and discussed by Singewald (1927, 1928).
Fossils collected during the field work were exam-
ined by Pilsbry (1944). Pilsbry made a comparison
at generic level between the fauna from the Rio
Pachitea and those from the Cuenca Basin, Ecua-
dor, and from the Magdalena Valley, Colombia. At
that time none of the species was known to occur
in more than one of the three basins, though some
of the [Aylacostoma] Longiverena are closely sim-
ilar. Pilsbry concluded that until further collecting
afforded a more definite clue to the age of the Pach-
itea deposits they could tentatively be considered
to be about the age of the La Cira deposits — Upper
Oligocene or Lower Miocene. Such differences as
appeared between the faunas of the La Cira, Pach-
itea, and Cuenca beds were more likely to be due
to their geographic separation than to any material
difference in age. The correlation of these three de-
posits, and Pilsbry’s age assignment of Upper Oli-
gocene or Lower Miocene, based as it was on fairly
slender evidence, has been corroborated by the
present study. The common presence of Pomacea
manco in the Middle Miocene Loyola Formation of
the Cuenca Basin and at Pachitea, suggests that the
deposits at the latter locality might be slightly youn-
ger than thought previously. Koch and Blissenbach
(1962:77), however, suggest that the Sol Formation,
in which the fauna occurs, lies close to the Creta-
ceous/Tertiary boundary, or may be wholly Creta-
ceous (Koch and Blissenbach, 1962: fig. 21 ; pi. 3).
The evidence for such a radical reassessment of the
age is not convincing, being based on several new
species of charophytes and new, unnamed, but
numbered ostracods. Two species of charophytes
( Tectochara supraplana sulcata and Kosmogyra
monilifera) from this area had earlier been consid-
ered as of Eocene or Oligocene age (Peck and Rek-
er, 1947).

It is worth noting at this point that Mitricaiilis
incarum Pilsbry, described from “Marine [?Eocene]
Tertiary of the Pachitea River,’’ is now known from
Maestrichtian sediments of the Cuenca Basin (R.
Cleevely determination in Bristow and Hoffstetter,
1977:351). The deposits with M. incarum underlie
the non-marine fossils mentioned above; a Creta-
ceous age is not inconsistent for this part of the
sequence.

Magdalena Embayment, Colombia

Fossiliferous strata have been found at three prin-
cipal levels in the Tertiary sediments of the Magda-
lena Valley between the Sogamoso and Carare
rivers (Wheeler, 1935). As with Tertiary non-marine
sediments elsewhere the fauna when described was
entirely new and thought to be endemic. No inde-
pendent dating of the associated strata was possible
and only tentative ages could be assigned to the
respective formations. However, some of the ar-
guments for the age determination are in part cir-
cular as having “fixed” the age of the oldest, Los
Corros, fauna at the top of the Esmeraldas For-
mation as Upper Eocene, the succeeding forma-
tions were tied into this chronology. Nevertheless,
the various ages have never been seriously ques-
tioned and are generally accepted at the present
day.

Esmeraldas Formation

Pilsbry and Olsson (1935:7) “based partly on stra-
tigraphy and partly on faunal evidence” concluded
that the Los Corros fauna belonged to the Upper
Eocene, equivalent to the Saman Formation of Peru
and the Jacksonian of the southern United States.
The faunal evidence was based on the somewhat
suspect comparison of two species of Diplocyma in
the Los Corros fauna with one species, Potamides
lagunitensis (Woods), in the Saman Formation. The
overall fauna had an “Eocene rather than an Oli-
gocene aspect.” The stratigraphic evidence appears
to be based on the fact that marine Upper Eocene
rocks are widespread in the coastal region of Co-
lombia and that therefore “it seems reasonable to
believe that the non-marine equivalent of these
rocks should occur in the Tertiary embayments, as
well exemplified by the deposits of the Magdalena
valley.”

More recently Van der Hammen (1957:65), based
on palynology, suggested an Upper Eocene date for
the Esmeraldas Formation. However, the evidence
is inconclusive because the pollen spectrum for the

18

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Table 2. — Distribution aj the non-encleinic. Tertiary non-marine Ecuadorian MoUusca. Pius sign indicates species is present.

Ecuador

Cuenca

Peru

Colombia

Rio

Esme-

La

Vene-

Species

Chota Loyola Azogues Mangan

Loja Pachitea Pebas

Canama raldas Mugrosa Cira

Real Caqueta zuela Brazil Florida

Diplodon (E.) guarani-

anus hihlianus

+

+

Fossula derbyi

+

+

Pomacea nianco

+

+

Aylacostoma browni

+

+

+

+

+

Aylacostoma dickersoni

+ +

Aylacostoma sulcata

+

+

To.xosoma ehoreum

+

+

Liris minuscula

+

+

Hydrobia ortoni

+

+

+

Dyris gracilis

+ +

+

+

Poteria bibliana

+

+

Puperita sphaerica

+

+

Erodona itiuitensis

+

+

Upper Eocene is matched almost exactly by one in
the Upper Oligocene (Van der Hammen, 1957; pi.
1). Additionally, the age of the Esmeraldas Eor-
mation to which the ‘Upper Eocene’ pollen zones
correspond is regarded unequivocably as Upper
Eocene because of the fossiliferous Upper Eocene
horizon [Los Corros] which is contained therein.
The circular argument for the age of this fauna is
thus nearly complete. Germeraad et al. (1968: Eig.
17) have indicated that the base of the Esmeraldas
Formation is of Middle-Upper Eocene age, but un-
fortunately the top of the formation in which Los
Corros fauna occurs, has not been palynologically
dated. The upper Eocene age of the Los Corros
Formation is not established with any degree of cer-
tainty. The available evidence from Colombia (see
Mugrosa below) does not preclude an Oligocene age
for the fauna.

Mugrosa Formation

The Mugrosa Formation succeeds the Esmeral-
das Formation with apparent conformity. The Mu-
grosa fossiliferous horizon varies from 360 to 1,350
m above the base. One species, Aylacostoma eu-
cosmius (Pilsbry and Olsson), is also known from
La Cira Formation, while the new species Paleoan-
culosa kennerleyi in the Upper Miocene Mangan
Formation of the Cuenca Basin, previously in-
dicated by Bristow and Hoffstetter (1977) as A. sig-
macliiliis (Pilsbry and Olsson) from the Mugrosa
Formation, is very similar to the latter species.

Pilsbry and Olsson (1935:8) thought that some of
the "Flemisinus" in the Mugrosa Formation were

closely related to forms known in the Middle Oli-
gocene of Cuba and Antigua. At the time these au-
thors were writing (1935), the Oligocene included
the Aquitanian and Burdigalian stages which at the
present day are considered as the lowest stages of
the Miocene.

Hoffstetter (1970), however, suggests that these
Cuban and Antiguan occurrences are of Lower Oli-
gocene age. Van der Hammen (1960), without giv-
ing details, dated the lower part of the formation on
pollen as Lower Oligocene, and the upper part as
Middle Oligocene.

There appears to be a general agreement for the
Oligocene age of the formation, but there is uncer-
tainty as to whether it is of Lower or Upper Oli-
gocene. Upper Oligocene is preferred.

La Cira Formation

This formation rests conformably on the Mugrosa
Formation. The La Cira fossil horizon occurs some
780 to 2,070 m above the base of the formation. To
date, only one of the molluscan species, A. eucos-
niius ( = ''Hemisinus" [Longiverena] laciranus Pils-
bry and Olsson, 1935) is known outside this horizon
in the Mugrosa Formation.

Wheeler thought that the fauna might be Upper
Oligocene or Lower Miocene in age. Porta (1974),
in a resume of the age of the formation, suggested
that the lower part was probably of Oligocene age,
and the upper part of Miocene age.

The author favors a Miocene age for the La Cira
fauna and, if the correlation first suggested by Pils-
bry (1944) with the Cuenca and Pachitea deposits

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

19

is substantiated, it may well be of Middle Miocene
age.

Caqueta Basin, Colombia

A recently described section (M. Eden, personal
communication) along part of the Rio Caqueta in

southern Colombia has revealed fossiliferous strata
of possible Pliocene age. These have been named
the La Tagua Beds. They consist of up to at least
25 m of clay, silt, claystone, and siltstone, locally
with sandy intercalations, and rest on the Upper
Cretaceous Roraima Formation.

The molluscs from the La Tagua Beds have been

20

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

identified by C. P. Nuttall (personal communica-
tion). The fauna includes Doryssa sp., Cochliopina
sp., Aylacostoma browni (Etheridge), Hydrobia cf.
ortoiii (Conrad), Dyris gracilis (Boettger), Aniso-
thyris erectus Conrad, Anisothyris sp., and uniden-
tifiable unionids. This freshwater fauna has close
affinities with that from Pebas and Iquitos (only the
Doryssa and Cochliopina are not known to occur
at Iquitos and Pebas), and has a modern aspect.
Nuttall concluded that the La Tagua Beds were late
Cenozoic, possibly of Pliocene age. A. browni is
known also from both the Middle and Upper Mio-
cene Loyola and Mangan formations of the Cuenca
Basin.

One described species of ostracod, Pelocypris
ziichi Triebel, has been recorded from the La Tagua
Beds (Sheppard and Bates, 1980). P. ziichi is known
only from San Salvador where it occurs in strata of
probable Pleistocene age (Triebel, 1953). Other new
species and genera of ostracods are common to
both the La Tagua and Pebas beds.

Canama, Peru

Brown (1879), who described the section at Can-
ama, and at several other localities downstream,
distinguished an upper unit of younger (?Pleisto-
cene -i- ?Holocene) river deposits, from the under-
lying, gently dipping. Tertiary rocks, of which a
maximum of 20 m was seen at any one locality.

One fossil (Aylacostoma browni) from Canama
occurs in the Middle Miocene Loyola Formation
and in the Upper Miocene Mangan Formation of
the Cuenca Basin. It also occurs in the Pebas Beds.
Other fossils from Canama suggesting a good cor-
relation with the Pebas Beds of Pebas, Pichua and
Iquitos include Pachydon carinata Conrad, P. ten-
nis Gabb (—hauxwelli Woodward), P. erectus Con-
rad (=tnmida Etheridge), P. cnneatiis (Conrad)
(=Corbnla canamaensis Etheridge). Former au-
thors used Pachydon as an alternate for Anisothy-
ris.

Iquitos and Pebas, Peru

Widespread fossiliferous strata have been found
around Iquitos (see de Greve, 1938:13-18), at Pe-
bas, some 30 mi below Pebas (Conrad, 1871), and
at Canama on the Rio Javary (Brown, 1879). Similar
strata, but unfossiliferous, were noted by Brown
(1879:76) at Sao Paulo, lower down the Amazon.
The Iquitos and Pebas faunas have much in com-
mon and may be treated as one. They are also sim-

ilar to the deposits and fauna from Canama which
are described separately in this account.

Although widely distributed, the thickness of the
fossiliferous strata and associated beds exposed at
any one locality is not very much. De Greve
(1938:16) described a partial 6.8 m section of blue
and yellow clay which was observed in the Rio
Itaya near Iquitos. Lignite is locally associated with
the fossiliferous beds.

Some 42 fossils, dominantly molluscs, have been
recorded from Pebas and it is the richest locality
discussed in that paper. The fauna, both in numbers
of species in common and in the method of pres-
ervation (as undistorted shells, commonly with the
nacreous layer intact, and from which the matrix
can be easily removed), is most closely linked to
that of the La Tagua and Canama localities where,
to date, more limited faunas have been obtained.
Two species (Aylacostoma browni and A. sulcata)
are common to the Cuenca deposits and to the col-
lective Pebas/La Tagua/Canama faunas. Of partic-
ular interest is A. browni, as it is known in both the
Loyola and Mangan formations of the Cuenca Ba-
sin, from La Tagua, from Canama, and from Tres
Unidos, Brazil (Parodiz, 1969:141 under the name
sidcatus) and is thus the most widely distributed of
the Tertiary non-marine molluscs in the area under
discussion. However, Anisothyris, which is an im-
portant element of the Pebas and La Tagua faunas,
is absent in the Cuenca Basin, but replaced by Er-
odona.

There is no independent method of dating this
fauna. Most workers are agreed that it is of Tertiary
age, although some deposits in nearby Brazil of sim-
ilar lithology to that at Pebas and Iquitos have
been shown to be Pleistocene (Simpson, 1961). The
shell preservation (see above) presents a “young”
aspect, and the fact that some shells retain their
coloring has been cited as an indication of no great
antiquity (Gabb, 1869:197). The difference in the
method of preservation between the Amazonian
faunas and those of the Andes may be no more than
a reflection of the differing tectonic environments
in which the deposits occur. Those of the Cuenca
area, for instance, are strongly folded and faulted
and have been uplifted by some 2,500 m (that is,
from near sea level, as witnessed by Necronectes
proavitus) since deposition. Marine shells similarly
preserved to those from Pebas are known from the
Upper Miocene/Lower Pliocene deposits of coastal
Ecuador (see for example, Pilsbry and Olsson,
1941). The retention of shell color cannot be taken

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

as an indicator of age because several of the Neri-
tina in the Cuenca Basin still retain their markings,
and the author has seen marine gastropods of Upper
Miocene age from coastal Ecuador with their color
preserved. Color pattern in the Neritacea is known
from the Devonian onwards.

The common occurrence of four species from
Cuenca and the Amazon clearly indicates no sig-
nificant time gap between the respective faunas. In
the author’s opinion the Pebas fauna is probably of
Upper Miocene or Lower Pliocene age.

Summary of Occurrences in Cuenca Basin

Twenty-nine species of molluscs are now known
from the Cuenca Basin, and one each from the Cho-

ta and Uoja basins. Of this fauna, 12 species have
been recorded outside Ecuador, principally from
localities in the western Amazon Basin, suggesting
that in earlier times, before the main Andean up-
lift, there was a closer connection between the
Cuenca Basin (the present day Atlantic/Pacific
watershed crosses the basin) and the western
Amazonian deposits. The maximum age for the
Cuenca fossils is uppermost Lower Miocene. It is
suggested that other occurrences of the Cuenca fau-
na outside Ecuador might also be of Miocene age.
The distribution of the non-endemic Cuenca mol-
luscs outside Ecuador is shown in Table 2.

PART 2. PALEONTOLOGY
J. J. Parodiz
INTRODUCTION

The majority of the Mollusca of the Tertiary, non-
marine deposits of Ecuador belong to the upper sec-
tion of the Lower Miocene, Loyola Lormation, with
brackishwater gastropods as Neritina, or purely
freshwater bivalves of the Mutelacea and Uniona-
cea. In the later, the Hyriidae in South America are
known in South America since the early Paleocene
in Patagonia.

Although fewer species are known from the Mid-
dle Miocene, Azogues Lormation, and the Upper
Miocene, Mangan Lormation, such genera as Neo-
corhicula are common to both and found abun-
dantly in all the Miocene. The freshwater cerithi-
aceans, Aylacostoma , Doiyssa, and Paleoanciilosa,
are also abundant in the Mangan Lormation and in
the last genus, P. kenneryleyi is found in great num-
bers and is a characteristic of that formation. In the
northern Chota Basin, the middle section contains
exclusively the lacustrine Liris minuscula. In the
Cayetano Lormation in the Loja Basin (most prob-
ably Upper Miocene in age), the common species
is Dyris gracilis.

The Miocene non-marine fauna of Ecuador dif-

SYSTEMATIC

Class Bivalvia
Superfamily Unionacea
Lamily Hyriidae Swainson, 1840
(Ortmann’s Hyriinae of “Mutelacea”)
Subfamily Hyriinae, restricted
Parodiz and Bonetto, 1965
Tribe Diplodontini (Diplodontidae Ihering, 1901)
Genus Diplodon Spix, 1827

IriJea Swainson, 1840.

Prodiplodon Marshall, 1928.

Eudiplodon Marshall, 1928.

Schleschiella Modell, 1950.

Type species. — Diplodon ellipticiis Spix (in Wag-
ner, 1827), on pi. 16, figs. 1-2.

Adams and Adams (1854) under the name Diplo-
don included an heterogenous conglomeration of
species from all over the world. The genus was geo-

fers, as a whole, from that known in Peru at Pebas
and Iquitos, which is younger. In the Peruvian Plio-
cene, the large freshwater mussels of the Unionacea
and Mutelacea are practically absent, but charac-
terized by more species of Neritina and Anisothyr-
is. Anisothyris is not found in the Ecuadorean Mio-
cene but replaced by Erodona, although it is not
common. Also the Neocorhicida and Paleoancii-
losa are absent in the Peruvian Pliocene. The pres-
ent knowledge of the Miocene malacofauna of Ec-
uador permits the reconsideration of the age of
Pebas — not only have some Pebas species been
found in Cuenca, Loja, and Chota basins, but also
very few of the known species from Pebas have
survived. Pebas is, probably, not younger than the
Lower Pliocene.

Table 3 includes all the species known from the
Ecuadorean Miocene. Only one species is con-
firmed as surviving in the Recent fauna, but as an
allochronic subspecies — Diplodon (Eciiadorea)
guaranianns guaranianus. The Ostomy a that is listed
here as O. ”cfr.” flnviatilis has affinities with this
living species, although it may not be the same.

ACCOUNTS

graphically restricted to South America by Ortmann
(1921); the variable characteristics of the shells
have recurrent features among the species groups
and subgenera.

Ortmann was the first to study the glochidia of
many species, but he did not use those embryolog-
ical features for subgeneric divisions. Accepting
Simpson’s two subgenera, Diplodon sensu stricto
and Cyclomya (now under the correct name Rhip-
idodonta), Ortmann recognized within the first a
group of species around Diplodon hylaeiis, which
subsequent authors included in Eciiadorea Mar-
shall and Bowles. When the condition of the larval
stages became better known, the main groups were
separated on that biological basis. According to
Bonetto (1965n and after), most of the better known
species belong to two groups: 1) with parasitic glo-
chidian larvae; shells are usually longer than high;

!982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

23

Table 3. — Species of MoUusca known from llie Miocene of Ecuador. A pins sign indicates that the species is known from the basin

Cuenca Basin

Species

Loyola Azogues Mangan

Chota Basin Formation Formation Formation Loja Basin

Bivalvia

Superfamily Unionacea
Family Hyriidae

Diplodon (Ecuadorea) guaraniunus hiblianus
(Marshall and Bowles)

Diplodon (E.) bristowi, new species
Diplodon (E.) liddlei (Palmer)

+

+

+

Supeifamily Mutelacea
Family Mycetopodidae
Subfamily Monocondylaeinae
Fossida derbyi (Ihering)
Monocondylaea azognensis (Palmer)
Monocondylaea pacchiana (Palmer)
Monocondylaea sp.

+

+

+

+

Subfamily Anodontitinae
Anodontites olssoni (Palmer)

+

Superfamily Sphaeriacea
Family Sphaeriidae
Pisidium sp.

+

Family Corbiculidae

Neocorbicida cojitamhoensis (Palmer)

+ + +

Superfamily Myacea
Family Corbulidae

Erodona iquitensis (de Greve)
Ostomya cf. fluviatUis (FI. Adams)

+

+

Gastropoda

Supeifamily Trochacea
Family Trochidae
ICaliostoma sp.

+

Supeifamily Neritacea
Family Neritidae
Subfamily Neritinae

Neritina pacchiana Palmer

Neritina loyolaensis, new species

Pnperita aff. sphaerica (Olsson and Harbison)

+ +
+

+

Supeifamily Cyclophoracea
Family Aperostomatidae

Poteria {Pseiidoaperastoma) hibliana (Marshall and Bowles)

+

Supeifamily Viviparacea
Family Ampullariidae

Pomacea (Limnopomus) manco Pilsbry

+

Supeifamily Rissoacea
Family Hydrobiidae

Hydrobia ortoni (Gabb)

Liris minnsciila (Gabb)

Dyris gracilis Conrad
Lyrodes sp.

Toxosoma eboreum Conrad
PotamoHthoides hiblianus Marshall and Bowles

+

+

+

+

+

+

24

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Table 3. — Continued.

Cuenca Basin

Species

Loyola Azogues Mangan

Chola Basin Formation Formation Formation Loja Basin

Supeifamily Cerithiacea
Family Thiaridae

Aylacostoma sulcatns (Conrad) +

Aylacostoma hrowni (Etheridge) + +

Aylacostoma sp. + +

Aylacostoma dickersoni (Palmer) + +

Family Pleuroceridae

Doiyssa bihliana (Marshall and Bowles) + +

Paleoancidosa kennerleyi, new species +

Superfamily Lymnaeacea
Family Planorbiidae

Gyraidus sp. +

Superfamily Succineacea
Family Succineaidae

Siiccinea sp. +

Inceilae sedis +

subgenus Diplodon s.s.; 2) with glochidia of direct
development; shells with greater height about the
middle of the valve, subrotund; subgenus Rhipi-
dodonta Morch, 1853 ( = Cvc/o/h_vu).

Although the glochidial development is still un-
known in a good number of species, and shell vari-
ation is considerable, the system is workable for the
main group of species. One of these groups, with
shells like those in the first group (to which the Mio-
cenic Ecuadorean species belong), have non-parasit-
ic glochidia as Rhipidodoiitcr, also they present re-
duction in size, and greater development of umbonal
sculpture forming radial V’s inserted one into the
next, reaching the center of the disc or, occasion-
ally, the ventral margin. These were included by
Modell (1950) in the genus Ecuadorea. The oldest
of this group were found in the Oligocene of Peru,
and several are still living in the northwest, middle,
and southwest of South America, but absent in east-
ern Brazil. Shell characteristics in this group seem
to be more constant than those found between the
subgenera Rhipidodonta and Diplodon s.s.

The parasitic glochidia are probably a more prim-
itive type, and non-parasitic glochidia are a second-
ary evolutionary development; this is suggested by
the oldest forms of Diplodon in the Paleocene,
while Ecuadorea appears later in the Tertiary (a
similar parallel evolution occurred also in other
Unionacea families of North America). T. R. Rob-
erts (1975) described remains of fossil fish of the
genus Hoplias in the Ecuadorian Miocene; parasitic

glochidia have been found with frequency in the
fish Hoplias malabaricus of the Parana, producing
cysts in the gills. If the fossil Ecuadorea had a tran-
sitional type of glochidia with any relationship with
the fossil Hoplias, it could not be verified; how-
ever, their direct descendant living species in Ec-
uadorea have all non-parasitic glochidia.

The distinction between the flat and elongated
Diplodon s.s. and the short, inflated and strongly
costulated Ecuadorea is already evident in the Pa-
leocene species; the Diplodon are extraordinarily
similar to their ancestors from the Triassic of Penn-
sylvania. Pilsbry (1921) described several species of
that age from York Co., Pennsylvania, which he
stated are “like Diplodon and Hyria and totally
unlike that of Unio and allied genera of the North-
ern Hemisphere.” Comparison of Diplodon penn-
sylvanicus Pilsbry from the Triassic with Diplodon
bondembenderi Doello-Jurado from the Paleocene
of Patagonia shows that they belong to the same
group. Richards (1948) also described Antediplodon
borealis (using Marshall’s generic name) from the
Newark series of York Co., which has a shape sim-
ilar to that of the living D. rhuacoicus group.

A full generic separation of Ecuadorea from Dip-
lodon, however, would obscure rather than clarify
taxonomic relationship. Obviously Modell (1950)
was not entirely acquainted with all the fossil and
living species involved. Eor practical purposes,
Ecuadorea can be accepted as a subgenus. In order
to avoid repetitious comments when describing the

1982

BRISTOW AND PARODIZ— ECU ADORIAN TERTIARY SEDIMENTS

23

fossil species from Ecuador, it is pertinent here, to
give a brief account of the other species that can be
assigned to Ecuadorea.

Diplodon latouri (Pilsbry and Olsson, 1935). —
Oligo-Miocene of La Cira Eormation, Colombia
(see Parodiz, 1969:62). Described as Triplodon, it
is the oldest species of the group and also the small-
est (known only by the type, it might be ajuvenile).
The V-shaped sculpture reaches only — as in piuir-
anianus bihlianus — the middle of the disc. The au-
thors already indicated that it may be referred to
Ecuadorea.

Diplodon (Ecuadorea) hylaeus (d’Orbigny ,
1842). — Type locality: Palometas River, north-cen-
tral Bolivia (specimens usually labelled as D. hy-
laeus from southeastern localities actually belong
to D. guaranianus). It is very rare, as its author
remarked. Extremely thin compared to D. gua-
ranianus, so much that the external ribs can be seen
strongly marked on the internal surface which is
very iridescent instead of white, and hinge and mus-
cle scars weaker.

Diplodon (Ecuadorea) hylaeus pazi (Hidalgo,
1868). — Described originally as Castalia. Type lo-
cality: Imbabura, Ecuador. Differs from D. hylaeus
hylaeus by its sculpture reaching the ventral margin
very regularly. It is a living subspecies.

Diplodon (Ecuadorea) guaranianus (d’Orbigny,
1835). — Described originally as Unio. Type locali-
ty: Parana River at Itaty, Corrientes, Argentina.
The most abundant living species, showing clinal
variation in its distribution along the Parana-Para-
guay drainage. D. asuncionis Marshall from Para-
guay, and D. hasemani Ortmann from the Guapore
River, Brazil, correspond to such clinal forms. This
species has a very strong and heavy shell, as well
as a strongly developed hinge, and the interior of
the valves is pure white.

Diplodon from the Miocene of Ecuador
Subgenus Ecuadorea Marshall, 1934

Castalioides Marshall, 1934.

Type species. — Ecuadorea hihliana Marshall and
Bowles.

Original description. — “Hyriinae with plentiful radial sculp-
ture similar to that of Diplodon and still more similar to that of
Hyria. The radial ribs are arranged in V pattern, each V nesting
in a succeeding one. Posterior dorsal area with several plicae
crossing it obliquely to the margin.”

Original description of Castalioides. — “Shell with strong
sculpture of radial ribs, several of the innermost pairs arranged

to form very long Vs. Ribs crossing the anterior and posterior
slopes form a divaricate pattern with the radial ribs.”

Type. — C. laddi, “Quaternary,” Venezuela.

These two diagnoses, although differently word-
ed, say the same thing. Castalioides cannot be sep-
arated from Ecuadorea. The authors of Ecuadorea
said that “it is difficult to decide the relationship of
this genus to Recent genera . . . in form the shell
is like Diplodon but that genus is not plentifully
sculptured.” The similarities of Castalioides laddi
are of the same order as those found in Diplodon
guaranianus in relation to D. bihlianus. Marshall
and Bowles also found resemblances of Ecuadorea
hihliana with ""Castalia" pazi Hidalgo, which now
is recognized as a Diplodon. The hinge of Casta-
lioides, which in Palmer’s (Liddle and Palmer, 1941)
opinion is apparently different from that of Ecua-
dorea, does not differ from some of the variations
present in the living D. guaranianus ,

Diplodon (Ecuadorea) guaranianus bihlianus
(Marshal! and Bowles)

Ecuadorea bibliana Marshall and Bowles, 1932:5, figs. 7-8;

Palmer, 1941:401, pi. 7, figs. 1-6.

Castalioides laddi Marshall, 1934:78, figs. 1-4; Palmer, 1941:402.
Ecuadorea laddi, Modell, 1950: 143.

Diplodon guaranianus bihlianus, Parodiz, 1969:66, pi. 6, figs. 1-

7, pi. 8, fig. 6; Bristow and Hoffstetter, 1977:183.

Type locality. — Biblian, northwest of Azogues,
Loyola Eormation, Ecuador. Type in NMNH.

Specimens observed. — Localities CRB 18 and 28
(sandstones and conglomerates), Basal Loyola Eor-
mation NW of Loyola.

Original description . — “Shell rather compressed, slightly nar-
rower in front. Concentric sculpture of fine growth striae, with
a few of the rest periods a little accentuated. Radial sculpture of
a number of riblets as to form a series of V’s, and with other
riblets at the front and back which if continued would form ad-
ditional V’s. The anterior prong of each V is narrow, clear-cut,
and nearly straight. The posterior prong is heavier and more
irregular and curves toward the front of the shell. At the lower
end where the radial sculpture dies out, the surface is somewhat
pimpled. The posterior dorsal area with several (five or six) dis-
tinct flutings running across it to the margin. The dorsal and
ventral margins both arquate. L. approx. 33 mm, H. 24 mm (half
1 1 mm).”

Remarks. — Complete discussion of this subspe-
cies is found in Palmer (Liddle and Palmer, 1941 )
and Parodiz ( 1969). Palmer suggested that the sculp-
tural pattern of E. hihliana has similarities with
Diplodon santamariae Simpson, but this species
(type in NMNH) has only one small V at the
umbo, the rest being short and corrugated, and

26

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

its position is within the group of Diplodon de-
lodonlits (Lam.) with parasitic glochidia (see Paro-
diz, 1973:264). Its distribution is in southern Brazil.

The specimens observed from Basal Loyola
(CRB 18) are higher, in relation to the length if com-
pared with the dimensions given for the type. The
half diameter of these specimens remains the same,
appearing flatter, and thus differing from the typical
D. gnaranianus giiaranianus, which is always in-
flated and stronger.

Diplodon (Ecuadorea) bristowi, new species

Holotype. — British Museum LL27820 is in a ma-
trix from locality CRB 18b of Basal Loyola Eor-
mation (Lat. 2°43'20"S, Long. 78°51'45"W) in same
deposit containing Monocondylaea azoguensis and
from an altitude of 2,530 meters.

Description. — Valve (right) relatively small.
Umbo well advanced toward the anterior margin.
Shape triangular-oval; the anterior half of the valve
is perfectly round, while the posterior is decidedly
triangular, more so than in other congeneric
species, especially at the posterior lower margin;
thus, the slope of the dorsal margin falls very
obliquely. Sculpture consisting of 30 radial ribs, not
chevroned but irregular on the umbonal area, and
becoming increasingly wider toward the lower mar-
gin and especially on the posterior side. The ribs
closer to the anterior margin are almost vertical,
becoming more oblique in proportion as they ap-
proach the posterior margin, then following the di-
rection of the slope, where they are also thicker and
more separated. Without coarse lines of growth but,
instead, being concentrically striated with very fine
incisions crossing the ribs as well as the interspaces
(5 or 6 of such incisions can be counted per mm).
Length 25 mm, height 19 mm. Anterior rounded
side less than half of the triangular posterior.

Comparisons . — The peculiar triangular shape
acutely angulated on the posterior side, its regularly
rounded anterior portion, the absence of coarse
concentric tines on the lower part of the valve, and
its sculpture distinguishes this new species from D.
bihlUinus. Compared with Recent species, D. (E.)
hristowi appears as an ancestor of D. (E.) pazi
from northwestern Ecuador, which has a similar
type of radial sculpture, but differs in the other de-
scribed features.

Diplodon {Ecuadorea) liddlei (Palmer)

Diplodon liddlei Palmer, in Liddle and Palmer, 1941:404, pi. 8,
figs. 1-5; Parodiz 1969:66, pi. 8, figs. 1-4; Bristow and Hoff-
stetter, 1977:183.

Fig. 1 (top). — Diplodon (Ecuadorea) gnaranianus hihlianus (BM
LL27807). From Loyola Formation. x2.

Fig. 2 (bottom). — Diplodon (E.) hristowi, new species (BM
LL27820). From Loyola Formation. x2.

Type locality. — Center of Azogues anticline,
northwest of Azogues, Canar, Ecuador. Type in
Paleontological Research Institute, Ithaca, New
York.

Specimens observed. — Loc. CRB 7, Biblian
sandstones and conglomerates (Basal Loyola For-
mation NW of Loyola).

Original description. — "Shell elongate-quadrate, plump, an-
terior end short flaring dorsally with a narrow wing above the
hinge line; posterior dorsal area concave; sloping ventrally;
hinge with two pseudocardinals in the right valve, the lower
tooth larger, with a large socket between; correspondingly a
large pseudocardinal in the left valve; anterior adductor and re-

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

27

Fig. 3 (left). — Fossula derbyi (BM LL27806). Mangan Formation. Natural size.

Fig. 4 (right). — Monocondykieu azoguensis (BM LL27813). Basal Loyola Formation. Natural size.

tractor muscle scars preserved in paratype; umbones ornament-
ed with radial ribs of short interlocking V-shaped pattern. The
lower portion of the shell is sculptured with coarse lines of growth
only. L. 35 mm, H. 28 mm, semidiameter 6 mm.”

Remarks. — The shell is approximately of the
same size as D. {£.) biblianus but more com-
pressed and with stronger hinge. The ornamenta-
tion scarcely reaches the center of the disc. The
ornamentation scarsely reaches the center of the
disc. Although Palmer assumed that liddlei has sim-
ilarities in the hinge and other shell characters with
Diplodon mogymirim Ortmann. D. mogymirim is
a southeastern form from Sao Paulo, Brazil, with
parasitic glochidia, and is a synonym of D. e.xpan-
siis (Kuster) (see Parodiz, 1973).

Superfamily Mutelacea
Family Mycetopodidae Gray, 1840
(restricted Conrad 1853)

Subfamily Monocondylaeinae Model), 1942
Tribe Fossulini Bonetto, 1966
Genus Fossula Lea, 1870

Fossicida Marshall, 1925, an error.

Type species. — Monocondylaea fossiciilifera
d’Orbigny, 1835. Subseq. designation by Ihering,
1893. =Fossida balzani Ihering, 1893.

The main character in this genus is the double
pseudocardinal in the right valve, instead of a single

tooth as in Monocondylaea, and the teeth are stum-
py, not spatuliform, with a narrow and sinuous hinge
plate bearing tooth-like irregularities on the poste-
rior side. The umbonal cavity is not so deep as in
Monocondylaea and the anterior adductor is shal-
lower. Fossula probably represents a primitive type
in the subfamily.

Fossula cf. derbyi (Ihering)

Fig. 3

Diplodon derbyi Ihering, 1907:466, pi. 18, fig. 128; Parodiz,

1969:82, pi. 8, fig. 3.

Remarks. — The one internal cast available from
Loc. CRB 36b, Mangan Formation, Upper Miocene
of the Cuenca Basin, is not well preserved. Its de-
termination as F. derbyi is only tentative by com-
parison with the incomplete type of the species
from strata of uncertain age (probably Upper Ter-
tiary) in Rio Grande do Sul, Brazil (see Parodiz,
1969). Most of the living and fossil Fossula are
known from southern South America, and only one
species, F. venezuelensis Pilsbry and Olsson, is
known from north of the Amazon. It must be added
that the very brief description of '"Diplodon" der-
byi does not conform with the figured type. New
discoveries are needed either in Ecuador or Brazil,
to clarify the position of these fossils. The specimen
is in British Museum, LL27806.

28

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 19

Fig. 5. — Monocondylaea sp. (BM LL278l3a). Basal Loyola For-
mation. Natural size.

Tribe Monocondylaeini

Characterized by a well developed tuberculiform
or spatuliform pseudocardinal, sometimes folded in
the appearance of a double one.

Genus Monocondylaea d’Orbigny, 1835

Aplodon Spix. 1827 (not Aplodon Rafinesque, 1818, a nomen
nudum in Pulmonata').

Spixiconclui Pilsbry, 1893, substitute for Aplodon.

Type species. — M. paraguayana d’Orbigny, 1835,
Remarks. — This genus is, according to Bonetto
(1966), not a primitive one but a more specialized
one among the South American Mutelacea, and it
is widely distributed. Although the hinges of the
fossil species described are mostly unknown, the
genus can be recognized by its rather solid,
subquadrate shell, without radial sculpture; the
cloth-like periostracum characteristic of the living
species is, of course not observable. The prismatic

' The name Aplodon Rafinesque is a nomvn nudum and therefore does not preoc-
cupy the name. Spix's name which was not in use for more than 50 years is according
to the ICZN (article 23h) a nomen ohiitum. Therefore the better known name of
Monocondylaea is maintained.

area is wide and rather thick, and the valves some-
times gaping. The oldest species known is Mono-
condylaea marshalliana Pilsbry, 1935, from the
Oligocene of Colombia.

Monocondylaea azoguensis Palmer
Fig. 4

1 Monocondylaea azoguensis Palmer, in Liddle and Palmer, 1941 :

405, pi. 9, fig. 8; Parodiz, 1969:78, pi. 8, fig. 5; Bristow and

Hoffstetter, 1977:183.

Type locality. — Biblian sandstone, 3 km NW
Azogues, Ecuador (Basal Loyola Formation), Ho-
lotype (a cast) in Paleontological Research Insti-
tute, Ithaca, New York.

Material e.\amined. — One specimen, Brit. Mus.
Nat. Hist. (LL27813) coll. CRB 18b.

Original description. — “Shell quadrate, short, anterior and
ventral margins rounded; dorsal margin high; posterior margin
obliquely inclined to about the middle line, then it turns and is
broadly rounded to the ventral margin. An obscure fold occurs
from the umbonal area to the point of angulation of the mid-
posterior junction. It resembles the posterior folds of Monocon-
dylaea. A second line is suggested from the posterior-ventral
junction toward the umbo but crushing in that area indicates that
the mark is not normal. The shell [sic] is smooth with no impres-
sion of radial sculpture on the beak. The anterior portion is not
produced just below the beak as in most Monocondylaea. The
hinge is not available and the specimens are poorly preserved.
L. 29 mm; H. 25 mm; semidiameter 5 mm (cast).’’

Remarks. — The hinge has not been observed be-
cause of the condition of the fossil specimen. The
species can be distinguished from M. pacchiana by
being shorter and more quadrate; it is also about
25% smaller than pacchiana.

Monocondylaea pacchiana Palmer

'I Monocondylaea pacchiana Palmer, in Liddle and Palmer, 1941;

49, pi. 9, figs. \-2{paccliensis in plate); Parodiz, 1969:80, pi. 8,

figs. 2, 7; Bristow and Hoffstetter, 1977:183.

Type locality. — Biblian sandstone at Quebrada
Paccha, Azuay, Ecuador ( = Basal Loyola Forma-
tion), Syntypes at Paleontological Research Insti-
tute, Ithaca, New York,

Original description. — “Shell medium, umbos low, dorsal line
straight; posterior end broad, straight or slightly rounded, anterior
end sloping, short; shell smooth; hinge unknown. L. 40 mm; H.
33 mm; semidiameter 10 mm."

Remarks. — Described from fragments of the shell
adhering to the cast. The concentric growth lines
define it better than azoguensis as a Monocondy-
laea as does the absence of radial sculpture. It is
possible that pacchiana represents a more mature
form of azoguensis, because both belong to Basal

1982

BRISTOW AND PARODIZ— ECU ADORl AN TERTIARY SEDIMENTS

29

Loyola; the differences in the currently known ma-
terials may be verified when better specimens be-
come available.

Monocondylaea sp.

Fig. 5

Remarks. — The specimen consists of the lower
half of an internal cast from locality CRB 18, Basal
Loyola. The middle line of the posterior dorsal mar-
gin and inflation toward the center reveals it to be
a Monocondylaea dissimilar to the above men-
tioned species, but the condition of the cast makes
it insufficient for description. The specimen is Brit-
ish Museum, LL27813.

Subfamily Anodontitinae Modell, 1942
Gtnus Anodontites Bruguiere, 1792

PatularUi Swainson, 1840.

Glaharis Gray. 1847.

Stygcinodan Martens, 1900.

Pachyanodon Martens, 1900.

Ruganodontites Marshall, 1931 .

Type species. — Anodontites crispata (Bruguiere,
1792 (orig. design.)

This genus is distributed in all South America

from Colombia to northern Patagonia, with excep-
tion of the Pacific slope from Ecuador southwards.
Its edentulous, Anodonta-like hinge is characteris-
tic, differing from Leila which has incipient cren-
ulations or articulations and more anteriorly point-
ed valves. The shells are elongated in most species
and not — or very slightly — gaping. In very gerontic
anodontitinoids, the hinge line may present sali-
ences in one valve that correspond to sinuses in the
other, but such condition does not constitute an ar-
ticulate hinge. In most fossils, however, the hinge
has not been observed. The larger and greener
species without surface ornamentation differ some-
what from the typical group of A. crispata, and
have been separated in Glaharis', however, because
there are intermediate forms, the subgeneric divi-
sions proposed are still very unsatisfactory. In Ec-
uador only a very rare species has been found liv-
ing, A. napoensis (Lea), which Marshall (1931)
included in his Ruganodontites and whieh accord-
ing to Haas (1931) is a form of the common and
variable A. crispata. However, the fossil speeies
from Ecuador cannot be compared with those now
living in the northern part of the continent, except
in regard to the light radial sculpture preserved only
on the lower part of the shell. The specimens are
assigned to the following species:

Anodontites olssoni Palmer

Anodontites olssoni Palmer, 1941:406, pi. 9, figs. 6-7; Parodiz,

1969:86; Bristow and Hoffstetter, 1977:183.

Anodontites sp. Marshall and Bowles, 1932:6.

Type locality. — Biblian sandstone, west of Azo-
gues.

Material examined. — From loc. CRB 18, Basal
Loyola.

Original description. — “Shell large, thick, umbos large, swol-
len; hinge line straight [in the remarks the author said that the
dorsal line is displaced]; posterior end slopes obliquely from the
posterior termination of the hinge line to the rounded posterior-
ventral margin. Hinge unknown; surface smooth with conspic-
uous, radiating, undulating lines over the anterior portion of the
shell from about the middle to almost the anterior margin,
stronger ventral; irregular stages of growth. L. 65 mm; H. 46
mm; thickness (both valves) 35 mm."

Remarks. — A noticeable condition in this species
is that in casts found with both valves together one
valve slips under the other. Palmer as well as Mar-
shall and Bowles remarked that “one valve has
slipped toward the ventral margin so that the beak
is beneath the beak of the other valve.” The species
apparently forms a transition between A. crispata
and those equally rounded and swollen species of
the southern regions which do not have conspicu-
ous ornamentation. An older species, A. lacirianits
Pilsbry and Olsson from the Oligocene of Colombia,
has a shape rather similar to A. olssoni, but its sur-
face is smooth.

SupeiTamily Sphaeriacea
Family Corbiculidae (=Cyrenidae of authors)
Genus Neocorbicula Fischer, 1887

Cyanocyclas Ferussac, 1811 (in part of authors).

Type species. — Tellina limosa Maton ( =Cyclas
variegata d’Orbigny).

A typical Neotropical genus, with long siphons
(these are about 10 mm long observed in live spec-
imens), having the same crenulations on the lateral
teeth, and concentric sculpture as the Old World
Corbie ala, but differs by the presence of a pallial
sinus and in the development of the embryos.

Neocorbicula cojitamboensis (Palmer)

Figs. 6-7

Corhicida {Cyanocyclas) cojitamboensis Palmer, 1941:408. pi.

9, fig. 6.

Corhicida (C. ) pacchiana Palmer, 1941:407, pi. 9, fig. 5.
Neocorbicula cojitamboensis. Parodiz, 1969:92, pi. 10, fig. II;

Bristow and Hoffstetter, 1977: 183, 194.

30

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Figs. 6-7. — Neocorhiculu cojitamhoensis (BM 278331). Gray shale conglomerate from Mangan Formation. Natural size.

Neocorhicula pacchiana, Parodiz, 1969:91, pi. 10, fig. I; Bristow

and Hoffstetter, 1977:183.

Type locality. — Arroyo Potrero west of Cojitam-
bo, near Cuenca, in bituminous limestone of
Mid-Miocene (the deposit corresponds to Loyola
Formation). Type in Paleontological Research In-
stitute, Ithaca, New York.

Localities. — Middle Miocene: Basal Loyola,
CRB 6, 7, 9, 10, 15, 18, 26, 34, 48; Middle Loyola,
11a, 11c; Upper Loyola, 11b; Basal Azogues, 1, 4,
13. Upper Miocene: Mangan, 20, 36, 42; Ayancay,
39.

Description. — Shell trigonal in shape; size as the average in
living species. Umbo high; hinge with narrow laterals; surface
with well marked concentric ribs and lines of growth. Height 19
mm (larger specimens 23), major diameter 21 mm (wider speci-
mens 26), lesser diameter 4 mm when well preserved (most fos-
sils are very compressed laterally).

Remarks. — The original descriptions of cojitam-
boensis and pacchiana are almost identical, except
for the size indicated as larger in pacchiana. In her
comments Palmer added that the "Paccha Cyano-
cyclas differ in being less concave beneath the
beaks”; this character is not very conspicuous in
the fossils and it is highly variable in living species.

The uncertainty about the deposit in which pac-
chiana was found, in relation to the limestone with
cojitamhoensis, was the main criterion used to sep-
arate them as different species. The numerous sam-
ples collected by Bristow from the Lower to the
Upper Miocene show all the intermediate variations
as belonging to one and the same species. Those of
localities 36 and 42 in the Mangan Formation are
relatively larger but still not specifically separable.
Most of the fossils, in the same shales containing
A. browni, underwent a diatrophic process of
compression in such a way that in many shells only
the upper disc remains elevated, while the lower
part down to the ventral margin is flattened. From
matrices of Basal Loyola, the numerous casts in the
yellowish sandstone are not well preserved, and are
mostly fragments.

The name pacchiana takes precedence in Palm-
er’s description being immediately above cojitam-
boensis\ however, the type of cojitamhoensis is a
better preserved specimen, with more conspicuous
characters. For this reason, and according to Arti-
cle 24A and recommendation of the ICZN, cojitam-
boensis can hold priority.

Biological remarks. — Living Neocorbicnla have

1982

BRISTOW AND PARODIZ— ECU ADORIAN TERTIARY SEDIMENTS

31

a brownish-green periostracum and the internal sur-
face of the shell is tinted with violet. They are vi-
viparous, whereas in the genus Corhicula there is
a late larval benthic stage. Eggs of Neocorhiciila
are 5 mm in diameter; after hatching inside the in-
ternal marsupial gill the embryos are uncommonly
long (some one-quarter the size of the maternal
shell) and of an advanced stage of development,
looking as exact miniatures of the adult. The indi-
viduals become sexually mature before they reach
the average adult size, and in the marsupium of very
small specimens (only one-third of the adult size)
I found embryos 1.5 mm long. Larger embryos are
located in the upper part of the gill and the smaller
at the bottom of the marsupium. After dissecting
more than a thousand individuals from different
populations and localities (see Parodiz, 1965), no
males were found and all individuals were gravid
females. The embryos inside the same individual
show an extraordinary resemblance to each other,
but differ from individuals of other demes, thus
forming not dines but clones. Such a characteristic
points to parthenogenesis. The fossil shells of
Neocorhiciila cojitamboensis show that the same
clonic condition existed in the Miocene.

Although most of the living Neocorhiciila are
found in river systems that empty into the Atlantic,
they occurred abundantly during the Tertiary on the
western part of the continent {N. stelzneri Parodiz
in northwestern Argentina) and also in the Paleo-
cene of Patagonia (N . pehiiencliensis Doello-Jura-
do).

Family Sphaeriidae Jeffreys, 1862

(According to Article 40 of the ICZN Code, Sphaeriidae Jef-
freys has priority over Sphaeriidae Erichson, 1845 in Insects.)

Cycladidcie Rafinesque, 1820, nomen oblitum.

Pisidiidae Gray in Turton, 1857.

Genus Pisidium Pfeiffer, 1821

This genus differs from other Sphaeriidae by its
inequilateral shell with the umbos not subcentral
but towards the front.

IPisidium sp.

Remarks. — The extremely small size (about 1
mm) of the ferrugineous or blackish casts in a hard
sandstone matrix made the assignation to Pisidium
as only tentative. The matrix corresponds to the
loc. CRB 7, of Basal Loyola and contains also re-
mains of neritoid brackishwater snails, as well as
Hydrobiidae and Planorbiidae, and also Erodona.

Other similar casts are found in loc. 26i. Roberts
( 1975:262) indicated 'IPisidiiim from CRB 1 1 (Mid-
dle Loyola) in deposits with numerous fragments of
Neocorhiciila. The deposits may represent a strand
in which embayment brackishwater materials are
mixed with drifted freshwater shells.

Superfamily Myacea
Family Corbulidae

Genus £'ro</o/i« Daudin in Bose, 1808

Potamomya Sowerby, 1839.

Azam d’Orbigny, 1839.

Type species. — Erodona mactroides Daudin,
1808 ( =Mya lahiata Maton, 1809).

Suter ( 1913) used Corhiila ( Erodona ) for species
from New Zealand which do not belong to this
group. E. mactroides is found in the marginal
brackishwaters of La Plata estuary and in similar
locations in southern Brazil. ‘ ^Potamomya ' '
ochreata Hinds, usually indicated as a synonym,
or possibly a subspecies, lives in the freshwaters of
the Amazon. In Pleistocene deposits of the south-
ern areas, E. mactroides is abundantly found. No
fossils forms were recognized for the northern and
northwestern part of the continent (in our present
contents, from Ecuador). Authors dealing with this
group of corbulids, Conrad, Dali, and de Greve,
remarked on its relationship with the Pachydon-
Anisotliyris group. It is significant that Pilsbry and
Olsson ( 1935), describing Corbiila magdalensis ,
from the Miocene of Colombia, indicated Erodona
as a possible subgenus. The history and taxonomy
of Erodona mactroides has been thoroughly studied
by Carcelles (1941) with photographs of the several
individual variations.

1 Erodona iquitensis (de Greve)

Fig. 8

Anisothyris iquitensis de Greve, 1938:46, pi. 5, figs. 38-41.

Remarks. — This species has more features that
resemble Erodona than species of Anisothyris. It
differs from A. obliqiiiis Gabb (the accepted type
of Anisothyris) and other species in the group,
in that the shell is wider than high and the umbos
are less tumid and not so strongly prosogy rated;
the resilium and corresponding chondrophore are
vertical, not oblique or almost horizontal as in
Anisothyris, and lacks the stumpy, mamillaeform
cardinal-like tooth; the tooth, divided, is the same,
but proportionally smaller, as in Erodona. The

32

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Fig. 8. — (upper left two) Internal and apical view of valve of Erodona iquitensis. (upper right three) Apical, internal views, and umbonal
asymmetry in valves of Erodona nuu troides. (lower five figures) Different views of umbo and hinge in Anisothyris ohliqiius. Scales
represent I cm.

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

33

valves are very expanded posteriorly and the
anterior end rounded; both adductor scars very
shallow. The character that best distinguishes
this species from E. mactroides is the contour of
the posterior end of the valves which, in a profile
view from the top (umbonal or dorsal view), is rath-
er concave instead of convex. Lunule and escutch-
eon are well-defined and an obsolete, filiform liga-
ment fits into the marginal line at both sides of the
umbo.

I have seen numerous living and fossil specimens
of E. mactroides showing great variability in shape,
some posteriorly truncated and with strong dorsal
Carina, and others very rostrate as in iquitensis.
Living Erodona have a brown dull periostracum,
with some longitudinal blackish markings, some-
times in form of V’s on the upper middle part of the
valve; such characters, of course, cannot be de-
tected on fossils.

Anisothyris iquitensis was described from a single
valve from Iquitos, Peru, and only one specimen
similar to it exists among the material from the Ec-
uadorian Miocene (CRB 7) of Basal Loyola. It
might be that this peculiar shell represents a tran-
sition, or a subgeneric difference between Erodona -
Anisothyris, and related to the true marine Corhida
of the subgenus Carycorbida, but in any case it is
closer to Erodona. True Anisothyris do not appear
in Miocene materials, and the several species
known of that genus are from the Pliocene of Peru.

Genus Ostomya Conrad, 1874

Himella H. Adams, 1860. Not Himella Dallas, 1854. Type H.

fluviatilis Ad., 1860; Maranon River.

Ostomya Conrad, 1874. Type O. papyria Conrad; fossil of the

“Pebas” Group.

Giiianadesma Morrison, 1943. Type G. sinuosum Morrison;

Cuyuni River at Kartabo, British Guiana.

Type species. — Ostomya fluviatilis (H. Adams).

Original description of Himella. — “Shell thin, inaequilateral,
inaequivalve, left valve larger, closed, with a light constriction
extending from the beaks to the ventral margin, covered with an
epidermis; beaks tumid, obtuse. Hinge composed of an obscure
tooth in the right valve, with a corresponding cavity in the left
valve; ligament external, cartilage internal contained in a nar-
row, almost horizontal cartilage-process in each valve. Pallial
line slightly sinuated.” The (unfigured) specimen was from H.
Cumings’ collection obtained from Mr. Bates.

Original description of Ostomya. — “Thin, concentrically pli-
cated; hinge with a spoon-shaped oblique fosset in the left valve
and a small tooth near the apex. Right valve cartilage fosset very
oblique, almost parallel with the hinge line.” The author gave
no dimensions of the type, but if the figured specimen (pi. I, fig.

6) is of natural size, then it must correspond to a rather young
specimen.

Original description of Guianadesma. — “Shell small, slightly
inaequivalve, rhomboidal. translucent, nacreous porcellana-
ceous within, furnished with a variously wrinkled epidermis.
Hinge anodont. Ligament short, internal, opisthodetic, that is
posterior to the beaks, on a narrow ledge; without an ossicle.
Pallial line broad, shallow. Right valve emarginate postero-ven-
traliy.”

Comparison of the preceding descriptions shows
that the three named genera belong to the same
thing. Although Adams did not illustrate H. fluvia-
tilis, it is, according to Pilsbry ( 1944), the same as G.
sinuosum figured by him (pi. 1 1 , figs. 42-44) for O.
fluviatilis, and probably O. papyria, is also the
same. This is also supported by Morrison's ( 1943)
illustrations (pi. 8, figs. 1-6) with drawing of the gen-
eral anatomy, and with the paratypes 1 observed in
the collection of the Carnegie Museum of Natural
History.

H. Adams (1860) referred his species to Corbulidae,
“with greater affinity with Azara" and Pilsbry said
it is one of the “several stocks of Corbulidae {Er-
odona, Anisothyris, Ostomya ) that invaded fresh-
waters during the middle and later Tertiary in South
America, but only Ostomya appear to have per-
sisted as a purely fresh-water genus.” We have
seen that Erodona ( —Azara) is also found in fresh-
water stations in the Amazon. Although Morrison
in his description of Guianadesma placed it with
the Lyonsiidae, it is significant that his first impres-
sion in the preliminary report on the molluscs of
Kartabo (a manuscript is in the file of the Section
of Invertebrates at Carnegie Museum of Natural
History) was that such a bivalve belongs to Cor-
bulidae. The inclusion of Ostomya in that family,
however, might present some questions, because it
is byssiferoLis, living on rocks, has certain char-
acteristics on the hinge line, and has a peculiar peri-
ostracum. One specimen from Tapajos River at
Santarem, Para, Brazil (CM, collected by Hase-
man, 1909) has more overlapping valves, as in Er-
odona, and the seminternal ligament is at both sides
of the umbo; these were found with Congeria septa
Conrad.

Ostomya . fluviatilis (H. Adams)

Remarks. — One possible specimen is a left valve
in Matrix CRB 26b of Basal Loyola that contains
numerous Neocorhicula. The valve is more elon-
gated than in the typical O. fluviatilis and the other
fossils above mentioned. It probably is an earlier

34

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

chronological and allopatric form, but only can be
fully described if new and better materials are
found. Other fossil Ostomy a are known from La
Cira Eormation of Pachitea River, Upper Oligo-
cene-Lower Miocene of Peru — P. terminales Pils-
bry and O. pacchiteana Pilsbry.

Class Gastropoda
Order Prosobranchia
Superfamily Trochacea
Eamily ?Trochidae

Matrix CRB 7 of Basal Loyola contains a cast
that is a piece of a helicoid shell, which is the largest
among other gastropod materials. It is only a basal
fragment, apparently a trochoid. The cast is 10 mm
wide, showing regular and fine spiral sulcations (5
in number), with the umbilicus very wide, deep,
and smooth. This fragment of the base of a shell
may correspond to Calliostoma, but no more can
be indicated beyond this observation until more
larger parts or complete specimens are found. (The
cast in the matrix was filled with plastiline in order
to obtain a better view of its shape.)

The matrix is probably the result of mixed marine
and freshwater deposits, which contains Hydro-
hia — though this can be also brackishwater— in
large quantities, and also the freshwater Gy raid us,
casts of small Neritids, the problematic, bivalve-
looking specimen mentioned on page 50, and un-
identified Crustaceans.

Superfamily Neritacea
Eamily Neritidae
Subfamily Neritinae
Genus Lamarck, 1816

Erom the Neogene of Peru and Ecuador several
species of Neritina have been described. The ear-
liest one (Gabb, 1869) was indicated as from the
Miocene (Boettger, 1878, assigned to the same de-
posits an Oligocene-Miocene age), but most authors
placed these deposits in the Upper Tertiary-Mio-
cene. Those from Pebas Eormation in Peru are
probably of Pliocene age, and some species sur-
vived in the Pleistocene of Brazil which was also
indicated as “Pebas” until G. G. Simpson (1961)
made the correction in 1961. Neritina pacchiana
Palmer and the species described herein from Ec-
uador are the oldest species of the genus, and prob-
ably N. pacchiana is the precursor of N. roxoi from
Pebas. T. R. Roberts (1975) also listed the genus
Theodoxus from the same deposits, but this genus
is only known from the Pliocene and Recent of Eu-

rope and Near East. Neritina aff. antazonensis de
Greve was listed by Roberts from the Loyola Eor-
mation, but this Peruvian species, of considerable
size, was not found in the materials studied here,
which were collected by Bristow.

All the species can be differentiated, briefly, as
follows (in older, Miocene, materials the external
patterns indicated for those of Pebas, may not be
observable):

Neritina roxoi de Greve. — Close to N. ortoni
Conrad, with oblong color patches over a reticulat-
ed background.

Neritina pacchiana Palmer. — Similar to roxoi in
coloration but thicker and with more pointed apex
(upper part of the shell almost triangular).

Neritina ortoni Conrad. — (New name for TV.
pupa Gabb not Linnaeus.) Smaller with well sepa-
rated ziczac axial lines.

Neritina etheridgei Roxo. — (New name for TV.
ziczac Etheridge not Linnaeus.) Larger, globose,
with columellar area concave and numerous waved
lines.

Neritina amazonensis de Greve. — Similar to Eth-
eridgei Roxo, but with flat apex and without color
lines.

Neritina puncta Etheridge. — Similar to amazo-
nensis, but with spiral lines of dots instead of waved
ziczac lines.

All the above listed species (except TV. pac-
chiana) are from the Pliocene of Peru, of globular
shape, and different from the Neritina of the Mio-
cene of Ecuador.

Neritina loyolaensis , new species
Eig. 9, 10, 11

Holotype. — CM 46790, collected by C. R. Bris-
tow (CRB 9) at Descanso, Entrance of Tahual Can-
yon, in Basal Loyola Formation; 6 paratypes, in
British Museum GG 19827, and 19862.

Description. — Shell higher (25%) than wide, with a spire that
corresponds to about one-third of the body whorl. Well defined
suture. Shape similar to the Recent N. virginea (L.) but with
more acute apex as in N. recUvata Say. Body whorl marked, in
dorsal view, by alternate dark and clear bands of equal width;
on the upper part of the whorl these bands are narrower, be-
coming slightly wider toward the base; the bands begin obliquely
toward the right, and by the middle of the whorl form an acute
angle, turning to the left in ziczac pattern; near the base they
become almost vertical and parallel. The bands are also strongly
marked as to give the appearance of a sulcated surface. The base
is oblong, not globose as in other fossil species. Length of
the shell 9.4 mm; diameter 7.3 mm; last whorl 6.3 mm from
suture to base.

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

35

Fig. 9 (left). — Neiitiiui ioyolaensis, new specie.s holotype (CM 46790). Basal Loyola Formation, at Descanso. x7>/2.
Fig. 10 (right). — Same, paratype in matrix (CM 46790a).

Variations in markings. — In other specimens besides the
above described holotype, one is larger (12.5 mm high) with the
bands near the suture running almost parallel. Another in the
same matrix is a fragment of the shell showing more of the por-
tion toward the lip, where the bands are broken into irregular
oblong spots in a manner similar to that in N. paccliiana.

Comparisons. — N. Ioyolaensis differs from N.
pacchiana by being considerably more elongated
{paccliiana is as wide as high) and in the color pat-
tern {N. pacchiana shows only triangular patches
without any axial banding). N. pacchiana has a
lower apex. From N. ortoni, N. Ioyolaensis differs
by its more numerous bands, not waving but of reg-
ular ziczac pattern, and being not globular, larger,
and with much higher apex.

N. Ioyolaensis can be easily separated from the
other Neritina listed above from the Neogene. Col-
or pattern in species of this genus is variable; if it
would be possible to assume that it represents a
peculiar variation of TV. pacchiana, the criterion,
then, to differentiate the other species would be in-
valid; all the species are distinguished by their pe-
culiar shape, and altitude of the spire as well as the
pattern (coloration pattern in Neritina is a feature
fairly well preserved in Tertiary fossils). TV. loyo-

laensis is stratigraphically older than pacchiana,
and even more so than the other Neogene species
from eastern Peru.

TV. Ioyolaensis belongs to the subgenus Vitta
Morch 1852, of which Neritina {Vitta) virginea (L.)
is the type species.

Neritina pacchiana Palmer 1941

N. pacchiana Palmer, 1941, Bull. Amer. Paleontol., 100:396, pi.

9, figs. 3-4.

Type locality. — (This locality is not indicated
with the description but it is inferred from the pre-
liminary remarks by Palmer in previous pages)
“Near Paccha,’’ Quebrada Paccha (float in stream
beds) Areniscas (sandstones) of Azogues. This is
above Loyola Formation (Guapan?) in the Mio-
cene-Pliocene of Ecuador. Holotype in Paleontol-
ogy Research Institute, Ithaca, New York, no.
4009.

Original description. — “Shell small, spire slightly elevated,
columella callus thickened, more so anteriorly. Surface with
bands of dark patches, irregular in size, some part of the shell
compactly covered with dark spots. The coloration is described
from one specimen, the only shell retaining such features. The

36

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Fig. II. — Neritinci loyoUiensix, new species (paratype BM GG
19827/1). x3.

coloration is closest to that of certain color variations of N.
virginea L. of the West Indies and Brazil. Since the species of
Neritina offer a wide variation in surface color markings, the
pattern of one shell would not be enough for specific identifi-
cation. fhe shape of the species diffeis from the 5 fossil forms
of the Upper Amazonian fauna in being more erect, i.e. the spire
is above the line of the posterior margin of the aperture and not
at the same level with it. N . pacchiana does not have the spire
elevated so much as the Recent N. virginea L. Height 6 mm;
greatest diameter 4.5 mm."

Remarks. — Although this species is identifiable
by its description and illustration, specimens from
other localities cannot be assigned correctly to it;
most of them belong to loyolaensis, which has a
different shape as well as color markings.

Neritina sp.

Fig. 12

Remarks. — Internal casts of globular small indi-
viduals, 4 by 4 mm, were found in the matrix of
coarse conglomerate of the samples of CRB 7, Bas-
al Loyola Formation, which also contains Hydro-
biidae and Neocorhiciila. It apparently belongs to
the N. ortoni Conrad complex, approaching that
species in size, but cannot be identified as that
species, which is from the Pliocene.

Other, unidentified casts from Upper Loyola at
Malpaso (CRB 30), which Roberts (1975) indicated
as Theodoxas (an Asiatic genus), are more likely
Neritina.

Gqvxus Puperita Gray, 1857

Type species. — Nerita papa Linnaeus.

Fig. 12. — Neritina sp. Matrix that contained the specimen iso-
lated below (of the N. ortoni complex). Basal Loyola Formation.
Carnegie Museum specimen. x6.

IPuperita aff. sphaerica (Olsson and Harbison, 1953)
Fig. 13

[Neritina] sphaerica Olsson and Harbison, 1953, Monogr. Acad.

Nat. Sci. Philadelphia, 8:340. pi. 60, fig. 6a-c.

Remarks. — Living Puperita can be distinguished
from Neritina by their more roundish, ‘‘naticoid”
shell, lack of denticulation on the inner lip, and the
surface patterns. In Tertiary fossils, especially
those of small size, the only dependable character-
istic is the shape. The ''Neritina" sphaerica de-
scribed from the late Pliocene of Saint Petersburg,
Florida, still shows a pattern of coloration with ob-
long spots, recalling that of the Miocenic N. pac-
chiana ; its authors suggested that most probably it
belongs to the subgenus Puperita (by others con-
sidered as a separate genus); also they said that
the species has weak columellar denticulations, a

1982

BRISTOW AND PARODIZ— ECU ADORI AN TERTIARY SEDIMENTS

37

Fig. 13. — Matrix from (CRB 48) Basal Loyola, with steinkern of
portion of a valve of Neocorhicula cojitamboensis (CM), x 1 1 .

character of Neritina. The transition of color pat-
tern— and somewhat also in the shape — from Ner-
itina pacchiana to sphaerica and the living species,
seems to indicate that, in the Neogene the generic
separation was still in the process.

Our specimens, from localities 46 and 48, of Basal
Loyola, are steinkerns, some still in the matrix. As
usual in cases of inner casts, the sutural groove ap-
pears deeper than in those better preserved speci-

Puperitu aff. sphaerica. On the lower part the matrix also shows a

mens of N. sphaerica', the difference is noticeable
enough to presume a specific distinction, even if
they are of the same shape and size; a formal de-
scription under such a conclusion can only be made
when better material may be obtained. On this oc-
casion they are just indicated as “affinis” sphae-
rica. One of the specimens (CRB 48 in matrix) is
about 5 by 4 mm, a little larger than the type of
sphaerica', all have three whorls, the last enlarging

38

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 19

rapidly. The matrix contains also Neocorbicula co-
jitamhoensis.

Superfamily Cyclophoracea
Family Aperostomatidae
Poteria Gray, 1850

PUitystoma Morch, 1850, not Meigen, 1803.

Subgenus Pseudoaperastoma Baker, 1943

Neocyclotus — in part — Crosse and Fischer, 1886.

Type species. — Cyclostoma inca d’Orbigny. (See
remarks in Parodiz, 1969:103.)

Poteria (Pseudoaperastoma ) bibliana
(Marshall and Bowles)

Fig. 14

Pomacea bihliiina Marshall and Bowles, 1932:4, figs. 4-5.
Poteria {Pseudoaperastoma) bibliana, Parodiz, 1969:103, pi. 16,

figs. 2-3.

Type locality. — Biblian sandstone of Cafiar ( =
Basal Loyola Formation) found together with Dor-
yssa bibliana and Diplodon (E.) giiaranianits bib-
lianas.

Remarks. — The shell of this species is very sim-
ilar to the Recent Poteria inca ; it has three to four
whorls rapidly increasing in size. Spire flat and the
body whorl composes most of the shell. Periphery
almost carinated. Aperture widely expanded.
Above the angulosity of the last whorl it is decid-
edly convex, and it descends obliquely below to-
wards the base of the aperture. It is two thirds wider
than high. Height 20 mm, width 30 mm.

Because Poteria bibliana is an operculated, ter-
restrial gastropod, the specimens must have been
drifted into deposits that contain truly freshwater
species. Our observed specimens are from locality
CRB 26, of Basal Loyola.

Superfamily Viviparacea

The two families, Viviparidae and Ampullariidae,
in this superfamily are both found with fossil rep-
resentatives in South America, but are of different
origins. While the Viviparidae (Vivipartis-Liopla-
codes) were found with certain abundance in the
early Paleocene of Patagonia and Brazil they be-
came extinct afterwards, there or elsewhere on the
continent. Ampullariidae, with fossils in the middle
and late Tertiary of the northwest, reached the
southern regions in more recent times, not older
than Pleistocene, and none living in Patagonia; they
probably belong to an ancient stock probably relat-
ed to the African genera, in the same manner as the

Mutelacea. The extinct South American Vivipari-
dae, on the other hand, were more closely related
to those of North America, or from the Old World.

Family Ampullariidae
Gexm^ Pomacea Perry and March, 1810

Ampullaria Lamarck, 1810, not Lamarck, 1799 = Pila Bolten,
1798.

AmpuUarius Montfort, 1810 (after March).

Pomiis H. and H. Adams, 1856.

AmpuUarius Parodiz, 1969:109.

Type species (of both AmpuUarius and Poma-
cea).— ^"Nerita" urceus Muller.

Subgenus Limnopomus Dali, 1904

Type species. — Ampullaria columellaris Gould.

Pomacea (Limnopomus) manco Pilsbry
Fig. 15

Pomacea manco Pilsbry, 1944:145, pi. 11. figs. 31-32. Boss and
Parodiz, 1977:116.

AmpuUarius (Limnopomus) manco, Parodiz, 1969:110.

Type locality. — Quebrada de Sungarillo, in strata
with "^Hemisinus" paleus, Oligocene of the Pachi-
tea River, Peru. Type in Academy of Natural Sci-
ences, Philadelphia.

Original description. — “The shell’s internal cast is globular,
umbilicate, with a moderately elevated spire. The last whorl is
very convex. The aperture is semilunar, rather narrow, L. 15.5
mm (Type), D. 14.5 mm. Another specimen D. 16 mm."

Remarks. — One specimen, CRB 34, of Basal
Loyola, has great affinities to, and it is, tentatively,
assignable to this species. The type of P. manco,
being umbilicated, shows a probable transition with
the subgenus Effusa. This species is small in size
in comparison with some of the other fossils (P.
prourceus Boss and Parodiz) and the living species
(P. maculata Perry = gigas Spix), which are giants
among the freshwater operculated snails.

The Oligocene age of P. manco is still question-
able; very similar species as P. guadasensis (An-
derson) from the Magdalena Valley in Colombia are
not older than Pliocene, and probably Pleistocene,
instead of early Tertiary as indicated by Anderson.

Superfamily Rissoacea
Family Hydrobiidae

The family name Hydrobiidae, as used by most
authors in a broad sense, contains an array of
subfamilies that may, or may not, correctly belong
to it, but may constitute distinct families — Hydro-

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

39

biinae s.s., Lyogyriniae, Littoridininae, Amnicoli-
nae, Benedictinae, Lithoglyphioae. Although the
Hydrobiidae are related to the marine family Ris-
soidae, they differ in certain features, as in the oper-
culae, the presence or absence of accessory tenta-
cles, the radula, and embryologically, because the
Hydrobiidae do not have the free-living larvae of
the marine forms.

In spite of Opinion 457 ( 1957) of the International
Commission of Zoological Nomenclature, in sup-
port of Bityynidae Gray, 1857, as the valid name
for the family, the consensus of most authors dis-
regards such rule, using BuHmidae ( = Bithynbiidae)
as a separated family from the real “Hydrobias.”
Recent anatomical studies help to clarify the rela-
tionship of the groups within or segregated from the
family. However, in fossils in which only the shell or
casts are known, the taxonomic difficulties persist.
The assignment to particular subfamilies or genera,
inferred by shell comparison with the living taxa,
cannot be certain in all cases.

Hydrobia is a genus considered to be restricted
to the Northern Hemisphere and to be replaced in
South America by groups more commonly placed
in the Littoridininae. How'ever, there are some
species in South America, especially those of
brackishwater environments that appear to be more
closely related to Hydrobia than to Littoridina or
others of its group.

Germs Hydrobia Hartmann, 1821 (sensu lato)

Hartmann in Sturm’s Fauna Deutschland, Abd.
6(5):46.

Type. — Helix acuta Draparnaud, subsequent des-
ignation by Gray, 1847.

By the shell alone, the differentiation in certain
species between Hydrobia and Littoridina is diffi-
cult, and this is even more of a problem in fossils.
These two genera are very similar in shape, and
both are devoid of sculpture in the adults. I have
compared several of the living species, abundant in
La Plata River system, identified as Littoridina, and
found that the named L. australis differs from the
others of the estuary, not living in freshwater but
always in brackishwater, and can be assigned better
to Hydrobia (this genus taken in a broad sense).
The similarity of shape, smooth surface, and flat-
tened sides of the whorls, of the species australis
with Hydrobia ortoni of the Neogene of Peru and
Ecuador, indicates that both must belong to the
same genus.

iHydrobia ortoni (Gabb, 1869)

Mesalia ortoni Gabb, 1869, Amer. J. Conchology, 4:198, p!. 16,

fig. 3. {Mesalia Gray, 1842 is a Turritellidae.)

Isaea ortoni Gabb, 1871.

Hydrobia (Conradia) ortoni, in Wenz, 1938. De Greve, 1938:90.
Hydrobia (Isaea) confusa Boettger, 1878. De Greve, 1938:90.
Hydrobia (Conradia) confusa, in Wenz, 1938. De Greve, 1938.

Type locality. — The type locality was indicated
in the introduction of the paper by Gabb (1869);
“high bluff at Pebas, on the Ambiyacu River, two
miles above its confluence with the Maranon, near
the southern border of Ecuador.”

Original description. — “Shell small, elongated, spire high,
whorls eight or nine, sometimes very plain, or in other cases
marked by two or more revolving carinae in the young state,
which alv/ays disappear as the shell grows older; the larger
whorls are smooth, flattened on the sides and round in above
and below, to the suture, which is deeply impressed; base of
body whorl rounded. Aperture subovate, acute behind, rounded
in advance; outer lip thin and straight, inner lip acute and slightly
reflected over the umbilical region. Dimensions. — Length .35 in.,
width .13 in.”

Remarks. — De Greve placed all the Pebas species
in Hydrobia (Conradia) but these correspond bet-
ter to Dyris, except ortoni which does not belong
to this group of sculptured shells. Its assignment to
Hydrobia is acceptable, sensu lato, provided it is
understood that the Conradia-Dyris group is ex-
cluded.

Numerous specimens and fragments of this
species are contained in the gray shales of loc. CRB
7 (Loyola Formation); some of these appear to have
more impressed sutures and the whorls more con-
vex because they are internal casts. However, these
differences are the same as seen in the original fig-
ures in Gabb (1869) in comparison with those pho-
tographed by de Greve ( 1938), and indicate also that
there is no clear distinction between ortoni and
confusa. The specimens from the Loyola Forma-
tion might correspond to a different form, due to
their age and location, but apart from some of them
being slightly more elongated, there are no other
features or characteristics that can be described.
For other IHydrobia sp. from the locality CRB 7
(Basal Loyola) reported by Roberts (1975;262) see
Ly codes.

In matrix of the locality CRB 7, of Basal Loyola,
a few casts of Hydrobia ortoni were observed. This
is the same matrix that contains the marine frag-
ment of the already mentioned ICaUiostoma, and
also the fish scale with a bivalve-like appearance.
This is another indication of the more salobre con-
dition in which Hydrobia ortoni lived.

40

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 19

Genus L/m Conrad, 1871

Type species. — Liris Uupieata Conrad, 1871,
Amer. J. Conchology, 6:194.

Original description. — “Elongated, subcylindrical, with con-
vex whorls and oblique longitudinal ribs; apex entire; aperture
suboval, small, peristome continuous, labium reflexed and prom-
inent. This may be only a subgenus of the former [Isaea =
Dyris], but the shell has a more general resemblance to Pupa
and is without an umbilicus. The aperture proportionally small-
er."

Wenz (1938) placed Liris within the subfamily
Littoridininae, although the taxonomic position of
certain genera of Hydrobiidae is still provisional.
Liris contains species which, as the type species,
have only longitudinal ribs, but others, as L. tu-
berculata, have a more decussated sculpture, and
are scalariform. This genus has also certain resem-
blances with Prososthenia Neumayr from the Plio-
cene of Europe, but Prososthenia is very irregular
in shape, has more oblique sutures, and has a strong
peristome. The two genera are typical, however, as
an example of great variability in many similar
groups of Hydrobiidae.

Liris minuscula (Gabb, 1869)

Tnrhonilla niinnsciila Gabb, 1869, Amer. J. Conchology, 4; 197,

pi. 16. fig. I .

Liris niinusciiUi. de Greve, I938;92, pi. I, figs. 31-35, pi. 2, figs.

I, 9, II. and text figs. 12-18. Parodiz, 1969:120.

Potamides “n.sp.,“ Bristow and Hoffstetter, 1977:337.

Type locality. — The type is from the Pliocene of
Iquitos, Peru.

Original description. — “Shell minute, elevated, slender;
whorls six or more, rounded, suture deep; surface marked by
about fifteen rounded, longitudinal ribs, with concave inter-
spaces; aperture subcircular, outer lip simple, straight, inner lip
slightly thickened.”

The figured type (in the Academy of Natural Sciences, Phila-
delphia) is a specimen with only the last 4 whorls, broken at the
apex; it measures 3.7 mm in length and 1.9 mm. wide.

Remarks. — The axial sculpture of this species
shows great variation in the development of the
ribs, which in some specimens are very strong, and
with the greater convexity of the whorls and deep
sutures, the shells take a scalariform aspect. Such
variations were profusely illustrated by de Greve
(1938). This species is characterized also by being
entirely devoid of spiral sculpture. It is close to the
type of L. laqiieata, but that species has a still
deeper suture, with the peristome completely de-
tached from the body whorl similar to that of the
Clausiliidae, and it is produced. The specimen of

lacpieata, in fig. 21 of de Greve (1938), looks like
an intermediate with minuscula.

The examined specimens are from the lots PHI
and PH2, localized at Eat. 0°29'N, Long. 78°3'W,
and Lat. 0°28'N, respectively, from the Tumbatii
Formation (Neogene) and also from San Cayetano
Formation. They are contained in a matrix of solid
“coquina” and are represented by numerous frag-
ments of crushed shells and also some complete
specimens. The coquina looks as if it had been
formed by brackishwater, rather than freshwater
sedimentation. Because the species was originally
described from the brackishwater fauna of Iquitos,
Peru, probably the older Ecuadorean populations
lived in similar environments, or might have been
accumulated on the strand by drift. Parts of the co-
quina are covered with hardened clay that may be
younger than the matrix, although it also contains
the species. One specimen in it is of considerably
larger size (about 8 mm long) and it is entirely cal-
cified.

Genus Dyris Conrad, 1871

Isaea Conrad, 1871, not Edwards, 1830.

Conradia Wenz, 1925, not A. Adams, I860.

Type species. — Dyris gracilis Conrad, 1871:195.

Original description. — “Subulate, with many volutions; ap-
erture ovate, labium reflexed. The mouth of this shell is similar
to that in the genus Melania, but the form and sculpture of the
shell are very different from those of Melania."

Conrad’s description of this genus is very insuf-
ficient, and it can only be identified by comple-
menting it with the description of the type species
from the Pebas Formation in Peru. The principal
characteristic of the genus is the lack of axial sculp-
ture, but it is also characterized by strong spiral ribs
carinae-like, the elongated spire, very deep suture,
and relatively small aperture. Among living genera
it resembles Calipyrgula and Lyrodes in shape.
Several species have been described from Pebas,
and recognized by de Greve under Hydrobia {Con-
radia). The variability of such species is so great
and their intergradations so many, that for the pur-
poses of correct identification only two species can
be considered — ortoni (Gabb) =confusa Gabb, and
Dyris gracilis Conrad, which includes tricarinata
Boettger and D. lintea Conrad. It has already been
seen that the larger species — ortoni — with less
marked suture and without sculpture does not be-
long to the same Hydrobia (Conradia) group in the
sense of de Greve (equivalent of Dyris), but to a
more Littoridina-like Hydrobia (sensu lato).

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

41

Dyris gracilis Conrad, 1871

Dyris gracilis Conrad, 1871, Amer. J. Conchology, 4: 195, pi. 10,

fig. 8.

Type locality. — (referred to in the introduction of
the paper) 30 miles belov/ Pebas, on the south side
of the Marahon, at Pichua, just west of Cochaqui-
nas; “the shells appear to be even more abundant
than at Pebas.” 30 miles below Pebas is in the di-
rection of Iquitos, from where the abundant mate-
rial studied by de Greve came, but still far from it.

Original description. — “Very slender and elongated; whorls
8, convex, revolving lines carinated, very regular, 4 on the pen-
ultimate and 5 on the last whorl; about the sutures there is rather
wide indented space, whorls minute and obliquely striated. The
figure is a rough outline, merely indicating the natural size.”
[The “indicated” size is 8 by 1 mm.]

Remarks. — In typical D. gracilis, the flat space
between the sutures is sometimes as wide as the
rest of the whorl that is ribbed, but in others that
space narrows. Those with wide spaces correspond
to the form tricarinata Boettger, with the three ribs
in the remainder of the whorl very conspicuous.
Between this condition and the normal gracilis
there are many intergradations and the two forms
cannot be separated clearly; gradual variations con-
tinue until the interspace completely disappears,
becoming the form that was named lintea by Con-
rad; this modification is accompanied by a widening
of the base.

In comparison with the abundance of specimens
found at Pebas, those of the Upper Miocene of Ec-
uador (San Cayetano Formation — Loja) are scat-
tered impressions on the surface of the shales, col-
lected by Bristow and Kennerley (JW 424) and are
mostly of the tricarinata form. This is perhaps the
ancestral form of gracilis from which the Pliocene
polytypic populations evolved.

Professor C. Carrion collected for the British
Museum, at the same locality of Loja, two small
slabs of very pale buff silty limestone with impres-
sions of the same species mixed with remains of
Characea (British Museum G 43325-6).

Genus Ljro*s Doering, 1885

Potamopyrgus (in part of authors for Neotropical species).
Pyrgophoms Ancey, 1888.

Type species. — Lyrodes guaranitica Doering,
1885, Bol. Acad. Nac. Cien. Cordoba, 7:461-462.

Type locality. — “Lagunas riberenyas” (small
lakes on side of the river) near Barrancas River

(tributary of Guayquiraro River, on the border of
the Corrientes-Entre Rios provinces, Argentina).
The type, from Doering’s collection at the Academy
in Cordoba, was lost; a neotype was selected by
Parodiz (1960:25) from Riachuelo, near Corrientes
city (CM 59-108).

Lyrodes sp.

Remarks. — From locality CRB 7 (Basal Loyola),
internal casts that are different from the other Hy-
drobiidae may be assignable to Lyrodes (the same
were indicated as Hydrobial by Bristow, 1973:23).
They measure about 3.5 mm in length and have
rounded whorls with deep sutures. Separated from
the matrix, one incomplete specimen has three
whorls which when entire, must have been more
than 4 mm; it shows very convex whorls of rapid
growth which is characteristic of the genus. For
their size and shape, the specimens can be com-
pared with Lyrodes lacirianus (Pilsbry and Olsson,
1935:9, pi. 5, fig. 6), but this is a much older species
from the Upper Oligocene of Colombia in La Cira
Formation (see also Parodiz, 1969:117). Although
belonging to the same group, the Ecuadorean sam-
ple must correspond to a different species, which
may be described when more and complete speci-
mens can be obtained. The matrix is the same one
that contains the “problematic fossil” (probably a
fish scale) that looks like a minute pectinid valve.

Genus Toxosoma Conrad

Pseudolacuna Boettger, 1878:496.

Type species. — Toxosoma eborea Conrad,
1874:31, pi. 1, fig. 7.

Original description. — “Conical, polished, the aperture pro-
jecting, subovate, direct, peristome continuous; coiumeila con-
cave with a plait or tooth in the middle, not oblique, base round-
ed, subumbilicated.”

Toxosoma eboreum Conrad (emend. Pilsbry, 1944)
Fig. 16

T. eborea Conrad, 1874; Tryon, 1883:270; Pilsbry, 1944:151, text
fig. 3a-b (name changed to eboreum because of the ending of
the genus name, soma, in this regard see ICZN Article 30a;
Parodiz, 1969:121.

Pseudolacima macroptera Boettger, 1878:496, pi. 13, figs. la-!5
(Pebas); Oliveiro Roxo, 1943:638, pi. 29, fig. 25; de Greve, 1938:
74, pi. 5, figs. 17-18, 24-29, 31-36.

Hydrobia (Paliidestrina) diibia Etheridge, 1879:86, pi. 7, fig. 11
(the author said that he named it nith much doubt): Oliveiro
Roxo, 1943:640-641, pi. 29, fig. 24 (as Isaea).

Type locality .—Pebas (Pliocene), Peru.

42

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 19

Fig. 14. — Poteria (Pseudouperastoma) hihliaiia (BM GG 19821). Basal Loyola Formation. x2.
Fig. 15. — Pomacea (Limnopomus) manco (BM GG 19819). Basal Loyola. x4.

Fig. 16. — Toxosoma ehoreum (BM GG 19816). Mangan Formation. x5.

Remarks. — Conrad’s diagnosis is very brief, in-
dicating that it is small (4.7 mm) with 5 whorls
rounded, aperture angular above, last whorl ex-
panded, and with a minute columellar tooth. The
figure, although very small, is recognizable as being
the same as P. macroptera, which Pilsbry verified
when he redescribed the type in the ANSP (161 152).
The dimensions are approximately the same, and
de Greve specimens are even smaller.

The fact that this species was frequently referred
to as macroptera does not make T. eborea a “no-
men oblitum," as the rule for that was not applied
before 1960 and, more important, it was not over-
looked; de Greve listed it (1938:5) among the Con-
rad species of 1874, but failed to recognize its iden-
tity. De Greve listed also Hydrobia diibia Etheridge
(1938:8, 10) but did not include it in the synonymy
or the systematic discussion of the species. The
figs. 24, 29, 33-34 on pi. 5 show specimens of the
variations of macroptera which are identical with
the form described by Etheridge.

T. eboreum is extremely variable in the devel-
opment of the last whorl and peristome; the colu-
mellar tooth sometimes is absent or so internally
placed (in fact it is a fold that continues on the in-
ternal columella) that it can be seen only in an
oblique view of the shell.

Our specimen (British Museum GG 19816) is
from locality CRB 26b, in matrix from Mangan For-
mation. It gives a front (apertural) view of the shell

which has a large body whorl; the penultimate
whorl has an adherence that gives the false impres-
sion of being carinated but the surface is smooth.
The last whorl occupies two-thirds of the length of
the shell (length 4.5 mm, width 2.5 mm). Length of
the aperture is approximately half of the shell di-
ameter. Width of the aperture is 1 mm. The speci-
men agrees with those illustrated by de Greve
(1938) in figs. 29 and 34. As for the spire, all the
specimens figured as P. macroptera are the same.

This species has similarities with ’’’'Lacuna"
(Ebora) crassilabris Conrad (which is probably a
To.xosoma) but it is much narrower, with a shorter
spire and angulated, instead of rounded, aperture.

Genus Potamolithoides Marshall and Bowles, 1932

Type. — Potamolithoides biblianus Marshall and
Bowles, 1932:4, figs. 1-3.

Original description . — “Shell small, resembling Potamolithus
but with spire depressed and base widely umbilicated or deeply
excavated."

Potamolithoides biblianus Marshall and Bowles, 1932
Fig. 17

Potamolithoides biblianus Marshall and Bowles, 1932:4, figs. 1-
3. Parodiz, 1969:116, pi. 17, figs. 1-2.

Holotype. — National Museum of Natural Histo-
ry, 312840.

Remarks. — The characters of the genus are better

1982

BRISTOW AND PARODIZ— ECU ADORIAN TERTIARY SEDIMENTS

43

Fig. 17. — PotamoHthoides biblianus (BM GG 19818). Basal Loy-
ola. x3.

indicated in the description of the type and only
species. It is of small size (3.5-5 mm by 5-7 mm),
apical whorls slightly sunken, whorls flatish above
and subanguiar at periphery, base flat, and aperture
oblique with thin lip.

The figures of the type do not show the flatness
on the upper part of the whorls, as the description
indicates, but the outline is rather regularly convex,
without angulosity. The whorls increase with reg-
ularity, and the last one is expanded on the outer
part near the peristome, a character for which the
authors found resemblance with Potamolithiis. In
most Potamolithus, however, the apex is rather
prominent; in very few forms, as in Potamolithus
lapidum paysanduanus (Pilsbry), the apex is very
low, but not sunken. By the characteristics of the
last whorl, PotamoHthoides is also comparable to
Eiibora Kadolsky, 1980 (new name for Ebora Con-
rad) but in this genus the apex is conspicuously el-
evated as in Potamolithus.^

Our specimen, from CRB 26b of Basal Loyola
(British Museum GG 19818), is embedded in the
matrix, showing only the last whorl, very similar to
the fig. 3 of the type of PotamoHthoides biblianus.
It is assignable to that species (Eig. 17).

^ When this manuscript was completed, an article by Dietrich Kadolsky appeared
in The Veliger, 22:364, April 1980, in which the taxonomic position of Euhora and
Toxosoma is discussed. It corroborates our previous indication of PseudoUuunu ma-
cropieru as synonym of Toxosoma eboreum and gives a revision of this species from
the Pliocene of Peru, with complete references. It also shows the great similarities of
Euhora {e\-Ehora) with some of the many species of PoiamoHthus now living in the
Uruguay River.

Supeifamily Cerithiacea

The living freshwater Cerithiacea, which former-
ly were included in the whole embracing “family"
group of the Melanians, offer some taxonomic dif-
ficulties for the separation of families, subfamilies,
and even genera. According to Morrison ( 1954) they
were derived from three different marine family
stocks. The two distinguishable groups in South
America belong to Thiaridae in which many species
are considered as being parthenogenetic, and Pleu-
roceridae, which are dioceous, and both differ also
in the mechanisms of oviposition. Because such
conditions are not possible to verify in fossils, the
placement of the genera within the families is in-
ferred only by characteristics and relative similarity
of the shells.

Family Pleuroceridae
Subfamily Potadominae ( = Melanatriinae)
Genus Doryssa H. and H. Adams, 1854

Sheppardiconcha Marshall and Bowles, 1932.

Type species. — Melania atra (Bruguiere, 1792).

Original description. — “Shell subulate, turreted, spire decol-
lated, whorls longitudinally plicate and decussated with trans-
verse ridges; aperture subcanaliculated in front; outer lip in-
crassated.”

Remarks. — This taxon was originally described
as a subgenus of Vibex [!] Oken, 1815, but its author
must have had in mind Vibex Gray, 1847, a syn-
onym of Pachymelania of West Africa, a Thiaridae.

When the genus Sheppardiconcha was described
with the species S. bibliana Marshall and Bowles
as type, the authors considered it to be of an age
no later than Pliocene or even earlier. They gave
only a brief diagnosis for Sheppardiconcha: "tur-
ritelliform spire, roundish aperture which is appar-
ently somewhat produced at the columellar side."
It was also suggested by the authors that ''Hemi-
sinus tuberculifera Conrad was an allied species
of Sheppardiconcha, but Conrad’s species is a
Thiaridae of the genus Aylacostoma (Hemisinus),
which is devoid of axial sculpture, which in Shep-
pardiconcha-Dory ssa appears as conspicuous sig-
moid riblets. The characters of both Dory ssa and
Sheppardiconcha are identical, and because of that
Morrison ( 1954) placed Sheppardiconcha as a syn-
onym of Dory ssa .

Older Dory ssa, D. maymarensis (Bonarelli),
were abundant in the early Tertiary of South Amer-

44

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Fig. 18. — Doryssa hihlianci (CM 46791). Mangan Formation.
x6.

ica in the Puca Formation. This species is also
known from the Tejon Formation, Eocene of Cali-
fornia. The living species of Doryssa replace
Pachychilus (most common Central American
Pleuroceridae) as you move southwards in South
America.

Doryssa bibliana (Marshall and Bowles)

Fig. 18

SheppiinUconcha hihliana Marshall and Bowles, 1932:3, pi. I,

fig. 6.

Hemisinus (S.) bihlianus. Palmer, 1941:40, pi. 6, figs. 1-2.
Doryssa bibliana, Parodiz, 1969:134, pi. 15, fig. 12.

Type locality. — Biblian, Prov. Cahar, Ecuador,
Lower Miocene. The type is in the National Mu-
seum of Natural History.

Original description. — “Shell turritelliform imperforate, whorls
numerous [the type specimen with broken apex has 6V^ whorls;
the entire shell must have had 10-11 whorlsj, slowly increasing
in size, somewhat flattened, longitudinal sculpture consisting of
sinuous, slightly protractive incremental striae. Spiral sculp-
ture of five strong, obscure nodulous lirae on the surface of the
whorls of the spire and one sunk in the suture. Base worn but
showing the remains of several lirae. Aperture roundish, colu-
mella curving forward.” Length (upper whorls missing) 20 mm.
Diameter 8.5 mm.

Remarks. — Palmer found this species to be abun-
dant in several localities between Azogues and Pac-
cha. Our observed specimens, especially those
from Loyola Formation, show a series of variations
from the type. In some the whorls are more convex,
the suture conspicuously canaliculated and fol-
lowed by a narrow shoulder, and the sigmoid lines
or riblets may be stronger so as to form a charac-
teristic axial-oblique and thick sculpture broken at
the top into tubercles. Other specimens show vari-
ations (see list below) among the series and also
from the type. Those figured previously (Parodiz,
1969: pi. 16) show also the variations in figs. 6 and
8, but that of fig. 12 (which corresponds to one fig-
ured by Palmer) is an extreme one, with the axial
riblets obsolete, and it is probably a 'Temporal di-
ne.”

The spiral sculpture begins about the middle of
the whorls, merging with the axial ridges and, on
the lower part becomes more regular, so that the
last three, above the suture, are stronger and par-
allel. The columellar lip at the aperture has in some
specimens a distinct callus and is concave at the
base. It seems that the sigmoid incremental ridges
became proportionally stronger from the popula-
tions of those of the Lower to the Upper Miocene.
It would not be of any taxonomic value to designate
by name the younger populations as if they were
chronological forms (what Simpson calls "succes-
sional” subspecies) because such variations are re-
current in all the Miocene strata and the extremes
are linked by numerous intermediate individuals.

A related species, derived probably from the

1982

BRISTOW AND PARODiZ— ECUADORIAN TERTIARY SEDIMENTS

45

same stock of the Pliocene at Iquitos, Peru,
ithium" coronatum Etheridge (v/hich is Doryssa),
is also highly variable, as was illustrated by de
Greve (1938), but, in general, the axial incremental
sculpture is reduced to the uppermost part of the
whorls, below suture, forming some elongated tu-
berculae; this is a tendency which becomes more
conspicuous in the living species.

The characteristics of the material collected by
Dr. Bristow at Loyola Formation (Lat. 78°52'W,
Long. 2°34'S) in CM 46791, are as follows: axial in-
cremental sculpture strong, 5 specimens; axial in-
cremental sculpture conspicuous, 1; axial incre-
mental sculpture less conspicuous, 10; axial
incremental sculpture conspicuous at base, 1; axial
incremental sculpture conspicuous on last whorl, 1;
axial incremental sculpture conspicuous at top, 3;
axial incremental sculpture not conspicuous, 4; spir-
al sculpture strong, 15; spiral sculpture conspicuous,
!; spiral sculpture not conspicuous (eroded), 8.

One full specimen with apex (37 by 15 mm, ap-
erture 14 by 9 mm) which is a little more than half
of the whorl, has the columellar lip slightly curved
to right, forming angular base. In one specimen, the
last whorls only are wider being 17 mm.

In the British Museum collection (G 55394-6)
were observed 36 specimens from (probably) Shep-
pard’s original lot, plus many others collected by
Dr. Bristow from Mangan CRB 36a, and from Basal
Loyola, 2, 5, 8, 14, 17, 18a-b, 28.

Doiyssa corrosensis (Pilsbry and Olsson, 1935:12,
pi. 2, fig. 89), originally described as Hemisinus
(see Parodiz, 1969: 136), was indicated as being from
Mangan Formation by Roberts (1975:261) and Bris-
tow and Hoffstetter (1977:194), according to a pre-
liminary report by Parodiz. Such identification must
be corrected here, because those Miocene speci-
mens belong to the common D. bibliana. D. cor-
rosensis is an ?Eocene species from Los Corros
Formation at Rio Sucio, Colombia (type in Acade-
my of Natural Sciences, Philadelphia). Recently
Boss and Parodiz (1977:118, figs. 10-11) reported
D. corrosensis from Peru in the vicinity of Yarina
on the Huallaga River, and its collector Dr. Bryan
Patterson indicated the strata as early Tertiary
(mid- or late Eocene), thus this species had a wide
range from Colombia to Peru in the early Tertiary
but not in the Miocene.

Genus Paleoanculosa Parodiz

Type species. — P . patagonica Parodiz, 1969:124,
pi. 14, figs. 3-4.

Fig. 19. — Paleoanculosa kennerleyi, new species (holotype, CM
46792). Near (east) Biblian, Mangan Formation. x6.

Shell with large body whorl, spire short and con-
ical. Sutures well defined. Body whorl depressed
about the middle. Upper part of the whorls subgra-
date or shouldered. Columella thickened by a cal-
lus. The genera! appearance recalls that of the living
""Anculosa" i=Leptoxis) from the southern United
States. The species previously known of this fossil
group were described as Melania'" or ^"Paludina"
from the early Tertiary of southern South America
(Argentina and Chile). The largest one, P. macro-
chilinoides (Doello-Jurado), has a certain resem-
blance to Hannatoma, but this genus is a Thiaridae.
The most abundant species, P. buUia (Ihering),
from the Paleocene of Patagonia is slender and less
shouldered, angulated at the base and perhaps rep-
resents a subdivision of the group.

The following new species is the only one known
from northern South America and the Miocene.

46

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Figs. 20-21. — Conglomerates of Mangan Formation, with Paleoanciilosn kennerleyi, new species (paratypes, CM 46793). Natural size.

Paleoanculosa kennerleyi, new species
Eigs. 19-21

Holotype.— CM 46792, CRB 20, Lat. 78°53'45"W,
Long. 2°45'I6"S (a short distance east of Biblian
and 6 km WSW Azogues) in strata of Mangan Eor-
mation. Upper Miocene-Eocene.

Paratypes. — 40 in CM and 60 in British Museum
from the type locality. Other paratypes from 20 km
north, Lat. 78°52'30"W, Long. 2°34'09"S, CM
46793. Apart from those individually separated
specimens, many others are contained in matrix
conglomerates.

Description. — Shell ovoid-conic, that is, the spire above the
last subsutural cord is decidedly pyramidal, while the last whorl
below that cord is very convex and ovate. Five and a half to six
whorls, the first one and a half regularly convex constitutes the
nucleus which is fiat; the following whorls increase rapidly and
regularly as to give a wide base to the spire which is one-third
of the total length. The sides of the whorls on the spire are flat,
although they might give the impression of being somewhat con-
vex because of a wide spiral ridge running across the lower half
of each whorl down to the penultimate one. On the last body
whorl, which is very wide and rounded, the ridge is at the top.

below the suture, ending at the lip, and leaves a marked cana-
liculation between it and the remainder of the whorl. The body
whorl is two-thirds of the total length, and as wide as high. The
suture is well impressed, giving to the spire a gradate aspect. All
the whorls, except at the apex, are noticeably marked by sigmoid
incremental lines; there are no other sculptural features. The
aperture occupies half of the length of the shell and begins at a
short distance from the ridge of the last whorl; peristomatic area
broken in most specimens but there is indication of the outer lip
being thin. The columellar wall is widely round and ends abrupt-
ly into a very short canal which is very slightly deflected to the
right. On the posterior side of the body whorl the incremental
sculpture is stronger, forming very elongated S's. The posterior
end of the canal is roundish and a little protractive. The umbi-
licus appears closed or very narrowly rimated.

Dimensions of holotype . — Length, 22.5 mm; diameter, 12 mm;
spire, 10.5 mm; length of last whorl, 12 mm; penultimate whorl,
2.9 mm high. The spire forms an angle of 65° in relation to the
axis.

Comparisons. — Compared with other species of
Paleoanculosa, P. kennerleyi has a more conical
spire which is wider at the base, the body whorl
more globose, and the incremental lines stronger,
with the depression on the upper part of the body
whorl deeper than in the older P. macrochilinoides.

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

47

Remarks. — The species is named in memory of
Brian Kennerley, leader of the British Geological
project in Ecuador, who died tragically in a car ac-
cident during a sojourn in Colombia. Specimens of
this species were previously reported by Pilsbry
and Olsson ( 1935, as Aykicostoma siginachiliis) and
Bristow and Hoffstetter (1977:195, as Aykicostoma
(H .) sigmachilus).

Some specimens are a little more elongated and
narrower than the type (22.5 mm by 9.5 mm) and
others are shorter and wider (13 mm). These pro-
portions are of specimens that still preserve parts
of the shell. Internal casts, of course, show whorls
apparently more convex and more gradated. In a
few specimens there are indications of three basal
cords, but this is a juvenile character which is ab-
sent in fully-developed specimens. Some specimens
show deformities caused by diastrophic pressure as
is seen in many other species in the same strata.

Family Thiaridae

The separation of the living Thiaridae from the
Pleuroceridae is based on the reproductive system,
the first being considered to be parthenogenetic,
although there are not enough studies to prove that
all forms in the family are so. The shells may show
characters convergent with the Pleuroceridae, in-
creasing the difficulty to distinguish the families
when the study is on fossils. The systems used by
Pilsbry ( 1944), Morrison ( 1954) and Wenz ( 1938) are
followed here.

Subfamily Aylacostominae
Hemisinueae Wenz, 1938.

Gtnxxs Aylacostoma Spix, 1827

Type species. — By subsequent designation Mor-
rison (1954) — Aykicostoma glahriim Spix, =Melan-
ia scalare Wagner, 1827.

Most of the shells in Aylacostoma (sensu stricto)
are turritelliform, with flat-sided whorls and spirally
ribbed, and angular at the base. The genus ranges
from Central America to South America in Brazil
and northern Argentina; they are divided in several
subgenera but the Ecuadorian Miocene species are
Aylacostoma (s.l.).

Aylacostoma sulcatus Conrad, 1871

Hemisinus sulcatus Conrad, 1871, Ainer. J. Conchology, 6:194,

pi. 10, fig. 2.

Seiuisinus sulcatus. de Greve, 1938: pi. 4, figs. 17-19, 21-24.

Type locality . — . . near Pebas (or 30 miles below
Pebas) in strata that cannot be later than Tertiary.”

Materials e.xamined. — Locality CRB 42, Mangan
Formation, middle Miocene. Abundant; found in
shales with Neocorbiciila cojitamhoensis.

Original description. — “Subulately turbinated, solid, pol-
ished, whorls slightly convex, revolving grooves or impressed
lines not closely arranged, about six on the penultimate whorl,
and two minute lines, one towards each boundary; last whorl
with about 23 lines, reach the base. An elegant species closely
allied to H. tenellus Reeve [from Pernambuco] but it has a longer
last whorl and narrower aperture.” Conrad gave no dimensions,
but the figure, which is of natural size, measures 26 mm long,
10 mm wide, and the aperture is 1 1 by 5 mm; the aperture is
noticeably angulated at both ends.

Remarks. — An extremely variable species of
which one variation is represented by Conrad’s fig-
ured type, for which he indicated a polished sur-
face, “elegant” appearance, and not too close and
not too impressed revolving lines. It is very clear
that Conrad did indicate real cords or ribs; 17 spec-
imens in our lot correspond to this smoothish form.
The largest is 25 mm long and 13 mm wide. The
brevity of the description and the not very satisfac-
tory figure, made de Greve comment ( 1938: 100),
”T. A. Conrad’s, pi. 10, f. 2, makes its distinction
not clear enough.”

A supplementary observation is here pertinent on
details not included in the original description. The
spiral lines are extremely fine and visible only under
good magnification (which distinguishes it from oth-
er named sulcatus with strong ribbed spirals). It
has numerous, regular but very fine, axial incre-
mental lines which run from the sutures becoming
undulate in the last whorl. Below the suture there
is one, sometimes two, better-marked spiral lines
crossing the axial lines of growth which at that point
are more conspicuous; this gives to the sutural zone
a marginated appearance. The columella is curved,
ending into a very short canal that does not extend
beyond the basal lip. Also, the last whorl appears
more elongated. It is possible that these specimens
represent an allochronic and allopatric form, but
because there is much variation in shape, as well as
in sculpture, any nomenclatorial distinction would
be of doubtful value.

Another form in the same lot is represented by
one specimen that differs more from Conrad’s de-
scription. It is more turreted with stronger spiral
lines so as to form, on the upper part of the whorls,
a cord-like sculpture, crossed by the weaker growth
lines. In the body whorl there are two cords on the

48

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 19

Fig. 22. — Two specimens of Aylacostoma hrowni from Middle Mangan Formation. x5. The species is also found at the Loyola
Formation, where most of the specimens are diastrophically deformed.

upper part below suture, a smoother middle area
with fine spirals and then, from the periphery,
which is somewhat angulated, to the base, about 10
spiral ridges crossed by extremely fine incremental
lines. This individual is similar to those figured by
de Greve (pi. 4, figs. 17, 19, 21-22, 24-25) as ^^Semi-
sinus" siilcatiis from Iquitos, Peru, in the Pliocene.

Other variations illustrated by de Greve have
strong axial riblets and revolving lines over the
whole surface. However, those of pi. 4, fig. 18 and
text fig. 23, are not sulcatus, and actually belong to
the following species, A. dickersoni (Palmer).

Aylacostoma dickersoni (Palmer)

Semisinus sukutus (Conrad), de Greve, 1938, pi. 4, fig. 18 and
text fig. 13 (non sulcatus Conrad).

Hemisinus peyeri dickersoni Palmer, in Liddleand Palmer, 1941;
398, pi. 6, figs. LS-18.

Aylacostoma (Longiverenii) peyeri dickersoni . Parodiz, 1969:149.

Type locality. — Arroyo Potrero, west of Cojitam-
bo, Cuenca vicinity, in bituminous limestone (this
corresponds to what is now recognized as Loyola
Formation, and the same loc. of Neocorhicula co-
jitamhoensis).

Remarks. — Described as a new variety of peyeri,
the author said (Palmer, in Liddle and Palmer 1941:
399) that ‘This ‘species’ is related to Hemisinus pey-
eri (de Greve)”; "Semisinus" peyeri — as originally
described — is from the Pliocene of Iquitos, Peru, and
it has less nodulous ribs with narrower interspaces.
Aylacostoma paleus (Pilsbry, 1950) is similar in
shape but differs by the absence of tubercles on the
ribs, and it is from the Oligocene of Peru. According
to Pilsbry A. peyeri and A. dickersoni are different
species, as well as of different age. For convenience
these species were assigned to the group Longi-
vereiur, however, subgeneric divisions in Aylaco-
stoma still need clarification.

The material of A. dickersoni collected by Bris-
tow is from the locality CRB 1, Basal Azogues.

Aylacostoma browni (Etheridge, 1879)

Fig. 22

IMelanopsis hrowni Etheridge, Quart. J. Geol. Soc., London,
35:87, fig, 5. Oliveira Roxo and Leonardos, Geol. Brasil,
1943:631, pi. 29, fig. 19.

Semisinus sulcatus, de Greve {in part), 1938: fig. 18, pi. 4, and
text fig. 19.

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

49

Aylacostoma sulcatus. Parodiz (in part], 1969:141, pi. 14, figs.

6-7.

Paleoancidosa cf. builia (Ihering), Bristov/, 1973:30.
Paleoanculosa sp. Bristow and Hoffstetter, 1977: i94.

Type locality. — Canama, Brazil. All the species
collected by C. B. Brown and described by Ether-
idge are from Canama. In the title of the paper Eth-
eridge indicated “Tertiary deposits of Solimoes and
Javary rivers”; Canama is not on the banks of these
rivers, but on the Curuga River v/hich is 85 mi east
of the Javary and the Peruvian border. Oliveira Roxo
and Leonardos ( 1 943) reported the species from Tres
Unidos, on the Javary. The Curuga River is a tribu-
tary of the Javary, but the confluence is farther north,
at Caixas.

Original description. — “Shell turreted, elongated; whorls five,
sides vertical; sulcus or suture at junction of the whorls de-
pressed, the sutural edge elevated; upper whorls double cari-
nated; body whorl concentrically banded by nearly equidistant
lines, slightly rugose at the base, here and there possessing a
varice; anterior canal slightly notched; outer lip toothed; colu-
mellar lip slightly reflected and thick. This shell resembles Me-
lania, and but for a siphonal notch could be referred to that genus
or the subgenus Plotia."

The author also remarked that he had several specimens but
nothing to compare with them, and that the species was, appar-
ently, like the other remains from the Amazon Valley, estuarine
in habit. The localities on the Javary and Solimoes are probably
Pliocene.

Remarks. — Twenty specimens, CM 46804, from
locality CRB 42 (the same that contains the many
A. sulcatus and Neocorbicuia) plus numerous frag-
ments in matrix or loose, made it possible to iden-
tify this species and correct the indication in my
paper of 1969 (p. 141) where it was synonymized
under A. sulcatus . The specimens referred to on
that occasion were received from the Geological
Service of Brazil from the locality Tres Unidos, on
the Javary River, and were labelled ''"Hydrobia
(Conradia) Untea Conrad.” Etheridge’s original il-
lustration of the species, although recognizable,
was not very accurate, and I figured the specimens
from Tres Unidos (pi. 14, figs. 6-7) as sulcatus, but
actually they are typical browni.

The spiral sculpture is very regular all over the
shell and the spire is very pointed and appears as
“inserted” on top of the body whorl, forming there
a well-defined flat shoulder. The crenulations indi-
cated for the outer lip correspond to the ending of
the spiral cords. There are no visible axial lines.
Younger individuals are rather slender, and the
spire slightly scalated; adults are always broader.

The Miocene specimens collected by Bristow are

Fig. 23. — Conglomerate from Loyola Formation with A) Dor-

ys.sa: B) Potamolitlioides', C) Aylacostoma", D) Gyrauliis{7)', E)
HydrohiaV?). Natural size.

generally larger than those known from the Pliocene
of Brazil, and the spiral sculpture may not be as
regular as in typical A. browni. As in other species,
they may represent an older ancestral race, or al-
lochronic subspecies, but until more and better ma-
terials for comparison are found, they remain indi-
cated, in a broad sense, as A. browni. Most of the
specimens have been deformed by diastrophic pres-
sure.

This species was also found by Bristow at Loc.
CRB 36, which corresponds to mid-Mangan, and in
sediments of Lower Miocene of Basal Loyola, CRB
18 (Fig. 23). Apparently, it had a wide range during
all the Miocene in Ecuador.

The average size of specimens of A. browni,
which had not been diastrophically deformed, is 20-
22 in length and 11 mm wide.

Aylacostoma sp.

From the locality CRB 42 (Mangan), there is a
very large and elongated specimen that is very dose
and possibly just a variation of A. browni. It mea-
sures 31 mm in length with the apex broken (entire

50

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Fig. 24. — Gyniulus sp. Basal Azogues Formation. Carnegie Museum specimens. x20.

it must have been more than 35) and the width of
the last whorl is 15 mm. The spiral sculpture on the
last whorl is very regular, and 17 spiral cords can
be counted from the sutural shoulder to the base.
Apart from the size and the appearance of being
more shouldered, all other characteristics are the
same as in A. browni. The reddish Mangan shales
that contain this form are also abundant in Neo-
corhicula.

Order Pulmonata
Superfamily Lymnaeacea
Eamily Planorbiidae
Genus CyroM/i/s Charpentier, 1837
Pig. 24

Prom locality CRB 7 (Basal Loyola), there are
casts of a very small planorbid (about 2.5 mm in
diameter) that probably corresponds to Gyraulus,
as Roberts indicated ( 1975, Gyraulis), but it can be
compared also with Armigenis, therefore the assig-
nation here is only tentative. There are Gyraulus
known from the Cretaceous-Paleocene of Europe
and North Africa, and from the Miocene of Ger-
many, G. trochifonnis (Stahl). Planorbids from the
Tertiary of South America, however, are known to
belong to the genus Taphius.

The shell is very flat, angulated at periphery, with
approximately four whorls, the last one (just broken

behind peristome) about three times as wide as the
penultimate; the growth from the nuclear whorls is
very rapid. The suture is well impressed. The ap-
erture is oval, elongated, and narrowed at the end.

Superfamily Succineacea
Eamily Succineaidae
Genus Sued nea Draparnaud, 1801

Several fragments of an unidentifiable Succinea
in limestone from locality KB 1, which contains
also small Sphaeriids bivalves. Roberts has indicat-
ed Succinea for CRB 7 (Loyola).

Incertae Sedis

In a piece of coarse conglomerate of Basal Loy-
ola (CRB 7) was observed a specimen of organic
remains, like a very tenue roundish valve com-
pressed into the matrix. Under the microscope it
looks like a left valve of some minute pectin-
id, like Propeamusium, on account of having 12
equally-spaced radial ribs and an auriculated ex-
pansion at the top. A more careful observation re-
veals the unlikeliness of such an assumption; it
probably represents an unidentified fish scale. It
measures 3.5 mm at its widest diameter. It was
probably deposited in shallow water into which oth-
er, freshwater organisms, like Hydrobiidae, and
others have been drifted.

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

51

LITERATURE CITED

Adams, H., and A. Adams. 1858. The genera of Recent Mol-
lusca. London, vol. 2, 497 pp.

Anthony, H. E. 1922. A new fossil rodent from Ecuador.
Amer. Mus. Novit., .^5; 1-4.

Baker, H. B. 1943. Two new subgeneric names in Poieriu.
Nautilus, 68: 109-1 12.

Berry, E. W. 1918. Age of certain plant-bearing beds and as-
sociated marine formations in South America. Bull. Geol.
Soc. Amer., 29:637-648.

. 1929. The fossil flora of the Loja Basin in southern

Ecuador. Johns Hopkins Studies. Geol., 10:79-136.

. 1934. Pliocene in the Cuenca Basin of Ecuador. J.

Washington Acad. Sci. , 24: 184-186.

. 1945. Eossil floras from southern Ecuador. Johns Hop-
kins Studies, Geo!., 14:95-150.

Boettger, O. 1878. Die Tertiarfauna von Pebas aus Oberen
Maranon. Jahrb. Geol. Reich., Vienna, 28:485-504.

Bonetto, a. a. 1965. Las especies del genero Diplodon en el
sistema del Rio de La Plata. Congreso Sudamericano Zool.,
Sao Paulo, 2:37-54.

. 1965. Especies de la subfamilia Monocondylaeinae en

las aguas del sistema del Rio de la Plata. Archiv f. Mollus-
kenkunde, 95:3-14.

Boss, K. J., and J. J. Parodiz. 1977. Paleospecies of Neotrop-
ical Ampullariidae and notes on other fossil non-marine
South American Gastropods. Ann. Carnegie Museum
46(91:107-127.

Bristow, C. R. 1973. Guide to the geology of the Cuenca Ba-
sin, southern Ecuador. Ecuadorian Geol. and Geophys.
Soc., Quito.

. 1976. On the age of the Nabon Eormation, Ecuador.

Newsletter, Stratigraphy, 5(2-3): 104-107.

Bristow, C. R., and R. Hoffstetter. 1977. Lexique strati-
graphique international, V. Amerique Latine, Ease. 5“2,
Ecuador. Centre National de Recherche Scientifique, Paris,
410 pp.

Brown, C. B. 1879. On the Tertiary deposits in the Solimoes
and Javary rivers in Brazil. Quart. J. Geol. Soc., London,
35:76-81.

Carcelles, a. 1941 . “Erodona mactroides” en el Rio de la
Plata. Physis, 19: 1 1-21 .

Collins, J. S., and S, E. Morris. 1976. Tertiary and Pleisto-
cene crabs from Barbados and Trinidad. Paleontology.
19:107-131.

Conrad, T. A. 1871, Descriptions of new fossil shells from
the upper Amazon. Amer. J. Conchoiogy, 6: 192-198.

. 1874fl. Remarks on the Tertiary clay of the Upper Am-
azon, with description of new shells. Proc. Acad. Nat. Sci.
Philadelphia, 26:25-32.

. 18746. Description of two new fossil shells of the Upper

Amazon. Proc. Acad. Nat. Sci. Philadelphia, 26:82-83.

Dale, W. H. 1890-1903. Contribution to the Tertiary fauna of
Elorida with special reference to the Miocene Silex-bed of
Tampa and Pliocene of the Caloosahatchie River. Trans.
Wagner Inst. Sci., Philadelphia, 3:1-199.

Daudin. E. M. 1808, Histoire Naturelle des Coquilles. Paris,
vol. 2, 329 pp.

Doering, A. 1875. Estudios sistematicos y anatomicos sobre
los Moluscos Pulmoniferos. Periodico Zoologico, Buenos
Aires 1 : 129-202.

Enoelhardt, H. 1895. Ueber Neue Tertiaifianzen Sud-Amer-
ikas. Ablh. Heurags. Senckenberg Naturf. Gesell., 19(2): I-
47.

Erazo, M. T. 1957. Apuntes sobre la geologia y estructu ra del
Valle de Cuenca. An. Univer. Cuenca, 13:157-197.

. 1965. Estudio de los deslizamientos del suelo en el aus-

tro. Cuenca, 26 pp.

E I HERiDGE, R. 1879. Notes on the Mollusca collected by C.
Barrington Brown from the Tertiary deposits of Solimoes
and Javary rivers, Brazil, Quart. J. Geol. Soc., London,
35:82-88.

Eields, W. R. 1957. Hystricomorph rodents from the Late
Miocene of Colombia, South America. Univ. California
Publ., Geol. Sci., 32:273-404.

Fischer, P. 1887. Manuel de Conchiliologie et de Paleontolo-
gie. F. Savy, Paris, vol. 1 , 1354 pp.

Gabb, W. M. 1869. Descriptions of fossils from the clay de-
posits of the Upper Amazon. Amer. J. Conchoiogy, 4: 197-
200.

(in Woodward, H.). 1871. The Tertiary shells of the

Amazon Valley. Ann. Mag. Nat. Hist., ser. 4, 7: 109.

Germeraad, J. H., C. A. Hopping, and J. Muller. 1968. Pal-
ynology of Tertiary sediments from tropical areas. Rev. Pa-
leobotan. Palynol., 6:188-348.

Gordon, W. A. 1966. Two crab species from the Middle Ter-
tiary of Puerto Rico. Trans. Caribbean Geol. Conf., 3: 184-
185.

Gray,J. E. 1847. A list of the genera of Recent Mollusca. Proc.
Zool. Soc. London, 15: 120-219,

Greve, L. de, 1938. Fine Molluskenfauna aus dem Neogen
von Iquitos am Oberen Amazonas in Peru. Abhandl.
Schweiz. Palaeontol. Gesell., 61: 133.

Haas, F. 1930-31. Versuch einer Kritischen Sichtung der Sud-
amerikanischen Najaden. Senckenbergiana, 12: 175-195;
13:30-52, 87.

Henderson, W. G. 1979. Cretaceous to Eocene volcanic ac-
tivity in the Andes of northern Ecuador. J. Geol. Soc. Lon-
don, 136:367-378.

Hoffstetter, R. 1970. Vertebrados Cenozoicos de Colombia.
Actas IV Congreso Latinoamer. Zool. , Caracas, 2:931-954.

Humboldt, A. D. 1823. Essai Geognostique sur le gisement
des roches dans les deux hemispheres. Levrault, Paris.

Ihering, H. von. 1893. Najaden von Sao Paulo und die geo-
graphische Verbreitung der Susswasser-Faunen von Sud-
amerika. Archiv Naturges., 1893:54-140.

. 1903. Les Mollusques des terrains cretaciques super-

ieures de I'Argentine orientale. An. Mus. Nac. Buenos
Aires, 9: 193-231.

. 1907. Les Mollusques fossiles du Tertiaire et du Cretace

Superieur de I’Argentine. An. Mus. Nac. Buenos Aires,
14:1-608.

Kadolsky, D. 1980. On the taxonomic position, the species,
and paleoecological significance of the genera Eubont, Tox-
osoma and Littondina(l) in the Pliocene Pebas Formation
of the Upper Amazon Region. The Veliger, 22:364-375.

Kennerley, j. B. 1973. Geology of Loja Province. Inst. Geol.
Sci., Overseas Div., London, Report 23.

Koch, E., and E. Blissenbach. 1962. Las capas rojas del Cre-
taceo Superior-Terciario en la region del curso medio del

52

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 19

Rio Uyacali, Oriente de Peru. Bol. Soc. Geol. Peru, 39:7-
141.

Lea, I. 1870. A synopsis of the family Unionidae. 4th edition,
Philadelphia, 184 pp.

Liddle, R. a., and R. V. W., Palmer. 1941. The geology and
paleontology of the Cuenca-Azogues-Biblian region, prov-
inces of Canar and Azuay, Ecuador. Bull. Amer. Paleontol.,
26:357-418.

Marshall, W. B. 1928. New fossil pearly-freshwater mussels
from deposits of the Upper Amazon of Peru. Proc. U.S.
Nat. Mus., 74(3): 1-7.

. 1931. Ruganodontites. a new genus of South American

pearly fresh-water mussels. Nautilus, 44:41-42.

. 1934. Two new species of pearly-freshwater mussels.

Washington Acad. Sci., 24:78-81.

Marshall, W. B., and E. O. Bowles. 1932. New fossil fresh
water mollusks from Ecuador. Proc. U.S. Nat. Mus., 82(5):
1-7.

Modell, H. 1950. Sudamerikanische Najaden der Gattungen
Castaliu Schelxchicdlu und Eciiadoreii. Archiv f. Mollus-
kenkunde, 79:135-146.

Morrison, J. P. E. 1954. The relationship of Old and New
World Melanias. Proc. U.S. Nat. Mus., 103:357-394.

Nunez Del Arco, E. 1971. Yacimiento de Bentonita de Char-
asol, Prov. de Canar, Ecuador. Direccion General de Minas
y Geologia, Quito.

Oliveira Roxo. M, G. de. 1924. Breve noticia sobre fosseis
terciarios Alto Amazonas. Bol. Serv. Geol. Miner. Brasil,
11:1.

Oi IVEIRA Roxo. M. G. DE, and O. H. Leonardos. 1943. Geo-
logia do Brasil, 2nd ed., Serv. Inform. Agricult. (Serie Di-
dactical, Rio de Janeiro, 2:1-813.

Olsson, a. a., and A. Harbison. 1953. Pliocene mollusks of
southern Elorida. Monogr. Acad. Nat. Sci. Philadelphia,
8:1-460.

Orbigny, a. d'. 1842. Voyage dans I'Amerique Meridionale.
Geologic, 3 (pt. 3): 1-290; Paleontology, 3 (pt. 4): 1-188;
Mollusques, 5 (pt. 3): 1-758.

Ortmann, a. E. 1921. South American naiades. Mem. Car-
negie Mus., 8:451-670.

Parodiz, J. J. 1960. Neotype of Lyrodes guaninitica Doering,
and description of a new species. Nautilus, 74:24-26.

. 1969. The Tertiary non-marine Mollusca of South

America. Ann. Carnegie Mus., 40:1-242.

. 1973. The species complex of Diplodoii delodontus

(Lamarck). Malacologia, 14:247-270.

Parodiz, J. J., and L. Hennings. 1965. The Neocorhicula of
the Parana-Uruguay Basin. Ann. Carnegie Mus., 38:69-96.

Peck, R. E., and C. C. Reker. 1947. Cretaceous and Lower
Cenozoic Charophyta from Peru. Amer. Mus. Novit.,
1369.

PiLSBRY, H. A. 1893. Notes on the genera of Unionidae and
Mutelidae. Nautilus, 7:30.

. 1921. Mollusks. ht Faunal Remains from the Trias, of

York County, Pennsylvania ( H. E. Wagner, ed.). Proc. Acad.
Nat. Sci. Philadelphia, 73:25-37.

. 1944. Molluscan fossils from the Rio Pachitea and vi-
cinity in eastern Peru. Proc. Acad. Nat. Sci. Philadelphia,
96:137-153.

. 1950. I’he name tiemisiniis (Longirerena) aviis. Nauti-
lus. 64:69.

PiLSBRY, H. A., and A. A. Olsson. 1935. Tertiary freshwater

mollusks of the Magdalena embayment, Colombia. Proc.
Acad. Nat. Sci. Philadelphia, 87:7-39.

. 1941. A Pliocene fauna from western Ecuador. Proc.

Acad. Nat. Sci. Philadelphia, 93:1-79.

Porta, J. de. 1974. Lexique stratigraphique international, 5
Amerique Latine, Ease. 4 Colombia (2d part). Centre Na-
tion. de la Recherche Scientifique, Paris.

PuTZER, H. 1968. Tertiare lignite im interandinen Graben von
Ecuador. Geol. Jb. Hannover. 85:461^88.

Repetto, F. 1977. Un mamifero fosil nuevo del Terciario del
Ecuador, ( Azuay-Canar). Tecnologica (Escuela Politecnica
Litoral), Guayaquil, l(2):33-38.

Richards, H. G. 1948. Fossil mollusks from the Trias of Penn-
sylvania. Noatulae Naturae, Acad. Nat. Sci. Philadelphia,
206:1-4.

Roberts, T. R. 1975. Characoid fish teeth from Miocene de-
posits in the Cuenca Basin, Ecuador. J. Zool., London,
175:258-271.

Savoyat, E., et al. 1970. Formaciones sedimentarias de la
Sierra tectonica andina en el Ecuador. Instituto Frances del
Petroleo, Serv. Nacional de Geologia y Minas, Quito.

Sheppard, G. 1934. Geology of the Interandine basin of Cuen-
ca, Ecuador. Geological Magazine, 7l(642):356-370.

Sheppard, G., and R. H. Bate. 1980. Pleistocene ostracods
from the Upper Amazon of Colombia and Peru.
Paleontologica 23.

SiGAL, J. 1968. Estratigraphia micropaleontologica del Ecua-
dor. Datos anteriores y nuevos. Instituto Frances del Petro-
leo, Serv. Nac. Geologia y Minas, Quito.

Simpson, G. G. 1961. The supposed Pliocene Pebas beds on
the Upper Jurua River, Brazil. J. Paleont., 35:620-624.

SiNGEWALD, J. T. 1927. Pongo de Manseriche. Bull. Geol. Soc.
Amer., 38:479-492.

. 1928. Geology of the Pichis and Pachitea rivers, Peru.

Bull. Geol. Soc. Amer., 39:447-464.

Spix, j. B. {ill Wagner, J. A.). 1827. Testacea Fluviatilia Bras-
iliam. Leipzig, 36 pp.

SuTER, H. 1913. Manual of the New Zealand Mollusca. Wel-
lington, 1020 pp.

Triebel, E. 1953. Fine fossile Pelocypris (Crustacea-Ostra-
coda) aus El Salvador. Senckenbergiana, 34(1-3): 1-4.

Tryon, G. W. 1883. Structural and systematic conchology.
Philadelphia, 2:1-430.

United Nations Development Program. 1969. Survey on
Metallic and non-Metallic Minerals. Technical Report No.
1, Coal Investigations (Operation Nol, Cuenca-Biblian and
Loja) Annex No. 1, Quito-New York.

. 1972. Survey on Metallic and non-Metallic Minerals

(Phase II), Geochemical, Geological and Geophysical In-
vestigations near San Miguel (Azogues). Technical Report
No. 12, New York.

Van DER Hammen, T. 1957. Periodicidad climatica y evolucion
de floras del Maestrichiano y del Terciario. Bol. Inst. Nac.
Geol., Colombia, 5(2): 5-46.

. I960. Estratigrafia del Maestrichiano y Terciario con-

tinetales y tectogenesis de los Andes Colombianos. Bol.
Inst. Nac. Geol., Colombia, 6( l-3):67-l28.

Van Eysinga, F. W. B. 1975. Geological time table. Elsevier
Scientific Co., Amsterdam.

Wenz, W. 1938. Handbuch der Palaozoologie, vol. 6 Gastro-
poda. Berlin, 1639 pp.

Wheeler, O.C. 1935. Tertiary stratigraphy of the Middle Mag-

1982

BRISTOW AND PARODIZ— ECUADORIAN TERTIARY SEDIMENTS

53

dalena Valley. Proc. Acad. Nat. Sci. Philadelphia, 87:29-
39.

White, E. 1. 1927. On a fossil cyprinodont from Ecuador. Ann.

Mag. Nat. Hist., ser. 9, 2:519-522.

Wolf, T. 1879. Viajes cientificos por la Republica del Ecuador.
2, Relacion de un viaje Geognostico por la provincia de
Azuay. Guayaquil, 78 pp.

. 1892. Geografia y geologia del Ecuador. Brockhaus,

Leipzig.

Woodward, H. 1871. The Tertiary shells of the Amazon Val-
ley. Ann. Mag. Nat. Hist., ser. 4, 7:59-64, 101-109.

/i

V.’ '

.V

*

■i

Copies of the following Bulletins of Carnegie Museum of Natural History may be obtained at the prices
listed from the Publications Secretary, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pitts-
burgh, Pennsylvania 15213.

1. Krishtalka, L. 1976. Early Tertiary Adapisoricidae and Erinaceidae (Mammalia, Insectivora) of North

America. 40 pp., 13 figs $2.50

2. Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the

late Pleistocene paleoecology of the central Appalachian Mountains of Virginia. 88 pp., 21
figs $12.00

3. Wetzel, R. M. 1977. The Chacoan peccary, Catagonus wagneri (Rusconi). 36 pp., 10 figs. . . $6.00

4. Coombs, M. C. 1978. Reevaluation of early Miocene North American Moropus (Perissodactyla,

Chalicotheriidae, Schizotheriinae). 62 pp., 28 figs $5.00

5. Clench, M. H., and R. C. Leberman. 1978. Weights of 151 species of Pennsylvania birds analyzed

by month, age, and sex. 87 pp $5.00

6. Schlitter, D. A. (ed.). 1978. Ecology and taxonomy of African small mammals. 214 pp., 48 figs.

$15.00

7. Raikow, R. J. 1978. Appendicular myology and relationships of the New World nine-primaried oscines

(Aves:Passeriformes). 43 pp., 10 figs $3.50

8. Berman, D. S, and J. S. McIntosh. 1978. Skull and relationships of the Upper Jurassic sauropod

Apatosaurus (Reptilia, Saurischia). 35 pp., 11 figs $3.00

9. Setoguchi, T. 1978. Paleontology and geology of the Badwater Creek area, central Wyoming. Part

16. The Cedar Ridge local fauna (Late Oligocene). 61 pp., 30 figs $4.50

10. Williams, D. F. 1978. Systematics and ecogeographic variation of the Apache pocket mouse (Roden-

tia: Heteromyidae). 57 pp., 23 figs $4.00

11. Guilday, J. E., H. W. Hamilton, E. Anderson, and P. W. Parmalee. 1978. The Baker Bluff Cave

deposit, Tennessee, and the Late Pleistocene faunal gradient. 67 pp., 16 figs $5.00

12. Swanepoel, P., and H. H. Genoways. 1978. Revision of the Antillean bats of the genus Brac/?_vp/zy//a

(Mammalia: Phyllostomatidae). 53 pp., 17 figs $4.00

13. Schwartz, J. H., and H. B. Rollins (eds.). 1979. Models and methodologies in evolutionary theory.

105 pp., 36 figs $6.00

14. Engesser, B. 1979. Relationships of some insectivores and rodents from the Miocene of North Amer-
ica and Europe. 68 pp., 12 figs., 20 plates $5.00

15. Bentz, G. D. 1979. The appendicular myology and phylogenetic relationships of the Ploceidae

and Estrildidae (Aves: Passeriformes). 25 pp., 5 figs $2.00

16. Mares, M. A. 1980. Convergent evolution among desert rodents; a global perspective. 51 pp., 25

figs $3.50

17. Lacher, T. E., Jr. 1981. The comparative social behavior of Kerodon rupestris and Balea spixii and

the evolution of behavior in the Caviidae. 71 pp., 40 figs $6.00

18. McIntosh, J. S. 1981. Annotated catalogue of the dinosaurs (Reptilia, Archosauria) in the collections

of the Carnegie Museum of Natural History. 67 pp., 22 figs $6.00

Qli

PHALLI OF RECENT GENERA AND SPECIES OF THE
FAMILY GEOMYIDAE (MAMMALIA: RODENTIA)

STEPHEN L. WILLIAMS

*

NUMBER 20 PITTSBURGH, 1982

BULLETIN

of CARNEGIE MUSEUM OF NATURAL HISTORY

PHALLI OF RECENT GENERA AND SPECIES OF THE
FAMILY GEOMYIDAE (MAMMALIA: RODENTIA)

STEPHEN L. WILLIAMS

Collection Manager, Section of Mammals

NUMBER 20

PITTSBURGH, 1982

BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY
Number 20, pages 1-62, 30 figures

Issued 16 June 1982

Price: $5.00 a copy

Craig C. Black, Director

Editorial Staff: Hugh H. Genoways, Editor; Duane A. Schlitter,
Associate Editor: Stephen L. Williams, Associate Editor;
Barbara A. McCabe, Technical Assistant.

(c) 1982 by the Trustees of Carnegie Institute, all rights reserved.

CARNEGIE MUSEUM OF NATURAL HISTORY, 4400 FORBES AVENUE
PITTSBURGH, PENNSYLVANIA 15213

CONTENTS

Abstract 5

Introduction 5

Methods and Materials 5

Results 7

Thomomys 8

Description 8

Thomomys bottae 8

Thomomys bidbivorous 14

Thomomys clusiiis 15

Thomomys idahoensis 16

Thomomys mazama 17

Thomomys monticola 20

Thomomys talpoides 21

Thomomys townsendii 22

Thomomys umbrinus 23

Statistics 24

Orthogeomys — Description 26

Zygogeomys — Description 28

Geomys 30

Description 30

Geomys arenarius 30

Geomys attwateri 34

Geomys bursarius 35

Geomys personatus 36

Geomys pinetis 37

Geomys tropicalis 38

Statistics 39

Pappogeomys — Description 42

Cratogeomys 47

Description 47

Cratogeomys castanops 47

Cratogeomys fumosus 49

Cratogeomys gymnurus 50

Cratogeomys merriami 51

Cratogeomys tylorhinus ■ 52

Cratogeomys zinseri 52

Statistics 53

Discussion 56

Summary 59

Acknowledgments 60

Literature Cited 60

t ^

•n .!

^ » i ■ /V • W

.'■At /

•'H ^

'A.

■*s

it

. :« Id

f *

I I

.t

■■■'jMaMkii'j»». '£S

ABSTRACT

A study of the external morphology of the phalli and bacula
of the family Geomyidae was conducted. All six genera and 24
species (including the nine species of Thomomys, the single
species of Zygogeomys, one of 1 1 species of Orthogeomys,
the six species of Geomys, one of two species of Pappogeomys,
and six of seven species of Cratogeomys) were examined. De-

tailed descriptions, with illustrations and measurements, were
provided for each taxa. Measurements were used in univariate
and multivariate statistics to examine individual, geographical,
and interspecific variation. Results of the study were compared
to systematic relationships proposed by previous investigators.

INTRODUCTION

Morphological studies of mammalian phalli have
been useful in differentiating related taxa (Geno-
ways, 1973; Hoffmeister and Lee, 1963; Hooper,
1958, 1959, 1960, 1961, 1962; Hooper and Hart,
1962; Hooper and Musser, 1964; Lidicker, 1968;
Williams et al., 1980). In spite of the numerous in-
vestigations that have been conducted on phalli of
various mammal groups, such studies dealing with
the New World family Geomyidae have been lim-
ited.

The phalli of the genus Thomomys are the best
documented in the family Geomyidae. However,
except for studies by Hill (1937) who described the
phallus of T. bulbivorous, all investigations of Tho-
momys have been concerned with only the baculum
and have excluded the soft anatomy of the phallus.
Bacula of species of Thomomys have received con-
siderably more attention because of their usefulness
in differentiating taxa, particularly in restricted geo-
graphical areas. Thaeler (1968) discussed bacular
differences between T. bottae, T. mazama, T. mon-
ticola, and T. talpoides in northern California. Bac-
ular differences were also examined between T.
mazama and T. talpoides in western Washington
by Johnson and Benson (1960). Long (1964) and
Thaeler (1972) discuss differences between the bac-
ula of T. idahoensis and T. talpoides in eastern
Idaho. Bacular differences between T. bottae and
T. iimbrinus in southern Arizona have been report-
ed by Hoffmeister (1969) and Patton (1973). The
only effort to make a more comprehensive exami-
nation of Thomomys bacula was made by Burt
(1960) in which bacula of T. bottae, T. bulbivorous,
T. talpoides, and T. umbrinus were described.

In the remaining genera of the family Geomyidae
{Orthogeomys, Zygogeomys, Geomys, Pappogeo-

mys, and Cratogeomys) nothing is known about the
external morphology of the phallus except for the
brief description of G. biirsarius by Hill (1937). Fur-
thermore, documentation concerning the bacula of
these taxa is very limited. Four of the six species
of Geomys (G. attwateri, G. bur sarins, G. person-
atus, and G. pinetis) have received brief descrip-
tions by Burt ( 1960), Kennedy (1958), and Sherman
(1940). Burt (1960) also briefly described the bacula
of Zygogeomys trichopus and two of the seven
species of Cratogeomys (C. merriami and C. tylor-
himis).

Therefore, the current knowledge of phalli in the
family Geomyidae is almost entirely restricted to a
few descriptions and dimensions of bacula of se-
lected taxa. Information about the soft anatomy of
the phallus of all geomyid species (except T. bul-
bivorous and G. bursarius) and the bacula of almost
half of the species is nonexistent. Furthermore,
there is essentially no phallic or bacular information
available concerning nongeographical and geo-
graphical morphometric variation within a species,
or among species of a common genus.

The purpose of this study is to examine and de-
scribe the external morphology of the phalli and the
bacula of members of the family Geomyidae. It is
intended to provide detailed descriptions, measure-
ments, and illustrations for each available taxon.
With these data, statistical analyses are performed
to assess nongeographical and geographical varia-
tion within selected species, followed by examina-
tion of variation among species. Information ob-
tained in this study is finally used to discuss
possible systematic relationships among the species
examined.

METHODS AND MATERIALS

races. Most penes were removed from freshly killed specimens.
Others were removed from study skins maintained in mammal

Penes were removed from 388 specimens of geomyid ro-
dents representing all six genera plus 24 species and 80 nominal

5

6

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

research collections, and subsequently rehydrated. Penes were
preserved in a solution of ethyl alcohol, formalin, and acetic acid
(AFA).

Because the validity of using dried penes has been previously
questioned (Lidicker, 1968), tests were performed to determine
the amount of restoration that might be expected from dried
penes. Eight penes were removed from fresh pocket gophers and
immediately illustrated, using a Wild camera-lucida dissecting
scope. Seven of the penes were allowed to dry completely. The
eighth sample was placed in AFA. These samples were main-
tained in this condition for over five years. The dried penes were
rehydrated and then all eight samples were illustrated again. The
outlines of both illustrations of each sample were placed over
each other to observe any alteration in shape and size caused by
their respective treatments. The seven dried samples did show
some alteration in shape after restoration. However, six of the
samples were not any different than similar alterations observed
in the sample that was kept in AFA. One dried sample was not
restored to its original shape. However, in this case the effects
caused by drying were easily detected without comparing illus-
trations. Therefore it was assumed that any serious distortions
caused by drying could be detected by careful examination of
the rehydrated sample. Distorted samples are generally com-
pressed laterally; the overall length in adult specimens is not
affected because of the support provided by the baculum. For
individual samples having distorted phalli, measurements of
width of the glans were omitted.

Although all samples were examined, for purposes of final
comparisons and analyses only adult individuals were used, thus
eliminating variation attributed to age. (To adequately examine
age variation considerably larger sample sizes should be avail-
able.) Determination of adult specimens was based on the fusion
of the basioccipital and basisphenoid bones. For some taxa, such
as Thomomys, development of sagittal and lambdoidal crests
and overall size were also used to determine mature, adult in-
dividuals.

Samples were first examined externally to determine charac-
teristic features. In addition to checking each sample from dor-
sal, ventral, and lateral views, special attention was given to
apical and epidermal structures. Within each species individual
samples were reexamined, compared, and a description was
written for the species. Anatomical terminology used in all de-
scriptions generally follows that of Hooper (1958). Next a rep-
resentative sample of each species was selected and schemati-
cally illustrated, using a Wild camera-lucida dissecting scope.
For illustrative purposes each sample was magnified about 12
times for lateral, ventral, and dorsal views, and about 50 times
for close-up views of epidermal structures. A metric scale was
always included with the lateral, ventral, and dorsal views so
that the amount of magnification could be accurately determined.

Individual penes were subsequently cleared and stained (Rus-
sell, 1973) so that the size, shape, and position of the baculum
could be determined. Because of potential loss of epidermal
structures (particularly on dried phalli) caused by clearing and
staining, such procedures were performed usually after thorough
examination of untreated samples was completed. The bacula
were examined and compared, and a description was written for
each species. The baculum corresponding to the illustrated glans
was also illustrated using the camera-lucida. In each case the
magnification was adjusted to the magnification used in the il-
lustration for the glans.

After clearing and staining was completed, measurements

were taken of the glans and baculum of each adult specimen.
This procedure was followed so that any differences that might
exist between fresh and dried penes would be minimized by their
common treatment. Because measurements were taken from
cleared and stained samples, all samples were examined while
they were submerged in glycerin, so that the shape would be
natural and problems of dessication would be eliminated. Mea-
surements were taken from magnified outlines of the samples,
using the camera-lucida, and dial calipers. Again a metric scale
was included in each diagram to determine the exact magnifi-
cation. Dimensions of the outline were then divided by the mag-
nification to provide the actual dimensions of the sample. Each
dimension was rounded to the nearest one-tenth millimeter. Us-
ing the described procedure the following measurements were
taken:

1. Length of distal tract. — Distance of the entire phallus be-
tween distalmost point of the glans and the ventral flexure, as
measured from the ventral aspect.

2. Length of glans. — Distance between the distalmost point
of the glans and a midventral point at the connection of the glans
and prepuce, as measured from the ventral aspect.

3. Length of protractile tip. — Distance between the distalmost
point of the glans and a midventral point on the distal edge of
the collar, as measured from the ventral aspect.

4. Width of glans across collar. — Greatest distance of the
glans across collar region, as measured from the ventral aspect.

5. Width of glans across base. — Distance of the glans at the
constriction where the glans and prepuce connect, as measured
from the ventral aspect.

6. Length of haculiim. — Distance between the distalmost and
proximalmost points of the baculum.

7. Width of bacillar base. — Greatest distance of the base of
the baculum as measured from the ventral aspect.

8. Height of bacidar base. — Greatest distance of the base of
the baculum, as measured from the lateral aspect.

Tests were made to determine if accurate bacular measure-
ments could be made without removing the baculum from the
glans, thus potentially destroying the glans. The samples were
measured within the glans, removed from the glans, and mea-
sured again. Percent differences in measurements of the length,
width, and height of 10 dissected bacula were (mean followed by
range in parentheses) 3.0 (0.1-10.9), 12.2 (0.0-40.8) and 9.1
(0.6-34.0), respectively. Based on these findings it was assumed
that accurate measurements could not be acquired from bacula
remaining in the glans. Therefore, all bacular measurements
were taken from dissected bacula.

Relative size relationships of phalli and bacula with body size
are often examined by utilizing the length of hind foot as a stan-
dard (Hooper, 1958, 1959, I960; Hooper and Hart, 1962; Hooper
and Musser, 1964). Because hind foot measurements often lack
consistency and precision in documentation, this study substi-
tuted condylobasal length as a comparative measurement. How-
ever, hind foot lengths were also documented for the benefit of
other investigators.

Standard statistics (mean, standard error, and coefficient of
variation) of individual nominal taxa were computed with a Tex-
as Instruments SR5I-A calculator. Analysis of age variation was
not included in this study. However, analyses of individual and
geographical variation were included, using adult individuals.
For series of samples with three or more individuals a Sum of
Squares Simultaneous Test Procedure (SS-STP), developed by
Gabriel (1964) and available in a univariate statistical program

1982

WILLIAMS— GEOMYID PHALLI

7

called UNIVAR (Power, 1970), was used to determine maxi-
mally nonsignificant subsets. Because sample sizes and geo-
graphical representation were limited in most taxa, the SS-STP
analysis was only applied to univariate analysis of geographical
variation in the genus Geomys.

Multivariate analyses were performed on a DEC- 10 computer
at Carnegie Mellon University, Pittsburgh. Means of one cranial
(condylobasal length), five phallic, and three bacular characters
of each species were used in a MINT multivariate analysis to
determine phenetic relationships. These relationships were de-
termined by cluster and principal component analyses. Matrices
of Q-Mode correlation (among OTU’s) and phenetic distance
coefficients were computed. Cluster analyses were conducted
using UPGMA (unweighted pair-group method using arithmetic
averages) on the correlation and distance matrices. A phenogram

was generated for each matrix. Phenograms were compared with
their respective matrices, and a coefficient of cophenetic cor-
relation was computed. A matrix of Pearson’s product-moment
correlation coefficient among characters was computed, and the
first three principal components extracted. Projections of the
OTU’s on the first three principal components were made.

Institutions from which material was used on this project, fol-
lowed by the respective acronym in parentheses, are as follows:
Carnegie Museum of Natural History (CM); Florida State Mu-
seum, University of Florida (FSM); Museum of Natural History,
University of Kansas (KU); Museum of Vertebrate Zoology,
University of California, Berkeley (MVZ); New Mexico State
University (NMSU); Texas Cooperative Wildlife Collection,
Texas A&M University (TCWC); The Museum, Texas Tech
University (TTU).

RESULTS

All six genera and 24 species of the family Geo-
myidae were examined. This study includes the
nine species of Thomomys — bottae, bulbivorous,
cliisius, idahoensis, mazama, monticola, talpoides,
townsendii, and umbrinus (see Hall, 1981, plus An-
derson, 1966, Hoffmeister, 1969, and Patton, 1973,
for use of T. bottae and T. umbrinus', Thaeler, 1979,
for use of T. clusius ; Thaeler, 1972, for use of T.
idahoensis ; and Hall and Kelson, 1959, and Thae-
ler, 1980, for use of T. townsendii)', the single
species of Zygogeomys — trichopus (see Hall, 1981);
one of the eleven species of Orthogeomys — hispi-
diis (see Hall, 1981); the six species of Geomys —
arenariiis, attwateri, bursarius, personatiis, pinetis,
and tropicalis (see Hall, 1981, and Williams and
Genoways, 1980, for use of G. pinetis and synon-
ymizing of G. colonus, G. cumberlandius, and G.
fontaneliis', Williams and Genoways, 1981, for taxo-
nomic changes in G. personatiis ', and Tucker and
Schmidley, 1981, for use of G. attwateri)', one of
the two species of Pappogeomys — bidleri (see Hall,
1981, and Honeycutt and Williams, 1982, for use of
Pappogeomys)', and six of the seven species of Cra-
togeomys (see Hall, 1981, and Honeycutt and Wil-
liams, 1982, for use of Cratogeomys).

The typical geomyid phallus is relatively simple
in form. The distal tract is about four to 20 times
longer than its width. Somewhere near the middle
of the distal tract exists a constriction where the
prepuce and proximal margin or base of the glans
connect. Proximal to the constriction, the distal
tract usually expands basally and is essentially fea-
tureless. The glans itself possesses several charac-

teristic features. The most conspicuous feature is
the collar which partially or completely encircles
the glans in the vicinity of the urethral opening. This
collar may be simple or very elaborate in shape.
The collar marks the proximal boundary of the pro-
tractile tip of the glans. On the ventral side of the
protractile tip is a single pair of urethral processes
which vary in shape and size and connect inside the
collar on the ventral wall of the urethral opening.
The region between the collar and constriction
often possesses more subtle features such as a mid-
ventral raphe and middorsal groove. Occasionally
lateral grooves are present and form one or two
dorsal protuberances, depending on the extent and
development of the middorsal groove. Between the
collar and constriction characteristic epidermal
structures are usually present. These structures may
also occur on the dorsal side of the protractile tip.
Each structure usually has one or more proximally
oriented projections which generally occur among
minute longitudinal folds and ridges. Epidermal fea-
tures can easily be seen with magnification. Indi-
vidual structures may be arranged in an uniform or
irregular pattern with respect to other structures.
The regions proximal to the constriction and on the
ventral and lateral sides of the protractile tip are
usually void of such epidermal structures.

The baculum of geomyid rodents is positioned
between the central axis and dorsal side of the phal-
lus. It extends into the protractile tip of the glans
and runs posteriorly into the area proximal to the
constriction. The baculum is entirely osseous and
usually consists of an enlarged base that tapers into

8

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

a long shaft that terminates with a distinct tip. The
baculum is typically curved with the dorsal side
being concave.

The following are more detailed descriptions
about the phalli and bacula of specific taxonomic
groups, at the generic level followed by descriptions
of represented species.

Thomomys

Description

Species belonging to the genus Thomomys show
remarkable variation in the size and shape of the
phallus. The length of the distal tract of adult spec-
imens examined ranged from 10.7 to 34.1 mm. The
length of the distal tract compared to relative body
size of individuals is very different among species,
with ratios of the condylobasal length to the length
of the distal tract ranging from 1.1 to 4.2. Basically
the shape of the phallus of individual species falls
into one of two types — either long and narrow, or
short to medium and broad. Depending on the
species, typical features, such as the constriction,
collar, and urethral processes, may vary from vis-
ible complexity to being simple and less distinct.
Other features such as raphe, middorsal grooves,
dorsal protuberances, and epidermal structures
may or may not exist — again depending on the
species considered.

The baculum is equally variable in size and is
directly related to the length of the distal tract. Ba-
sically the shape and structure are typical for geo-
myid rodents. The ratios of the condylobasal length
and length of the distal tract to the length of the
baculum ranges from 1.1 to 5.7 and 1.0 to 1.5, re-
spectively. Bacular lengths ranged from 9.1 to 32.9
mm in adult specimens examined.

The following are detailed descriptions of the
nine recognized species of Thomomys — T. bottoe,
T. bulbivorous, T. clnsiiis, T. idahoensis, T. ma-
zama, T. monticola, T. talpoides, T. townsendii,
and T. umbrinus. Table 1 provides the measure-
ments that apply to each description.

Thomomys bottae. — The shape and size of the
phallus of T. bottae (Fig. 1) agrees with the general
description of the phallus of geomyid rodents. How-
ever, compared to other species of Thomomys the
size of the phallus would be considered small to
medium-sized, with the length of the distal tract of
adult specimens examined ranging from 12.4 to 16.7
mm. The length of the distal tract is four to five
times greater than its width. The length of the glans

is less than half the length of the distal tract. The
sides of the glans are more or less parallel between
the collar and constriction, and are generally
straight or slightly recurved. The collar region tends
to be slightly expanded and is distinct from all as-
pects. The pair of urethral processes are exposed
and well-developed. Other features present on the
glans include a raphe that may vary in distinctness
among individuals, and a well-developed middorsal
groove that may extend almost the full length of the
glans. Between the collar and constriction there is
a pair of conspicuous dorsal protuberances.

Epidermal structures have a single proximally
oriented projection. The structures are generally
fairly uniform in shape, size, and pattern.

The shape of the baculum (Fig. 1) is typical of
most geomyid rodents, with a distinct base and tip
and distally tapering, and slightly curved shaft. The
length of the baculum among adults examined
ranged from 10.2 to 14.6 mm. The width of the base
tends to be less or equal to the height, and is slightly
tapered proximally when viewed dorsoventrally.
The description and measurements of the baculum
of T. bottae examined in this study correspond
closely to those found by other investigators (Burt,
1960; Hoffmeister, 1969; Ingles, 1965; Long and
Frank, 1968; Patton, 1973; Thaeler, 1968). How-
ever, this study examined only eastern subspecies
of T. bottae in which bacular measurements tended
to be larger than those reported of western races of
T. bottae (Burt, 1960; Ingles, 1965).

The ratios of condylobasal length to the length of
distal tract and to length of baculum ranged from
2.2 to 3.3 and 3.0 to 4.0, respectively. The ratio of
the length of distal tract to the length of baculum
ranged from 3.0 to 4.0.

Specimens examined. — Total (26).

T. b. actiiosus (1). — New Mexico: Torrance Co.: Red Canyon,
1 (TTU).

T. h. connectens (4). — New Mexico: Bernalillo Co.: Jet. Hwy.
500 and Hwy. 85, 1 (TTU); Torrance Co.: Willard, 2 (TTU);
Valencia Co.: Los Lunas, 1 (TTU).

T. h. limitaris (12). — Texas: Brewster Co.: 18.6 mi N, 1.2 mi E
Marathon, 4,300 ft, 6 (TTU); 17.9 mi N, 0.3 mi E Marathon,
4,300 ft, 1 (TTU); Crockett Co.: 15 mi N, 11 mi W Ozona, 1
(TTU); 17 mi (by road) NW Ozona, 1 (TTU); Pecos Co.: 17.0
mi N, 18.5 mi E Marathon, 4,500 ft, 1 (CM); Reagan Co.: 6
mi (by road) SE Stiles, 1 (TTU); Sutton Co.: Sonora cemetery,

1 (TTU).

T. b. limpiae (2). — Texas: Jeff Davis Co.: 10 mi NE Eort Davis,

2 (TTU).

T. h. opulentus (5). — New Mexico: Socorro Co.: 1 mi E San
Antonio, 5 (TTU).

1982

WILLIAMS— GEOMYID PHALLI

9

Table 1. — Standard statistics for adult specimens of 31 samples representing the nine species of Thomomys.

Taxon

N

Mean

(Range) ± 2 SE

cv

Length of distal tract

Thomomys bottae actuosus

!

14.5

Thomomys bottae connectens

4

15.7

(15.0-16.7) ± 0.73

4.6

Thomomys bottae limitaris

8

13.4

(12.4-15.8) ± 0.73

7.5

Thomomys bottae limpiae

2

15.1

(14.6-15.7) ± 1.10

5.1

Thomomys bottae opidentiis

5

12.6

(12.5-12.8) ±0.11

1.0

Thomomys bottae ruidosae

2

14.!

(13.9-14.3) ± 0.40

2.0

Thomomys bulbivorous

7

13.5

(11.3-14.5) ± 0.82

8.1

Thomomys clusius
Thomomys idahoensis confinus
Thomomys idahoensis idahoensis

5

12.3

(11.9-12.7) ± 0.31

2.8

Thomomys idahoensis pygmaeus
Thomomys mazarna mazama

6

29.8

(27.2-34.1) ± 2.11

8.7

Thomomys monticola

9

15.6

(15.0-16.4) ± 0.39

3.7

Thomomys talpoides bridgeri

5

17.2

(16.4-18.0) ± 0.60

3.9

Thomomys talpoides devexus

i

16.7

Thomomys talpoides fossor

3

24.4

(23.3-26.2) ± 1.82

6.4

Thomomys talpoides fuscus

1

14.6

Thomomys talpoides quadratus

2

14.1

(14.0-14.3) ± 0.30

!.5

Thomomys talpoides rostralis

I

21.9

Thomomys talpoides rufescens

!

20.0

Thomomys talpoides saturatus

1

20.2

Thomomys talpoides talpoides

1

19.1

Thomomys talpoides tenellus

1

21.8

Thomomys townsendii bachmani

3

13.5

(13.0-14.0) ± 0.58

3.7

Thomomys tovmsendii similis

3

14.9

(14.8-15.0) ± 0.13

0.8

Thomomys townsendii townsendii

1

15.5

Thomomys iimbrinus intermedius

2

10.9

(10.7-11.2) ± 0.41

9.2

Thomomys iimbrinus jiintae

!

10.3

Thomomys iimbrinus madrensis

1

10.5

Thomomys umbrinus sheldoni

3

11.1

(11.0-11.3) ± 0.20

1.6

Thomomys umbrinus spp.

2

12.1

(11.9-12.3) ± 0.40

2.3

Length of glans

Thomomys bottae actuosus

1

7.5

Thomomys bottae connectens

4

7.9

(7.4-S.3) ± 0.42

5.3

Thomomys bottae limitaris

8

6.9

(5. 9-8. 7) ± 0.59

12.1

Thomomys bottae limpiae

2

8.1

(7. 8-8. 3) ± 0.50

4.4

Thomomys bottae opulentiis

5

6.8

(6.3-7.7) ± 0.51

8.4

Thomomys bottae ruidosae

2

6.6

(6.4-6.S) ± 0.40

4.3

Thomomys bulbivorous

7

7.7

(6.6-8.S) ± 0.51

8.8

Thomomys clusius
Thomomys idahoensis confinus
Thomomys idahoensis idahoensis
Thomomys idahoensis pygmaeus

5

7.5

(6.6-S.3) ± 0.56

8.4

Thomomys mazama mazama

6

17.7

(16.0-19.3) ± 1.09

7.6

Thomomys monticola

7

8.3

(7. 6-9.4) ± 0.49

7.8

Thomomys talpoides bridgeri

5

8.3

(7.4-9.3) ± 0.73

9.8

Thomomys talpoides devexus

1

9.7

Thomomys talpoides fossor

3

11.7

(10.6-13.0) ± 1.39

10.3

Thomomys talpoides fuscus

1

8.7

Thomomys talpoides quadratus

2

7.1

(6.6-7. 7) ± 1.10

10.9

Thomomys talpoides rostralis

i

12.4

Thomomys talpoides rufescens

1

9.6

Thomomys talpoides saturatus

1

9.4

Thomomys talpoides talpoides

!

9.9

Thomomys talpoides tenellus

1

11.0

Thomomys townsendii bachmani

3

7.3

(6.6-7.7) ± 0.68

8.0

Thomomys townsendii similis

3

7.5

(7. 5-7. 5) ± 0.00

0.0

10

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

Table 1 . — Continued.

Taxon

N

Mean

(Range) ± 2 SE

cv

Thomomys townsendii townsendii

2

7.8

(7.7-7.8) ± 0.20

1.8

Thomomys iimhrinus intermedins

3

5.7

(4. 7-6.3) ± 0.98

14.9

Thomomys iimhrinus juntae

1

5.1

Thomomys umhrinus madrensis

1

5.7

Thomomys umhrinus sheldoni

3

6.2

(5. 7-6.5) ± 0.48

6.7

Thomomys umhrinus spp.

2

6.5

16.5-6.6) ± 0.10

1.1

Length of protractile tip

Thomomys hottae actuosus

1

3.3

Thomomys hottae connectens

4

3.5

(2.7-4. 1) ± 0.60

17.1

Thomomys hottae limitaris

8

3.1

12.9-3.5) ± 0.17

7.7

Thomomys hottae limpiae

2

3.7

13.6-3.9) ± 0.30

5.7

Thomomys hottae opulentus

5

3.1

12.6-3.7) ± 0.37

13.4

Thomomys hottae ruidosae

2

2.5

12.5-2.6) ± 0.10

2.8

Thomomys hulhivorous

7

3.3

12.3-4.2) ± 0.55

22.0

Thomomys clusius

5

3.8

13.4-4.3) ± 0.34

10.1

Thomomys idahoensis conjinus
Thomomys idahoensis idahoensis
Thomomys idahoensis pygmaeus
Thomomys mazama maz.ama

6

2.7

11.8-3.3) ± 0.56

25.4

Thomomys monticola

7

3.5

12.8-4.0) ± 0.35

13.2

Thomomys talpoides hridgeri

5

4.8

14.3-5.9) ± 0.60

14.1

Thomomys talpoides devexus

1

3.7

Thomomys talpoides fossor

3

4.4

13.7-5.4) ± 1.05

20.6

Thomomys talpoides fuscus

1

3.1

Thomomys talpoides quadratus

2

3.7

13.2^. 1) ± 0.90

17.2

Thomomys talpoides rostralis

1

5.3

Thomomys talpoides rufescens

1

4.9

Thomomys talpoides saturatus

1

3.9

Thomomys talpoides talpoides

1

4.5

Thomomys talpoides tenellus

1

3.7

Thomomys townsendii hachmani

3

3.2

12.9-3.1) ± 0.48

13.0

Thomomys townsendii similis

3

3.4

13.2-3.6) ± 0.24

6.1

Thomomys townsendii townsendii

2

3.2

12.7-3.7) ± 1.00

22. 1

Thomomys umhrinus intermedins

3

2.5

(2.0-2. 9) ± 0.55

18.9

Thomomys umhrinus juntae

1

2.6

Thomomys umhrinus madrensis

1

2.2

Thomomys umhrinus sheldoni

3

2.7

(2.3-3. 1) ± 0.46

14.8

Thomomys umhrinus ssp.

2

3.1

(3. 0-3. 3) ± 0.30

6.8

Width of glans across collar

Thomomys hottae actuosus

1

2.3

Thomomys hottae connectens

4

2.7

(2. 3-3.0) ± 0.30

11.1

Thomomys hottae limitaris

8

2.3

(1. 9-2.9) ± 0.26

15.9

Thomomys hottae limpiae

2

3.0

(2.9-3. 1) ± 0.20

4.7

Thomomys hottae opulentus

5

2.8

(1.9-3. 5) ± 0.61

24.3

Thomomys hottae ruidosae

2

1.9

(1.6-2. 2) ± 0.60

22.3

Thomomys hulhivorous

7

3.7

12.9-4.5) ± 0.41

14.6

Thomomys clusius

5

3.2

12.6-3.7) ± 0.42

14.6

Thomomys idahoensis conjinus
Thomomys idahoensis idahoensis
Thomomys idahoensis pygmaeus
Thomomys mazama mazama

6

2.5

(1. 9-3.0) ± 0.43

21.1

Thomomys monticola

5

2.4

11.7-3.2) ± 0.54

25.3

Thomomys talpoides hridgeri
Thomomys talpoides devexus

5

2.2

(1.8-2. 5) ± 0.25

13.0

Thomomys talpoides fossor

3

2.2

(2. 0-2. 4) ± 0.24

9.5

Thomomys talpoides fuscus

1

1.9

Thomomys talpoides quadratus

2

2.6

12.6-2.6) ± 0.00

0.0

1982

WILLIAMS— GEOMYID PHALLI

1 1

Table 1. — Coniinued.

Taxon

N

Mean

(Range) :!

; 2 SE

CV

Thomomys talpoides rostralis

1

3.2

Thomomys talpoides nifescens

1

2.8

Thomomys talpoides saturatus

1

1.7

Thomomys talpoides talpoides

1

1.8

Thomomys talpoides tenellas

1

1.8

Thomomys townsendii bachmani

3

3.0

(2.9-3. 1)

±0.11

3.3

Thomomys townsendii similis

3

2.7

(2. 6-2. 8)

± 0.13

4.3

Thomomys townsendii townsendii

2

2.7

(2.2-3.21

± 1.00

26.2

Thomomys nmhriniis intermedins

3

2.2

(1. 5-2.5)

± 0.67

26.2

Thomomys umhrinns jnntae

1

4.1

Thomomys umhrinns madrensis

1

2.0

Thomomys nmbrinus sheldoni

2

2.7

(2.2-3. 1)

± 0.90

23.6

Thomomys nmhriniis ssp.

2

3.3

(2.9-3.81

± 0.90

19.3

Width of glans across base

Thomomys hottae actnosns

1

1.7

Thomomys hottae connectens

4

2.5

(2. 2-2. 7)

± 0.21

8.2

Thomomys hottae limitaris

8

1.9

(1. 7-2.4)

± 0.16

12.3

Thomomys hottae limpiae

2

2.4

(2.3-2.51

± 0.20

5.9

Thomomys hottae opnlentns

5

2.1

(1.5-2. 8)

± 0.47

24.8

Thomomys hottae rnidosae

2

1.7

(1. 7-1.7)

± 0.00

0.0

Thomomys htdhivorons

7

3.7

(3.2-4.31

± 0.38

13.5

Thomomys cinsins

5

2.1

(1.9-2. 3)

± 0.13

7.1

Thomomys idahoensis confinns

Thomomys idahoensis idahoensis

Thomomys idahoensis pygntaens

Thomomys mazama mazama

6

3.0

(2.6-3.31

± 0.20

8.2

Thomomys monticola

5

1.8

(1.5-2.31

± 0.35

21.7

Thomomys talpoides bridgeri

5

2.2

(2. 0-2. 4)

± 0.13

6.7

Thomomys talpoides devexns

Thomomys talpoides fossor

3

1.9

(1. 7-2.0)

± 0.18

8.0

Thomomys talpoides fuscns

1

1.7

Thomomys talpoides qnadratns

2

2.1

(2.0-2. 1)

± 0.10

3.4

Thomomys talpoides rostralis

1

2.4

Thomomys talpoides nifescens

1

2.3

Thomomys talpoides saturatus

1

1.7

Thomomys talpoides talpoides

1

1.9

Thomomys talpoides tenellns

1

2.0

Thomomys townsendii bachmani

3

2.7

(2.5-2.91

± 0.24

7.7

Thomomys townsendii similis

3

2.1

(2.0-2. 3)

± 0.18

7.3

Thomomys townsendii townsendii

2

2.5

(2.1-2.91

± 0.80

22.6

Thomomys umhrinns intermedins

3

2.1

(1.4-2.51

± 0.73

30.2

Thomomys umhrinns jnntae

1

2.6

Thomomys umhrinns madrensis

1

2.0

Thomomys umhrinns sheldoni

2

2.3

(1.8-2. 7)

± 0.90

27.7

Thomomys umhrinns ssp.

2

2.5

(2. 2-2. 9)

± 0.70

19.8

Length of hacninm

Thomomys hottae actnosns

1

11.3

Thomomys hottae connectens

4

13.1

(11.1-14.6)

± 1.54

11.8

Thomomys hottae limitaris

8

1 1.4

(10.3-13.1)

± 0.69

8.6

Thomomys hottae limpiae

2

11.2

(11.0-11.4)

± 0.40

2.5

Thomomys hottae opnlentns

5

10.8

(10.2-11.7)

± 0.55

5.7

Thomomys hottae rnidosae

1

11.9

Thomomys hnihivorons

5

10.1

(9.6-10.3)

± 0.26

2.9

Thomomys cinsins

5

11.3

(10.0-12.1)

± 0.71

7.0

Thomomys idahoensis confinns

1

16.2

Thomomys idahoensis idahoensis

6

21.5

(20.0-23.4)

± 1.04

5.9

Thomomys idahoensis pygmaens

4

19.6

(17.8-21.0)

± 1.34

6.8

12

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

Table 1. — Continued.

Taxon

N

Mean

(Range) ± 2 SE

cv

Thomomys mazania mazama

5

28.2

(24.9-32.9) ± 2.46

10.7

Thomoniys monticola

5

14.6

(13.8-15.2) ± 0.48

3.7

Thomomys talpoides hridgeri

5

15.8

(14.7-16.5) ± 0.63

4.4

Thomomys talpoides devexus

1

15.7

Thomomys talpoides fossor

3

21.7

(21.1-22.6) ± 0.92

3.7

Thomomys talpoides fusciis

1

13.1

Thomomys talpoides quadratus

2

12.8

(12.3-13.3) ± 1.03

5.7

Thomomys talpoides rostralis

1

20.1

Thomomys talpoides nifescens

1

18.7

Thomomys talpoides saturatus

1

18.0

Thomomys talpoides talpoides

1

18.2

Thomomys talpoides teneiius

1

20.5

Thomomys townsendii hachmani

3

12.5

(12.2-12.7) ± 0.29

2.0

Thomomys townsendii similis

3

13.6

(13.3-13.9) ± 0.35

2.2

Thomomys townsendii townsendii

2

13.5

(13.5-13.6) ± 0.10

0.5

Thomomys umbrinus intermedins

1

8.0

Thomomys umbrinus juntae

1

9.1

Thomomys umbrinus madrensis

Thomomys umbrinus sheldoni

3

9.6

(9.1-10.6) ± 0.97

8.7

Thomoniys umbrinus ssp.

2

10.5

(10.0-11.1) ± 1.10

7.4

Width of hacular base

Thomomys bottae actuosus

1

1.4

Thomomys bottae connectens

4

1.7

(1. 6-2.0) ± 0.17

10.2

Thomomys bottae limitaris

8

1.5

(1.0-1. 8) ± 0.19

17.6

Thomomys bottae limpiae

2

1.5

(1.4-1. 6) ± 0.20

9.4

Thomomys bottae opulentus

5

1.3

(1.0-1. 6) ± 0.19

16.7

Thomomys bottae rudiosae

i

1.3

Thomomys bidhivorous

5

2.2

(1.7-2. 5) ± 0.26

13.4

Thomomys clusius

5

1.6

(1. 4-1.9) ± 0.19

13.5

Thomomys idahoensis confinus

1

1.1

Thomomys idahoensis idahoensis

6

1.4

(1.3-1. 5) ± 0.61

5.4

Thomomys idahoensis pygmaeus

3

1.2

(1.1-1. 3) ± 0.13

9.9

Thomomys mazama mazama

5

1.6

(1. 3-2.1) ± 0.23

17.7

Thomomys monticola

5

1.6

(1. 2-2.0) ± 0.26

18.2

Thomomys talpoides bridgeri

5

1.8

(1. 4-2.0) ± 0.22

13.9

Thomomys talpoides devexus

1

1.9

Thomomys talpoides fossor

3

1.7

(1.7-1. 8) ± 0.07

3.4

Thomomys talpoides fuscus

1

1.1

Thomomys talpoides quadratus

2

1.4

(1. 4-1.4) ± 0.00

0.0

Thomomys talpoides rostralis

1

2.4

Thomomys talpoides rufescens

1

1.6

Thomomys talpoides saturatus

1

1.7

Thomomys talpoides talpoides

1

1.7

Thomomys talpoides teneiius

1

1.4

Thomomys townsendii bachmani

3

1.1

(0.9-1. 2) ± 0.18

13.9

Thomomys townsendii similis

3

1.5

(1.5- 1.5) ± 0.00

0.0

Thomomys townsendii townsendii

2

2.3

(2. 2-2.4) ± 0.20

6.1

Thomomys umbrinus intermedins

1

0.9

Thomomys umbrinus juntae

1

2.0

Thomomys umbrinus madrensis

Thomomys umbrinus sheldoni

3

1.7

(1.4-2.0) ± 0.35

18.0

Thomomys umbrinus ssp.

2

1.7

(1.6-1. 7) ± 0.10

4.1

Height of bacular base

Thomomys bottae actuosus

1

1.4

Thomomys bottae connectens

4

2.0

(1.6-2. 4) ± 0.34

16.8

Thomomys bottae limitaris

8

1.8

( 1.4-2. 0) ± 0.13

10.6

Thomomys bottae limpiae

2

1.5

(1. 2-1.7) ± 0.50

23.6

1982

WILLIAMS— GEOMYID PHALLI

13

Table I.

ontiniied.

Taxon

N

Mean

(Range) ± 2 SE

CV

Thomomys hottae opulentus

5

1.3

(1.2-1. 8) ± 0.23

20.1

Thomomys holtae ruidosae

1

1.5

Thomomys hidhivoroiis

5

2.1

(1.9-2. 3) ± 0.15

7.8

Thomomys clusius

5

1.7

(1.4-1. 9) ± 0.19

12.5

Thomomys idahoensis confinus

1

0.9

Thomomys idahoensis idahoensis

6

1.4

(1.2-1. 8) ± 0.17

15.1

Thomomys idahoensis pygmaetis

3

1.1

(0.9-1. 2) ± 0.18

14.3

Thomomys mazama mazama

5

1.4

(1. 0-1.9) ± 0.24

20.9

Thomomys monticola

5

1.6

(1. 4-2.0) ± 0.22

15.6

Thomomys lalpoides bridgeri

5

1.5

(1. 4-1.7) ± 0.10

7.6

Thomomys talpoides devexus

1

1.7

Thomomys talpoides fossor

3

1.5

(1. 1-1.8) ± 0.44

25.2

Thomomys talpoides fnscits

1

1.6

Thomomys talpoides quadratiis

2

1.5

(1.5-1. 6) ± 0.10

4.7

Thomomys talpoides rostralis

1

1.8

Thomomys talpoides rufescens

1

1.9

Thomomys talpoides satiiratiis

1

1.5

Thomomys talpoides talpoides

1

1.6

Thomomys talpoides tenellns

1

1.8

Thomomys townsendii bachmanii

3

2.1

(2.0-2. 1) ± 0.07

2.7

Thomomys townsendii similis

3

1.9

(1.9-1. 9) ± 0.00

0.0

Thomomys townsendii townsendii

2

1.9

(1. 9-2.0) ± 0.10

3.7

Thomomys umbrinus intermedins

1

0.9

Thomomys umbrinus juntae

1

2.0

Thomomys umbrinus madrensis

Thomomys umbrinus sheidoni

3

1.7

(1. 6-1.9) ± 0.18

9.0

Thomomys umbrinus ssp.

2

1.7

(1.6-1. 8) ± 0.20

8.3

Condylohasal length

Thomomys bottae actuosus

1

42.1

Thomomys hottae connectens

4

43.9

(37.1-48.9) ± 5.71

13.0

Thomomys hottae limitaris

8

37.3

(34.0-40.9) ± 1.74

6.6

Thomomys hottae limpiae

2

38.7

(38.1-39.2) ±1.10

2.0

Thomomys hottae opulentus

5

40.8

(38.8-42.0) ± 1.19

3.3

Thomomys hottae ruidosae

2

37.7

(37.1-38.3) ± 1.20

2.3

Thomomys huihivorous

7

54.6

(42.4-57.4) ± 1.29

3.1

Thomomys clusius

5

31.6

(30.4-33.0) ± 0.91

3.2

Thomomys idahoensis confinus

1

34.5

Thomomys idahoensis idahoensis

6

32.2

(31.0-33.3) ± 0.34

2.6

Thomomys idahoensis pygmaeus

4

30.1

(29.4-30.5) ± 0.23

1.5

Thomomys mazama mazama

6

35.9

(34.7-36.7) ± 0.73

2.5

Thomomys monticola

9

35.1

(33.9-36.4) ± 0.47

2.0

Thomomys talpoides bridgeri

5

34.5

(32.5-35.7) ± 1.07

3.5

Thomomys talpoides devexus

1

36.1

Thomomys talpoides fossor

3

39.1

(37.2-40.3) ± 1.92

4.3

Thomomys talpoides fuscus

1

35.6

Thomomys talpoides quadratiis

2

34.5

(34.2-34.7) ± 0.50

1.0

Thomomys talpoides rostralis

1

36.7

Thomomys talpoides rufescens

1

39.6

Thomomys talpoides saturatus

1

34.9

Thomomys talpoides talpoides

1

38.5

Thomomys talpoides tenellus

1

37.3

Thomomys townsendii bachmani

3

47.6

(47.0-48.2) ± 0.70

1.3

Thomomys townsendii similis

3

51.0

(49.8-52.0) ± 1.28

2.2

Thomomys townsendii townsendii

2

50.3

(48.2-52.5) ± 4.31

6.0

Thomomys umbrinus intermedins

3

36.8

(36.3-37.6) ± 0.81

5.4

Thomomys umbrinus juntae

1

38.7

Thomomys umbrinus madrensis

1

37.7

Thomomys umbrinus sheidoni

3

39.0

(38.2-40.2) ± 1.20

2.7

Thomomys umbrinus ssp.

2

40.3

(39.6-40.9) ± 1.30

2.3

14

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

Table 1 . — Continued.

Taxon

N

Mean

(Range) ± 2 SE

cv

Length of hind foot

Thomomys hottae actiiosus

1

27

Thomomys hottae connectens

4

30.7

(29-33) ± 2.06

6.7

Thomomys hottae limitaris

8

26.0

(24-29) ±1.19

6.5

Thomomys hottae limpiae

2

28.0

(27-29) ± 2.01

5.1

Thomomys hottae opulentus

4

31.0

(30-32) ± 0.82

2.6

Thomomys hottae ruidosae

2

25.0

(24-26) ± 2.01

5.7

Thomomys hulhivorous

7

41.1

(39^5) ± 1.47

4.7

Thomomys clusius

5

20.4

(20-21) ± 0.49

2.7

Thomomys idahoensis confinus

1

25

Thomomys idahoensis idahoensis

6

23.3

(23-24) ± 0.42

2.2

Thomomys idahoensis pygmaeus

4

21.3

(21-22) ± 0.50

2.3

Thomomys muzama mazama

6

28.0

(27-29) ± 0.52

2.3

Thomomys monticola

9

28.7

(27-30) ± 0.58

3.0

Thomomys talpoides hridgeri

5

28.4

(27-29) ± 0.80

3.1

Thomomys talpoides devexus

1

29

Thomomys talpoides fossor

3

26.7

(24-31) ± 4.38

14.2

Thomomys talpoides fiiscus

1

28

Thomomys talpoides quadratus

2

26.5

(26-27) ± 1.00

2.7

Thomomys talpoides rostralis

1

27

Thomomys talpoides rufescens

1

30

Thomomys talpoides saturatus

1

28

Thomomys talpoides talpoides

1

31

Thomomys talpoides tenellus

1

28

Thomomys townsendii hachmani

2

36.0

(36-36) ± 0.00

0.0

Thomomys townsendii similis

3

38.7

(38-39) ± 0.67

1.5

Thomomys townsendii townsendii

2

38.5

(37-40) ± 3.01

5.5

Thomomys umbrinus intermedins

3

29.7

(29-30) ± 0.67

5.0

Thomomys umbrinus juntae

1

29

Thomomys umbrinus madrensis

I

29

Thomomys umbrinus sheldoni

3

28.7

(28-29) ± 0.67

2.0

Thomomys umbrinus ssp.

2

30.5

(29^32) ± 3.01

6.9

T. b. ruidosae (2). — New Mexico: Otero Co.: 14 mi S, 1 mi E

Cloudcroft, 2 (TTU).

Thomomys hulbivoroiis. — Although the phallus
of T. hulbivoroiis (Fig. 2) would be considered short
and broad like T. hottae, the shape and character-
istic features are distinctly unique from all other
geomyid species. The length of the distal tract, av-
eraging 13.5 mm and ranging from 11.3 to 14.5 mm
in seven adult specimens examined, is small com-
pared to other species. Measurements of the length
of distal tract in this study differ from the 18.0 mm
reported by Hill (1937), but the phallus of his spec-
imen was evidently over-extended and then mea-
sured. The length of the distal tract is three to four
times greater than its width, with the glans being
more than half the length of the distal tract. Proxi-
mal to the constriction the phallus has a bulbous
shape. Distally, the glans has converging concave
sides, when viewed from a dorsoventral aspect.
When viewed from a lateral aspect the sides are

more or less straight and parallel. Except for the
midventral region, the collar of the glans is unusu-
ally large and well-developed, and distinct from all
aspects. The urethral processes are also unique
among geomyid rodents. The main portion of the
urethral processes is exposed, well-developed, and
partially connected along their common side. Un-
like other members of the family each process pos-
sesses a smaller, yet conspicuous, projection on the
ventral surface. The region between the collar and
constriction also possesses some unique features.
For instance, instead of possessing minute longi-
tudinal grooves as in most other geomyid rodents,
T. biilbivoroiis possesses an erratic pattern of crev-
ices. The distribution of these crevices is predom-
inantly lateral and ventral, except midventrally
where the absence of such crevices gives an indi-
cation of the existence of a midventral raphae. Dor-
sally the crevices tend to form grooves that con-
verge, but then fade out toward the lower middorsal

1982

WILLIAMS— GEOMYID PHALLI

15

Fig. I . — Phallus and baculum of Thomomys bottae opuientus from New Mexico: Socorro Co., I mi E San Antonio (TTU 26433).
Illustrated are the lateral (A), ventral (B), and dorsal (C) views of the phallus, magnified view of epidermal structures (enclosed in
rectangle) occurring on the glans, and ventral (D) and lateral (E) views of the baculum. The dotted lines in A, B, and C indicate the
position of the baculum in the phallus (scale is five millimeters).

region of the glans. In this area the separation of
the glans from the proximal portion of the distal
tract is not evident. Similarly on the ventral side
the midventral region possesses a smaller but more
conspicuous “bridge” between the proximal and
distal portions of the distal tract. The middorsal
groove is restricted to only the collar. Dorsal pro-
tuberances do not occur in T. bulbivorous.

Epidermal structures have a small, single proxi-
mally oriented projection. The size and distribution
of these structures are irregular. The structures may
be distributed in small clusters.

Although the general shape of the baculum of T.
bulbivorous (Fig. 2) is typical of geomyid rodents,
it does possess a broad tip and a large bulbous base.
The width of the base is about the same size as the
height. The length of the baculum among seven
adult individuals examined averaged 10.1 mm and
ranged from 9.6 to 10.3 mm. Burt (1960) and Ingles
(1965) reported shorter bacular measurements of
8.5 and 8.8 mm, respectively.

For five specimens the mean and range (in paren-

theses) of ratios of the condylobasal length to the
length of the distal tract and the length of baculum
were 4.0 (3. 7-4. 2) and 5.5 (5. 3-5. 7), respectively.
The ratio of the length of the distal tract to length
of baculum averaged 1 .4 with a range from 1.3 to
1.5.

Specimens examined. — Total ( 1 1).

T. bulbivorous (1 1). — Oregon; Benton Co.: 5 mi N, 5 mi E
Corvallis, 200 ft, 2 (CM); 4.5 mi N, 4.5 mi E Corvallis, 1
(TTU); 4 mi NE Corvallis, 7 (CM); Corvallis, 1 (TTU).

Thomomys clusius. — The phallus of T. clusius
(Fig. 3) is similar to that of T. bottae, except that
the length of the glans tends to be longer and make
up a greater portion of the distal tract. The length
of the distal tract, ranging from 1 1.9 to 12.7 mm in
five adults examined, is about four to five times
greater than the width, and about two times greater
than the length of the glans. The sides of the glans
gradually expand distally and flare outward at the
well-developed collar. The urethral processes are
large and conspicuous. A midventral raphe is pres-

16

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

Fig. 2. — Phallus and baculum of Thoniomys hulbivoroiis from Oregon: Benton Co., Corvallis (TTU 22816), illustrated as in Fig. I.

ent as well as a middorsal groove. However, the
latter does not extend the entire length of the glans
and is most evident in the vicinity of the collar. The
dorsal protuberances are not evident.

Epidermal structures, examined on cleared phal-
li, have single proximally oriented projections. The
size, shape, and pattern of the structures are fairly
uniform.

The baculum of T. clusius (Fig. 3) resembles that
of T. hottae, although the size of the former tends
to be smaller. The length of the baculum of five
adults examined ranged from 10.0 to 12.1 mm.
The width of the base is about equal to, or slightly
smaller than the height of the base. The base is
distinct and is slightly tapered proximally.

Comparing five individuals, the average ratios,
followed by ranges (in parentheses), of the condy-
lobasal length to length of distal tract and length of
baculum, and length of distal tract to length of bac-
ulum are 2.6 (2. 5-2. 7), 2.8 (2.6-3. 1), and 1.1
(1.0-1. 2), respectively.

Specimens examined. — Total (5).

T. clusius (5). — Wyoming: Carbon Co.: 14.4 mi S, 6.0 mi E
Rawlins, 6,900 ft, TI9N R86W Sec 3 1 , 2 (NMSU); Sweetwater
Co.: 1.0 mi S, 0.6 mi W Bitter Creek, 6,850 ft, TI8N R99W
Sec 15, I (NMSU); 6.9 mi S, 2.0 mi E Bitter Creek, 7,000 ft.

TI7N R99W Sec 13, I (NMSU); 7.1 mi S, 0.9 mi E Bitter

Creek, 7,100 ft, TI7N R99W Sec 14, I (NMSU).

Thomomys idahoensis. — The only soft anatomi-
cal parts of phallic material of T. idahoensis ex-
amined in this study were from subadult individu-
als. Therefore, the following comments regarding
the external phallus may be subject to some dis-
crepancies due to age variation.

The phallus of T. idahoensis (Fig. 4) is long and
slender, with a length eight to nine times greater
than the width. The length of the glans is about half
the length of the distal tract. The sides of the glans
are somewhat expanded apically to the region of
the urethral opening. Distal to the urethral opening
a pair of long slender urethral processes are ex-
posed. The protractile tip is relatively long com-
pared to other species of Thomomys. In the sub-
adult specimens examined there was no indication
of a collar, midventral raphae, middorsal groove,
or dorsal protuberances. Although epidermal struc-
tures were weakly-defined in specimens examined,
T. idahoensis probably conforms to other members
of the genus by having a single proximally ori-
ented projection. It is possible that phallic features
may become more prominent with maturity.

The length of the baculum among adults exam-

1982

WILLIAMS— GEOMYID PHALLI

17

Fig. 3. — Phallus and baculum of Thornomys clusius from Wyoming: Carbon Co., 14.4 mi S, 6.0 mi E Rawlins (NMSU 5988), illustrated
as in Fig. 1.

ined ranged from 16.2 to 23.4 mm. There were no
consistent size relationships between height and
width of the base. The baculum of T. idahoensis
(Fig. 4) differs from previously discussed taxa by
being relatively long, slender, and having a base
and tip that are not as distinct. The ratio of the
condylobasal length to the length of the baculum
ranged from 1.4 to 1.7, except for the single speci-
men of r. I. confirms which had a ratio of 2.1. Ob-
servations of bacula of T. idahoensis in this study
are in agreement with those of Long (1964) and
Thaeler (1972).

Specimens examined. — Total (16).

T. i. confinus (1). — Montana: Ravalli Co.: 3.5 mi N, 6.0 mi E
Stevensvilie, 3,650 ft, T9N R19W Sec 10, 1 (NMSU).

T. i. idahoensis (il). — Idaho: Bingham Co.: 0.5 mi N, 4.0 mi
E Shelley, 4,670 ft, TIN R37E Sec 25, ! (NMSU): Bonneville
Co.: 1/2 mi N, 3‘/2 mi W Idaho Falls, 5 (TTU); 5.3 mi N, i.O
mi W Shelley, 4,650 ft, T2N R37E Sec 31, i (NMSU); 0.4 mi
N, 1.0 mi W Swan Valley, 5,400 ft, T2N R43E Sec 35, 3
(NMSU); 0.6 mi S, 2.3 mi W Swan Valley, 5,400 ft, TIN R43E
Sec 3, 1 (NMSU).

T. i. pygmaeus (4). — Wyoming: Uinta Co.: 0.8 mi N, 3.6 mi W
Ft. Bridger, 7,000 ft, T16N R116W Sec 35, 4 (NMSU).

Thornomys mazama. — Like T. idahoensis, the
phallus of T. mazama (Fig. 5) is considered long
and slender. The distal tract of T. mazama is the
longest of all geomyid species, with a length ranging
from 27.2 to 34. 1 mm among six adult specimens
examined. The length is about 10 to 12 times greater
than the width. The glans is between one-half and
two-thirds the length of the distal tract. The sides
of the glans are essentially straight and parallel.
Similar to T. idahoensis, this species lacks many of
the features found on other geomyid rodents. For
instance, T. mazama does not possess a distinct
collar, midventra! raphae, or middorsal groove, and
the constriction is not distinct. A dorsal protuber-
ance may occur apically. The urethral processes are
unique from other geomyid species in that they tend
to be a continuation and a part of the ventral margin
of the urethra! opening. The protractile tip is shorter
proportionally than that of T. idahoensis.

Epidermal structures are sparsely and erratically
distributed from the urethral opening proximaliy for
half the length of the glans. The structures, having
single proximaliy oriented projections, are small,
uniform in size, and have an irregular pattern.

18

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

Fig. 4. — Phallus (A, B, C) of subadult Thomomys iJahoensis from Idaho: Bonneville Co., Vi mi W Idaho Falls (TTU 19393), and
baculum (D, E; same scale) of adult T. idahoensis from Idaho: Bonneville Co., 0.4 mi N, 1.0 mi W Swan Valley (NMSU 4812),
illustrated as in Fig. I.

Although the baculum of T. mazaina (Eig. 5) re-
sembles that of T. idahoensis, it differs by being
larger and having a slightly more conspicuous base.
Basically the baculum is long, slender, and has a
weakly-defined tip and base. The length of the bac-
ulum among adult individuals examined ranged
from 24.9 to 32.9 mm. Other investigators (Ingles,
1965; Johnson and Benson, 1960; Long and Erank,
1968; Thaeler, 1968) have placed the lower limits of
bacular length at 21 and 22 mm. Although there is
a tendency for the base of the baculum to be some-
what flattened dorsoventrally, the width of the bac-
ulum may equal or be slightly less than the height
of the baculum base.

Thomomys mazama is unique from other geo-
myid species by having the longest distal tract and
baculum, both absolutely and in proportion to the
body size. An examination of six adult individuals
showed a mean ratio of 1.2 with a range from 1.1
to 1.3, and a mean ratio of 1.3, with a range from
1.1 to 1.5, when comparing condylobasal length to
length of distal tract and length of baculum, respec-
tively. The ratio between the length of the distal
tract and length of baculum averaged 1 .0 and ranged
from 1.0 to 1.2. (Every specimen had a distal tract
that was longer than the baculum thus resulting in
ratios greater than 1.0. Because the difference be-
tween both dimensions was often only detectable

1982

WILLIAMS— GEOMYID PHALLI

Fig. 5. — Phallus and baculum of Thomomys mazama niazama from California: Siskiyou Co., 3.6 mi S,
129482), illustrated as in Fig. 1 (scale is three millimeters).

D E

1 .2 mi E Bartle, 4,250 ft (MVZ

20

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY NO. 20

Fig. 6. — Phallus and baculum of Thonioniys monlicola from California: Siskiyou Co., V2 mi N Mount Shasta City, 3,600 ft (TTU 19407),
illustrated as in Fig. 1 .

beyond a single decimal position, procedures of
rounding-off the ratio often resulted in a value of
1.0.)

Specimens examined. — Total (23).

T. m. mazama (23). — California: Shasta Co.: Dickson Flat,
4,200 ft, 4 (MVZ); Red Mountain, 5,400 ft, I (MVZ); Siskiyou
Co.: Colby Meadow, 3.6 mi S, 1.2 mi E Bartle, 3 (MVZ);
Lombardi Ranch, M mi N, IV2 mi W Mount Shasta City, 3,525
ft, 3 (TTU); Lombardi Ranch, 1.5 mi WNW Mt. Shasta, 3,525
ft, 2 (MVZ); 1 mi W Mount Shasta City, 9 (CM). Oregon;
Jackson Co.: 20 mi E Ashland, I (TTU).

Thomomys monticola. — The long and narrow-
type phallus of T. monticolo (Eig. 6) has a length
six to seven times its width, with the length of the
glans being about half the length of the distal tract.
The length of the distal tract of nine adults exam-
ined ranged from 15.0 to 16.4 mm. Erom all aspects
the sides of the glans are more or less straight and

tend to expand apically. Thomomys monticola has
a distinct collar that may be extended more distally
on the dorsal side. A pair of well-developed urethral
processes are exposed and may connect at the low-
er end of their common side. The midventral raphe,
if present, is not distinct. Thomomys monticola
possesses a middorsal groove and a pair of dorsal
protuberances, but neither feature is well-devel-
oped. The constriction is well-defined.

Like other members of the genus Thomomys , the
epidermal structures have a single proximally ori-
ented projection. The structures are irregular in
size, fairly uniform in pattern, and distributed be-
tween the collar and constriction and on the dorsal
side of the protractile tip.

The baculum (Eig. 6) is distally tapered, possess-
es a distinct bulbous basal region, and is well
curved. The length of the baculum of nine adults

1982

WILLIAMS— GEOMYID PHALLI

21

Fig. 7. — Phallus and baculum of Thomomys talpoides fossor from New Mexico: Taos Co., Wi mi SE Taos Ski Area (TTU 9292),
illustrated as in Fig. 1.

examined ranged from 13.8 to 15.2 mm. Descrip-
tions and dimensions of the baculum of T. monti-
cola closely agree with those of Long and Frank
(1968) and Thaeler (1968). However, Ingles (1965)
suggests the bacular length may range from 12 to 17
mm. There was no trend found between the width
and height of the base of the baculum; the dimen-
sions of one character may be greater than, less
than, or equal to the other.

Among five adult individuals the ratio of the con-
dylobasal length to the length of the distal tract and
length of baculum averaged 2.2 (range, 2.1 to 2.3)
and 2.4 (range, 2.3 to 2.5), respectively. In each
case the ratio of the length of the distal tract to the
length of the baculum was 1.1.

Specimens examined. — Total (15).

T. monticola (15). — California: Madera Co.: Agnew Meadow,
9.5 mi W Mammoth Lakes, 8,300 ft, 1 (MVZ); Nevada Co.:
T17N RilE Sec 25, Bear Valley, 4,520 ft, 3 (MVZ); Sagehen
Creek, 3 mi NW Hobart Mills, 6,400 ft, 5 (MVZ); Alder Creek,
4 mi NNE Truckee, 5,700 ft, 1 (MVZ); Siskiyou Co.: 3 mi S
McCloud, 2 (TCWC); V2 mi N Mount Shasta City, 3,600 ft, 1
(TTU); Mount Shasta City, 2 (CM).

Thomomys talpoides. — The phallus of T. tal-
poides (Fig. 7) is considered to be long and narrow
like that of T. idahoensis, T. mazama, and T. mon-
ticola although its size is quite variable compared
to other species of Thomomys. The distal tract in
T. talpoides is about eight to nine times longer than
its width. The length of the glans is less than half

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

22

the length of the distal tract which ranges from 14.0
to 26.2 mm among adult specimens examined. The
phallus of T. talpoides more closely resembles that
of T. monticola than any other species of Thomomys
because of its similar size, shape, and features. The
sides of the glans are essentially parallel but curve
toward the dorsal side to conform with the shape
of the baculum. In the region of the collar, the sides
expand slightly. The glans of T. talpoides is basi-
cally featureless by lacking a well-developed mid-
ventral raphe, middorsal groove, and dorsal protu-
berances. However, the collar and constriction of
T. talpoides tend to be somewhat more distinct.
The urethral processes are long and narrow.

The epidermal structures have a single proxi-
mally oriented projection and are relatively small
and uniform in size and pattern. The epidermal
structures are distributed between the collar region
and the constriction.

The baculum (Eig. 7) of T. talpoides, which is
also variable within the species, consists primarily
of a simple, slightly curved shaft with the tip and
base being barely discernable. The height and width
of the bacular base may be greater than, less than,
or equal to the other. The length of the baculum
among adult specimens examined ranged from 12.3
to 22.6 mm. Data presented by other investigators
(Burt, 1960; Ingles, 1965; Johnson and Benson,
1960; Long, 1964; Long and Erank, 1968; Thaeler,
1968) indicate considerable variation of bacula
among geographical races of T. talpoides, with
bacular lengths ranging from 10.4 to 31.0 mm.
Although variation occurs within the species, data
presented in this study, combined with data from
other published sources, indicate broad geographi-
cal trends in baculum sizes. Eor instance, subspe-
cies reported having smaller bacula tend to be clus-
tered on the western part of the range, whereas the
races reported to have larger bacula occur primarily
in the eastern half of the range.

The ratios of the condylobasal length to the
length of the distal tract and length of baculum
ranged from 1.4 to 2.5 and 1.6 to 2.9, respectively.
The ratio between the length of the distal tract and
length of baculum ranged from 1.0 to 1.1.

Specimens examined. — Total (35).

T. t. hridgeri (16). — Idaho: Bear Lake Co.: Home Canyon, 5 mi

NE Montepelier, 1 (TCWC); V2 mi S, 4V4 mi E Palisades, 6,400

ft 15 (CM).

T. t. hallatus (3). — Montana: Caster Co.: 10 mi S, 3 mi E

Locate, 3 (CM).

T. t. devexus (1). — Washington: Grant Co.: 5 mi N Coulee
City, 1,600 m, 1 (TCWC).

T. t.fossor (3). — Colorado: Chaffee Co.: Trout Creek, 8V2 mi
NE Buena Vista, 1 (TCWC). New Mexico: Taos Co.: IV2 mi
SE Taos ski area, 1 (TTU); beaver pond SE Taos ski area, 1
(TTU).

T. t. fascus (1). — Idaho: Blaine Co.: Alturas Lake, 7,000 ft, 1
(TCWC).

T. t. quadratus (3). — Idaho: Owyhee Co.: 12 mi S, 7 mi W
Murphy, 3 (CM).

T. t. rostralis (2). — Wyoming: Albany Co.: 5 mi N Laramie, 1
(TTU); 3.5 mi N Laramie, 1 (TTU).

T. t. rufescens (3). — Manitoba: 11 mi E Dauphin, 1 (TCWC).
North Dakota: Morton Co.: 8 mi S, 15 mi W Mandan, 1
(CM); Williams Co.: 1 mi N Buford, 2,100 ft, 1 (CM).

T. t. saturatus (1). — Idaho; Bonner Co.: 2'/i mi N Nordman,
2,700 ft, 1 (TCWC).

T. t. talpoides (1). — Saskatchewan: 13 mi S Blaine Lake, 1,600
ft, 1 (TCWC).

T. t. tenellas (1). — Colorado: Rio Blanco Co.: 22 mi SW Meek-
er, 6,200 ft, 1 (TCWC).

Thomomys townsendii. — The shape and features
of the phallus of T. townsendii (Eig. 8) are most
similar to T. bottae and T. itmbrinns with the pri-
mary difference being a larger size in T. townsendii.
The length of the distal tract, which ranged from
13.0 to 15.5 mm among adult specimens examined,
is four to five times greater than the width. The
length of the glans is about half the length of the
distal tract. The sides of the glans expand to the
collar and are recurved. The collar is distinct from
all aspects. The pair of urethral processes are ex-
posed and well-developed. Standard features, in-
cluding the midventral raphae, middorsal groove,
dorsal protuberances, and distinct constriction, are
present. All of these features, except perhaps the
dorsal protuberances, are well defined. The mid-
dorsal groove extends from about the middle of the
protractile tip to about halfway between the collar
and constriction.

The epidermal structures have single, relatively
large, proximally oriented projections. The pattern
and size is fairly uniform. The structures occur be-
tween the collar and constriction and on the dorsal
side of the protractile tip.

The baculum of T. townsendii (Fig. 8) is strongly
curved with the dorsal side being concave as in oth-
er geomyid bacula. The baculum has a distinct tip
and bulbous base. The length of the baculum ranged
from 12.2 to 13.9 mm among adult individuals ex-
amined. Ingles (1965) and Thaeler (1968) reported
bacular lengths of T. townsendii to be as short as
9.2 mm and 10.7 mm, respectively. In adult speci-
mens of 7. t. baclimani and T. t. similis examined.

Fig. 8. — Phallus and baculum of Thomomys townsendii townsendii from Idaho: Canyon Co., 1 mi N Caldwell (TTU 19420), illustrated
as in Fig. 1.

the width of the baculum base was always less than
the height of the base; the opposite was found in T.
t. townsendii.

The ratios of condylobasal length to the length of
the distal tract ranged from 3. 1 to 3.6; condylobasal
length to length of baculum ranged from 3.6 to 3.9;
length of the distal tract to the length of the baculum
was I.l in ail cases.

Specimens examined. — Total (9)

T. t. bachmani (2). — Oregon: Harney Co.: Harney Lake Dunes,
Malheur National Wildlife Refuge, T26S R30E Sec 28, 2 (CM).
T. t. similis (3). — Idaho: Bingham Co.: IVz mi S Springfield,
4,400 ft, 3 (CM).

T. t. townsendii (4). — Idaho: Canyon Co.: 1 mi N Caldwell, 4
(TTU).

Thomomys umbrinus. — The shape and size of the
phallus of T. umbrinus (Fig. 9) is most similar to T.
bottae. Because the available material for both T.
boitae and T. umbrinus is restricted to limited areas
of the extensive geographic ranges of both species,
further study of geographic and nongeographic vari-

ation is needed before specific differences can be
defined.

The length of the phallus of T. umbrinus is about
four times greater than the width. The glans is about
half the length of the distal tract. The length of the
distal tract ranged from 10.3 to 12.3 mm among
adult individuals examined. Viewed dorsoventrally,
the sides of the glans may be straight or recurved
but tend to be more or less parallel. The collar re-
gion is expanded and distinct. The urethral pro-
cesses are well developed but connected for most
of the length along the common side. Other features
noted include a well-defined midventral raphe,
poorly-developed middorsal groove, and the ab-
sence of dorsal protuberances.

The epidermal structures of T. umbrinus have the
characteristic single, proximally oriented projec-
tion, common to other members of the genus. The
structures are relatively large and uniform in size
and pattern.

The baculum of T. umbrinus (Fig. 9) is similar to
that of T. bottae but tends to be shorter. A definite

Fig. 9. — Phallus and baculum of Thomomys umhhnus from Tlaxcala: La Malinche Mt., 8 km S, 7 km W Huamantla (CM 55893),
illustrated as in Fig. 1.

base and tip are present; the shaft is curved. The
length of the baculum ranged from 8.0 to 11.3 mm
among adult specimens examined. In most individ-
uals examined the width and height of the base of
the baculum were very close to being equal. Like
T. hottae and T. talpoides, the baculum of T. iim-
briniis demonstrates geographical variability. Bac-
ula from Arizona reported by Burt (1960), Hoff-
meister (1969), and Patton (1973) were smaller (9.0
to 10.2 mm), had a bulbous base, and had a well
curved shaft. Data presented here suggest these
characters are variable when other subspecies,
throughout the range of T. umhrinus, are consid-
ered.

The ratios of condylobasal length to the length of
the distal tract and length of the baculum ranged
from 3.2 to 3.7 and 3.6 to 4.5, respectively. The
ratio between the length of the distal tract and the
length of the baculum ranged from 1.0 to 1.3.

Specimens examined. — Total (28).

T. II. airiageiiis (1). — San Luis Potosi: 4 mi E Villa Arriaga,

1 (MVZ).

T. II. intermedins (8). — Arizona: Santa Cruz Co.: Gardner Can-
yon, 3.0 mi N, 3.7 mi W Sonoita, 3 (TTLf); Patagonia Mts.,
Italian Canyon 1 (MVZ); Patagonia Mts., Sycamore Canyon,
4,500 ft, 4 (MVZ).

T. u.juntae (I). — Chihuahua: I mi S Delicias, 1 (TTU).

T. II. madrensis (2). — Chihuahua: 2.4 mi NE Colonia Garcia,
I (MVZ): 1 mi W Colonia Garcia, 1 (MVZ).

T. II. sheldoni (4). — Durango: 3 mi NW El Salto, 4 (MVZ).

T. II. ssp. (12). — Tlaxcala: 10 km N, 9 km E Apizaco, 2,500 m,
1 (CM); 9 km N, 7 km E Apizaco, 2,500 m, 3 (CM); 8 km S,

7 km W Calpulalpan, 2,900 m, 4 (CM); La Malinche Mt., 7 km
S, 11 km W Huamantla, 4,800 m, I (CM); La Malinche Mt.,

8 km S, 7 km W Huamantla, 4,000 m, 2 (CM); 4 km N, 19 km
W Tlaxcala, 2,200 m, 1 (CM).

Statistics

Age variation in phallic characters of Thomomys
was observed in this study and has been reported
by other investigators (Burt, 1960; Johnson and
Benson, 1960). However, proper analysis and doc-
umentation of this form of variation is beyond the
scope of the present study. Table 1 provides stan-
dard statistics (sample size, mean, range, standard
error, and coefficient of variation) for adult speci-
mens of Thomomys examined.

Examination of individual variation was con-
ducted on four species of Thomomys , thus provid-
ing a general assessment of the individual variation
in the genus. Each sample included individuals tak-
en from a restricted geographical area that would
be assumed to be part of a continuous population.

1982

WILLIAMS— GEO'MYID PHALLI

25

The taxa examined for individual variation were T.
bottae opidentus, T. bulbivorous, T. mazama ma-
zama, and T. talpoides bridgeri. Coefficients of
variation in these samples ranged from 1.0 to 25.4
for phallic and bacular measurements, with the
above taxa averaging 14.3, 11.4, 15.0, and 9.2, re-
spectively. Generally, the length of the distal tract
and the length of the baculum had the lowest val-
ues, and the length of the protractile tip and the
width of the glans at the collar had the highest val-
ues. The characters having the lowest and highest
coefficients of variation for T. bottae were length
of distal tract (1.0) and width of glans across the
base (24.8); for T. bulbivorous , length of baculum
(2.9) and length of protractile tip (22.0); for T. ma-
zama, length of glans (7.6) and length of protractile
tip (25.4); for T. talpoides, length of distal tract (3.9)
and length of protractile tip (14.1). Compared to
investigations of the baculum by Long and Frank
(1968) coefficients of variation acquired in this
study (Table 1) were high for T. mazama and T.
monticola, low for T. townsendii, and overlapping
in T. bottae and T. talpoides.

Sample sizes and limited geographical coverage
of specimens representing species of Thomomys
precludes any meaningful analysis of geographic
variation within a species from being conducted.
However, examination and comparison of measure-
ments of individuals strongly suggests that varia-
tion, at least in size, does occur between geograph-
ical races, or subspecies, of Thomomys . In four out
of eight characters T. bottae and T. umbrimis had
subspecies with nonoverlapping measurements.
Thomomys talpoides and T. townsendii had sub-
species that did not overlap in five and six charac-
ters, respectively. The greatest variation between
geographical races was observed in T. talpoides.

Examination of variation among species of Tho-
momys was restricted to multivariate analysis, us-
ing sample means with the MINT statistical com-
puter program. Univariate analysis was not used
because the sample sizes of most geographical races
v/ere limited, and combining samples of the same
species could lead to a biased representation of the
species, particularly if geographic variation occurs
within the species. The samples of Thomomys used
in the multivariate analysis, with corresponding
number for reference purposes, are six subspecies
of T. bottae {actuosus, 1; connectens, 2; limitaris,
3; lirnpiae, 4; opulentus, 5; riddosae, 6), T. bulbi-
vorous (7), T. clusius (8), T. mazama mazama (9),
T. monticola (10), ten subspecies of T. talpoides

(bridgeri, U;fossor, 12; fuscus, 13; quadratus, 14;
rostralis, 15; rufescens, 16; saturatus, 17; tal-
poides, 18; tenellus, 19), three subspecies of T.
townsendii (bachmani, 20; similis, 21; townsendii,
22), and four subspecies of T. umbrinus (interme-
dins, 23; juntae, 24; sheldoni, 25; ssp. from Tiax-
cala, 26).

The distance phenogram produced by the MINT
program is illustrated in Fig. 10. The cophenetic
correlation value of this phenogram was 0.820. Al-
though clustering of samples shows mixing between
species, some trends are evident. T. mazama
formed a group distinct from all other taxa. The
next two groups to split away from the remaining
samples were T. bulbivorous and T. u. juntae which
clustered together but remained distinct from each
other. The next subdivisions isolate T. u. > erme-
dius and then break down into two large clusters.
One cluster consists of seven subspecies of T. tal-
poides {bridgeri, rufescens, fossor, saturatus, tal-
poides, tenellus, and rostralis). The second cluster
subdivides to form a cluster consisting of the three
subspecies of r. townsendii {similis, bachmani, and
townsendii) and T. b. connectens. The other cluster
consists of a mosaic of taxonomic groups. In the
remaining clusters T. b. actuosus, T. b. opulentus,
T. b. riddosae, and T. t. fuscus form one group; T.
b. limitaris, T. monticola, T. t. quadratus, T. b.
lirnpiae, and T. clusius form one group; T. u. shel-
doni and T. umbrinus from Tlaxcaia form the final
group.

The first three principal components extracted
from the matrix of correlation among characters is
illustrated in Fig. 11. The first vector separates
three general groups. The first group consists of T.
mazama (9) which is plotted to the far left. The
second grouping includes T. monticola (10) and the
samples of T. talpoides (il-19). In the second
grouping the westernmost subspecies of T. tal-
poides, represented by T. t. fuscus and T. t. qua-
dratus, cluster together on the right side of the
group with T. monticola. T. monticola falls within
the range of T. talpoides with all three vectors. The
third group consists of the samples of T. bottae
(1-6), T. bulbivorous (7), T. clusius (8), T. town-
sendii (20-22), and T. umbrinus (23-26). In the third
group T. bottae and T. clusius are separated from
T. bulbivorous and T. umbrinus with the first vec-
tor. T. bulbivorous and T. clusius are separated
with the second and third vectors, respectively. T.
townsendii overlaps with T. bottae and T. umbrinus
with all three vectors.

26

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

T .

bottae actuosus

I.

bottae ODulentus

T.

bottae ru

i dosae

I.

talDO i des

f uscus

I.

bottae limitaris

I.

mont i cola

I.

taloo i des

Quadr atus

1.

bottae 1 i 1

mp i ae

1.

clus i us

T.

umbr i nus

Sheldon i

1.

umbr i nu s

(Tlaxcala)

1.

bottae connectens

T.

townsend i

1 s I m 1 1 1 s

T.

townsend i

i bachmani

J.

townsend i

i townsend i i

I.

talpo i des

br i doer i

T.

talpo 1 des

ruf escens

1.

talpo i des

f ossor

I.

taloo i des

sat ur at u s

T.

taloo i des

taloo ides

I.

talpo i des

tenellus

T.

taloo 1 des

r ostr al i s

T.

umbr i nus

i ntermed i us

1.

bulbivorus

1.

umbr i nus

iuntae

T.

mazama mazama

3.0

2.0

1 .0

0.0

Fig. 10. — Distance phenogram of 26 samples of Thomomys resulting from clustering by unweighted pair-group method using arithmetic
averages (UPGMA). The cophenetic correlation coefficient for the phenogram is 0.820.

The amount of phenetic variation explained by
components I, II, and III, is 39.2%, 31.7%, and
11.1%, respectively. Results of principal compo-
nent analyses, showing the influence of each char-
acter for the first three components, are given in
Table 2. For component I the length of the distal
tract, length of glans, and the length of the baculum
are the most heavily weighted. The width of the
glans across the base, width of bacular base, and
height of bacular base are the most heavily weight-
ed characters in component II. For component III

the length of the protractile tip is the most heavily
weighted character.

Orthogeomys

Description

Orthogeomys hispidiis. — In this study the genus
Orthogeomys is only represented by samples of O.
hispidus. The length of the phallus of O. hispidus
(Fig. 12) is about four times longer than the width,
with the length of the glans being slightly more than
half the length of the distal tract. The length of the
distal tract of two adult specimens examined was

1982

WILLIAMS— GEOMYID PHALLI

27

Fig. li. — Three-dimensional projection of 26 samples of Thomomys {T. bottae, 1-6; T. bulbivoroiis, 7; T. cliisins, 8; T. mazama, 9;
T. monticola. 10; T. talpoides, 11-19; T. townsendii, 20-22; T. umbrinus, 23-26) onto the first three principal components based upon
a matrix of correlation among one cranial, five phallic, and three bacular measurements. Components I and II are indicated in the plots
and component 111 is represented by height.

15.5 and 15.7 mm. The sides of the glans, from all
aspects, are more or less straight and parallel, and
expanded slightly at the collar. The collar is distinct
on all sides except in the middorsal region. On the
dorsal side the collar is extended more distally than
on the other sides. The urethral processes are small
and partially concealed by the collar. The midven-
tral raphe is well-developed. The dorsal side is fea-
tured with a single prominant protuberance which
may have a shallow middorsal depression that con-
stitutes the middorsal groove found in other geo-
myid species. The dorsal protuberance is evident

from the protractile tip to the lower portion of the
glans where it fades out.

The epidermal structures of O. hispklns are large
and consist of a single (occasionally two) proxi-
mally oriented projection. The size and pattern of
the epidermal structures is fairly uniform. The
structures are restricted to the area between the
constriction and collar.

The baculum of O. hispUiiis (Fig. 12) is massive
compared to bacula of most other geomyid rodents.
The base is somewhat bulbous and compressed dor-
soventrally. The base gradually tapers to the thick

28

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

D

NO. 20

E

Fig, 12. — Phallus and baculum of Orthogeomys hispidus negatus from Tamaulipas: 67 km S Ciudad Victoria (CM 55787), illustrated
as in Fig. 1.

shaft which terminates with a weakly defined tip.
The entire baculum is well-curved. The length of
the baculum of two adult specimens examined was
12.8 and 13.7 mm.

Eor two adult specimens, the ratios of the con-
dylobasal length to the length of the distal tract
were 3.9 and 4.1; ratios of the condylobasal length
to the length of baculum were 4.7 and 4.8; ratios of
the length of the distal tract to the length of baculum
were 1.1 and 1.2.

Samples of O. hispidus are inadequate for statis-

Table 2. — Factor matrix from correlation among nine characters
of eight species (26 samples o/Thomomys examined.

Character

Component I

Component II Component III

Condylobasal length

0.469

-0.547

0.287

Length of distal tract

-0.930

-0.300

0.180

Length of glans

-0.878

-0.341

0.289

Length of protractile tip

-0.544

-0.308

-0.681

Width of glans across collar

0.577

-0.622

-0.043

Width of glans across base

0.279

-0.789

0.389

Length of baculum

-0.933

-0.293

0.139

Width of bacular base

-0.045

-0.790

-0.362

Height of bacular base

0.300

-0.729

-0.185

tical analyses. Table 3 provides measurements of
specimens of O. hispidus examined in this study.

Specimens examined. — Total (8).

O. h. chiapensis (1). — Chiapas: 7.5 mi (by road) NW Pueblo
Nuevo, 6,000 ft, 1 (KU).

O. h. isthmicus (1). — Veracruz: 5 mi SE Lerdo de Tejada, 1
(KU).

O. h. negatus (4). — Tamaulipas: 45 mi S Cd. Victoria, 2 (TTU);

67 km S Cd. Victoria, 2(1 CM, 1 TTU).

O. h. torridus (1). — Veracruz: 1214 mi N Tihuatlan, 300 ft, 1
(KU).

O. h. yucatanensis (1). — Quintana Roo: 6.5 km NE Playa del
Carmen, 1 (TTU).

Zygogeomys

Description

Zygogeomys trichopus. — In this study the phal-
lus of Zygogeomys (Fig. 13) is represented by a
juvenile and a young adult specimen. Because the
basioccipital and basisphenoid are not well fused in
the young adult specimen, the following comments
based on that single specimen may be subject to
some discrepancies due to age variation. Measure-
ments for the specimen are given in Table 3.

1982

WILLIAMS— GEOMYID PHALLI

D

29

E

Fig. 13. — Phallus and baculum of Zygogeomys trichopus triciiopiis from Michoacan: 5.3 km E Tacitaro (CM 55902), illustrated as in
Fig. 1.

The distal tract is four to five times longer than
wide, and measures 14.7 mm. The glans is less than
half the length of the distal tract and is somewhat
narrower when viewed dorsoventrally than when
viewed from a lateral aspect. From a lateral aspect
the ventral side is straight, whereas the dorsal side
is recurved with the widest part being the collar.
Viewed dorsoventrally the lateral sides show a
bulge in the midsection of the glans and a slight

expansion at the collar. The collar is distinct from
all aspects. The pair of urethral processes and the
protractile tip are reduced in size compared to other
members of the family. The urethral processes are
almost hidden by the collar. Other features include
a well-defined constriction, prominent midventral
raphe, a middorsal groove, and a pair of well-de-
veloped dorsal protuberances. The middorsal
groove in this specimen begins at the lower portion

Table 3. — Dimensions of external phallic characters, bacular characters, condylobasal length, and hind foot length of individuals of
Orthogeomys and Zygogeomys examined. Young adult specimens had well-developed sagittal and lamhdoidal crests and kicked

complete fusion between the hasioccipital and basisphenoid.

Taxon

Catalog

number

Age

Length

of

distal

tract

Length

of

glans

Length
of pro-
tractile
tip

Width
of glans
across
collar

Width
of glans
across
base

Length

of

baculum

Width

of

bacular

base

Height

of

bacular

base

Condylo-

basal

length

Length
of hind
foot

Orthogeomys hispidus
isthmicus

KU 67625

Young Adult

16.5

9.4

3.5

14.6

2.5

1.3

62.6

52

Orthogeomys hispidus negatus

CM 55787

Adult

15.5

8.7

3.5

5.0

3.4

13.7

2.7

2.0

63.9

49

Orthogeomys hispidus negatus

TTU 10248

Adult

15.7

—

—

—

—

12.8

2.1

1.5

61.9

41

Orthogeomys hispidus negatus

TTU 14339

Young Adult

14.8

7.5

2.9

3.6

3.5

11.7

1.9

1.3

54.7

39

Orthogeomys hispidus torridus

KU 88521

Young Adult

15.7

7.9

2.7

—

—

13.1

2.7

1.3

61.3

46

Zygogeomys trichopus
trichopus

CM 55902

Young Adult

14.7

7.0

2.1

3.3

2.9

12.7

2.4

1.7

58.3

46

30

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

of the protractile tip, crosses the collar and contin-
ues for half the length of the glans where it fades
out. The dorsal protuberances are more or less dis-
tinct for the entire distance between the collar and
constriction.

The epidermal structures are large and disk-
shaped with three proximally oriented projections
lying side-by-side. The epidermal structures are
somewhat variable in size and may be restricted to
the region of the glans between the collar and con-
striction.

The baculum of the specimen examined (Fig. 13)
is 12.7 mm in length and has a distinct base and tip.
The shaft is constricted on both ends where it joins
the base and tip. A specimen examined by Burt
( 1960) was tapered between the base and shaft. The
baculum is well curved and possesses a small
depression on the ventral side at the junction of the
base and shaft. The base is dorsoventrally com-
pressed. Measurements of the baculum are similar
to those given by Burt (1960).

The ratios of the condylobasal length to the
length of the distal tract and length of baculum were
4.0 and 4.6, respectively. The ratio of the length of
the distal tract to the length of the baculum was 1.1.

Specimens examined. — Total (2).

Z. /. trichopus (2). — Michoacan: 5.3 km E Tancitaro, 1 (CM);

7 km E Tancitaro, 1 (TTU).

Geomys

Description

Unlike Thomomys, species belonging to the ge-
nus Geomys show very little variation in the size
and shape of the phallus. The length of the distal
tract ranged from 11.7 to 16.8 mm in adult individ-
uals examined. The size relationship between the
length of the distal tract and body size of individuals
is much more conservative than in Thomomys, with
ratios of the condylobasal length to the length of
the distal tract ranging from 2.8 to 4.5. In all sam-
ples of Geomys examined the length of the distal
tract was three to four times greater than the width
and the short, broad glans was about half the length
of the distal tract. The sides of the glans are usually
recurved, terminating apically with a definite and
often elaborate collar. The collar is distinct from all
aspects. The pair of urethral processes vary in size
and shape. Some individuals have urethral process-
es that may be folded and/or have serrated margins.
Although the general shape of the glans appears
complex compared to other geomyid rodents, fea-

tures such as the midventral raphe, middorsal
groove, and dorsal protuberances are usually not
well-developed and often not distinct. The connec-
tion of the prepuce and glans is always distinct and
often extended more apically on the dorsal side.

The epidermal structures of Geomys are similar
to those of Zygogeornys in that each structure pos-
sesses multiple projections instead of a single pro-
jection as in Thomomys and Orthogeomys. The
number of projections on the epidermal structures
in Geomys is usually three but the number may
vary. Epidermal structures occur between the col-
lar and constriction and on the dorsal side of the
protractile tip in all species of Geomys.

The baculum of all samples of Geomys examined
consisted of a gently curving shaft which terminated
in a weakly-defined tip and a weakly to strongly-
defined base. The dorsal side is concave; the ventral
side, although convex, often has a slight depression
just anterior to the basal region. The baculum of
Geomys shows very little variation in relative size
compared to the size of the body. In adult speci-
mens examined the length of the baculum ranged
from 8.3 to 12.8 mm. The ratios of the condylobasal
length to the length of the baculum and length of
the distal tract to the length of the baculum ranged
from 3.7 to 6.0 and 1.1 to 1.6, respectively. Like
other geomyid species, the baculum is positioned
between the central axis and the dorsal side of the
phallus. The most discernable difference in the bac-
ulum of species of Geomys is variation in the shape
and size of the base.

The following are detailed descriptions of the six
recognized species of Geomys — G. arenariiis, G.
attwateri, G. bur sarins, G. personatus, G. pinetis,
and G. tropicalis. Table 4 provides measurements
that apply to each description.

Geomys arenariiis. — The phallus of G. arenariiis
(Fig. 14) is typical of the genus. The length of the
distal tract ranged from 12.7 to 13.4 mm in five adult
specimens examined. The sides of the glans tend to
be strongly recurved. The collar is always expanded
and distinct. Elaborate folds or convolutions along
the collar are typical in G. arenariiis. The urethral
processes are well developed. Other features such
as the midventral raphe, middorsal groove, and dor-
sal protuberances are often barely discernable. The
middorsal groove is mainly evident as it passes
through the collar.

Epidermal structures usually have three proxi-
mally oriented projections. However, some struc-
tures may have more projections. The shape and

1982

WILLIAMS— GEOMYID PHALLI

31

Table 4. — Standard statistics for adult specimens representing 17 samples of the five species of Geomys.

Taxon

N

Mean

(Range) ± 2 SE

cv

Length of distal tract

Geomys arenarius arenarius

5

13.0

(12.7-13.4) ± 0.25

2.1

Geomys attwateri

4

13.6

(13.4-13.7) ± 0.15

1.1

Geomys bursarius jugosicidaris

2

13.3

(12.7-13.9) ± 1.20

6.4

Geomys bursarius knoxjonesi

12

13.8

(12.7-15.9) ± 0.57

7.1

Geomys bursarius llanensis

2

13.3

(12.7-13.8) ± 1.10

5.8

Geomys bursarius iutescens

3

14.0

(13.7-14.2) ± 0.33

2.1

Geomys bursarius major

7

14.1

(12.5-15.8) ± 0.78

7.3

Geomys bursarius sagitallis

2

15.3

(15.0-15.5) ± 0.50

2.3

Geomys bursarius texensis

1

13.5

Geomys personatus davisi

2

14.3

(14.0_14.7) ± 0.70

3.5

Geomys personatus maritimus

2

14.9

(14.9-14.9) ± 0.00

0.0

Geomys personatus megapotamus

7

15.8

(14.3-16.8) ± 0.63

5.3

Geomys personatus personatus

4

16.9

(15.6-19.1) ± 1.60

9.7

Geomys personatus streckeri

1

13.0

Geomys pinetis fontanelus

Geomys pinetis pinetis

7

12.8

(11.7-14.5) ± 0.72

7.4

Geomys tropicalis

4

13.8

(13.2-14.4) ± 0.61

4.4

Length of glans

Geomys arenarius arenarius

5

7.5

(6.5-8. 4) ± 0.63

9.4

Geomys attwateri

4

7.1

(6.9-7.51 ± 0.25

3.5

Geomys bursarius jugosicidaris

2

7.3

(6.5-8. 0) ± 1.50

14.5

Geomys bursarius knoxjonesi

12

7.2

(6.7-9.0) ± 0.35

8.3

Geomys bursarius llanensis

2

6.9

(6.8-6.91 ± 0.10

1.0

Geomys bursarius iutescens

3

1.4

(7.2-7.61 ± 0.24

2.8

Geomys bursarius major

7

7.4

(6.5-8.31 ± 0.45

8.!

Geomys bursarius sagitallis

2

7.5

(7.3-7.61 ± 0.30

2.8

Geomys bursarius texensis

1

6.7

Geomys personatus davisi

2

7.5

(7. 1-7.8) ± 0.70

6.6

Geomys personatus maritimus

2

7.9

(7.9-7.91 ± 0.00

0.0

Geomys personatus megapotamus

7

8.2

(6.9-9.31 ± 0.59

9.6

Geomys personatus personatus

4

8.7

(8.5-8.91 ± 0.17

2.0

Geomys personatus streckeri

1

6.8

Geomys pinetis fontanelus

Geomys pinetis pinetis

7

6.3

(5.7-6.81 ± 0.28

5.9

Geomys tropicalis

4

6.6

(6.3-7. 0) ± 0.30

4.5

Length of protractile tip

Geomys arenarius arenarius

4

2.7

(2.2-3. 1) ± 0.39

14.3

Geomys attwateri

4

2.5

(1.9-2. 9) ± 0.42

16.8

Geomys bursarius jugosicidaris

2

3.3

(2.9-3.71 ± 0.80

17.1

Geomys bursarius knoxjonesi

12

2.6

(2.0-3.0) ± 0.16

10.9

Geomys bursarius llanensis

2

2.5

(2.4-2.61 ± 0.20

5.7

Geomys bursarius Iutescens

3

3.0

(2.7-3.51 ± 0.50

14.5

Geomys bursarius major

6

2.4

(1. 7-3.1) ± 0.47

23.8

Geomys bursarius sagitallis

2

2.5

(2.2-2.91 ± 0.70

19.8

Geomys bursarius texensis

1

2.4

Geomys personatus davisi

2

3.4

(3. 2-3. 6) ± 0.40

8.3

Geomys personatus maritimus

2

3.3

(3. 1-3.6) ± 0.50

10.7

Geomys personatus megapotamus

7

3.4

(2.7-4.21 ± 0.38

15.0

Geomys personatus personatus

3

3.1

(2.8-3.31 ± 0.29

8.1

Geomys personatus streckeri

1

2.8

Geomys pinetis fontanelus

Geomys pinetis pinetis

6

2.2

(1.8-2. 7) ± 0.29

16.1

Geomys tropicalis

4

2.3

(2. 0-2. 7) ± 0.35

15.3

Width of glans across collar

Geomys arenarius arenarius

5

4.4

(3.5-4.81 ± 0.55

22.7

Geomys attwateri

4

3.9

(3.5-4.21 ± 0.38

9.7

32

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

Table 4.

on tinned.

Taxon

N

Mean

(Range) ± 2 SE

cv

Geoniys bnrsarins jiigosicidaris

2

2.9

(2.7-3. 0) ±

0.30

7.3

Geoniys hursarius kno.xjonesi

11

4.3

(3.9-5. 1) ±

0.22

8.6

Geoniys bnrsarins llanensis

2

3.9

(3.9-4.0) ±

0.10

1.8

Geoniys bnrsarins Intescens

3

4.1

(3.8-4.41 ±

0.35

7.5

Geoniys bnrsarins major

7

4.1

(3.2-4.71 ±

0.39

12.7

Geoniys bnrsarins sagilallis

2

3.9

(3.6-4.11 ±

0.50

9.1

Geoniys bnrsarins te.xensis

1

3.7

Geoniys personatns davisi

2

3.4

(3.2-3.61 ±

0.40

8.3

Geoniys personatns niaritinins

2

4.5

(4.5-4.51 ±

0.00

0.0

Geoniys personatns niegapotanins

7

4.3

(3.4-4.71 ±

0.36

11.1

Geoniys personatns personatns

4

5.2

(4.7-5.81 ±

0.45

8.7

Geoniys personatns streckeri

1

2.9

Geoniys pinetis fontanelns

Geoniys pinetis pinetis

7

3.5

(3.1-3.91 ±

0.22

8.2

Geoniys tropicalis

4

3.9

(3.5-4.41 ±

0.40

10.4

Width of glans across base

Geoniys arenarins arenarins

5

3.5

(3. 1^.21 ±

0.45

16.3

Geoniys attwateri

4

2.9

(2.5-3.31 ±

0.34

11.6

Geoniys bnrsarins jiigosicidaris

2

2.9

(2.6-3.11 ±

0.50

12.2

Geoniys bnrsarins kno.xjonesi

11

3.4

(2.9-3.81 ±

0.16

7.7

Geoniys bnrsarins llanensis

2

3.5

(3.5-3.51 ±

0.00

0.0

Geoniys bnrsarins Intescens

3

3.4

(3.1-4.01 ±

0.60

15.3

Geoniys bnrsarins major

7

3.7

(3.4-4.21 ±

0.25

8.8

Geoniys bnrsarins sagitallis

2

3.3

(3.2-3.51 ±

0.30

6.4

Geoniys bnrsarins te.xensis

1

3.3

Geoniys personatns davisi

2

2.5

(2.5-2.51 ±

0.00

0.0

Geoniys personatns niaritinins

2

3.5

(3.1-4.01 ±

0.90

18.2

Geoniys personatns niegapotanins

7

3.7

(3.3-4.31 ±

0.29

10.6

Geoniys personatns personatns

4

4.2

(4.0-4.41 ±

0.17

4.1

Geoniys personatns streckeri

1

2.8

Geoniys pinetis fontanelns

Geoniys pinetis pinetis

7

2.8

(2.3-3.21 ±

0.28

13.5

Geoniys tropicalis

4

3.1

(2.7-3.41 ±

0.31

10.0

Length of hacnlnm

Geoniys arenarins arenarins

5

10.8

(9.9-11.41 ±

0.54

5.6

Geoniys attwateri

4

9.9

(9.3-10.71 ±

0.63

6.4

Geoniys bnrsarins jiigosicidaris

2

10.5

(10.0-10.91 ±

0.90

6.1

Geoniys bnrsarins kno.xjonesi

12

11.3

(10.3-11.91 ±

0.34

5.2

Geoniys bnrsarins llanensis

2

9.7

(9.4-9.91 ±

0.50

3.6

Geoniys bnrsarins Intescens

3

10.9

(9.8-12.11 ±

1.34

10.6

Geoniys bnrsarins major

7

10.9

(10.4-11.51 ±

0.30

3.6

Geoniys bnrsarins sagitallis

2

11.0

(10.8-11.21 ±

0.40

2.6

Geoniys bnrsarins te.xensis

1

9.2

Geoniys personatns davisi

2

11.3

(11.1-11.51 ±

0.40

2.5

Geoniys personatns niaritinins

2

11.5

(10.5-12.61 ±

2.09

12.9

Geoniys personatns niegapotanins

7

11.6

(10.6-12.11 ±

0.46

5.2

Geoniys personatns personatns

4

12.1

(11.2-12.81 ±

0.73

6.0

Geoniys personatns streckeri

1

9.6

Geoniys pinetis fontanelns

1

9.9

Geoniys pinetis pinetis

7

9.3

(8.3-10.51 ±

0.62

8.8

Geoniys tropicalis

4

9.9

(9.3-10.41 ±

0.50

5.0

Width of bacillar base

Geonivs arenarins arenarins

5

1.6

(1.2-1.81 ±

0.21

14.9

Geonivs attwateri

4

2.1

(1.8-2.41 ±

0.26

12.6

Geoniys bnrsarins jiigosicidaris

2

2.3

(2.2-2.31 ±

0.10

3.1

Geoniys bnrsarins kno.xjonesi

12

1.9

(1.6-2.41 ±

0.15

14.1

Geoniys bnrsarins llanensis

2

1.9

(1.8-2.01 ±

0.20

7.4

1982

WILLIAMS— GEOMYID PHALLI

33

Table 4. — Continued.

Taxon

N

Mean

(Range) ± 2 SE

cv

Geoniys hursariiis httescens

3

2.1

(1.8-2. 3) ± 0.55

12.6

Geomys bursarius major

7

2.1

(1.5-2. 8) ± 0.35

21.8

Geomys bursarius sagitallis

2

2.2

(2. 1-2.3) ± 0.20

6.4

Geoniys bursarius texensis

1

2.2

Geomys personatus davisi

2

1.7

(1.5-1. 8) ± 0.30

12.5

Geomys personatus maritimus

2

2.1

(1.8-2. 3) ± 0.50

16.8

Geomys personatus megapotamus

7

1.9

(1. 6-2.0) ± 0.14

10.0

Geomys personatus personatus

4

2.1

(1.9-2. 3) ± 0.17

8.2

Geomys personatus streckeri

!

1.6

Geomys pinetis fontanelus

1

2.4

Geomys pinetis pinetis

7

2.3

(1.9-2. 7) ± 0.22

12.7

Geomys tropicalis

4

1.6

(1. 3-1.9) ± 0.27

17.2

Height of bacillar base

Geomys arenarius arenarius

5

1.6

(1.1-1. 8) ± 0.27

19.0

Geomys attwateri

4

1.8

(1. 7-2.0) ± 0.15

8.3

Geomys bursarius Jugosicularis

2

1.5

(1.5-1. 6) ± 0.10

4.7

Geomys bursarius knoxjonesi

12

1.9

(1.5-2. 3) ± 0.15

13.7

Geomys bursarius llanensis

2

1.2

(1. 1-1.3) ± 0.20

11.8

Geomys bursarius lutescens

3

1.6

(1.4-1. 8) ± 0.23

12.5

Geomys bursarius major

7

1.6

(1.4-2.0) ± 0.16

12.9

Geomys bursarius sagitallis

2

1.9

(1. 7-2.0) ± 0.30

11.2

Geomys bursarius texensis

I

1.6

Geomys personatus davisi

2

1.7

(1.7-1. 8) ± 0.05

4.1

Geomys personatus maritimus

2

2.0

(1.7-2. 3) ± 0.60

21.2

Geomys personatus megapotamus

7

1.8

( 1.5-2. 1) ± 0.17

12.4

Geomys personatus personatus

4

2.0

(1.7-2. 2) ± 0.22

10.8

Geomys personatus streckeri

1

1.3

Geomys pinetis fontanelus

1

1.9

Geomys pinetis pinetis

7

1.3

(1.2-1. 5) ± 0.09

8.7

Geomys tropicalis

4

1.5

(1.3-1. 8) ± 0.22

14.8

Condylobasal length

Geomys arenarius arenarius

5

46.0

(45.4-47.0) ± 0.55

1.3

Geomys attwateri

4

43.2

(41.2-44.4) ± 1.49

3.5

Geomys bursarius jugosicularis

2

44.6

(43.4_45.8) ± 2.41

3.8

Geomys bursarius knoxjonesi

12

43.9

(40.5-45.2) ± 0.79

3.1

Geomys bursarius llanensis

2

42.2

(41.2^3.2) ± 2.01

3.3

Geomys bursarius lutescens

3

48.1

(44.3-51.1) ± 4.00

7.2

Geomys bursarius major

7

47.0

(44.2-50.1) ± 1.86

5.2

Geomys bursarius sagitallis

2

44.7

(44.2-45.2) ± 1.00

1.6

Geomys bursarius texensis

1

40.5

Geomys personatus davisi

2

47.6

(44.8-50.5) ± 5.68

8.5

Geomys personatus maritimus

2

52.9

(51.8-54.1) ± 2.31

3.1

Geomys personatus megapotamus

7

50.4

(47.0-52.5) ± 1.52

4.0

Geomys personatus personatus

4

56.7

(55.7-57.8) ± 0.93

1.6

Geomys personatus streckeri

1

38.5

Geomys pinetis fontanelus

1

43.7

Geomys pinetis pinetis

7

50.3

(46.7-52.4) ± 1.65

4.3

Geomys tropicalis

4

46.6

(44.1-49.4) ± 2.18

4.7

Length of hind foot

Geomys arenarius arenarius
Geomys attwateri

5

33.2

(32-35) ± 0.98

3.3

Geomys bursarius jugosicularis

2

32.5

(32-33) ± 1.00

2.2

Geomys bursarius knoxjonesi
Geomys bursarius llanensis

9

32.2

(30-35) ±1.19

5.5

Geomys bursarius lutescens

3

33.3

(30-35) ± 3.34

8.7

Geomys bursarius major

3

33.7

(31-35) ± 2.67

6.9

Geomys bursarius sagitallis

1

30

34

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

Table 4. — Continued.

Taxon

N

Mean

(Range) ± 2 SE

cv

Geomys hursarius texensis
Geomys personatus davisi

2

36.5

(35-37) ± 3.01

5.8

Geomys personatus maritimus

1

37

Geomys personatus megapotamus

4

36.7

(36-39) ± 1.50

2.7

Geomys personatus personatus
Geomys personatus streckeri

4

39.7

(39-42) ±1.71

4.3

Geomys pinetis fontanelus

1

33

Geomys pinetis pinetis

7

33.7

(32-37) ±1.36

5.3

Geomys tropicalis

4

31.5

(30-33) ± 12.9

4.1

pattern of the structures is relatively uniform, but
the size may be variable.

The baculum (Fig. 14) is typical of the genus with
a distinct base, shaft, and tip. The length of the
baculum ranged from 9.9 to 11.4 mm among five
adult specimens examined. The baculum is slightly
curved. The base is bulbous with the sides, viewed
dorsoventrally, gradually tapering to the shaft. The
width of the base may be larger, smaller, or equal
to the height of the base.

For five specimens of G. arenarius the mean ra-
tios (range in parentheses) of the condylobasal

length to the length of the distal tract and length of
baculum were 3.5 (3. 4-3. 6) and 4.2 (4. 0-4. 6), re-
spectively; the length of the distal tract to the length
of the baculum was 1.2 (1.1-1. 3).

Specimens examined. — Total (7).

G. a. arenarius (7). — New Mexico; Dona Ana Co.: W Las
Cruces on Rio Grande, 1 (TTU). Texas: El Paso Co.: 2.5 mi
N Rio Grande, 30 mi E El Paso, 1 (TTU); 1 mi S Eabens, 5
(TTU).

Geomys attwateri. — The phallus of G. attwateri
(Fig. 15) is similar to that of G. arenarius except

Fig. 14. — Phallus and baculum of Geomys arenarius arenarius from Texas: El Paso Co., 1 mi S Eabens (TTU 17554), illustrated as in
Fig. 1.

Fig. 15. — Phallus and baculum of Geomys attwateri from Texas: Milam Co., 0.8 mi N, 1.3 mi E Milano (TTU 19137), illustrated
as in Fig. 1 .

that it tends to be slightly larger. The length of the
distal tract ranged from 13.4 to 13.7 mm in four
adult specimens examined. The collar, constriction,
urethral processes, midventral raphae, and epider-
mal structures resemble that of G. arenarius. How-
ever, G. attwateri tends to have a middorsal groove
and dorsal protuberances that are more distinct.

The baculum of G. attwateri (Fig. 15) has fea-
tures, shape, and size similar to that of G. arenarius.
The length of the baculum of four adult specimens
examined ranged from 9.3 to 10.7 mm. The base tends
to be dorsoventrally compressed. The description
and length of the bacula examined correspond to
those given by Kennedy ( 1958), except that the av-
erage width of the base was greater in the series of
adults examined in this study.

The mean ratio (range in parentheses) of the con-
dylobasal length to the length of the distal tract for
four adult G. attwateri was 3.2 (3. 1-3.3); condylo-
basal length to the length of the baculum was 4.4
(4. 0-4. 8); length of the distal tract to the length of
the baculum was 1.4 (1.3-1. 5).

Specimens examined. — Total ( 10).

G. attwateri (10). — Texas: Gonzales Co.: 2.2 mi N Nixon, !
(TTU); Milam Co.: 1.3 mi N, 3 mi E Milano, 1 (TTU); 1.1 mi
N, 2.5 mi E Milano, 3 (TTU); 0.8 mi N, 1.3 mi E Milano, 2
(TTU); 1.3 mi S, 3.3 mi W Milano, ! (TTU); San Patricio Co.:
between Aransas Pass and Ingleside on Hwy. 361, 2 (TTU).

Geomys bursarius. — The phallus of G. hursarius
(Fig. 16) is similar to that of G. attwateri. The
length of the distal tract ranged from 12.5 to 15.9
mm in adult specimens examined. Basic features,
such as the collar, constriction, urethral processes,
midventral raphe, and epidermal structures are sim-
ilar to those of G. attwateri.

The baculum of G. bursarius (Fig. 16) is very
similar to G. arenarius in size and shape. Bacular
length ranged from 9.2 to 11.9 mm in adult speci-
mens examined. The base of the baculum of G. bur-
sarius is almost always dorsoventrally compressed.
The tip is distinct and slightly expanded laterally.
Description and dimensions of bacula of G. bursar-
ius examined in this study closely correspond to
findings of Burt (1960).

36

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

Fig. 16. — Phallus and baculum of Geomys hursarius knoxjonesi from Texas: Gaines Co., 0.8 mi S, 15 mi E Seminole (TTU 19157),
illustrated as in Fig. 1.

In adult specimens of subspecies of G. bursarius
examined, the ratio of the condylobasal length to
the length of the distal tract ranged from 2.8 to 3.7;
condylobasal length to the length of the baculum
ranged from 3.7 to 4.8; length of the distal tract to
the length of the baculum ranged from 1.1 to 1.5.

Specimens examined. — Total (60).

G. b. indiistriiis (1). — Kansas: Ford Co.: 13 mi S Dodge City,
1 (KU).

G. b. Jugossicularis (2). — Colorado: Prowers Co.: Lamar, 3,663
ft, 2 (TTU).

G. b. knoxjonesi (26). — Texas: Cochran Co.: 0.5 mi N, 1.8 mi
W Morton, 1 (TTU); 4 mi S, 0.2 mi E Morton, 1 (TTU); 2.5
mi N, 2.5 mi W Whiteface, 1 (TTU); Gaines Co.: 5 mi SW
Seagraves on Hwy. 385, 2 (TTU); 3 mi N Seminole, 1 (TTU);
0.8 mi S, 15 mi E Seminole, 5 (TTU); Winkler Co.: 10 mi NE
Kermit on Hwy. 115, 3 (TTU); 4.1 mi N, 5.1 mi E Kermit, 10
(TTU); 0.3 mi N, 2.5 mi E Kermit, 2 (TTU).

G. b. llanensis (5). — Texas: Llano Co.: 2.6 mi N, 1.8 mi E
Castell, 2 (TTU); 9.2 mi S, 1 1 mi E Kingsland, 1 (TTU); 10 mi
S, 1.8 mi E Kingsland, 1 (TTU); 0.2 mi N, 8.7 mi W Llano,
1 (TTU).

G. b. lutescens (3). — Kansas; Ellis Co.: 3 mi S Antonio, 1
(TTU). Nebraska: Garden Co.: 30 mi N, 2.0 mi W Oshkosh,
3,850 ft, 2 (CM).

G. b. major (12). — New Mexico: Guadalupe Co.: 1 mi SW San-
ta Rosa, 1 (TTU). Texas: Collingsworth Co.: 2 mi N, 9 mi W

Wellington, 4 (TTU); McLennon Co.: 1 mi S Waco, 3 (TTU);
Mitchell Co.: 3 mi N, 0.3 mi E Colorado City, 4 (TTU).

G. b. missouriensis (2). — Missouri; St. Louis Co.: 3.2 mi N, 2.3
mi E Manchester, 1 (TTU); 1.9 mi N, 0.3 mi E Manchester,
1 (TTU).

G. b. sagittalis (3). — Texas: Galveston Co.: 2 mi N Texas City,
1 (TTU); Montague Co.: 3.1 mi E Jet. 59 and Fm. Rd. 1758
on Fm. Rd. 1758, 1 (TTU); Robertson Co.: 3.4 mi N, 2 mi W
Calvert, 1 (TTU).

G. b. texensis (6). — Texas: Mason Co.: 0.3 mi S, 1.5 mi W
Castell, 3 (TTU); 0.3 mi S, 0.8 mi W Castell, 1 (TTU); 0.7 mi
S, 2.1 mi W Castell, 1 (TTU); 1 mi N, 1.1 mi W Mason, 1
(TTU).

Geomys personatus. — The phallus of G. person-
atus (Eig. 17) is basically shaped like that of other
members of the genus. Subspecies of G. personatus
include some of the largest (G. p. personatus) and
smallest (G. p. streckeri) members of the genus
Geomys. Proportionally the size of the phallus fol-
lows this pattern. The length of the distal tract is
three to four times the width, and the length of the
glans is about half the length of the distal tract. The
length of the distal tract in G. personatus ranged
from 13.0 to 19.1 mm in adult specimens examined.
The sides of the glans, viewed dorsoventrally, are
straight or slightly recurved and gradually expand

Fig. 17. — Phallus and baculum of Geomys personatus personatus from Texas; Nueces Co., Mustang Island, Access Road No. 2 (TTU
15356), illustrated as in Fig. 1.

apically to the collar. The presence and distinctness
of the collar, constriction, urethral processes, mid-
ventral raphe, middorsal groove, dorsal protuber-
ances, and epidermal structures are basically the
same as G. bursarius with the primary distinction
being proportionally different sizes.

The baculum of G. personatus (Fig. 17) is basi-
cally the same shape as G. arenarius, G. attwateri,
and G. bursarius. The size of the baculum varies
proportionally among subspecies of G. personatus
with the length ranging from 9.6 to 12.8 mm among
adult specimens examined. The range of these mea-
surements is in agreement with material examined
by Kennedy (1958). Although there is a tendency
for the base of the baculum to be dorsoventrally
compressed, this trend is not always true because
the width of the base may occasionally be less than
or equal to the height.

Among the subspecies of G. personatus exam-
ined the ratios of the condylobasal length to the
length of the distal tract and to the length of the
baculum ranged from 2.9 to 3.7 and 4.0 to 5.0, re-
spectively. The ratio of the length of the distal tract
to the length of the baculum ranged from 1.2 to 1.6.

Specimens examined. — Total (29).

G. p. davisi (2). — Texas: Zapata Co.: 3 mi N, 2.8 mi W Zapata,
2 (CM).

G. p.fallax ( 1 ). — Texas: Bee Co.: 0.8 mi N, 4.3 mi W Beeville,

1 (TTU).

G. p. maritimus (3). — Texas: Nueces Co.: Flour Bluff, 8.0 mi
S, 8.3 mi E Corpus Christi, 3 (TTU).

G. p. megapoianms (13). — Tamaulipas: barrier island, approx.
70 mi S Rio Grande, I (TTU); barrier island, 10 mi N Boca
Santa Maria, 6 (TTU); barrier island, 5 mi S road at Wash-
ington Beach, I (TTU). Texas: Duval Co.: 3 mi S, 24.6 mi E
Hebbronville, 5 (TTU).

G. p. personatus (8). — Texas: Nueces Co.: Mustang Island,
Access Rd. No. 2, 7 (TTU); Mustang Island, 7 mi S, 4 mi W
Port Aransas, 1 (TTU).

G. p. streckeri (2). — Texas: Dimmit Co.: near Carrizo Springs,

2 (TTU).

Geomys pinetis. — The phallus of G. pinetis (Fig.
18) maintains the general shape and relative size of
G. arenarius, G. attwateri, and G. bursarius. The
length of the distal tract, ranged from 11.7 to 14.5
mm in seven adult specimens examined. The length
of the glans tends to be less than half the length of
the distal tract. The shape of the glans has minor
differences from that of other Geomys. The sides
of the glans, viewed dorsoventrally, are more or

38

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

Fig. 18. — Phallus and baculum of Gconiys pinetis pinetis from Florida: Duval Co., 1.0 mi NW Bayard on Hwy. 1 (TTU 16619),
illustrated as in Fig. I.

less straight and parallel with a slight flare at the
collar. Viewed laterally, the ventral side may be
somewhat convex or straight and the dorsal side
recurved. Although the collar is less flared than oth-
er Geomys, it is still a distinct feature, as is the
constriction which is often strongly extended api-
cally on the dorsal side. In G. pinetis the urethral
processes often appear reduced. The midventral
raphe is well developed. The development of the
middorsal groove and dorsal protuberances is sim-
ilar to that of G. attwateh, G. hursarins, and G.
personatiis. The size and pattern of epidermal
structures is relatively uniform and is similar to that
of other members of the genus.

The baculum of G. pinetis (Fig. 18) ranged in
length from 8.3 to 10.5 mm in seven adult specimens
examined. The width of the base is always greater
than the height of the base. One of the most striking
characteristics found in G. pinetis, that sets it apart
from other Geomys, is the shape of the base of the
baculum. Viewed dorsoventrally, the base of the
baculum has sides that expand apically and then
converge sharply to join the shaft, instead of grad-
ually tapering to the shaft as in other members of
the genus. This study found this character to be
consistent in all specimens examined, including the

holotype and a paratype of G. p. fontanelus. Sher-
man (1940) reported no differences between the
baculum of G. p. fontanelus and other populations
of G. pinetis.

In seven adult specimens of G. p. pinetis the
mean ratios (range in parentheses) of the condylo-
basal length to the length of the distal tract and
length of the baculum were 3.9 (3. 5-4. 5) and 5.4
(4. 9-6.0), respectively. The ratio of the length of
the distal tract to the length of the baculum aver-
aged 1.4 and ranged from 1.2 to 1.6.

Specimens examined. — Total (19).

G. p. fontanelus (2). — Georgia: Chatham Co.: 8 mi NW Sa-
vannah, 2 (FSM).

G. p. pinetis (17). — Florida: Alachua Co.: 1 mi N, 2.8 mi W
Alachua, 1 (CM); 0.5 mi S, 2 mi E Alachua, 1 (CM); 1.4 mi
NW Jet. Hwy. 24 and Hwy. 41, 1 (TTU); 0.9 mi SW Jet. 1-75
and Hwy. 24, 1 (TTU); 3.4 mi SW Jet. 1-75 and Hwy. 24, 1
(TTU); 3.4 mi SW Jet. 1-75 and Hwy. 24, 1 (TTU); Calhoun
Co.: 1.2 mi N Clarksville, 1 (TTU); Duval Co.: 1.0 mi NW
Bayard on Hwy. 1, 1 (TTU); Highlands Co.: Sebring, 2 (CM);
Hillsborough Co.: Tampa, vicinity University of South Flor-
ida, 5 (TTU); Walton Co.: 1.1 mi E county line on Hwy. 90,

1 (TTU). Georgia: Camden Co.: 1.1 mi NE Kingsland, 1
(TTU).

Geomys tropicalis. — The phallus of G. tropicalis
(Fig. 19) resembles G. personatiis except for size

1982

WILLIAMS— GEOMYID PHALLI

39

Fig. 19. — Phallus and baculum of Geomys tropicaUs from Tamaulipas: 2.5 mi SE Altamira (TTU 14318), illustrated as in Fig. 1.

variation. The length of the distal tract ranged from
13.2 to 14.4 mm in four adult specimens examined.
Like G. personatus, the sides of the glans of G.
tropicaUs, viewed dorsoventrally, are more or less
straight and are parallel or expanded apically to the
collar. The collar is distinct but may not share the
elaborate convolutions found in other species of
Geomys. The development of the midventral raphe,
middorsal groove, dorsal protuberances, and epi-
dermal structures resembles that of G. personatiis.

The baculum of G. tropicaUs (Fig. 19) is basically
the same as the other species of Geomys, except
G. pinetis. The length of the baculum of four adult
specimens examined ranged from 9.3 to 10.4 mm.
In all four specimens the width of the base almost
equalled the height of the base.

The mean ratio (range in parentheses) of the con-
dylobasal length to the length of the distal tract for
four adult G. tropicaUs was 3.4 (3. 1-3.5); condy-
lobasal length to the length of the baculum was 4.7
(4. 5-5.0); length of the distal tract to the length of
the baculum was 1.4 (1. 3-1.5).

Specimens examined. — Total (10).

G. tropicaUs (10). — Tamaulipas: 2.5 mi SE Altamira, 9 (TTU);

2.5 mi SSE Altamira, 1 (TTU).

Statistics

Age variation in Geomys was observed in this
study and reported by Kennedy (1958). However,
the amount of age variation was not analyzed in the
present study. Table 4 provides the standard statis-
tics (sample size, mean, range, standard error, and
coefficient of variation) for adult specimens of Geo-
mys examined.

Examination of individual variation was con-
ducted on a representative sample from different
species of Geomys. Each sample used included in-
dividuals taken from a restricted geographical area
that would be assumed to be part of a continuous
population. The taxa examined for individual vari-
ation were G. arenarius arenarius, G. attwateri,
G. personatus personatiis, G. pinetis pinetis, and
G. tropicaUs. Coefficients of variation for phallic
and bacular measurements ranged from 1.1 to 22.7.
The average coefficients of variation for the above
taxa were 13.0, 8.7, 7.2. 10.2, and 10.2, respective-
ly. The characters that generally had the lowest val-
ues for coefficients of variation were the length of
the distal tract, length of the glans, and length of
the baculum. The highest values were found among
different characters for different species. However,

40

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

the length of the protractile tip consistently had high
values, as indicated by a mean coefficient of vari-
ation of 14. 1 for the five samples examined. The
characters having the lowest and highest coeffi-
cients of variation in G. arenarius were length of
distal tract (2.1) and width of glans across collar

(22.7) ; in G. attwateri, length of distal tract (1.1)
and width of bacular base (12.6); in G. personatus,
length of glans (2.0) and height of bacular base

( 10.8) ; in G. pinetis, length of glans (5.9) and length
of protractile tip (16.1); in G. tropicalis, length of
distal tract (4.4) and width of bacular base (17.2).

Because several nominal subspecies of G. bur-
sarius were available for this study, an attempt was
made to determine if geographical variation of phal-
lic and bacular characters does occur in Geomys.
The limited geographical coverage of samples pre-
cluded quantifying the amount of geographical vari-
ation that might exist. Examination of ranked
means of samples of G. bursarius generally re-
vealed no consistencies in size for all of the char-
acters. Taxa ranked among the largest for one or
more characters may rank among the smallest in
other characters. However, taxa from south-central
Texas (G. b. llanensis and G. b. texensis) tended to
group together and have smaller-sized individuals.
The only other trend observed among ranked means
of G. bursarius was that G. b. lutescens and G. b.
major were grouped together in six of eight char-
acters. Based on these observations it is assumed
that better geographical coverage of samples, and
larger sample sizes may reveal greater geographical
variation than was found in these samples of G.
bursarius. However, it is unlikely the amount of
variation would approach that observed in members
of the genus Thomomys.

Examination of variation among species of Geo-
mys was conducted, using samples with at least
three individuals. Subspecies within a species were
not combined to form a single group because of
possible problems of geographical variation and
biased representation. The samples used were G.
arenarius, G. attwateri, G. bursarius knoxjonesi, G.
b. lutescens, G. b. major, G. personatus megapo-
tamus, G. p. personatus, G. pinetis, and G. tropi-
calis. Univariate and SS-STP analyses of these
samples were performed with the UNIVAR (Pow-
er, 1970) computer program. Analysis of variance
revealed significant differences (P ^ 0.05) for all
phallic and bacular characters. G. p. personatus
had the highest mean for all phallic and bacular

characters except for the length of the protractile
tip and width of the bacular base, where it ranked
second; G. p. megapotamus had the highest mean
for the length of the protractile tip and was ranked
second, third, and fourth in highest means for three,
two, and one characters, respectively. The lowest
mean for all characters, except the width of bacular
base, was found in G. pinetis-, G. pinetis had the
highest mean for width of the baculum thus sup-
porting this character as being diagnostic for the
species (see description). G. tropicalis was ranked
among the smallest in the samples examined with
rankings of first, second, and third smallest in one,
five, and one characters, respectively. G. attwateri
also ranked among the smallest with five of seven
characters ranking either second or third smallest
of the samples examined. G. arenarius and G. bur-
sarius showed no specific trends in relationships of
means. The SS-STP analysis provided broadly
overlapping nonsignificant subsets for all phallic
and bacular characters. Only the length of distal
tract and length of glans had cases of nonoverlap-
ping subsets. In the three subsets formed with the
length of the distal tract the two nonoverlapping
subsets were formed with the two samples of G.
personatus in one subset and the remaining samples
in the other subset. An intermediate subset con-
sisting of G. p. megapotamus, G. b. major, and G.
b. lutescens overlapped slightly with the other two
subsets. For the length of the glans four nonsignifi-
cant subsets were formed. A subset consisting of
G. p. personatus, G. p. megapotamus, and G. ar-
enarius did not overlap with the subset of the small-
er-sized (in ranked order) G. b. lutescens, G. b.
knoxjonesi, G. attwateri, G. tropicalis, and G. pi-
netis. Two intermediate subsets broadly overlapped
with each other and the two subsets previously dis-
cussed. In the remaining characters two subsets
were formed with the length of protractile tip; three
subsets were formed with width of glans across col-
lar, and the bacular characters; four subsets were
formed with the width of glans across base.

Variation between species of Geomys was ex-
amined by multivariate analysis, using sample
means with the MINT statistical computer pro-
gram. The samples of Geomys used in the multi-
variate analysis, followed by the corresponding ref-
erence number, were G. arenarius (1), G. attwateri
(2), seven subspecies of G. bursarius {jugossicularis,
3; knoxjonesi, 4; llanensis, 5; lutescens, 6; major,
7; sagittalis, 8; texensis, 9), five subspecies of G.

1982

WILLIAMS— GEOMYID PHALLI

41

G . arenar i us arenar i us

G.

bursar i us

knox ionesi

i •

bursar i us

lutescens

G-

bursar i us

ma ior

G.

bursar i us

saq i tt al i s

G.

attwater i

G.

bursarius

texensis

G.

bursar i us

Ilanensis

t r 0 p i c a 1 i s

bursar i us iuqoss i cular i s
per sonat us d a v i si
personat us strecker i
p i ne t i s p i net i s
per sonat us mar i t i m u s
per sonat us meqapot amus
per sonat us per sonatus

3.0 2.0 1.0 0.0

Fig. 20. — Distance phenogram of 16 samples of Geornys resulting from clustering by unweighted pair-group method using arithmetic
averages (UPGMA). The cophenetic correlation coefficient for the phenogram is 0.727.

personatiis (davisi, 10; maritimus, 11; megapota-
mus, 12; personatiis, 13; streckeri, 14), G. pinetis
(15), and G. tropicalis (16).

The distance phenogram produced by the MINT
program is illustrated in Fig. 20. The cophenetic
correlation value for this phenogram is 0.727. The
phenogram forms two groups. One group consists
of the larger members of G. personatiis — persona-
tus, megapotamiis, and maritimus. The second
group subdivides into a group represented by G.
pinetis and series of additional clusters including
samples of the other taxa of Geornys. In the next
subdivision G. p. streckeri is distinctly separated
and followed by a group consisting of G. p. davisi

and G. h.jugossicitlaris which remain relatively dis-
tinct from each other. The remaining subspecies of
G. bur sarins form two clusters. The first cluster
includes G. b. knoxjonesi, G. b. liitescens, G. b.
major, and G. b. sagittalis. The second cluster in-
cludes G. b. texensis and G. b. Ilanensis. Geornys
arenarius was placed with the first cluster of G.
bursarius', G. attwateri and G. tropicalis were
placed with the second cluster.

The first three principal components extracted
from the matrix of correlation among characters is
illustrated in Fig. 21. Four of the five samples of G.
personatiis (10, 11, 12, 13) form a loose grouping
toward the right side of the plot. Samples of G.

42

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

Fig. 21. — Three-dimensional projection of 16 samples of Geomys (G. arenarius, 1; G. attwateri, 2; G. bursariiis, 3-9; G. personatus.
10-14; G. pinetis, 15; G. tropicalis, 16) onto the first three principal components based upon a matrix of correlation among one cranial,
five phallic, and three bacular measurements. Components I and II are indicated in the plots and component III is represented by
height.

bitrsanus, (3, 4, 5, 6, 7, 8, 9) and G. attwateri (2)
also form a loose, but unified group in the upper left
side of the plot. The remaining taxa — G. arenarius
(1), G. pinetis (15), G. tropicalis (16), and G. p.
streckeri (14) — are situated toward the left side of
the plot. G. pinetis (15) and G. p. streckeri (14) are
separated from all other taxa in component I,
and separated from each other in component II. G.
pinetis plots closer to G. biirsarius than any other

Table 5. — Factor matrix from correlation among nine characters
of the five species (16 samples) of Geomys examined.

Character

Component I

Component II

Component III

Condylobasal length

0.785

-0.085

0.155

Length of distal tract

0.908

-0.012

0.065

Length of glans

0.950

0.196

-0.010

Length of protractile tip

0.565

0.743

0.193

Width of glans across collar

0.807

-0.439

-0.322

Width of glans across base

0.723

-0.499

-0.269

Length of baculum

0.918

0.262

-0.046

Width of bacular base

0.135

-0.451

0.870

Height of bacular base

0.817

0.009

0.121

taxa. G. arenarius (1), G. p. streckeri (14), and G.
tropicalis (16) are separated from each other with
all three components and separated from the re-
maining taxa in component III.

The amount of phenetic variation explained by
components I, II, and III are 59.6%, 14.6%, and
11.3%, respectively. Results of principal compo-
nent analyses showing the influence of each char-
acter for the first three components are given in
Table 5. All characters except the width of the bac-
ulum are heavily weighted in the first factor. The
length of the distal tract, length of the glans, and
length of the baculum had the highest values in
component I. In components II and III the char-
acters with heavy weightings were the length of the
protractile tip and width of bacular base, respec-
tively.

Pappogeomys

Description

Pappogeomys bulleri. — In this study the genus
Pappogeomys is represented only by samples of P.

1982

WILLIAMS— GEOMYID PHALLI

43

Fig. 22. — Phallus and baculum of Pappogeomys biiileri hulleri from Jalisco: 14 mi NW Mascota, 6,400 ft (KU 111714), illustrated as
in Fig. 1.

buUeri. All phalli of P. bnUeri examined were re-
moved from museum specimens and normal resto-
ration was assumed.

The phallus of P. bulleri (Fig. 22) has a distal
tract about five times longer than its width and
about twice as long as the glans. The length of the
distal tract ranged from 8.7 to 10.4 mm in adult
specimens examined. The sides of the glans, viewed
dorsoventrally, are more or less parallel or slightly
converging apically. From a lateral view, the dorsal
and ventral sides tend to be more recurved. The
collar is most apparent on the dorsal side, but loses
its distinctness as it extends laterally and apically.
Occasionally traces of the collar can be observed
on the dorsal side. The pair of urethral processes
are barely visible because of their position under
the collar. Typical features of geomyid rodents are
poorly developed in P. bulleri. A midventral raphe
is present but not always distinct. The middorsal
groove and dorsal protuberances may be present
but never well-developed. These features might be
more obvious if fresh samples are examined.

Epidermal structures in P. bulleri have a single,
proximally oriented projection. The size and pat-

tern of structures are fairly uniform. The structures
are present on the entire glans, except distal to the
collar on the ventral side and along the extreme
margins of the apex.

The baculum of P. bulleri (Fig. 22) is typical of
geomyid rodents. It consists of a bulbous base,
slender shaft, and tip. Viewed dorsoventrally, the
base gradually tapers to the shaft. The length of the
baculum in adult specimens examined ranged from
7.8 to 9.4 mm. The baculum is slightly curved.

The ratio of the condylobasal length to the length
of the distal tract and the length of the baculum
ranges from 3.4 to 4.4 and 4.0 to 4.9, respectively.
The ratio of the length of the distal tract to the
length of the baculum ranged from 1.1 to 1.2.

Samples of P. bulleri were included with samples
of Cratogeomys for statistical analysis because of
the close relationship of the taxa as suggested by
Russell (1968/?). Table 6 provides measurements of
P. bulleri examined in this study.

Specimens examined. — Total (5).

P. b. bulleri (3). — Jalisco: 20 mi SE Autlan, 7,700 ft, I (KU);

18 mi (by Carranza Rd.) W Ciudad Guzman, 1 (TTU); 14 mi

NW Mascota, 6,500 ft I (KU).

44

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

Table 6. — Standard statistics for adult specimens of 17 samples representing one

geomys.

species of Pappogeomys and six

species of Crato-

Taxon

N

Mean

Range ± 2 SE

CV

Length of distal tract

Pappogeomys biilleri hulleri

2

9.5

(8.7-10.4) ± 1.71

12.7

Pappogeornys hulleri infiiscus

1

10.1

Pappogeomys hulleri lutulentus

1

10.2

Cratogeomys castanops castanops

1

11.4

Cratogeomys castanops perplanus

12

11.4

(10.3-12.9) ± 0.43

6.5

Cratogeomys castanops pratensis

1

11.5

Cratogeomys castanops ruhellus

1

11.7

Cratogeomys fumosus

2

16.3

(15.3-17.3) ± 2.01

8.7

Cratogeomys gymnurus gymnurus

1

20.2

Cratogeomys gymnurus russelli

3

15.9

(15.3-16.3) ± 0.59

3.2

Cratogeomys gymnurus tellus

1

16.0

Cratogeomys merriami irolonis

2

15.5

(15.3-15.8) ± 0.50

2.3

Cratogeomys merriami merriami

2

16.1

(15.8-16.3) ± 0.50

2.2

Cratogeomys tylorhinus angustirostris

5

16.9

(15.6-17.8) ± 0.73

4.9

Cratogeomys tylorhinus planiceps

1

12.5

Cratogeomys tylorhinus tylorhinus

1

15.8

Cratogeomys zinseri

2

14.1

(13.2-14.9) ± 1.71

8.5

Length of glans

Pappogeomys hulleri hulleri

2

4.9

(4. 7-5. 2) ± 0.50

7.2

Pappogeomys hulleri infuscus

1

5.5

Pappogeomys hulleri lutulentus

1

5.4

Cratogeomys castanops castanops

1

6.3

Cratogeomys castanops perplanus

12

6.3

(5.8-7. 1) ± 0.23

6.4

Cratogeomys castanops pratensis

1

6.5

Cratogeomys castanops ruhellus

1

6.4

Cratogeomys fumosus

2

8.5

(8.4-8. 6) ± 0.20

1.7

Cratogeomys gymnurus gymnurus

1

10.9

Cratogeomys gymnurus russelli

3

8.8

(8.2-9. 1) ± 0.57

5.6

Cratogeomys gymnurus tellus

1

9.2

Cratogeomys merriami irolonis

Cratogeomys merriami merriami

2

9.3

(8. 8-9.9) ± l.IO

8.4

Cratogeomys tylorhinus angustirostris

6

9.4

(8.7-10.1) ± 0.44

5.8

Cratogeomys tylorhinus planiceps

1

6.4

Cratogeomys tylorhinus tylorhinus

1

8.5

Cratogeomys zinseri

2

7.3

(6.8-7.7) ± 1.42

8.7

Length of protractile tip

Pappogeomys hulleri hulleri

2

1.5

(1.5-1. 5) ± 0.00

0.0

Pappogeomys hulleri infuscus

1

1.9

Pappogeomys hulleri lutulentus

1

1.8

Cratogeomys castanops castanops

1

2.0

Cratogeomys castanops perplanus

12

2.0

(1.6-2. 3) ±0.11

10.4

Cratogeomys castanops pratensis

1

1.9

Cratogeomys castanops rehellus

1

1.9

Cratogeomys fumosus

2

2.5

(2. 4-2. 5) ± 0.10

2.8

Cratogeomys gymnurus gymnurus

1

3.0

Cratogeomys gymnurus russelli

3

2.8

(2.5-3. 3) ± 0.50

15.6

Cratogeomys gymnurus tellus

1

2.8

Cratogeomys merriami irolonis

Cratogeomys merriami merriami

2

3.4

(3. 3-3. 5) ± 0.20

4.1

Cratogeomys tylorhinus angustirostris

6

2.8

(2.3-3. 1) ± 0.28

12.3

Cratogeomys tylorhinus planiceps

1

2.0

Cratogeomys tylorhinus tylorhinus

1

2.6

Cratogeomys zinseri

2

2.0

(2.0-2.0) ± 0.00

0.0

1982

WILLIAMS— GEOMYID PHALLI

45

Table 6. — Continued

Taxon

N

Mean

Range ± 2 SE

cv

Width of glans

across collar

Pappogeomys bulleri bulleri

2

1.8

(1. 6-2.0) ± 0.40

15.7

Pappogeomys bulleri infuscus

1

2.0

Pappogeomys bulleri liilulentus

1

1.5

Cratogeomys castanops castanops

1

2.8

Cratogeomys castanops perplanus

12

2.5

(2. 3-2. 7) ± 0.10

7.1

Cratogeomys castanops pratensis

1

2.8

Cratogeomys castanops rebellus

1

2.5

Cratogeomys fumosus

2

3.0

(3.0-3.0) ± 0.00

0.0

Cratogeomys gymnurus gymnurus

1

3.7

Cratogeomys gymnurus russelli

2

2.9

(2.9-3.0) ± 0.10

2.4

Cratogeomys gymnurus tellus

1

3.1

Cratogeomys merriami irolonis

Cratogeomys merriami merriami

2

2.3

(2. 2-2.4) ± 0.20

6.1

Cratogeomys tylorhinus angustirostris

6

3.6

(3. 2-4.0) ± 0.27

9.1

Cratogeomys tylorhinus planiceps

Cratogeomys tylorhinus tylorhinus

1

2.9

Cratogeomys zinseri

2

3.5

(3.4-3. 6) ± 0.20

4.0

Width of glans

across base

Pappogeomys bulleri bulleri

2

1.6

(1.5-1. 7) ± 0.20

8.8

Pappogeomys bulleri infuscus

1

1.5

Pappogeomys bulleri lutulentus

1

1.7

Cratogeomys castanops castanops

1

2.8

Cratogeomys castanops perplanus

12

2.5

(2.3-2. 7) ± 0.10

6.9

Cratogeomys castanops pratensis

1

2.7

Cratogeomys castanops rubellus

1

2.3

Cratogeomys fumosus

2

3.1

(3.0-3. 3) ± 0.30

6.8

Cratogeomys gymnurus gymnurus

1

3.7

Cratogeomys gymnurus russelli

2

3.3

(3. 2-3.4) ± 0.20

4.3

Cratogeomys gymnurus tellus

1

4.2

Cratogeomys merriami irolonis

Cratogeomys merriami merriami

2

2.9

(2. 8-2.9) ± 0.10

2.4

Cratogeomys tylorhinus angustirostris

6

3.9

(3. 5^.5) ± 0.28

8.7

Cratogeomys tylorhinus planiceps

Cratogeomys tylorhinus tylorhinus

1

3.2

Cratogeomys zinseri

2

3.2

(3. 2-3. 2) ± 0.00

0.0

Length of baculum

Pappogeomys bulleri bulleri

2

8.6

(7. 8-9.4) ± 1.60

13.1

Pappogeomys bulleri infuscus

1

8.7

Pappogeomys bulleri lutulentus

1

8.7

Cratogeomys castanops castanops

1

9.5

Cratogeomys castanops perplanus

11

10.2

(9.1-10.8) ± 0.31

5.0

Cratogeomys castanops pratensis

I

10.5

Cratogeomys castanops rubellus

2

9.9

(9.9-10.0) ± 0.10

0.7

Cratogeomys fumosus

2

13.1

(12.9-13.3) ± 1.40

19.0

Cratogeomys gymnurus gymnurus

1

16.3

Cratogeomys gymnurus russelli

3

14.2

(13.9-14.3) ± 0.27

1.6

Cratogeomys gymnurus tellus

1

14.2

Cratogeomys merriami irolonis

2

14.4

(14.1_14.7) ± 0.60

2.9

Cratogeomys merriami merriami

2

14.4

(14.0-14.8) ± 0.80

3.9

Cratogeomys tylorhinus angustirostris

6

13.7

(12.4-15.5) ± 1.02

9.1

Cratogeomys tylorhinus planiceps

1

12.2

Cratogeomys tylorhinus tylorhinus

1

12.9

Cratogeomys zinseri

2

12.1

(11.6-12.6) ± 1.00

5.8

46

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

Table 6.

Continued

Taxon

N

Mean

Range ±

2 SE

cv

Width of bacular base

Pappogeomys bulleri hulleri

3

1.7

(1. 3-2.2)

± 0.53

26.9

Pappogeomys hulleri infusctis

1

1.6

Pappogeomys bulleri lutulentus

1

2.0

Cratogeomys castanops castanops

1

2.4

Cratogeomys castanops perplanus

12

2.3

(2.0-2.8)

± 0.14

10.7

Cratogeomys castanops pratensis

1

2.3

Cratogeomys castanops ruhellus

2

2.1

(2. 1-2.2)

± 0.10

3.4

Cratogeomys fumosus

2

2.5

(2. 5-2. 5)

± 0.00

0.0

Cratogeomys gymniirus gymnurus

1

3.5

Cratogeomys gymnurus russelli

3

3.1

(2.7-3. 5)

± 0.47

13.0

Cratogeomys gymnurus tellus

1

3.4

Cratogeomys merriami irolonis

2

3.0

(2. 8-3. 2)

± 0.40

9.4

Cratogeomys merriami merriami

2

2.9

(2.7-3.0)

± 0.30

7.3

Cratogeomys tylorhinus angustirostris

6

3.3

(3.0-3.7)

± 0.20

7.5

Cratogeomys tylorhinus planiceps

1

2.8

Cratogeomys tylorhinus tylorhinus

1

2.7

Cratogeomys zinseri

2

2.7

(2.7-2. 8)

± 0.10

2.6

Height of bacular base

Pappogeomys bulleri hulleri

3

1.4

(1. 1-1.8)

± 0.42

25.7

Pappogeomys bulleri infusctis

1

1.5

Pappogeomys hulleri lutulentus

1

1.4

Cratogeomys castanops castanops

1

1.4

Cratogeomys castanops perplanus

12

1.7

(1.4-1.9)

± 0.08

8.5

Cratogeomys castanops pratensis

1

1.7

Cratogeomys castanops ruhellus

2

1.6

(1. 5-1.7)

± 0.20

8.8

Cratogeomys fumosus

2

2.0

(1. 8-2.2)

± 0.40

14.1

Cratogeomys gymnurus gymnurus

1

2.8

Cratogeomys gymnurus russelli

3

2.3

(2.0-2. 7)

± 0.41

15.3

Cratogeomys gymnurus tellus

1

2.4

Cratogeomys merriami irolonis

2

1.9

(1. 9-2.0)

± 0.73

3.7

Cratogeomys merriami merriami

2

1.2

(1.0-1. 4)

± 1.42

23.6

Cratogeomys tylorhinus angustirostris

6

2.1

(1. 9-2.4)

± 0.17

10.2

Cratogeomys tylorhinus planiceps

1

1.9

Cratogeomys tylorhinus tylorhinus

1

1.9

Cratogeomys zinseri

2

1.7

(1. 5-2.0)

± 1.50

20.8

Condylobasal length

Pappogeomys bulleri hulleri

3

39.3

(37.7-41.6)

± 2.36

5.2

Pappogeomys bulleri infuscus

1

39.0

Pappogeomys bulleri lutulentus

1

35.0

Cratogeomys castanops castanops

1

56.3

Cratogeomys castanops perplanus

12

58.7

(55.2-64.1)

± 1.47

1.1

Cratogeomys castanops pratensis

1

52.3

Cratogeomys castanops rubellus

2

48.5

(46.5-50.4)

± 3.91

5.7

Cratogeomys fumosus

2

59.9

(59.0-60.8)

± 1.80

2.1

Cratogeomys gymnurus gymnurus

1

72.7

Cratogeomys gymnurus russelli

3

64.2

(62.9-66.2)

± 2.01

2.7

Cratogeomys gymnurus tellus

1

71.9

Cratogeomys merriami irolonis

2

66.5

(66.2-66.8)

± 0.60

0.6

Cratogeomys merriami merriami

2

67.3

(67.6-67.9)

± 1.21

1.3

Cratogeomys tylorhinus angustirostris

4

62.4

(62.3-67.6)

± 2.05

3.3

Cratogeomys tylorhinus planiceps

1

61.9

Cratogeomys tylorhinus tylorhinus

1

63.6

Cratogeomys zinseri

2

63.8

(62.2-65.4)

± 3.31

3.5

1982

WILLIAMS— GEOMYID PHALLI

47

Table 6. — Continued

Taxon

N

Mean

Range ± 2 SE

cv

Length of hind foot

Pappogeomys hiilleri bulleri

3

29.5

(29-30.5) ± 1.00

2.9

Pappogeomys bulleri infuscus

1

30

Pappogeomys bulleri lutulentus

1

27.5

Cratogeomys castanops castanops

1

36

Cratogeomys castanops perplanus

II

39.5

(35-45) ± 1.65

6.9

Cratogeomys castanops pratensis

1

38

Cratogeomys castanops rubellus

2

32

(32-32) ± 0.00

0.0

Cratogeomys fumosus

2

47.0

(46-48) ± 2.01

3.0

Cratogeomys gymnurus gymnurus

1

56

Cratogeomys gymnurus russelli

3

48.7

(47-51) ± 2.41

4.3

Cratogeomys gymnurus tellus

1

45

Cratogeomys merriami irolonis

2

46

(45-47) ± 2.01

3.1

Cratogeomys merriami merriami

2

51.5

(51-52) ± 1.00

1.4

Cratogeomys tylorhinus angustirostris

6

44.8

(41-47) ± 1.89

5.2

Cratogeomys tylorhinus planiceps

1

46

Cratogeomys tylorhinus tylorhinus

1

43

Cratogeomys zinseri

2

48

(48-48) ± 0.00

0.0

P. b. infuscus (1). — Jalisco: 7 mi SSWTequila, 9,000ft, 1 (KU).
P. b. lutulentus (1). — Jalisco: Sierra de Cuale, 7,300 ft, 1 (KU).

Cratogeomys

Description

The size and shape of the phallus of Cratogeomys
is more variable among species than Geomys, but
less variable than Thomomys. The length of the dis-
tal tract in adult specimens examined ranged from
10.3 to 17.8 mm. Like Geomys, the size relationship
between the length of the distal tract and body size
of individuals of Cratogeomys is less variable than
that of some species of Thomomys . The ratio of the
condylobasal length to the length of the distal tract
ranges from 3.6 to 5.5. Basically the glans in Cra-
togeomys is unique from other geomyid rodents
(except Pappogeomys) in that the sides of the
glans, viewed dorsoventrally, tend to be parallel or
converge apically. Viewed laterally the dorsal side
tends to be straight and the ventral side may be
straight or recurved with a slight flaring at the col-
lar. The collar is also unique from most other geo-
myid species. On the ventral side, the collar is dis-
tinct and often exhibits some degree of folding.
However, as the collar passes around the glans the
distinctness of the collar is reduced laterally and is
often nonexistent dorsally. In all cases the collar is
extended more apically on the lateral sides. In all
samples of Cratogeomys the constriction is a dis-
tinct feature and is often extended apically on the
ventral side. Most species also possess well-devel-

oped urethral processes, midventral raphe, middor-
sal groove, and a pair of dorsal protuberances.

The epidermal structures of Cratogeomys have
a single, proximally oriented projection. In all
species of Cratogeomys, the epidermal structures
occur between the collar and constriction and distal
to the collar on the dorsal side.

The length of the baculum in adult specimens ex-
amined ranged from 9.1 to 15.5 mm. The baculum
of Cratogeomys, like Geomys, shows little varia-
tion in relative size compared to body size. The
ratio of the condylobasal length and the length of
the distal tract to the length of the baculum ranges
from 4.4 to 6.3 and 1.0 to 1.4, respectively. Like
other geomyid species, the baculum is positioned
in the center, or slightly dorsal to the center, of the
phallus, and consists of a distinct base, shaft, and
tip. In all specimens examined the base was dor-
soventrally compressed.

The following are detailed descriptions of six of
the seven recognized species of Cratogeomys. Ta-
ble 6 provides measurements that apply to each de-
scription.

Cratogeomys castanops. — The phallus of C. cas-
tanops (Fig. 23) is medium-sized compared to other
members of the genus. The length of the distal tract,
which ranged from 10.3 to 12.9 mm in adult speci-
mens examined, is about five times greater than the
width and about twice as long as the length of the
glans. The sides of the glans, viewed dorsoventral-
ly, tend to be parallel or converge to the apex.

48

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

A B C

D

NO. 20

E

Fig. 23. — Phallus and baculum of Cratogeornys castanops perplanus from Texas: Gaines Co., 4.4 mi N, 9.3 mi W Seminole (TTU
25999), illustrated as in Fig. 1.

Viewed laterally the sides are more or less straight
and parallel, with a slight flare on the ventral side
in the region of the collar. The collar is most distinct
on the ventral side and less distinct as it extends
apically on the lateral sides. The collar, although
poorly developed, may be apparent on the dorsal
side. The protractile tip is relatively long. The ure-
thral processes and midventral raphe are well de-
veloped. The middorsal groove tends to be short,
and the dorsal protuberances are often not distinct.

The epidermal structures of C. castanops have
a single proximally oriented projection. The size
and pattern of structures are uniform.

The length of the baculum ranged from 9. 1 to 10.8
mm in adult specimens examined. The baculum of
C. castnops (Fig. 23), is unique from other geomyid
rodents because of its massive appearance. Viewed
dorsoventrally there is a distinct base that is straight
or slightly concave on the end. The base tapers to
a relatively broad shaft that has straight parallel
sides. There is a slight indentation of the shaft just

proximal to the broad tip. Viewed laterally the base
and tip are less evident as they join the shaft to
form a relatively straight baculum.

The ratio of the condylobasal length to the length
of the distal tract ranged from 4.0 to 5.5; ratio of
the condylobasal length to the length of the baculum
ranged from 4.7 to 6.3; ratio of the length of the
distal tract to the length of the baculum ranged from
1.1 to 1.2.

Specimens examined. — Total (25).

C. c. castanops (1). — Colorado: Prowers Co.: 2.0 mi S, 1.0 mi
E Lamar, 1 (TTU).

C. c. perplanus (16). — New Mexico: Guadalupe Co.: 1 mi S
Santa Rosa, 2 (TTU); Lea Co.: 3.3 mi W Crossroads, 1 (TTU);
2.9 mi W Crossroads, I (TTU); 2.7 mi W Crossroads, 1 (TTU);
1.4 mi N, 0.5 mi E Maljamar, 1 (TTU); Roosevelt Co.: 12.8
mi W Floyd, 1 (TTU). Texas: Bailey Co.: 5.8 mi S, 0.7 mi W
Needmore, 1 (CM); Deaf Smith Co.: 1 mi N, 17.9 mi W Here-
ford, 1 (TTU); 1 mi N, 15.5 mi W Hereford, 1 (TTU); Gaines
Co.: 4.4 mi N, 9.3 mi W Seminole, 3 (TTU); 4.4 mi N, 6.2 mi
W Seminole, 1 (TTU); Hockley Co.: 1 mi N, 4.3 mi W Level-

Fig. 24. — Phallus and baculum of Cnitogeomys fumosus from Colima: 2 mi W Colima (CM 55806), illustrated as in Fig. 1.

land, I (TTU); Randall Co.: 0.2 mi N, 6.5 mi E Canyon, I
(TTU).

C. c. pratensis (4). — Texas: Brewster Co.: 1 1.8 mi N, 2.0 mi E
Marathon, 4,900 ft, 1 (CM); Jeff Davis Co.: 9 mi NE Eort
Davis, 1 (TTU); Pecos Co.: 17.0 mi N, 18.5 mi E Marathon,
4,500 ft, 1 (CM); 14.4 mi N, 18.3 mi E Marathon 4,450 ft, I
(CM).

C. c. siniulans (1). — Texas; Lubbock Co.: Lubbock airport, I
(TTU).

C. c. riibellus (3). — San Luis Potosi: 5.7 mi E Jet. Hwy. 80 and
Hwy. 101 near Tepeyac, 1 (TTU). Zacatecas: 45 km (by
road) NE Morelos, 2 (TTU).

Cnitogeomys fumosus. — The phallus of C. fu-
mosus (Fig. 24), although larger than C. castanops,
is still considered to be medium-sized. The length
of the distal tract is about five or six times greater
than the width, and less than twice the length of the
glans. The length of the distal tract in two adult
specimens examined was 17.3 and 15.3 mm. The
shape of the glans and its collar are similar to that
of C. castanops. Also, as in C. castanops, the pro-
tractile tip appears relatively long compared to
some of the other species of Cnitogeomys. The

midventral raphe and urethral processes are distinct
and well-developed. The middorsal groove and dor-
sal protuberances of C. fumosus tend to be more
distinct than C. castanops.

The epidermal structures of C. fumosus have a
single, proximally oriented projection. The size of
individual structures is variable.

The baculum of C . fumosus (Fig. 24), viewed dor-
soventrally, has distally converging sides that tend
to obscure the position of the tip and base of the
baculum. Viewed laterally, the tip and base are
more apparent. The baculum is slightly curved and
measured 13.3 and 12.9 mm in two adult specimens
examined.

For the two adult specimens the ratio of the con-
dylobasal length to the length of the distal tract was
3.7 and 4.0; condylobasal length to the length of the
baculum was 4.7 and 4.8; length of the distal tract
to the length of the baculum was 1.2 and 1.3.

Specimens examined. — Total (4).

C. fumosus (4). — Colima: 2 mi W Colima, 4 (3 CM, 1 TTU).

Fig. 25. — Phallus and baculum of Cratogeomys gymnunis gymminis from Jalisco: 3 km S, 14 km W Ciudad Guzman (CM 55812),
illustrated as in Fig. I .

Cratogeomys gymnunis . — The phallus of C.
gymnunis (Fig. 25) is the largest of the species of
Cratogeomys examined. The length of the distal
tract ranged from 15.3 to 20.2 mm among adult
specimens examined. The length of the distal tract
is about four to five times greater than its width.
The length of the glans is a little more than half the
length of the distal tract. The sides of the glans,
viewed dorsoventrally, tend to be rather straight
and parallel. Viewed laterally the dorsal side is
more or less straight and the ventral side is recurved
with a flare at the collar. The collar of C. gymnunis
differs from previously discussed species of Cra-
togeomys in that it appears to be constricted around
the urethral opening. Compared to C. castanops
and C . fiimosus, C. gymnunis tends to have poorly-
developed urethral processes. A middorsal groove
and dorsal protuberances are present. The shape of
the dorsal protuberances are unique from previ-

ously discussed Cratogeomys in that the pair of
protuberances appear to exist for the entire length
of the glans. The only indication of a collar on the
dorsal side is a slight expansion of the dorsal pro-
tuberances in that area.

The epidermal structures of C. gymnunis have a
small, single, proximally oriented projection. The
size tends to be uniform, but not the pattern.

The length of the baculum of C. gymnunis (Fig.
25) ranged from 13.9 to 16.3 mm in adult specimens
examined. The baculum consists of a large bulbous
base, a well-curved shaft that is expanded in the
middle, and a distinct tip.

The range of ratios of the condylobasal length to
the length of the distal tract and the length of the
baculum was 3.6 to 4.5 and 4.4 to 5.1, respectively.
The ratio of the length of the distal tract to the
length of the baculum ranged from 1.1 to 1.2.

1982

WILLIAMS— GEOMYID PHALLI

51

Fig. 26. — Phallus and baculum of Cratogeomys meniami merriami from Mexico: 1 km S, 214 km W Rio Frio, 3,100 m (CM 55815),
illustrated as in Fig. 1.

Specimens examined. — Total (10).

C. g. gymnums (5). — Jalisco: 18 mi W Ciudad Guzman, 2
(TTU); 3 km S, 14 km W Ciudad Guzman, 3 (CM).

C. g. imparilns (1). — Michoacan: 8 mi E Opopeo, 1 (TTU).

C. g. msselli (3). — Jalisco: 12 mi S Toleman, 7,700 ft, 3 (KU).
C. g. tellns (1). — Jalisco: 15 mi E Ameca, 1 (TTU).

Cratogeomys merriami. — The phallus of C. mer-
riami (Fig. 26) is medium-sized and differs from oth-
er species of Cratogeomys by its slender appear-
ance. The distal tract is about six times longer than
it is wide and less than twice the length of the glans.
The length of the distal tract ranged from 15.3 to
16.3 mm in adult specimens examined. The sides of
the glans, viewed dorsoventrally, tend to be tapered
distally. Viewed laterally, the glans appears narrow
in the vicinity of the constriction, where it steadily

expands to the region of the collar. The collar is
similar to that of other species of Cratogeomys but
is quite prominent and often lacks convolutions.
The long protractile tip is similar to that of C. cas-
tanops and C . fumosus. The urethral processes are
unique in that they are mostly attached to the ure-
thral wall, thus restricted to movement. The mid-
ventral raphe, if present, is not very distinct. The
middorsal groove is usually shallow and broad. A
pair of reduced dorsal protuberances is present.

Epidermal structures of C. merriami are small
and consist of a small single projection. The distri-
bution of the structures tends to follow that of other
Cratogeomys, but the pattern consists of short, di-
agonal rows.

The baculum of C. merriami (Fig. 26) has a base

52

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

that tends to be more dorsoventrally compressed
than other species of Cratogeomys. The specimen
illustrated in Eig. 26 demonstrates this feature, but
may represent an extreme situation. Viewed dor-
soventrally, the baculum is similar to C. fumosus
in having distally converging sides that obscure the
position of the base and tip. The length of the bac-
ulum of adult specimens examined ranged from 14.0
to 14.8 mm. These measurements are greater than
the baculum length of 13.9 and 13.8 mm reported
by Burt (1960). Except for this difference, the de-
scriptions of this study generally agree with those
of Burt ( 1960).

The ratios of the condylobasal length to the
length of the distal tract and length of the baculum
ranged from 4.2 to 4.3 and 4.5 to 4.8, respectively.
The ratio of the length of the distal tract to the
length of the baculum ranged from 1.1 to 1.2.

Specimens examined. — Total (12).

C. m. fidvescens (1). — Veracruz: Guadalupe Victoria, 8,300 ft,
1 (TCWC).

C. m. irolonis (3). — Hidalgo: 4 km W Apan, 2 (CM); 2 km N,
3 km W Tepeapulco, 1 (CM).

C. m. meniami (8). — Distrito Federal: 1-8 mi E Sn. Gregoria
Altapulco, 2,270 m, 1 (KU); Santa Cruz Acalpixca, 2,270 m,
I (KU) 4 km W Xochimilco, 2,270 m, 1 (KU). Mexico: 55 km
SE Mexico City, 10,500 ft 1 (TCWC); I mi SSW Rio Erio, 1
(KU); 1 km S, IVi km W Rio Erio, 3,100 m, 1 (CM).Tlaxcala;
8 km S, 7 km W Calpulalpan, 2,900 m, 2 (CM).

Cratogeomys tylorhiniis. — The phallus of C. ty-
lorhimis (Eig. 27) is larger than all other species of
Cratogeomys, except C. gynmnriis. The dimen-
sions and features of C. tylorhinus are also more
similar to C. gymuurus than other Cratogeomys.
The length of the distal tract is four to five times
greater than its width. The length of the distal tract
ranged from 12.5 to 17.8 mm in adult specimens
examined. The length of the glans is more than half
the length of the distal tract. Like C. gymnurus the
sides of the glans, viewed dorsoventrally, are more
or less straight and parallel. Viewed laterally the
dorsal side is straight for most of its length and the
ventral side is recurved with a flaring in the collar
region. Also as in C. gymnurus, the collar of C.
tylorhinus is constricted around the urethral open-
ing. The shape and distinctness of most features,
including collar, constriction, midventral raphe,
middorsal groove, dorsal protuberances, and epi-
dermal structures, are very similar to C. gymnurus.

The baculum of C. tylorhinus (Eig. 27) is similar
to that of C. gymnurus by having a large bulbous
base, a well-curved shaft that is expanded in the

middle, and a distinct tip. The main bacular differ-
ence between these species is that C. tylorhinus has
a slightly smaller baculum with a length ranging
from 12.2 to 15.5 mm among adult specimens ex-
amined. This range of measurements is greater than
the range (10.0-12.5 mm) reported by Burt (1960).
However, descriptions of the bacula of this species
in this study correspond with those of Burt (1960).

The ratios of the condylobasal length to the
length of the distal tract and to the length of the
baculum range from 3.7 to 4.9 and 4.5 to 5.1, re-
spectively. The range of ratios of the length of the
distal tract to the length of the baculum was 1.0 to
1.4.

Specimens examined. — Total (12).

C. t. angustirostris (9). — Jalisco: 4 mi W Mazamitla, 6,600 ft,
4 (CM); 3 mi WSW Mazamitla, 1 (KU); 5 mi SW Mazamitla,
2 (TTU). Michoacan: Jesus Diaz, W slope Sierra Patamba,
7,500 ft, 1 (KU); 2 mi N Tarequato, 7,200 ft, 1 (KU).

C. t. planiceps (1). — Mexico: El Rio ( = San Bernabe), 14 mi
NW Toluca, 1 (KU).

C. t. tylorhinus (2). — Hidalgo: ca. 6 mi S Pachuca, 1 (TTU).
Mexico: Templo del Sol, Piramida de San Juan Taotihuacan,
1 (KU).

Cratogeomys zinseri. — The phallus of C. zinseri
(Fig. 28) is considered to be medium-sized com-
pared to the other species of Cratogeomys. The
length of the distal tract measured 14.9 and 13.2 mm
in two adult specimens examined. The length of the
distal tract is about four times the width of the distal
tract and about twice the length of the glans. The
general shape and features of the glans of C. zinseri
resemble those of C. gymnurus and C. tylorhinus,
with the main difference being a proportionally
smaller size. However, the collar region does not
appear to be as constricted as in C. gymnurus and
C. tylorhinus. Other features, including the mid-
ventral raphe, middorsal groove, and dorsal protu-
berances, show the same degree of development as
those two species. The dorsal protuberances are
different in that they lack the expanded region near
the collar and apex.

The epidermal structures of C. zinseri have a sin-
gle proximally oriented projection. Although the
size of the epidermal structures is uniform, the pat-
tern is not.

The bacula (Fig. 28) of two adult C. zinseri mea-
sured 12.6 and 1 1.6 mm. Viewed dorsoventrally the
baculum resembles that of C. gymnurus and C. ty-
lorhinus in having a large base, a shaft expanded in
the middle, and a distinct tip. However, the bacu-

Fig. 27. — Phallus and baculum of Cratogeomys tylorhiiuis angustirosiris from Jalisco: 4 mi W Mazamitla, 6,600 ft (CM 55831), illustrated
as in Fig. 1.

lum of C. zinseri differs from these species by being
relatively straighter.

The ratio of the condylobasal length to the length
of the distal tract was 4.4 and 4.7; condylobasal
length to the length of the baculum was 5.2 and 5.4;
length of the distal tract to length of the baculum
was 1.1 and 1.2.

Specimens examined. — Total (7).

C. zinseri (7). — Jalisco: Lagos de Moreno, 6,150 ft, 7 (CM).

Statistics

Age variation was observed in phallic characters
of Cratogeomys. However, analysis and documen-
tation of age variation is beyond the scope of the
present study. Table 6 provides standard statistics
(sample size, mean, range, standard error, and coef-

ficient of variation) for adult specimens of Crato-
geomys examined.

Examination of individual variation was con-
ducted on the largest samples of Cratogeomys hav-
ing individuals from the same geographical area that
would be assumed to be part of a continuous pop-
ulation. The samples used to examine individual
variation were C. castanops perplanus from Lea
Co., New Mexico, and Gaines Co., Texas (calcu-
lated separately from sample of C. c. perplanus pre-
sented in Table 6 that included a larger geographical
area), and C. tylorhinus angiistirostris. Coefficients
of variation of phallic and bacular characters in
these two samples ranged from 4.9 to 12.3. The av-
erage coefficient of variation for C. castanops and
C. tylorhinus was 7.1 and 8.5, respectively. In both

Fig. 28. — Phallus and baculum of Crutogeomys zinseri from Jalisco: Lagos de Moreno, 6,150 ft (CM 55848), illustrated as in Fig. 1.

species the length of the distal tract and length of
the glans had the lowest values with C. castonops
having 5.4 and 5.5, and C. tylorhinus having 4.9 and
5.8, respectively. The length of the baculum, which
had low values in Thomomys and Geomys, did not
rank among the lower values of Cratogeomys (6. 1
in C. castanops and 9.1 in C. tylorhinus). In C.
casta nops the width of the glans across base (5.8)
was less than the value for the length of the bacu-
lum; in C. tylorhinus the width of the glans across
base (8.7) and width of baculum base (7.5) were
less. In both species the length of the protractile tip
had the greatest coefficient of variation (8.8 in C.
castanops and 12.3 in C. tylorhinus).

Sample sizes and limited geographical coverage
of specimens representing Pappogeomys and Cra-
togeoniys species examined were inadequate to
provide a meaningful analysis of geographic varia-
tion within a species. Examination and comparison
of measurements of individuals did not reveal the
type of variation observed in the genus Thomomys.

Examination of variation between Pappogeomys
and species of Cratogeomys was restricted to multi-
variate analysis, using sample means with the
MINT statistical computer program. Samples of a
common species were not pooled because the sam-
ple sizes were often limited, and combining samples
of the same species could lead to a biased repre-
sentation of the species if geographical variation
occurs within the species. The samples of Pappo-
geomys and Cratogeomys used in the multivariate
analysis, with corresponding number for reference
purposes, are three subspecies of P. bulleri (bulleri,
1; infuscus, 2; lutulentus, 3), four subspecies of C.
castanops (castanops, 4; perplanus, 5; pratensis,
6; rubellus, 7), C . fumosus (8), three subspecies of
C. gymnurus (gymnurus, 9; russelli, 10; tellus, 11),
C. merriami merriami (12), two subspecies of C.
tylorhinus (angustirostris, 13; tylorhinus, 14), and
C. zinseri (15).

The distance phenogram produced by the MINT
program is illustrated in Fig. 29. The cophenetic

1982

WILLIAMS— GEOMYID PHALLI

55

buller i buller i

bulleri i i

nf uscus

buller i li

utulentus

castanops

castanops

castanops

perplanus

castanops

pratens i s

castanops

rubellus

z i nser i
f umosus

t vlorh i nus tvlorhinus

qymnur us r ussell i

gymnur us tellus

tylorh i nus anqust i r ostr i s

gymnur us gymnur us

me r r i am i mer r i am i

Fig. 29. — Distance phenogram of 15 samples, representing Pappogeomys and Cratogeomys, resulting from clustering by unweighted
pair-group method using arithmetic averages (UPGMA). The cophenetic correlation coefficient for the phenogram is 0.716.

correlation value of the phenogram was 0.716. The
phenogram is split into two major groups — one
group consisting of P. biilleri, C. castanops, and C.
zinseri, and a second group consisting of C. fumo-
siis, C. gymnums, C. merriami, and C. tylorhinus.
In the first group all samples of P. bulleri cluster
together and are separate from all samples of C.
castanops which also cluster together. C. zinseri
clusters with the samples of C. castanops but is still
separate and distinct. In the second group C. mer-
riami separates from a cluster including C. fnmo-
sns, C. gymnnrns, and C. tylorhinus. This final clus-
tering does not conform to taxonomic groups as it
did in other parts of the phenogram; C. fiimosus
clusters very closely with C. t. tylorhinus and both
are grouped with C. g. russelli\ C. g. tellus and C.
t. angustirostris also form a group together; C. g.
gymnums forms a group by itself which is distinct
from the rest.

The first three principal components extracted
from the matrix of correlation among characters are
illustrated in Fig. 30. The three-dimensional projec-
tions cluster into three distinct groups along com-

ponent I. All of the samples of P. bulleri (1, 2, 3)
form a definite group on the left side. Next all sam-
ples of C. castanops (4, 5, 6, 7) form a distinct
group toward the middle of the plot. The last group
consists of broadly scattered projections that rep-
resent C. fumosus (8), C. gymnums (9, 10, 11), C.
merriami (12), C. tylorhinus (13, 14), and C. zinseri
(15). C. zinseri (15) can be separated from the oth-
ers with the first component but certainly not by

Table 7. — Factor matrix from correlation among nine characters
of one species (three samples) of Pappogeomys and six species
( 14 samples) of Cratogeomys examined.

Character

Component \

Component II

Component III

Condylobasal length

0.939

-0.015

0.251

Length of distal tract

0.970

-0.101

-0.118

Length of glans

0.979

-0.160

-0.080

Length of protractile tip

0.869

-0.480

-0.026

Width of glans across collar

0.842

0.442

0.223

Width of glans across base

0.945

0.166

0.187

Length of baculum

0.975

-0.163

-0.084

Width of bacular base

0.971

-0.003

0.049

Height of bacular base

0.795

0.401

-0.440

56

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 20

Fig. 30. — Three dimensional projection of 15 samples representing Pappogeomys (P. biilleri, 1-3) and Cratogeomys (C. castanops,
4-7; C. fumosus, 8; C. gymnurus, 9-11; C. merriami, 12; C. tylorhinus, 13-14; C. zinseri, 15) onto the first three principal components
based upon a matrix of correlation among one cranial, five phallic, and three bacular measurements. Components I and II are indicated
in the plots and component III is represented by height.

the magnitude of P. bulleri and C. castanops. C.
zinseri (15) is also separated with component II. C.
merriami (12) is also separated from the other sam-
ples with component II. C. gymnurus (9, 10, 11) is
distinct from C. fumosus along component I but
overlaps with C. tylorhinus (13, 14) in the first two
components and is barely separated with compo-
nent III. C . fumosus (8) and C. tylorhinus (13, 14)
are separated with component III.

The amount of phenetic variation explained by

components I, II, HI are 85.2%, 7.5%, and 4.1%,
respectively. Results of principal components anal-
yses showing the influence of each character for the
first three components are given in Table 7. All
characters are heavily weighted in the first factor.
In the second factor the characters having the great-
est weighing were length of protractile tip of glans,
width of glans across collar, and height of bacular
base. The height of the bacular base also had the
heaviest weighing in the third factor.

DISCUSSION

This study found the phallus and baculum of
members of the family Geomyidae to be relatively
simple and comparable to Liomys (Genoways,
1973), Reithrodontomys, and some species of Pero-
myscus (Hooper, 1958, 1959). Throughout the fam-
ily the basic phallic and bacular features remain uni-
form. The phallus always includes a glans that is
featured with a v/ell-defined collar that marks a con-
spicuous protractile tip; urethral processes are al-

ways present and are situated on the ventral side of
the protractile tip; epidermal structures occur on
the surface of the glans proximal to the collar and
on the dorsal surface of the protractile tip. The bac-
ulum always has a simple osseous shaft with some
terminal differences indicating the presence of a tip
and basal region. These characters typify the family
Geomyidae. Additional modification and elabora-
tion of the phallus and baculum, similar to that ob-

1982

WILLIAMS— GEOMYID PHALLI

57

served in other rodent families (for example, Sci-
uridae, Cricetidae, Muridae, or Echymidae), is
reduced in geomyid rodents (Burt, 1960; Hooper,
1958, 1959, 1960, 1961, 1962; Hooper and Hart,
1962; Hooper and Musser, 1964; Lidicker, 1968;
Williams et al., 1980).

Because the basic phallic and bacular features are
relatively uniform in the family Geomyidae, it is
difficult to define unique characters that would dif-
ferentiate the genera. Thomomys is the only genus
that has representatives (L. clusius, T. idahoensis,
T. mazama, T. monticola, and T. talpoides), with
long, narrow phalli and bacula, and phalli and bac-
ula that are large in proportion to body size (as in-
dicated by condylobasal length). However, there
are other representatives of the genus Thomomys
(T. bottae, T. hidbivorous, T. townsendii, and T.
umbrinus) that have similar dimensions and pro-
portions to most members of other genera. At this
point, characters (excluding size) such as the width
of the collar, distinctness of the collar (particularly
from a dorsal view), and number of projections on
epidermal structures, are the most useful in further
differentiating genera. Eor instance, Geomys and
Zygogeomys differ from Thomomys, Orthogeomys,
Cratogeomys, and Pappogeomys by having more
than a single projection on each epidermal struc-
ture. Pappogeomys , Cratogeomys , and some
species of Thomomys typically have a glans that
tapers distally. The collar of these species often
lacks flaring and is not well-defined dorsally. Geo-
mys typically has a collar that is distinct from all
aspects and is often the widest part of the glans.
Although various characters and combinations of
characters can be used for generic identification, it
is not possible to select specific features that will
typify all members of a genus and differentiate them
from all other members of the family. However, in
some species, characteristics unique to that species
can be used for differentiation from all other mem-
bers of the family. Eor instance, the phallus of Tho-
momys bulbivoroits and the bacula of Geomys pi-
netis and Cratogeomys cast an ops have features
that are distinct and characterstic to the respective
taxon.

At the species level, nongeographic and geo-
graphic variation contribute to many differences
observed in features and dimensions. It is possible
that physical problems of working with soft, deli-
cate, anatomical parts could account for some of
the variation observed. Because of the variation
occurring within a species, identifying phallic and

bacular characteristics of any particular species
should be based on bacular dimensions and overall
morphological features.

In spite of the limitations imposed on selecting
discernible characteristics, this study revealed
some interesting relationships among species of the
same genus. However, it is believed that no system-
atic conclusions should be based solely on findings
in this study. Instead, it is intended that for any
taxonomic group, data presented should be used in
conjunction with other types of data published by
other investigators. In this sense this study presents
an approach to systematic problems among pocket
gophers that can be useful in providing supportive
or refutive arguments for previous systematic stud-
ies.

The systematic relationships among the species
of Thomomys have been speculated by several in-
vestigators. Elliot (1903) recognized the uniqueness
of T. hidbivorous and erected the new subgenus
Megascapheus to separate it from the other
species. Bailey (1915) distinguished species groups
of Thomomys based on whether the shape of the
rostrum was heavy or slender. Although T. hidbi-
vorous remained unique it was grouped in the
heavy-rostrum species along with bottae, town-
sendii, and umbrinus. However, Russell (1968u)
maintained that T. bidbivorous is a group distinct
from the heavy-rostrum and slender-rostrum species
groups. At the present time arguments presented by
Thaeler (1980) probably best represent the relation-
ships among the species at the subgeneric level.
Based on chromosome data two species groups are
apparent. These data as well as characters used in
other studies, such as sphenoid fissure (Durrant,
1946; Hall, 1946; Hall and Kelson, 1959), infraor-
bital canals (Durrant, 1946), enamel configuration
of lower fourth premolar and angular process of ra-
mus (Thaeler, 1980), prompted Thaeler ( 1980) to
propose two subgenera of Thomomys — Thomomys
and Megascapheus. The former includes T. clusius,
T. idahoensis, T. mazama, T. monticola, and T.
talpoides', the latter includes T. bottae, T. bulbi-
vorous, T. townsendii, and T. umbrinus.

Data presented in this study concerning the phal-
lus and baculum of Thomomys corresponds to the
systematic relationships proposed by Thaeler
(1980). The subgenus Thomomys is characterized
by a long, slender and simple phallus, a long bac-
ulum with a less distinct base, and a relatively long
phallus and baculum compared to body size; the
subgenus Megascapheus has a shorter, broader.

58

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

and more complex phallus, a shorter baculum with
a distinct base, and a relatively short phallus and
baculum compared to body size. Mean ratios of the
condylobasal length to the length of the distal tract
and length of the baculum, rank the species of Tho-
momys in the following order: mazama, talpoides,
monticola, idahoensis, clitsiits, hottae, townsendii,
iimhrinus, and bidbivorous. Clearly the two sub-
genera do maintain distinct differences in the phal-
lus and baculum. As in previous studies (Elliot,
1903; Russell, 1968«) T. bidbivorous remains rela-
tively distinct from other species. The phallus is
unique in shape, features, and ratio with baculum
length. However, the similarities between T. bid-
bivorous and other members of the subgenus Me-
gascapheus, as well as data presented by Thaeler
(1980), indicate that this species is a member of the
subgenus in spite of its uniqueness.

The dimensions of the phallus and baculum of the
genus Thomomys are extremely variable among
species. Although examination of variation within
a species was restricted by the few number of sub-
species available, this study indicates that consid-
erable variation persists to the subspecific level. It
is possible that examination of the phallus and bac-
ulum of specimens throughout the range of a
species could serve as an indicator of subspecific
groups — particularly in T. talpoides. In view of the
variation in the dimensions of the phallus and bac-
ulum of T. talpoides, it is interesting that both mor-
phological and metrical data indicate a similarity
between T. monticola and T. talpoides. The simi-
larity is even more intriguing because one of the
subspecies of T. talpoides that is most similar to T.
monticola is T. t. quadratiis which occurs in the
same geographical region as T. monticola (Thaeler,
1968).

The systematic relationships of species of Geo-
mys have been discussed by several investigators
with a degree of uncertainty resulting from conflict-
ing data. It is generally accepted that two different
groups of Geomys exist — the eastern group and the
western group. The eastern form, or the pinetis-
species group previously included G. colonus, G.
cumberlandius, G. fontanelus, and G. pinetis until
they were all synonymized under G. pinetis by
Williams and Genoways ( 1980). There is some ques-
tion as to which western species of Geomys is most
closely related to G. pinetis. Russell (1968u) sug-
gests that G. bursarius and G. pinetis differentiated
from a common ancestor during the Sangamon
time. This assumption is based on examination of

the fossil record and cranial characteristics. The
fossil record in Elorida has led Martin (1974a,
1974/j) and Martin and Webb (1974) to suggest that
G. personatus and G. pinetis are more closely re-
lated and could be conspecific. Karyotypic data
presented by Davis et al. (1971) and Williams and
Genoways (1975) confirms the specific status of
each species but does not determine whether G.
bursarius or G. personatus is more closely related
to G. pinetis. Eurther questions of systematic re-
lationships of species of Geomys occur among the
western species — G. arenarius, G. attwateri, G.
bursarius, G. personatus, and G. tropicalis. Mor-
phological and karyological data support the taxo-
nomic status of these species (Davis et al., 1971;
Davis, 1940; Hart, 1978; Honeycutt and Schmidly,
1979; Kennedy, 1958, 1959; Merriam, 1895; Tucker
and Schmidly, 1981; Williams and Genoways, 1977,
1978, 1981), but different relationships among many
of these have been suggested. Cranial characteris-
tics examined by Alvarez (1963) and electrophoretic
data presented by Selander et al. ( 1973) suggest that
G. arenarius is closely related to G. personatus and
G. tropicalis. Although G. arenarius externally and
cranially resembles G. personatus of the lower Rio
Grande Valley, Davis (1940) speculated that G. cir-
enariiis is more closely related to G. lutescens ( =
bursarius) of Texas and New Mexico for physio-
graphic reasons. After examining the fossil record
and cranial characteristics, Russell (1968a) suggest-
ed that G. arenarius and G. personatus both dif-
ferentiated independently from G. bursarius, prob-
ably during the Wisconsin glaciation. Penney and
Zimmerman (1976) suggested that both of these
species, as well as G. tropicalis, differentiated in-
dependently and at different times from G. bursar-
ius. However, cranial characters (Alvarez, 1963),
zoogeography (Selander et al., 1962), and parasite
data (Price and Emerson, 1971) indicate G. tropicalis
has closer affinities to G. personatus than G. bur-
sarius. Therefore the main systematic questions in
Geomys concern the relationships of G. arenarius
and G. pinetis to G. bursarius or G. personatus.
Also, the relationship of the recently resurrected G.
attwateri (Tucker and Schmidly, 1981) to other
species of Geomys is unknown.

Comments regarding the relationships among
species of Geomys, based on phallic and bacular
characters, are difficult to make because these char-
acters are eonservative (particularly when com-
pared to other genera, such as Thomomys and Cra-
togeomys). The primary difference observed

1982

WILLIAMS— GEOMYID PHALLI

59

between species was size, but this feature cannot
be considered indicative of any particular species
because of geographical variation. Size character-
istics might be most useful in comparisons at the
level of subspecies. The width of bacular base was
particularly useful in differentiating the eastern
pocket gophers (G. pinetis) and the western pocket
gophers (G. arenarius, G. attwateri, G. bursariiis,
G. personatus, and G. tropicalis). In this study oth-
er similarities among species were best indicated by
multivariate analysis, which were able to distin-
guish some of the species. This analysis supports
the recognized subspecific differences of G. p.
streckeri (Williams and Genoways, 1981), and in-
dicates that both G. pinetis and G. arenarius are
more similar to G. bursariiis than to G. personatus,
thus supports work done by Russell (1968a). Fur-
thermore, G. arenarius is most similar to the closest
geographical race of G. bursariiis, G. b. knoxjonesi.
This similarity agrees with comments made by Da-
vis (1940) about the relationships of G. arenarius.
G. attwateri was most similar to G. bursariiis.

The first comprehensive effort to determine the
systematic relationships of the genera Cratogeoniys
and Pappogeomys was done by Merriam (1895).
The genera (followed by the currently recognized
species in parentheses) that Merriam (1895) recog-
nized were Pappogeomys (biilleri), Cratogeoniys
(castanops and merriami), and Platygeoniys ifii-
mosiis, gymniiriis, and tylorhiniis). It was suggested
in that revision that gymniiriis and tylorhiniis were
very closely related, and that castanops could be
subgenerically distinct from other Cratogeoniys.
Goldman ( 1939a, \939b) reviewed Platygeoniys and
Pappogeomys. In these reviews several new taxa
were described, including the currently recognized
species neglectiis and zinseri that were named un-
der Platygeoniys. Hooper ( 1946) later synonymized
Platygeoniys under the genus Cratogeoniys. Rus-
sell (1968/?) conducted one of the most comprehen-
sive studies of this group. In that revision Crato-
geomys (including the Platygeoniys group) was
synonymized under Pappogeomys. Under this ge-

neric designation the subgenus Pappogeomys in-
cluded alcorni (Russell, 1957) and biilleri and the
subgenus Cratogeoniys was split into the castanops
species-group (castanops and merriami) and the
gymniiriis species-group {fiimosiis, gymniiriis, neg-
lectiis, tylorhiniis, and zinseri). In this arrangement
Russell (1968/?) suggested that gymniiriis, tylorhin-
iis, and zinseri are the closest related species of the
subgenus Cratogeoniys. However, electrophoretic
data presented by Honeycutt and Williams (1982)
create several questions concerning the systematic
arrangements suggested by Russell (1968/?). Ac-
cording to Honeycutt and Williams (1982), the sub-
genera Pappogeomys and Cratogeoniys (as estab-
lished by Russell, 1968/?) deserve generic
recognition; C. castanops and C. merriami are not
monophyletic; C. gymniiriis and C. tylorhiniis are
similar enough to each other to suggest a conspe-
cific relationship. Characteristics and statistical
analyses of the phalli and bacula of Cratogeoniys
and Pappogeomys found in this study support many
of the relationships suggested by other investi-
gators. Pappogeomys (represented by P. biilleri) is
most easily distinguished from Cratogeoniys by
size. Within Cratogeoniys, each species of the cas-
tanops species-group (C. castanops and C. nier-
rianii), maintain characteristics (particularly bacular
characteristics) and dimensions that differentiate
them from other members of the genus. However,
the anatomical and metrical differences observed
between C. castanops and C. merriami support
Merriam ( 1895) and Honeycutt and Williams ( 1982)
in that C. castanops could be subgenerically dis-
tinct from other members of the genus. In the gyni-
niiriis species-group anatomical characters tend to
be more subtle and the main difference among
species is size. This is particularly true for C. gym-
niiriis, C. tylorhiniis, and C. zinseri, thus supporting
the close similarity of these species (particularly C.
gymniiriis and C. tylorhiniis) as discussed by Mer-
riam (1895), Russell (1968/?), and Honeycutt and
Williams (1982).

SUMMARY

This study involving the phalli and bacula of the
family Geomyidae was intended to serve as an ap-
proach to the systematics of the family. Material
used in this study included 388 specimens repre-

senting all six genera and 24 of the currently rec-
ognized species in the family.

The methods used and material available pre-
cluded further examination of other problems that

60

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

should be investigated. One such problem is the
growth and development of the phallus and bacu-
lum. It would be very interesting to know about the
size and feature changes that occur with maturity
to a reproductively active age. Another interesting
problem is the detailed documentation of geograph-
ic variation of the phallus and baculum of geomyid
species that have extensive distributions. It is
hoped that data presented in this study will be use-
ful in approaching such problems.

Based on data presented in this study several sys-
tematic relationships are suggested, most of which
are in agreement with previous investigators. Phal-
lic and bacular data reveal two groups of Thonio-
mys which agree with current subgeneric designa-
tions. The subgenus Thomomys includes T. clusius,
T. idahoensis, T. mazcima, T. monticola, and T.
talpoides: the subgenus Megascapheus includes T.
hottae, T. bidbivorous, T. townsendii, and T. um-
briniis. Although the phallus of T. bidbivorous is
unique, it is considered to be an extreme case of
variation existing in the subgenus Megascapheus.

The systematic relationships of Geomys indicat-
ed by phallic and bacular data reveal two species
groups which correspond to geographic distribu-
tions. One species-group includes G. pinetis; the
other species-group includes the western species
(G. arenarius, G. attwateri, G. bursarius, G. per-
sonatus, and G. tropicalis). Although data are pre-

sented to suggest that G. pinetis, G. arenarius, and
G. attwateri have a closer phenetic similarity to G.
bursarius than any other species of Geomys, addi-
tional data are needed for a more substantiated sys-
tematic arrangement of all species of Geomys.

Phallic and bacular differences of the genus Pap-
pogeomys (used by Russell, 1968/;) indicate species
groups that are in agreement with taxonomic
changes suggested by Honeycutt and Williams
(1982). These changes include generic recognition
of Pappogeomys and Cratogeomys. From material
examined in this study Pappogeomys includes bul-
leri\ Cratogeomys includes castanops, merriami,
fiimosus, gymnurus, tylorhinus, and zinseri. Within
the genus Cratogeomys there are probably three
distinct lineages of differentiation that may deserve
subgeneric recognition. These lineages include one
group consisting of C. castanops, a second group
consisting of C. merriami, and a third group which
includes the other species of Cratogeomys. In the
third group the relationship between C. gymnurus
and C. tylorhinus is considered to be extremely
close.

It is reiterated that data presented in this study
are not intended to be a sole basis for systematic
conclusions. Instead it is hoped that this new infor-
mation will be useful in determining the systematics
of the family Geomyidae when it is used in con-
junction with data from other studies.

ACKNOWLEDGMENTS

I sincerely appreciate the assistance provided by those indi-
viduals and institutions that helped acquire material for this proj-
ect. Those individuals that assisted in fieldwork include Dr. R.

J. Baker, Ms. L. E. Carroll, Ms. C. H. Carter, Mr. R. C. Dowler,
Dr. H. H. Genoways, Dr. I. F. Greenbaum, Ms. J. A. Groen,
Mr. D. R. Harvey, Ms. J. A. Homan, Mr. T. Kasper, Ms. M.
E. McGhee, Dr. J. C. Patton, Mr. E. F. Pembleton, Dr. J. Ra-
mirez-Pulido, Mr. M. J. Smolen, Mr. S. L. Tennyson, and Ms.

K. D. Williams. I am also grateful to the curators of mammal
collections for making specimens in their care available to me.
These curators include Dr. R. J. Baker (The Museum, Texas
Tech University), Dr. R. S. Hoffmann (Museum of Natural His-
tory, University of Kansas), Dr. S. R. Humphrey (Florida State
Museum, University of Florida, Gainesville), Dr. J. F. Patton
(Museum of Vertebrate Zoology, University of California,
Berkeley), Dr. D. J. Schmidly (Texas Cooperative Wildlife Col-
lection, Texas A&M University), and Dr. C. S. Thaeler (De-

partment of Biology, New Mexico State University). I am es-
pecially grateful to H. H. Genoways for his encouragement and
making several of the collecting trips possible. I am further in-
debted to my wife, K. Williams for assisting in the final prepa-
ration of the manuscript, and to Ms. B. A. McCabe and Ms. N.
J. Parkinson for typing the manuscript. Material used in this
paper was partly acquired from specimens collected in conjunc-
tion with other research projects. Funding for some of these
research projects was received from the Theodore Roosevelt
Memorial Fund of the American Museum of Natural History,
Research Funds of Texas Tech University, and the M. Graham
Netting Research Fund (established by a grant from the Cordelia
S. May Charitable Trust) of Carnegie Museum of Natural His-
tory. Funds for computer utilization at Carnegie Mellon Uni-
versity in Pittsburgh were made available by Carnegie Museum
of Natural History.

LITERATURE CITED

Alvarez, T. 1963. The Recent mammals of Tamaulipas, Mex-
ico. Univ. Kansas Publ., Mus. Nat. Hist., 14:363-473.

Anderson, S. 1966. Taxonomy of gophers, especially Tho-
momys in Chihuahua, Mexico. Syst. Zool., 15:189-198.

1982

WILLIAMS— GEOMYID PHALLI

61

Bailey, V. 1915. Revision of the pocket gophers of the genus
Thomomys. N. Amer. Fauna, 39:1-136.

Burt, W. H. 1960. Bacula of North American mammals. Misc.
Publ. Mus. Zool., Univ. Michigan, 113:1-76.

Davis, B. L., S. L. Williams, and G. Lopez. 1971. Chromo-
somal studies of Geomys. J. Mamm., 52:617-620.

Davis, W. B. 1940. Distribution and variation of pocket gophers
(genus Geomys) in the southwestern United States. Bull.
Texas Agric. Exper. Sta., 590:1-38.

Durrant, S. D. 1946. The pocket gophers (genus Thomomys)
of Utah. Univ. Kansas Publ., Mus. Nat. Hist., 1:1-82.

Elliot, D. G. 1903. A list of mammals obtained by Edmund
Heller from the coast region of northern California and Or-
egon. Publ. Field Columbia Mus., Zool. Ser., 76:175-197.

Gabriel, K. R. 1964. A procedure for testing the homogeneity
of all sets of means in analysis of variance. Biometrics,
20:459-477.

Genoways, H. H. 1973. Systematics and evolutionary relation-
ships of spiny pocket mice, genus Liomys. Spec. Publ.
Mus., Texas Tech Univ., 5:1-368.

Goldman, E. A. 1939a. Review of the pocket gophers of the
genus Platygeomys. J. Mamm., 20:87-93.

. 19396. The pocket gophers of the genus Pappogeomys.

J. Mamm., 20:93-98.

Hall, E. R. 1946. Mammals of Nevada. Univ. California Press,
Berkeley, 710 pp.

. 1981. The mammals of North America. John Wiley and

Sons, New York, l:xv -I- 1-600 -I- 99.

Hall, E. R., and K. R. Kelson. 1959. The mammals of North
America. The Ronald Press Co., New York, 1:1-546 -I- 79.

Hart, E. B. 1978. Karyology and evolution of the plains pocket
gopher, Geomys hursariiis. Occas. Papers Mus. Nat. Hist.,
Univ. Kansas, 71:1-20.

Hill, J. E. 1937. Morphology of the pocket gopher mammalian
genus Thomomys. Univ. California Publ. Zool., 42:81-171.

Hoffmeister, D. F. 1969. The species problem in the Tho-
momys bottae-Thomomys umbrinus complex of gophers in
Arizona. Misc. Publ., Univ. Kansas Mus. Nat. Hist.,
51:75-91.

Hoffmeister, D. F., and M. R. Lee. 1963. The status of the
sibling species Peromyscus merriami and Peromyscus ere-
micits. J. Mamm. 44:201-213.

Honeycutt, R. L., and D. J. Schmidly. 1979. Chromosomal
and morphological variation in the plains pocket gopher,
Geomys bur sarins, in Texas and adjacent states. Occas.
Papers Mus., Texas Tech Univ., 58:1-54.

Honeycutt, R. L., and S. L. Williams. 1982. Genic differ-
entiation in pocket gophers of the genus Pappogeomys with
comments on intergeneric relationships in the Subfamily
Geomyinae. J. Mamm., 63:208-217.

Hooper, E. T. 1946. Two genera of pocket gophers should be
congeneric. J. Mamm., 27:397-399.

. 1958. The male phallus in mice of the genus Peromys-
cus. Misc. Publ. Mus. Zool., Univ. Michigan, 105:1-24.

. 1959. The glans penis in five genera of cricetid rodents.

Occas. Papers Mus. Zool., Univ. Michigan, 613:1-11.

. 1960. The glans penis in Neotoma (Rodentia) and allied

genera. Occas. Papers Mus. Zool., Univ. Michigan,
618:1-21.

. 1961. The glans penis in Proechimys and other cavio-

morph rodents. Occas. Papers Mus. Zool., Univ. Michigan,
623:1-18.

. 1962. The glans penis in Sigmodon, Sigmomys, and

Reithrodon (Rodentia, Cricetinae). Occas. Papers Mus.
Zool., Univ. Michigan, 625: 1-1 1 .

Hooper, E. T., and B. S. Hart. 1962. A synopsis of Recent
North American microtine rodents. Misc. Publ. Mus. Zool.,
Univ. Michigan, 120:1-68.

Hooper, E. T., and G. G. Musser. 1964. The glans penis in
Neotropical cricetines (family Muridae) with comments on
classification of muroid rodents. Misc. Publ. Mus. Zool.,
Univ. Michigan, 123:1-57.

Ingles, L. G. 1965. Mammals of the Pacific states. Stanford
Univ. Press, Stanford. California, xii -I- 506 pp.

Johnson, M. L., and S. B. Benson. 1960. Relationship of the
pocket gophers of the Thomomys mazama-talpoides com-
plex in the Pacific Northwest. Murrelet, 41:17-22.

Kennerly, T. E., Jr. 1958. The baculum in the pocket gopher.
J. Mamm., 39:445—446.

. 1959. Contact between the ranges of two allopatric

species of pocket gophers. Evolution, 13:247-263.

Lidicker, W. Z., Jr. 1968. A phylogeny of New Guinea rodent
genera based on phallic morphology. J. Mamm., 48:609-643.

Long, C. A. 1964. The baculum of pocket gophers of south-
western Wyoming. Trans. Kansas Acad. Sci., 66:754-756.

Long, C. A., and T. Frank. 1968. Morphometric variation and
function in the baculum with comments of correlation of
parts. J. Mamm., 49:32-43.

Martin, R. A. 1974<;. Fossil mammals from the Coleman IIA
Fauna, Sumter County. Pp. 35-99, in Pleistocene mammals
of Florida (S. D. Webb, ed.), Univ. Presses Florida, Gaines-
ville, X -I- 270 pp.

. 19746. Fossil vertebrates from the Haile XIVA Fauna,

Alachua County. Pp. 100-113, in Pleistocene mammals of
Florida (S. D. Webb, ed.), Univ. Presses Florida, Gaines-
ville, X -I- 270 pp.

Martin, R. A., and S. D. Webb. 1974. Late Pleistocene mam-
mals from the Devil’s Den Fauna, Levy County. Pp.
114-145, in Pleistocene mammals of Florida (S. D. Webb,
ed.), Univ. Presses Florida, Gainesville, x -I- 270 pp.

Merriam, C. H. 1895. Monographic revision of the pocket go-
phers, family Geomyidae (exclusive of the species of Tho-
momys). N. Amer. Fauna, 8:1-258.

Patton, J. L. 1973. An analysis of natural hybridization be-
tween the pocket gophers, Thomomys hottae and Thomo-
mys umhrinus, in Arizona. J. Mamm., 54:561-584.

Penney, D. F., and E. G. Zimmerman. 1976. Genic divergence
and local population differentiation by random drift in the
pocket gopher genus Geomys. Evolution, 30:473-483.

Power, D. M. 1970. Geographic variation of red-winged black-
birds in central North America. Univ. Kansas Publ., Mus.
Nat. Hist., 19: 1-83.

Price, R. D., and K. C. Emerson. 1971. A revision of the genus
Geomydoecus (Mallophaga:Trichodectidae) of the New
World pocket gophers (Rodentia: Geomyidae). J. Med.
Ent., 8:228-257.

Russell, E. L. 1973. Improved methods for staining bones of
small fetuses and vertebrates in alizarin red S. BioScience,
23:366-367.

Russell, R. J. 1957. A new species of pocket gopher (genus
Pappogeomys) from Jalisco, Mexico. Univ. Kansas Publ.,
Mus. Nat. Hist., 9:357-361.

. 1968(7. Evolution and classification of the pocket go-

62

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 20

phers of the subfamily Geomyinae. Univ. Kansas Publ.,
Mus. Nat. Hist., 16:473-579.

. 1968/7. Revision of pocket gophers of the genus Pap-

pogeomys. Univ. Kansas Publ. Mus. Nat. Hist., 16:581-776.

Selander, R. K., D. W. Kaufman, R. J. Baker, and S. L.
Williams. 1973. Genic and chromosomal differentiation in
pocket gophers of the Geoniys hursahus group. Evolution,
28:557-564.

Sherman, H. B. 1940. A new species of pocket gopher (Geo-
mys) from eastern Georgia. J. Mamm., 21:341-343.

Thaeler, C. S., Jr. 1968. An analysis of the distribution of
pocket gopher species in northeastern California (genus
Thamomys). Univ. California Publ. Zook, 86:1-46.

. 1972. Taxonomic status of the pocket gophers, Tho-

momys idahoensis and Thamomys pygmaeus (Rodentia,
Geomyidae). J. Mamm., 53:417-428.

. 1980. Chromosome numbers and systematic relations

in the genus Thamomys (Rodentia: Geomyidae). J. Mamm.,
61:414-422.

Thaeler, C. S., Jr., and L. L. Hinesley. 1979. Thamomys
chisiiis, a rediscovered species of pocket gopher. J. Mamm.,
60:480-488.

Tucker, P. K., and D. J. Schmidly. 1981. Studies of a contact
zone among three chromosomal races of Geomys bursarius
in East Texas. J. Mamm., 62:258-272.

Williams, S. L., and H. H. Genoways. 1975. Karyotype of
Geomys pinetis (Mammalia: Geomyidae), with a discussion
of the chromosomal relationships within the genus. Exper-
ientia, 31:1 141-1142.

. 1977. Morphometric variation in the tropical pocket go-
pher (Geomys tropicalis). Ann. Carnegie Mus., 46:245-264.

. 1978. Review of the desert pocket gopher, Geomys ar-

enariiis (Mammalia: Rodentia). Ann. Carnegie Mus.,
47:541_570.

. 1980. Morphological variation in the southeastern pock-
et gopher, Geomys pinetis (Mammalia: Rodentia). Ann.
Carnegie Mus., 49:405-453.

. 1981. Systematic review of the Texas pocket gopher,

Geomys personatiis (Mammalia: Rodentia). Ann. Carnegie
Mus., 50:435-473.

Williams, S. L., J. C. Hafner, and P. G. Dolan. 1980. Gians
penis and bacula of five species of Apodemiis (Rodentia:
Muridae) from Croatia, Yugoslavia. Mammalia, 44:245-258.

ST

/!

/

(

I

{

1

;5

ti..W

V:!'- V'

r:':.

TJ?*' .''r?'

•5

■\

iim

’1

d

• '( iiff%

Copies of the following Bulletins of Carnegie Museum of Natural History may be obtained at the prices
listed from the Publications Secretary, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pitts-
burgh, Pennsylvania 15213.

1. Krishtalka, L. 1976. Early Tertiary Adapisoricidae and Erinaceidae (Mammalia, Insectivora) of North

America. 40 pp., 13 figs $2.50

2. Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the

late Pleistocene paleoecology of the central Appalachian Mountains of Virginia. 88 pp., 21
figs $12.00

3. Wetzel, R. M. 1977. The Chacoan peccary, Catagonus wagneri (Rusconi). 36 pp., 10 figs. . . . $6.00

4. Coombs, M. C. 1978. Reevaluation of early Miocene North American Moropus (Perissodactyla, Chal-

icotheriidae, Schizotheriinae). 62 pp., 28 figs $5.00

5. Clench, M. H., and R. C. Leberman. 1978. Weights of 151 species of Pennsylvania birds analyzed by

month, age, and sex. 87 pp $5.00

6. Schlitter, D. A. (ed.). 1978. Ecology and taxonomy of African small mammals. 214 pp., 48 figs.

$15.00

7. Raikow, R. J. 1978. Appendicular myology and relationships of the New World nine-primaried oscines

(Aves: Passeriformes). 43 pp., 10 figs $3.50

8. Berman, D. S., and J. S. McIntosh. 1978. Skull and relationships of the Upper Jurassic sauropod

Apatosaurus (Reptilia, Saurischia). 35 pp., II figs $3.00

9. Setoguchi, T. 1978. Paleontology and geology of the Badwater Creek area, central Wyoming. Part

16. The Cedar Ridge local fauna (Late Oligocene). 61 pp., 30 figs $4.50

10. Williams, D. E. 1978. Systematics and ecogeographic variation of the Apache pocket mouse (Rodentia:

Heteromyidae). 57 pp., 23 figs $4.00

11. Guilday, J. E., H. W. Hamilton, E. Anderson, and P. W. Parmalee. 1978. The Baker Bluff Cave

deposit, Tennessee, and the Late Pleistocene faunal gradient. 67 pp., 16 figs $5.00

12. Swanepoel, P., and H. H. Genoways. 1978. Revision of the Antillean bats of the genus Brachyphylla

(Mammalia: Phyllostomatidae). 53 pp., 17 figs $4.00

13. Schwartz, J. H., and H. B. Rollins (eds.). 1979. Models and methodologies in evolutionary theory.

105 pp., 36 figs $6.00

14. Engesser, B. 1979. Relationships of some insectivores and rodents from the Miocene of North America

and Europe. 68 pp., 12 figs., 20 plates $5.00

15. Bentz, G. D. 1979. The appendicular myology and phylogenetic relationships of the Ploceidae

and Estrildidae (Aves: Passeriformes). 25 pp., 5 figs $2.00

16. Mares, M. A. 1980. Convergent evolution among desert rodents: a global perspective. 51 pp., 25

figs $3.50

17. Lacher, T. E., Jr. 1981. The comparative social behavior of Kerodon rupestris and Galea spixii and

the evolution of behavior in the Caviidae. 71 pp., 40 figs $6.00

18. McIntosh, J. S. 1981. Annotated catalogue of the dinosaurs (Reptilia, Archosauria) in the collections

of the Carnegie Museum of Natural History. 67 pp., 22 figs $6.00

19. Bristow, C. R., and J. J. Parodiz. 1982. The stratigraphical paleontology of the Tertiary non-marine

sediments of Ecuador. 52 pp., 24 figs $5.00

• f

I

J

I

I

\

I

BULLETIN

of CARNEGIE MUSEUM OF NATURAL HISTORY

THE PALEOGENE MAMMALS OF CHINA

LI CHUAN-KUEI

Resident Museum Specialist, Section of Vertebrate Fossils
{permanent address: Institute of Vertebrate Paleontology and Paleoanthropology ,
Academia Sinica, Beijing, People’s Republic of China)

TING SU-YIN

Institute of Vertebrate Paleontology and Paleoanthropology ,

Academia Sinica, Beijing, People’s Republic of China

NUMBER 21

PITTSBURGH, 1983

BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY
Number 21, pages 1-98, 16 figures, 5 tables

Issued 25 March 1983

Price: $15.00 a copy

Mary R. Dawson, Acting Director

Editorial Staff; Hugh H. Genoways, Editor; Duane A. Schlitter,
Associate Editor; Stephen L. Williams, Associate Editor;
Mary Ann Schmidt, Technical Assistant.

© 1983 by the Trustees of Carnegie Institute, all rights reserved.

CARNEGIE MUSEUM OF NATURAL HISTORY, 4400 FORBES AVENUE
PITTSBURGH, PENNSYLVANIA 15213

Dedicated to:

Dr. Minchen M. Chow, our professor, who is the founder
in China of the study of Paleogene mammals;

Our colleagues of the Laboratory of Paleomammalogy of
IVPP, who devote labor in this field;

Our American colleagues, who have helped us so eagerly
and have been so friendly.

’W!S

CONTENTS

Preface 9

Introduction 11

Acknowledgments 12

Chapter 1. The Chinese Paleogene Mammalian Faunas
Section 1. Paleocene

I. Guang-dong (Kwantung) Province

1. Nan-xiong (Nanhsiung) Basin 13

II. Jiang-xi (Kiangsi) Province

2. Chi-jiang (Chihkiang) Basin 14

III. Hu-nan Province

3. Cha-Iing Basin 17

IV. An-hui (Anhui) Province

4. Qian-shan (Chienshan) Basin 17

5. Xuan-cheng (Hsuancheng) Basin 20

V. He-nan (Honan) Province

6. Tan-tou Basin 20

VI. Shaan-xi (Shensi) Province

7. Shi-men Basin 20

VII. Xin-jiang (Sinkiang) Region

8. Turpan (Turfan) Basin 20

VIII. Nei-mong-gol (Inner Mongolia) Region

9. Nao-mu-gen (Nomogen) Area 21

Table 1. A comparative table of the Nao-mu-gen, Bayan Ulan, Gashato, and Naran Bulak

faunas 22

Table 2. An analyzed table of Chinese Paleocene mammals 23

Section 2. Eocene

I. Nei-mong-gol (Inner Mongolia) Region

1 . Bayan Ulan Formation 22

2. Arshanto Formation 22

3. Irdin Manha Formation:

3A. Irdin Manha Area 24

3B. Camp Margetts Area 25

3C. Shara Murun Area 26

4. Shara Murun Formation 27

Table 3. A comparison of the fossils from the Irdin Manha Formation and Shara Murun Formation 29

II. Xin-jiang (Sinkiang) Region

5. Turpan (Turfan) Basin 28

6. Jung-gur (Dzungar) Basin 30

III. Shaan-xi (Shensi) Province

7. Lan-tian District 30

IV. Bei-jing (Peking) and He-bei (Hopei) Province

8. Chang-xin-dian (Changsintien) Locality 30

9. Ling-shan Locality 30

V. Shan-dong (Shantung) Province

10. Wu-tu Basin 30

11. Niu-shan Basin 30

12. Xin-tai (Sintai) Basin 31

VI. Shan-xi (Shansi) Province

13. Yuan-chu Basin 31

VII. He-nan (Honan) Province

14. Tan-tou Basin 33

15. Ji-yuan (Chiyuan) Basin 33

16. Ling-bao Basin 33

1 7. Lu-shi Basin 34

18. Wu-cheng Basin 34

19. Xi-chuan (Sichuan) Basin 36

VIII. Hu-bei (Hupei) Province

20. Jun-xian Basin 37

21. Yi-chang (Ichang) District 37

IX. An-hui (Anhui) Province

22. Lai-an Basin 37

X. Jiang-xi (Kiangsi) Province

23. Chi-jiang (Chihkiang) Basin 38

24. Yuan-shui Basin 38

XL Hu-nan (Hunan) Province

25. Heng-yang Basin 38

XII. Guang-xi (Kwangsi) Region

26. Bo-se (Bai-se) Basin 40

XIII. Yun-nan (Yunnan) Province

27. Lu-nan Basin 41

28. Li-jiang Basin 43

Section 3. Oligocene

I. Nei-mong-gol (Inner Mongolia) Region

1 . Shara Murun-Irdin Manha Area 44

lA. Ulan Gochu and Urtyn Obo Formation 44

IB. Houldjin Formation 44

Table 4. The list of the mammalian fauna of Hsanda Gol Formation (Middle Oligocene),

Mongolia 46

Table 5. The list of the mammalian fauna of Ardyn Obo Formation (Early Oligocene), Mongolia 47

2. Deng-kou (Saint Jacques) Area 45

3. Hos Burd (Suhaitu) Basin 45

II. Ning-xia (Ninghsia) Region

4. Ling-wu Basin 47

5. Others 48

III. Xin-jiang (Sinkiang) Region

6. Turpan (Turfan) Basin 48

7. Ha-mi Basin 48

8. Jung-gur (Dzungar) Basin 48

IV. Gan-su (Kansu) Province

9. Taben buluk Area 48

10. Shargaltein Gol Area 49

11. Shih-ehr-ma-cheng Locality 50

V. Shaan-xi (Shensi) Province

12. Lan-tian District 50

VI. Shan-xi (Shansi) Province

13. Yuan-chu Basin 50

VII. Guang-xi (Kwangsi) Region

14. Bo-se Basin 50

15. Yong-le Basin 50

VIII. Gui-zhou (Kweichow) Province 51

IX. Yun-nan Province

16. Qu-jing (Chu-ching) Basin 51

17. Lu-nan Basin 51

18. Luo-ping Basin 51

Chapter 2. The Systematic and Stratigraphic Distribution of Chinese Paleogene Mammals 52

Chapter 3. The Index

I. Systematic index 65

II. Stratigraphic formation index 76

Chapter 4. The Bibliography 78

Appendix I. Comparative list of Chinese authors’ names in English and Pinyin 86

Appendix II. Comparative table of the localities in “Conventional English” or Wade-giles and

Pinyin 87

Appendix III. Map of Chinese Paleogene mammalian fossil localities 95

Appendix IV. A tentative correlation of the Chinese Paleogene formations containing mammalian

fossils 97

I

f«

yl

'*si

-“ Sr. r.K

W*v- i v4>.', 'VVjg ' 7

'S' ^*^7 ■.-

V:. 'll ...

fyS:= -3

-^"'l££,i' 1.;.'

-7 7?'

i~ ::* 1^1. ■_..,.

■ :;f.>^ «

PREFACE

Fossil mammals from China have been known to
science for well over one hundred years. British
explorers found Late Quaternary mammals in a
locality at Niti Pass in southern Tibet, which were
illustrated by Falconer and Cautley in Fauna Anti-
gua Sivalensis (1845) and later (1868) described by
Falconer. Some other early publications on Chinese
fossil mammals, such as those by Owen (1870) and
Koken (1885), were based on Late Tertiary and
Quaternary materials purchased from Chinese drug
stores. Although there were other important works
on Chinese fossil mammals in the early part of the
1900s, most notably Schlosser’s (1903) Die fossilen
Sdugetiere Chinas, it was not until the 1920s when
field projects by a number of different groups,
including the Geological Survey of China, the Cen-
tral Asiatic Expedition of the American Museum of
Natural History, and paleontologists from Munich,
Upsala, and Paris, led to a great expansion in knowl-
edge of the Chinese record. The first Paleogene
mammals of China were discovered by the Central
Asiatic Expedition and reported by Matthew and
Osborn.

In 1942 Teilhard de Chardin and Leroy sum-
marized the Chinese record that had been published
up to the end of 1941. At about that time world
events interfered with the regular development of
paleontology in China (as well as elsewhere) but by
1950 Chinese mammalian paleontologists were back
at work under the auspices of the Cenozoic Labo-
ratory of the Ministry of Geology. Their work, in
the organization that later became known as the
Institute of Vertebrate Paleontology and Paleoan-
thropology of the Academia Sinica, was under the
inspired and dedicated leadership of Chung-chien
Young until his death in 1979, when our esteemed
colleague Minchen Chow undertook the direction
of the Institute. Under these two eminent directors,
the paleontologists of the IVPP have greatly
expanded the Chinese record for the entire Ceno-
zoic. Perhaps most noteworthy, however, has been
the growth of knowledge of the Paleogene, which
was very poorly known even in 1940 when Teilhard
de Chardin and Leroy could report only various Late
Eocene localities in Inner Mongolia as representing
the Paleogene within the boundaries of China. In

the 40 years since Teilhard de Chardin and Leroy’s
report, a succession of Paleocene faunas have been
described, the Eocene record has been filled in, and
the Oligocene faunas have become better known.
The record is a fascinating and significant one, bear-
ing not only such groups as the endemic but highly
important Anagalida but also various perissodac-
tyls, condylarths, and others that facilitate intercon-
tinental correlations. Unfortunately for those pale-
ontologists without a knowledge of the Chinese
language, however, much of the literature on these
faunas has been published in Chinese alone or with
only short summaries in another language. Locality
and stratigraphic information is scattered through-
out the Chinese literature and it has been difficult
to obtain a comprehensive understanding of strati-
graphic relationships for the Chinese materials. For-
tunately this paper will do much to remedy this
situation so far as the Paleogene is concerned.

It came about as a result of the coincidentally
concordant visits to the United States of two com-
petent specialists in the Chinese Paleogene, Chuan-
kuei Li (to Carnegie Museum of Natural History)
and Su-yin Ting (to Louisiana State University).
These scholars were inundated with requests from
American and western European colleagues for
information on Chinese Paleogene mammals, on
locality and stratigraphic relationships. Our igno-
rance made the need for some sort of summary work
very apparent. Li and Ting took time from their
busy schedules and their own research projects to
provide this masterly summary. It is patterned on
Teilhard de Chardin and Leroy’s arrangement, but
goes much further, especially with details of local-
ities and with carefully executed maps. It illustrates
not only the comprehensive knowledge of these two
colleagues but also the tremendous advances result-
ing from the work of all our colleagues at the IVPP
during the past 40 years.

We at Carnegie Museum of Natural History are
proud to make this summary available in the cer-
tainty that it will greatly improve understanding of
the Chinese Early Tertiary and will stimulate further
research in mammalian paleontology.

Mary R. Dawson

9

L

i* I

INTRODUCTION

Our best opening to this review of Chinese Paleo-
gene mammals turns back the pages of history to
the words, “Our knowledge of Chinese fossil mam-
mals has been advancing rapidly since 1920 . .
from the book “Chinese Fossil Mammals” pub-
lished by Pierre Teilhard de Chardin and Pierre
Leroy in 1 942. From his experience of working with
Chinese fossil mammals for twenty years and his
wide paleontological knowledge, Teilhard de Char-
din reviewed evidence on the fossil mammalian fau-
nas of Eastern Asia that had been published up to
the 1 940s, presenting it in a synoptic form helpful
to further scientific progress.

Looking back upon the progress of the past nearly
forty years since Teilhard de Chardin’s work, some
important breakthroughs and exciting discoveries
have occurred in China, especially with respect to
Early Tertiary mammals.

The most highly significant advance that should
be emphasized is the discovery of an essentially
complete stratigraphic sequence of Paleocene age,
containing three or four mammalian assemblages,
or faunal zones. Not only has it broadened our
knowledge of mammalian faunas and their geolog-
ical distribution in Paleocene time, but also with its
high numbers, rich diversity, and specific endemic
color, it has given rise to some speculations about
the origin and migration of mammals at the begin-
ning of the age of mammals.

Mammalian faunas of Eocene age were docu-
mented earlier, but were largely limited to the later
Eocene. Great progress has been made on the pre-
viously little known Middle Eocene faunas and on
Early Eocene faunas, which were entirely unknown
in Teilhard de Chardin’s time. Most notably, the
discovery of fossils of Homogalax, Heptodon, Co-
ry phodon, and Hyopsodus, all typical of the North
American Early Eocene (Wasatchian), are of con-
siderable stratigraphic and zoogeographic signifi-
cance.

Mammalian faunas of Oligocene age were much
less common than those of the Eocene, but large
collections have been found recently.

At the time of this writing, more than 300 genera
and 560 species from the Chinese Paleogene have
been reported. The great number of new collections
found in China, the hundreds of new publications
on these finds, and especially the visit to the U.S.

of a Chinese delegation of vertebrate paleontologists
and the paper, “The Mammal-bearing Early Ter-
tiary Horizons of China”, by Chow and Zheng
(1980), strongly attracted the attention of many for-
eign colleagues and friends. Although our foreign
colleagues have difficulty discussing and exchanging
information because of the language barrier, they
are eager to know and understand the advances in
Chinese vertebrate paleontology and discuss them.

We have been deeply impressed by the enthu-
siastic reception of Chinese progress in vertebrate
paleontology from foreign colleagues. As Chinese
vertebrate paleontologists, we would like to make
more readily available the newest advances in the
study of Chinese Paleogene mammals, results of the
devoted labors of our teacher and colleagues, to our
foreign friends, even though it is very difficult for
us to make it as complete as we would like to because
of the limitations of our knowledge thus far. By
writing this paper we would also like to give thanks
to our American colleagues for receiving and helping
us so eagerly.

We realize that some mistakes and omissions are
unavoidable, no matter how diligently we checked
this work. We would greatly appreciate corrections
and constructive criticism of our work.

We have gathered together material on Chinese
Paleogene mammals published up to the end of 1 980
in this synopsis. The catalogue is divided into the
following four parts:

Chapter 1. “The Chinese Paleogene Mammalian
Fauna” is the main body of this work. In order to
give our reader a somewhat generalized but clear
conception of each fossil site and fauna, we divide
the chapter into three sections, Paleogene, Eocene,
and Oligocene, and give the details from each basin.
The history, location, dimensions, stratigraphic
sequence, and list of mammalian fauna were given
as completely as possible for each site. By using the
coordinate system and providing translated maps,
we have made it possible for the reader to get an
idea of the location. In some cases, we have used
the coordinates of nearby counties or towns instead
of small localities when we could not get their def-
inite coordinates. Using basically G. G. Simpson’s
classification, we have listed the mammalian fauna.
Also we respect the original author’s idea about the
classification for the fossils studied by him (her). To

12

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

check fossils conveniently with the original author’s
localities and reports, we add their held number and
reference literature number following each fossil.

Chapter 2, “The Systematic and Stratigraphic
Distribution of Chinese Paleogene Mammals”.

Chapter 3, “The Index” includes both systematic
and stratigraphic indexes.

Chapter 4 is a bibliography. About 280 papers are
listed in the bibliography. More than 60 percent of
them (170 papers) were completed by Chinese col-
leagues and mostly published in Chinese with or

without an English summary. So we give the note
“in Chinese” or “In Chinese with English sum-
mary” after each Chinese publication, so that the
reader can quickly determine whether or not trans-
lation is worthwhile.

We have four special appendices at the end of the
paper, a comparative list of Chinese authors’ names,
a list of fossil localities in English and Pinyin, a map
of 6 1 localities of Chinese Paleogene mammals, and
a tentative correlation of the Chinese Paleogene for-
mations.

ACKNOWLEDGMENTS

Drs. Craig C. Black and Mary R. Dawson of Carnegie Museum
of Natural History encouraged us to undertake this project. We
gratefully acknowledge their support and making funds available
for this purpose. Especially we would like to thank Dr. Mary R.
Dawson for her preface, correcting the entire English manuscript,
and giving much constructive advice on this paper. We are deeply
indebted to Dr. Judith A. Schiebout of Louisiana State University
for helping and supporting this work. Dr. Chow Minchen, Direc-
tor of the Institute of Vertebrate Paleontology and Paleoanthro-
pology, Academia Sinica, permitted us to publish this book and
offered valuable advice, for which we are deeply indebted. During
the preparation of this paper, Drs. Malcolm C. McKenna and
Richard H. Tedford, American Museum of Natural History (New
York), Henry L. Snyder, R. E. Ferrell, Louisiana State University
(Baton Rouge), Robert W. Wilson, University of Kansas (Law-

rence), William D. Turnbull, Field Museum of Natural History
(Chicago), Leonard Krishtalka, William W. Korth, Frederick H.
Utech, and David E. Boufford, Carnegie Museum of Natural
History, provided encouragement and help. We are grateful to
them. Most sincere thanks are given to Ms. Elizabeth A. Hill,
Carnegie Museum of Natural History, for carefully typing and
proofreading the whole manuscript, so we avoided many mis-
takes. Thanks also to Ms. Nancy J. Perkins of Carnegie Museum
of Natural History and Mr. Clifford Duplechin, James Kennedy
and Ms. Mary L. Eggort of Louisiana State University for drawing
the beautiful maps. We acknowledge also Ms. Lori Lewis and
Hwei-ling Cecilia Wang of Louisiana State University, Mr. Clif-
ford J. Morrow and Mrs. Anna R. Tauber of Carnegie Museum
of Natural History for their kind help.

CHAPTER 1-THE CHINESE PALEOGENE MAMMALIAN FAUNAS

Section I. Paleocene

I. Guang-dong (Kwantung) Province
1 . Nan-xiong (Nanhsiung) Basin

Young Chung-chien and Chow Min-chen, 1962'“'*''
Chang Yu-ping and Tung Yung-sheng, 1963"^*
Tang Xin and Chow Min-chen, 1964"’"

Chen Chia-chien, Tang Ying-jun, Chiu Chan-siang,
and Yeh Hsiang-kuei, 1973"'"

Zhou Ming-zhen, Zhang Yu-ping, Wang Ban-yue,
and Ding Su-yin, 19771259)

South China “Redbeds” Research Group, IVPP,

19 7 7(164)

a. Location: Nan-xiong County, Guang-dong Prov-
ince.

Coordinates: 25°00'-25°10'N; 1 14°15'-1 14°30'E;
25°07'N; 1 14°18'E (Nan-xiong city)

(Fig. 1)

b. Dimensions: 80 km long (NE-SW), 1 8 km wide (at
maximum).

c. Stratigraphic Sequence:

Dan-xia Formation (Eocene)

Dark red sandy-conglomerate rock, sandy mud-
stone and sandstone (100-550 m)’.

— ?Conformity —

Nung-shan Formation (Late Paleocene) (200 m)
Da-tang Member: Purplish dusky red marly sand-
stone and sandy marls with intercalating grey-
ish green sandy marls [73139, 73138, 73059]^
Zhu-gui-keng Member: Greyish green and pur-
plish red silt mudstone [73143] (etc.).

—Conformity or Disconformity—

Shang-hu Formation (Early-Middle Paleocene)
Purplish red marls and mudstone with interca-
lated thin sandstone and conglomerates [62 1 7,
6219e, 6233, 63081, 63082, 63084, 63087,
63088, 73057, 73150].

— Disconformity —

Nan-xiong Group (Late Cretaceous).

d. The list of the mammalian fauna:

Nung-shan Formation (Late Paleocene)

Da-tang Member:

Order Edentata Cuvier, 1798
Family Emanodontidae Ding, 1979
Ernanodon antelios Ding, 1979'”' [73139]
Order Primates Linnaeus, 1758
Family Adapidae Trouessart, 1879

Petrolemur brevirostre Tong, 1979"*^)
[73059]

' The number in parentheses refers to the paper listed in Chapter 4 (The Bibliography)
in which the appropriate fossils and/or stratigraphy are described and discussed.

^ The thickness of the formation in meters.

^ The numbers in brackets are the main field numbers of the IVPP.

Order Anagalida Szalay and McKenna, 1971
Family Anagalidae Simpson, 1931
cf Huaiyangale leura Ding and Tong,
1979(73, [73138.C]

Family Pseudictopidae Sulimski, 1968
Haltictops mirabilis Ding and Tong, 1 979'”’
[73138.C]

Haltictops meilingensis Ding and Tong,
1979'”’ [73059.C]

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875

Yantanglestes datangensis (Wang,

1976)(i94,(97, [73059.d]

Order Notoungulata Roth, 1903
Family Arctostylopidae Schlosser, 1923
Arctostylopidae gen. et sp. nov.''*""

Order Pantodonta Cope, 1873
Family Archaeolambdidae Flerov, 1952
Archaeolambdidae gen. et spp. nov."'"*’
Family Pastoralodontidae Chow and Qi, 1978
Altilambda pactus Chow and Wang, 1978*’*’
[73138]

Family Phenacolophidae Zhang, 1978
Minchenella grandis Zhang, 1978'2‘*5)(246)
[73059.b]

Yuelophus validus Zhang, 1 978”'”’
[73138.d]

Zhu-gui-keng Member:

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
IPachyaena sp."'’'" [73143]

Order Pantodonta Cope, 1873
Family Archaeolambdidae Flerov, 1952
Archaeolambdidae gen. et spp. nov.'"’'"
Family Phenacolophidae Zhang, 1978
Phenacolophidae gen. et sp. nov.'""*’
Shang-hu Formation (Early-Middle Paleocene)
Order Anagalida Szalay and McKenna, 1971
Family Anagalidae Simpson, 1931
Linnania lofoensis Chow et al., 1973''*”'”’’
[63081]

Order Carnivora Bowdich, 1821
Family Miacidae Cope, 1880
Pappict idops acies^ ung, 1978"“”’ [73057. a]
Pappictidops obtusus Wang, 1978'’'”’
[73057.b]

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Yantanglestes feiganensis (Chow et al.,

1973)(43X2S9K97, [6233]

Dissacusium shanghoensis Chow et al.,

1973(43X259, [63087]

Hukoutherium ambigum Chow et al.,

19 7 3(43X259, [63082]

Family Hyopsodontidae Lydekker, 1 889
Yuodon protoselenoides Chow et al.,

1973(43,(259, [63087]

Palasiodon siurenensis Chow et al.,

19 7 3(43,(259, [6217]

13

14

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Family Periptychidae Cope, 1882
lEctoconus [63084]

Order Tillodontia Marsh, 1875
Family Esthonychidae Cope, 1883 II.

Lofochaius brachyodus Chow et al.,

1973(43)(259) (Zeng-de-ao)

Order ?Tillodontia Marsh, 1875
Family incertae sedis

Dysnoetodon minuta Zhang, 1980'^'*^’

[73150]

Order Pantodonta Cope, 1873

Family Bemalambdidae Chow et al., 1973
Bemalambda nanhsiungensis Chow et al.,

1973(43,(259, [63084]

Bemalambda pachyoesteus Chow et al.,

1973(43,(259, [62.19e]

Bemalambda crassa Chow et a!., 197303,(259,
[63088]

Bemalambda [63087]

Jiang-xi (Kiangsi) Province
2. Chi-jiang (Chihkiang) Basin

Tang Xin and Chow Min-chen, 1964''^"

Zheng Jia-jian, Tong Yong-sheng, Ji Hong-xiang,
and Zhang Fa, 1973'-^’^’

Tong Young-sheng, Zhang Yu-ping, Wang Ban-yue,
and Ding Su-yin, 1976"’’“’

South China “Redbeds” Research Group, IVPP,

1977(164,

Tong Young-sheng, Zhang Yu-ping, Zheng Jia-jian,
Wang Ban-yue, and Ding Su-yin, 1979"''"
a. Location; Da-yu and Nan-kang counties, Jiang-xi
Province.

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

15

Fig. 2. — Map of the Late Cretaceous and Lower Tertiary vertebrate fossil localities of Chijiang Basin.

Coordinates: 25°23'-25°40'N; 1 14°20'-1 14°45'E;

25°29'N; 1 14°34'E (Chi-jiang town)

(Fig. 2)

b. Dimensions: 40 km long (NE-SW), 1 0 km wide (at
maximum).

c. Stratigraphic Sequence:

Ping-hu Formation (Early Eocene)

Purplish red politic sandstone and marls with
intercalating yellowish green siltstone [72027],
(200-300 m).

—Conformity—

Chi-jiang Formation (Late Paleocene) (473 m)
Wang-wu Member: Purplish red marls with inter-
calating red sandstone and green thin calcar-
eous mudstone [73041].

Lan-ni-keng Member: Purplish red mudstone with
intercalating dusky red and greyish green sand-
conglomerates [72034, 72035, 73039, 73046,
73048, 73052, 73055].

— Disconformity—

Shi-zi-kou Formation (Early Middle Paleocene)
Brick-red sandy marls with intercalating greyish
green sandstone [73042, 73043].

— Disconformity—

Nan-xiong Group (Late Cretaceous)

d. The list of the mammalian fauna:

Chi-jiang Formation (Late Paleocene)

Wang-wu Member:

Order Notoungulata Roth, 1903

16

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Fig. 3. — Map of the Paleocene fossil localities of Chaling Basin.

Family Arctostylopidae Schlosser, 1923
Allostylops periconotus Zheng, 1979'^^^’
[73041]

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Jiangxia chaotoensis Zhang, Zheng, and
Ding, 1979'"^»> [73041.1]

Order Pantodonta Cope, 1873

Family Archaeolambdidae Flerov, 1952
Archaeolambda sp.*'*'*' [73041]
Lan-ni-keng Member:

Order Insectivora Bowdich, 1821

Insectivora gen. et sp. indet. 1979'’'’' [73052]
Order Anagalida Szalay and McKenna, 1971
Family Anagalidae Simpson, 1931

Hsiuannania minor Ding and Zhang,

1979(74, [73055]

Family Pseudictopidae Sulimski, 1968
cf. Pseudictops tenuis Ding and Zhang,
1979,74, [73046]

Order Notoungulata Roth, 1903

Family Arctostylopidae Schlosser, 1923
Asiostylops spanios Zheng, 1979'“” [73039]
Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
IHapalodectes sp.'’"*' [73048]

Family Hyopsodontidae indet.

Hyopsodontidae indet.”"*' [73052]

Family Periptychidae Cope, 1882
Pseudanisonckus antelios Zhang, Zheng,
and Ding, 1979'’"*' [73039]

Order Pantodonta Cope, 1873
Family Archaeolambdidae Flerov, 1952
Archaeolambda cf. planicanina Flerov,

1952,(84X272, [73039]

Archaeolambda dayuensis Tong, 1979"*"'
[72034]

Nanlingilambda chijiangensis Tong,
1979"*"'

Family Harpyodidae Wang, 1979
Harpyodus decorus Wang, I979"'’’i
[73048.b]

Family Phenacolophidae Zhang, 1978

Ganolophus lanikenensis Zhang, 1979'2‘*'’>

[72035]

Mammalia, Order indet.

Family Didymoconidae Kretzoi, 1943
Archaeoryctes notialis Zheng, 1979'’*"
[72035]

Shi-zi-kou Formation (Early Middle Paleocene)
Order Pantodonta Cope, 1873
Family Bemalambdidae Chow et al., 1973

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

17

Bemalambda shizikouensis Wang and Ding,
1979(198) [73043, 73042.b]

Family Archaeolambdidae Rerov, 1952
Archaeolambdidae indet."*"** [73042]

III. Hu-nan Province

3. Cha-ling Basin

Tang Xin and Chow Min-chen, 1964"’"

Gao Hung-Hsiang, 1975'”’

Wang Ban-yue, 1975"’^’

South China “Redbeds” Research Group, IVPP,

1977(164,

a. Location: Cha-ling County, Hu-nan Province.
Coordinates: 26°38'-26°48'N; 1 !3°24'-l 13°34'E;

26°48'N; 1 13°33'E (Cha-ling city)

(Fig. 3)

b. Dimensions: 90 km long (NNE-SSW), 24 km wide
(at maximum) (Meso-Cenozoic Basin).

c. Stratigraphic Sequence:

Zao-shi Formation (Early Middle Paleocene)

Purplish red sandy claystone with intercalating
conglomerate, gypsum, and greyish green
marls [72137, 72139, 72135, 72130, 72136,
72133, 72132] (more than 53 m).

— Disconformity —

Dai-jia-ping Formation (Late Cretaceous)

d. The list of the mammalian fauna:

Zao-shi Formation (Early-Middle Paleocene)

Order Anagalida Szalay and McKenna, 1971
Family Anagalidae Simpson, 1931
Stenanagale xiangensis Wang, 1975"'*^)
[72137]

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
IDissacus rotundus Wang, 1 975"”^' [721 39]
Order Tillodontia Marsh, 1875
Family Esthonychidae Cope, 1883
Meiostylodon zaoshiensis Wang, 1975"’^’
[72135]

Order Pantodonta Cope, 1873
Family Bemalambdidae Chow et al., 1973
Bemalambda nanhsiungensis Chow et al.,

1975(193) [72139]

Bemalambdidae indet."’^’ [72139, 72130,
72136]

Hypsilolambda chalingensis Wang, 1975"'’^'
[72133]

Hypsilolambda impensa Wang, 1975"'*^’
[72132]

Hypsilolambda spp."’^' [72130]

Mammalia indet."”' [72133, 72130]

IV. An-hui (Anhui) Province

4. Qian-shan (Chienshan) Basin

Qiu Zhan-xiang and Li Chuan-kuei, 1972"^"

Qiu Zhan-xiang, Li Chuan-kuei, Huang Xue-shi,
Tang Ying-jun, Xu Qin-qi, Yen De-fa, and Zhang
Hong, 1977"”'

a. Location: Qian-shan and Huai-ning counties, An-
hui Province.

Coordinates: 30°35'-30°45'N; 1 16°28'-1 16°38'E;
30°36'N; 1 16°36'E (Qian-shan city)

(Fig. 4)

b. Stratigraphic Sequence:

Dou-mu Formation (Late Paleocene) (600 m)

Upper Member: Thick purplish red conglomerates
interbedding with sandstone and conglom-
eratic sandstone [71017].

Lower Member: Thick purplish red medium-fine
sandstone with intercalating thin conglom-
erates and silt mudstone [71015, 71079].

— Unconformity—

Wang-hu-dun Formation (Early-Middle Paleocene)
(1,800 m)

Upper Member: Purplish red fine sandstones with
intercalating thin greyish white sandstone
[70020, 70022, 71008, 71009, 71010, 71012,
71016, 71075].

Middle Member: Purplish red conglomerates and
conglomeratic gritstones interbedding with
fine sandstone.

Lower Member: Purplish red middle-fine sand-
stone with intercalating conglomerates and
greyish white arkose-sandstone [71001, 7 1 002,
71004, 71005, 71006].

Wang-he Formation (Late Cretaceous) — (7 50 m)

c. The list of the mammalian fauna:

Dou-mu Formation (Late Paleocene)

Upper Member:

Order Insectivora Bowdich, 1821
Family Indet.

Hyracolestes ermineus Matthew and Gran-
ger, 1925"”'"^'’' [71017]

Order Anagalida Szalay and McKenna, 1971
Family Anagalidae Simpson, 1931
Hsitiannania sp.'“*’ [71017]

Family Eurymylidae Matthew, Granger, and
Simpson, 1929

Heomys orientalis Li, 1977"°'" [71017]
Eurymyloidea indet.'""" [71017]

Family Mimotonidae Li, 1977
Mimotona wana Li, 1977"°'" [71017]
Order Notoungulata Roth, 1903
Family Arctostylopidae Schlosser, 1923
Sinoslylops promissus Tang and Yan,
1976"’-" [71017]

Order Pantodonta Cope, 1873
Family Archaeolambdidae Flerov, 1952
Archaeolambda tabiensis Huang, 1977'''"
[71017]

Lower Member:

Order Anagalida Szalay and McKenna, 1971
Family Anagalidae Simpson, 1931
Hsiuannania tabiensis Xu, 1976'’°*’ [71079]
Family Mimotonidae Li, 1977
Mimotona robusta Li, 1977"°'" [71079]
Family Pseudictopidae Sulimski, 1968
Allictops inserrata Qiu, 1977"”’ [71015]
Mammalia, Order indet.

Obtususdon hanhuaensis Xu, 19771210)
[71079]

Wang-hu-dun Formation (Early-Middle Paleocene)

Upper Member:

Order Anagalida Szalay and McKenna, 1971

18

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Fig. 4. — Map of the Paleocene fossil localities of Qianshan Basin.

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

19

Family Anagalidae Simpson, 1931

Huaiyangale chianshanensis Xu, 1976‘“®’
[70020]

Huaiyangale [70020]

Diacronus wanghuensis Xu, 1976'^°'*'
[71016]

Diacronus anhuiensis Xu, 1976'-“''’ [71009]
Family Eurymylidae Matthew, Granger, and
Simpson, 1929
Heomys sp."”’ [71010]

Family Mimotonidae Li, 1977
Mimotona wana Li, 1977*"’’’ [71016]
Mimotona sp."*”’ [71008]

Family Pseudictopidae Sulimski, 1968
Anictops tabiepedis Qiu, 1977" 5*” [71008,
7101 1, 70021, etc.]

Anictops alf. tabiepedis Qiu, 1977*"""
[70022]

Paranictops majuscula Qiu, 1977*'^*"
[70022]

1 Paranictops sp.*'^*" [71014]

Order Carnivora Bowdich, 1821
Family Miacidae Cope, 1 880

Pappictidops orientalis Qiu and Li, 1977"
[71008]

Order Condylarthra Cope, 1 88 1

Family Hyopsodontidae Lydekker, 1 889
Decoredon elongetus Xu, 1977*^'*" [71009]
Order Pantodonta Cope, 1873

Family Pastoralodontidae Chow and Qi, 1 978
Altilambda pactus Chow and Wang, 1978*^*’
[71075]

Altilambda tenuis Chow and Wang, 1 978*^*’
[71016]

Family Harpyodidae Wang, 1979
Harpyodus euros Qiu and Li, 1977*'“’
[71012]

Mammalia, Order indet.

Family Didymoconidae Kretzoi, 1 943
Zeuctherium niteles Tang and Yan, 1 976*"'^’
[71009]

Mammalia, Order indet.

Obtususdon hanhuaensis Xu, 1977*^'*"
[70020]

Lower Member:

Order Anagalida Szalay and McKenna, 1971
Family Zalambdalestidae Gregory and Simp-
son, 1926

Anchilestes impolitus Qiu and Li, 1977*'“’
[71001]

Family Anagalidae Simpson, 1931

Wanogale hodungensis Xu, 1976*-*”’
[71001]

Anaptogale wanghoensis Xu, 1976*-*”’"*^’
[71001]

Chianshania gianghuaiensis Xu, 1976*-*”’
[71001]

Family Pseudictopidae Sulimski, 1968
Cartictops canina Ding and Tong,

1979(I50„IS2,(73) [71005]

Anictops tabiepedis Qiu, 1977"^°’ [71001]
Paranictops sp.*'“’ [71005]

Order Condylarthra Cope, 1 88 1
Family Mesonychidae Cope, 1875

20

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Yantanglestes (Lestes) conenxus Yan and
Tang, 1976'^^3)(97) [71006]

Order Pantodonta Cope, 1873
Family Bemalambdidae Chow et al., 1973
Bemalambda [71001]
Bemalambdidae indet.'^'’’ [71002]

5. Xuan-cheng (Hsuancheng) Basin

Qiu Zhan-xiang and Li Chuan-kuei, 1972"-'"

Qiu Zhan-xiang, Li Chuan-kuei, Huang Xue-shi,
Tang Ying-jun, Xu Qin-qi, Yan De-fa, and Zhang
Hong, 1977"”'

a. Location; Xuan-cheng and Guang-de counties, An-
hui Province.

Coordinates: 30°58'N; 118°45'E (Xuan-cheng city)
(Fig. 5)

b. Stratigraphic Sequence;

Shuang-ta-si Group (Late Paleocene)

Purplish red silt mudstones and marls with inter-
calating conglomerates and lime nodules
[71071, 71073] (213-760 m).

— Unconformity—

Xuan-nan Formation (Late Cretaceous)

c. The list of the mammalian fauna:

Shuang-ta-si Group (Late Paleocene)

Order Anagalida Szalay and McKenna, 1971
Family Anagalidae Simpson, 1931

Hsiuannania maguensis Xu, 1976'-°*'
[71071]

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Dissacus magushanemis Yan and Tang,

1976<223' [71071]

Order Notoungulata Roth, 1903
Family Arctostylopidae Schlosser, 1923
Sinoslylops progressus Tang and Yan,
1976"’'' [71071]

Order Pantodonta Cope, 1873
Family Archaeolambdidae Flerov, 1952
Archaeolambda yangtzeensis Huang,
1978<‘"» [71073]

Mammalia, Order indet.

Family indet.

Wanotherium xuanchengensis Tang and
Yan, 1976'”'' [71073]

V. He-nan (Honan) Province

6. Tan-tou Basin

Tong Yong-sheng and Wang Jing-wen, I979<i8"<>88)
Tong Yong-sheng and Wang Jing-wen, 1980"*'"

a. Location: Luan-chuan County, He-nan Province.
Coordinates: 34°00'N; 1 1 1°46'E (Tan-tou town)

b. Stratigraphic Sequence:

Tan-tou Formation (Early Eocene)

Greyish green and greyish white mudstones
interbedding with greyish black marls and
kerogen shales (136^58 m).

Da-zhang Formation (Late Paleocene)

Dark red and greyish green mudstones interbed-
ding with greyish white marls (104-375 m).

—Conformity—

Gao-yu-gou Formation (Middle Paleocene)

Purplish red mudstones with intercalating con-
glomerate and greyish green silt band (302-
366 m).

— Disconformity—

Qiu-ba Formation (Late Cretaceous)

c. The list of the mammalian fauna:

Da-zhang Formation (Late Paleocene)

Order Anagalida Szalay and McKenna, 1971
Family Pseudictopidae Sulimski, 1968
Pseudictopidae indet."*’'

Order Pantodonta Cope, 1873
Family Pastoralodontidae Chow and Qi, 1978
Pastoralodontidae indet."*’'

Gao-yu-gou Formation (Middle Paleocene)

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Mesonychidae indet."*”

Order Pantodonta Cope, 1873
Family Bemalambdidae Chow et al., 1973
Bemalambdidae indet."*’'

VI. Shaan-xi (Shensi) Province

7. Shi-men Basin

Xue Xiang-xi, 1978”’’'

a. Location: Luo-nan County, Shaan-xi Province.
Coordinates: 34°06'N; 1 lO’lO'E (Luo-nan city)

b. Stratigraphic Sequence:

Red sandy mudstones (Paleocene)

c. The list of the mammalian fauna:

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Mesonychidae indet.””'

Order Pantodonta Cope, 1873
Family Bemalambdidae Chow et al., 1973
Bemalambdidae indet.'””

VII. Xin-jiang (Sinkiang) Region

8. Turpan (Turfan) Basin

Chow Min-chen, I960'”"”’

Zhai Ren-jie, Zheng Jia-jian, and Tong Yong-sheng,
1978'”"

Tong Yong-sheng, 1978"*’'

a. Location: Turpan and Shan-shan counties, Xin-jiang.
Coordinates: 42°45'-43’l 2'N; 89°05'-9 1°36'E;

42°52'N; 90°10'E (Shan-shan city)

(Fig. 6)

b. Dimensions: about 200 km (E-W), 50 km (N-S)
(Cenozoic exposure area).

c. Stratigraphic Sequence:

Tai-zi-cun Formation (Late Paleocene)

Purplish and brownish red arenaceous mudstone
with intercalating greyish-green, fine sand-
stone and marls [6402 1 , 64022, 6403 1 , 660 1 3]
(65 m).

— Disconformity —

Su-ba-shi Formation (Late Cretaceous)

d. The list of the mammalian fauna:

Tai-zi-cun Formation (Late Paleocene)

Order Multituberculata Cope, 1884
Multituberculata indet."*’' [64022]

Order Anagalida Szalay and McKenna, 1971

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

21

Family Eurymylidae Matthew, Granger, and
Simpson, 1929

?Eurymylidae gen. et sp. nov., 1978"*^’
[66013-2]

Family Pseudictopidae Sulimski, 1968
Pseudictops chaii Tong, 1978"*-' [66013]
Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Mesonychidae indet."*^' [64021-5]

Order Pantodonta Cope, 1873
Family Archaeolambdidae Rerov, 1952
Archaeolambda cf. planicanina Flerov,

1952< 1 821,272, [64031]

Family Pantolambdodontidae Granger and
Gregory, 1934

Dilambda speciosa Tong, 1978"*^'
[66013-1]

Family Phenacolophidae Zhang, 1978
Tienshanilophus subashiensis Tong,
197g,l82,,243, [64031]

Tienshanilophus lianmuqinensis Tong,
197g(i82„243, [66013-3]
Tienshanilophus shengjinkouensis Tong,

197g(182,|243, [64021]

Order Dinocerata Marsh, 1873

Family Uintatheriidae Flower, 1876
Prodinoceras diconicus Tong, 1978"*^’
[66013-4]

Jiaoluotherium turfanense (Chow) Tong,

197g,l82„33„32, [64022]

Houyanotherium primigenum Tong,
1978"*-' [64021]

Houyanotherium simplum Tong, 1978"*-'
[66013-6]

VIIL Nei-mong-gol (Inner Mongolia) Region
9. Nao-mu-gen (Nomogen) Area

Zhou Min-zhen, Qi Tao and Li Yong, 1976*^'*’
Chow Min-chen and Qi Tao, 1978*”'

a. Location: Formerly Nom Khong Shireh or Nomo-

gen Ora (see Fig. 7) Si-zi-wang Qi, Inner Mongolia.
Coordinates: around 43°00'N; 1 1 1°30'E

b. Stratigraphic Sequence:

Nao-mu-gen Formation (Late Paleocene)

Dark red and greyish green argillaceous sandstone,
and dark red sand clay (8 m visible thickness).

c. The list of the mammalian fauna:

Nao-mu-gen Formation (Late Paleocene)

Order Multituberculata Cope, 1884

Family Taeniolabididae Granger and Simp-
son, 1929

Prionessus lucifer Matthew and Granger,

19 2 5,53'"2"

Sphenopsalis nobilis Matthew, Granger, and
Simpson, 1928*”'"^*'

Family Lambdopsalidae Chow and Qi, 1978
Lambdopsalis bulla Chow and Qi, 1978'”'
Order Insectivora Bowdich, 1821

Family Deltatheridiidae Gregory and Simp-
son, 1926

Sarcodon pygmaeus Matthew and Granger,
1 925*55’"^"

Order Anagalida Szalay and McKenna, 1971
Family Mimotonidae Li, 1977
Mimotona borealis Chow and Qi, 1977*”'
Family Pseudictopidae Sulimski, 1968
Pseudictops lophtodon Matthew, Granger,
and Simpson, 1929*”*'*'-“"

Order ?Rodentia
Rodentia indet.****’

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
?Dissacus sp.'”’'

Plagiocristodon serratus Chow and Qi,
1978*”'

Order Notoungulata Roth, 1903
Family Arctostylopidae Schlosser, 1 923
Paleostylops iturus Matthew and Granger,

1925<53'*121'

22

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Table l.—A comparative table of the Nao-mu-gen, Bayan Ulan,
Gashato, and Naran Bulak faunas.

Taxa

Gashato

Faunas

Naran Nomo-
Bulak gen

Bayen

Ulan

Prionessus lucifer M. et G.

+

?

+

+

Sphenopsalis nobilis M. et al.

+

-

-f

-

Lambdopsalis bulla C. et Q.

-

-

+

-

Hyracolestes ermineus M. et G.

-f

-

-

-

Sarcodon pygmaeus M. et G.

+

-

+

-

Praolestes nanus M. et G.

-f

-

-

-

Altanius orlovi D. et M.

-

+

-

-

Eurymylus laticeps M. et G.

3-

+

-

-

Mimotona borealis C. et Q.

-

-

+

+

Gomphos elkema Sh.

-f

-

-

-

Pseudictops lophiodon M. et G.

+

+

+

+

Khashanagale zofiae S. et M.

+

-

-

-

?Khashanagale sp. nov. (S. et M.)

-f

-

-

-

?Dissacus sp. 1

+

-

-

-

?Dissacus sp. 2

-

-

+

-

Pachyaena sp.

-

+

-

-f

Plagiocristodon serratus C. et Q.

-

-

+

+

Phenacolophus falla.x M. et G.

-f

-

-

-

Archaeolambda planicanina F.

-

-f

-

-

Pastoralodon lacustris C. et Q.

-

-

-f

-f

Convallisodon convexus C. et Q.

-

-

-t-

-

Convallisodon haliutensis C. et Q.

-

-

-f

-

Prodinoceras martyr M. et G.

+

-

-

-

Mongolotherium efremovi F.

-

-f

-

-f

Mongolotherium plantigradum F.

-

+

-

-

Pyrodon sp.

-

-

-

-f

Palaeostylops Iturus M. et G.

-t-

-1-

+

+

Palaeostylops macrodon M. et al.

-1-

+

-f

+

Lambdotherium sp.?

-

-

-

-I-

?Heptodon sp.?

-

-

-

+

Hyracotherium gabuniai D.

-

-f

-

-

Rodentia indet.?

-

-

+

-

Total number of taxa

15

10

14

12

Paleostylops macrodon Matthew, Granger,
and Simpson,

Order Pantodonta Cope, 1873

Family Pastoralodontidae Chow and Qi, 1978
Pastoralodon lacustris Chow dindQi, 1978'”’
Convallisodon convexus Chow and Qi,
1978'”’

Convallisodon haliutensis Chow and Qi,
1978'”’

Section 2. Eocene

I. Nei-mong-gol (Inner Mongolia) Region
1. Bayan Ulan Formation (Early Eocene)

Qi Tao, 1979"'*^’

Chow Min-chen and Zheng Jia-jian, 1980"’'”

a. Location: Southwest of Camp Margetts, Camp
Margetts Area, or about 20 km east of the Ai-li-
ge-miao.

Coordinates: around 43°20'N; 1 1 1°45'E

b. Stratigraphic Sequence: unpublished; thickness
7 m.

c. The list of the mammalian fauna:

Bayan Ulan Formation (Early Eocene)

Order Multituberculata Cope, 1884

Family Taeniolabididae Granger and
Simpson, 1929

Prionessus lucifer Matthew and Granger,

1 925"‘*’”"-”

Order Anagalida Szalay and McKenna, 197 1
Family Mimotonidae Li, 1977

Mimotona borealis Chow and Qi,

1 978*''’'”'^^’

Family Pseudictopidae Sulimski, 1968
Pseudictops lophiodon Matthew, Grang-
er, and Simpson, 1929"'*”"’'”

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Pachyaena sp."'*’"

Plagiocristodon serratus Chow and Qi,

1978<'47)(53’

Order Notoungulata Roth, 1903

Family Arctostylopidae Schlosser, 1923
Palaeostylops itiirus Matthew and
Granger, 1925"“”'””
Palaeostylops macrodon Matthew,
Granger, and Simpson, 1929"“”"-'’’
Order Pantodonta Cope, 1873

Family Pastoralodontidae Chow and Qi,
1978

Pastoralodon lacustris (?) Chow and Qi,

19 7 8(1471(53’

Order Dinocerata Marsh, 1873

Family Uintatheriidae Flower, 1876
Mongolotherium efremovi Flerov,

1957'I47|(271’

Pyrodon sp.'

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
?Lambdotherium sp."“”

Family Helaletidae Osbom, 1892

?Heptodon sp."“”

2. Arshanto Formation (Middle Eocene)

Berkey, C. P., and E. K. Morris, 1927'“’
Radinsky, L. B„ 1974"”’

Chow Min-chen, Qi Tao and Li Yong, 1976'”’
Qi Tao, 1979"“”

Qi Tao, 1980"“*’

a. Location: About 20 miles south-southeast of Iren
Dabasu, a mile east of Irdin Manha type locality.
Coordinates: around 43°30'N; 112°15'E

(Fig. 7)

b. Stratigraphic Sequence: “prevailingly red clays and
fine silts.”

c. The list of the mammalian fauna:

Order Perissodactyla Owen, 1848

Family Lophialetidae Matthew and
Granger, 1925

Schlosseria magister Matthew and
Granger, 1926'””"”’

In 1977 about 40 species of fossil mammals,
including a diversity of perissodactyls, were col-
lected by Qi Tao from the Arshanto Formation.
Most of the materials have not been published

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

23

Table 2.-

-An analyzed table of Chinese Paleocene mammals (E— early: M— middle: L

— late).

Taxa

Basin or area

Tola!

Nan-xiong

(Nanhsiung)

Chi-jiang

(Chikiang)

Qian-shan

(Chienshan)

Xuan-cheng

Cha-Img (Hsuancheng) Nao-mu-gen

Turpan

(Turian)

E.

L.

E.

L.

E.

L.

E.-M. L.

L.

L.

Multituberculata

_

-

_

-

_

_

_ _

3

1

4

Insectivora

-

-

-

1

-

1

-

1

-

3

?Primates

-

1

-

-

-

-

-

-

-

1

Anagalidae

i

1

-

1

6

2

1 1

-

-

13

Pseudictopidae

-

2

-

1

5

1

-

1

1

11

Eurymylidae

-

-

-

-

2

3

-

1

1

7

Zalambdalestidae

-

-

-

-

1

-

-

-

-

1

Didymoconidae

-

-

-

1

1

-

-

-

-

2

Mesonychidae

3

2

-

2

1

-

1 1

2

1

13

Condylarthra

3

-

-

2

-

-

-

-

-

5

Edentata

-

1

-

-

-

-

-

-

-

1

Notoungulata

-

1

-

2

-

1

1

2

-

7

Carnivora

2

-

-

-

1

-

-

-

-

3

Tillodontia

2

-

-

-

-

-

1

-

-

3

Phenacolophidae

-

3

-

1

-

-

-

-

3

7

Pantodonta

3

3

2

5

5

1

6 1

3

2

31

Dinocerata

-

-

-

-

-

-

-

-

4

4

Others

-

-

-

-

2

1

1 1

1

-

6

Total

14

14

2

16

24

10

10 5 1

14

13

122

1. Numbers of species, calculated up to the end of 1980.

2. Anagalida — 26%; Pantodonta = 25%; Condylarthra = 15% (percentage calculated as ratio of number of species of each order to total number of species).

yet, but Qi (1979)"'‘^’ tentatively gave the follow-
ing list;

Order Insectivora Bowdich, 1821

gen. et sp. nov. Sinosinopa sinen-

147,4

Order Rodentia Bowdich, 1821

paramyid gen. et sp. nov. ['^Asiamys

Tamqiiammys wilsoni Dawson, Li, and
Qi, in press Y'^Maodengomys

Order Carnivora Bowdich, 1821
gen. et sp. indet.'“'^>

Order Condylarthra Cope, 1881

Mongolonyx sp. nov. [“Mongolonyx
prominentis”]"'''’'

?Mesonyx sp. nov. Y'Mesonyx obtusi-

Order Tillodontia Marsh, 1875
gen. et sp. nov. [“Ulanius
Order Pantodonta Cope, 1873

Coryphodontidae, gen. et sp. nov.

["Metacoryphodon luminis"Y'‘*^^
Coryphodontidae, gen. et sp. nov.
[“?Metacoryphodon minor”Y'''^'

The new genus and species names placed between brackets are provisional and
may be subject to revision; thus they cannot yet be introduced into the scientific
nomenclature.

?Pantolambdodon sp. nov. Y'Pan-
tolambdodon minoP'Y''*^'

Order Dinocerata Marsh, 1873

Gobiatherium mirificum Osborn and
Granger, 1934""»'«'

Gobiatherium sp. nov. ['' Gobiatherium
way'or”]"'*'''

Gobiatherium sp. nov. [“Gobiatherium

monolabotum“Y'‘*^'

Order Perissodactyla Owen, 1848

Hyrachyus sp. nov. [“Hyrachyus nei-
mongoliensis’'Y'*^^

Hyrachyus sp. nov. [“Hyrachyus
Hyrachyus sp. nov. [“Hyrachyus me-

dius“Y'^^'

Hyrachyus sp. nov. [“?Hyrachyus
minof'Y'''^^

Colodon cf inceptud'’'''’

Tapiroidea gen. et sp. nov. [“ Euryletes

magnus“Y'''^'

Tapiroidea gen. et sp. nov. [“Euryletes

minimus''Y'‘'^'

Tapiroidea gen. et sp. nov. [“Euryletes

medius“Y'*^^

Schlosseria magister Matthew and
Granger, 1926"”'"'”’

Schlosseria sp. nov. [“Schlosseria

dimera^'Y'*^^

Schlosseria sp. nov. [“Schlosseria mas-

24

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Fig. 7. — Map of a portion of Inner Mongolia, showing Eocene and Oligocene collecting localities of the Central Asiatic Expedition of
the American Museum of Natural History.

Lophialetes expeditus Matthew and
Granger,

?Breviodon minutus (Matthew and
Granger, 1925)*''*^*''^^’*'^®’
Teleolophus sp. nov. ['‘Teleolophus pri-
[“T.

Microtitan sp. nov. [“?Microtitan elon-

Microtitan sp."'*’'

Desmatotitan sp.""*”

Forstercooperia sp. nov. [“?Forstercoop-
eria grandis”]"*'''

Forstercooperia sp. nov. [“Forstercoop-
eria elongata"]"'^'”

Urtinotherium sp. nov. [“?Urtinother-
ium m/«or”]"'*’'

3. Irdin Manha Formation (Early Late Eocene)

3A. Irdin Manha Area

Granger, W., and C. P. Berkey, 1922'*’'

Berkey, C. P., and F. K. Morris, 1927''*’
Radinsky, L. B., 1964"”'

Chow Min-chen and A. K. Rozhdestvensky,
I960'”'

Qi Tao, 1979"-"'

a. Location; 20 miles south-southeast of Iren Dabasu
(Iren-hot).

Coordinates: around 43°30'N; 1 12°15'E
1922: “23 miles south of Iren Dabasu.”

1923: “Telegraph line camp.”

1979: Su-ji-deng-en-ji Mesa

b. Stratigraphic Sequence: Grey sand-clays, sands
and gravels (30 ft).

c. The list of the mammalian fauna:

Irdin Manha Formation (Early Late Eocene)
Order Rodentia Bowdich, 1821
Family Paramyidae Miller and Gidley,
1918

paramyid spp."

Order Creodonta Cope, 1875
Family Oxyaenidae Cope, 1877
Sarkastodon mongoliensis Granger,
1938'*"

Family Hyaenodontidae Leidy, 1869
Paracynohyaenodon morrisi Matthew
and Granger, 1924""’’

Propterodon irdinensis Matthew and
Granger, 1925''’*’

Order Carnivora Bowdich, 1821
Family Miacidae Cope, 1880
Miacis invictus Matthew and Granger,
1925"’*'

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Hapalodectes serus Matthew and Grang-
er, 1925"’*'"'’*'

Andrewsarchus mongoliensis Osborn,

1924"36)'16*)

mesonychid gen. indet."**’

Pachyaena sp. (very large form)''**’
Mesonyx sp."**’

Order Pantodonta Cope, 1873
Family Coryphodontidae Marsh, 1876
Eudinoceras mongoliensis Osborn,
1924"**’

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

25

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
Metatelmatheriurn parvuni Granger and
Gregory, 1943*®’’

Protitan grangeri (Osborn, 1925)*'®®’*®”
Protitan robustus Granger and Gregory,

1943187.

Protitan obliquidens Granger and Greg-
ory, 1943*®”

Microtitan mongoliensis (Osborn,
1925)*'®®)*®”

Gnathotitan berkeyi (Osbom, 1925)*'®®)*®”
Family Lophialetidae Radinsky, 1965
Lophialetes expeditus Matthew and
Granger, 1925*'®*”*'®®’

?Lophialetes sp.*'®®’

Breviodon? minutus (Matthew and
Granger, 1925)"®®’*'®®’
?Rhodopagus pygmaeus Radinsky,
1965*'®®’

Simplaletes sujiensis Qi, 1980"'''”
Family Depereteilidae Radinsky, 1965
Teleolophus medius Matthew and
Granger, 1925"®®’"®®’

Family Helaletidae Osbom, 1892

Helaletes mongoliensis (Osborn,
1923)"®®’"®®’

Family Hyracodontidae Cope, 1879
Triplopus? proficiens (Matthew and
Granger, 1925)"®®’"®”’

Family Rhinocerotidae Owen, 1845
Forstercooperia totadentata (Wood,
1 938)*®*’®’"®”’

Order Artiodactyla Owen, 1 848
Family Choeropotamidae Owen, 1845
Gobiohyus orientalis Matthew and
Granger, 1925"®®’

Gobiohyus pressidens Matthew and
Granger, 1925*'®®’

Gobiohyus robustus Matthew and
Granger, 1925*'®®’

Family Entelodontidae Lydekker, 1883
achaenodont indet."®®’

Mammalia, Order indet.

Family Didymoconidae Kretzoi, 1943
Mongoloryctes auctus (Matthew and
Granger, 1925)"®®’"”®"’

Addition to the list above:

Insectivora

?Pantolestes sp.*”’

Artiodactyla

cf. Archaeomeryx indet.*"”

(From Geology of Mongolia, by C. P. Berkey and
F. K. Morris, 1927. Amer. Mus. Nat. Hist.,
p. 360-361.)

3B. Camp Margetts Area (Huhebolhe Cliff)®

^ “The available evidence thus suggests that the relationship between the beds called
“Irdin Manha” in the Camp Margetts area and the type Irdin Manha beds is complex
and not yet fully understood . . (Radinsky, 1964:5).

According to Qi (1980:28, 31 )"‘*^’the fossil mammals described from Camp Margetts
area (Huhebolhe Cliff) may represent different horizons. Qi divided the Early Tertiary
strata of this area into three zones:

Radinsky, L. B., 1964"®”

Qi Tao, 1979"''”

Qi Tao, 1980"”®’

a. Location: 25 miles south-southwest of Iren
Dabasu; 18 miles west-southwest of Irdin Manha
type locality.

Coordinates: around 43°25'N; 1 1 1°50'E
1923: AMNH field locality no. 147; a few miles
north of Camp Margetts, 1 930.

1930: Six principal localities of AMNH around
Camp Margetts.

1980: IVPP field no. 77037.

b. Stratigraphic Sequence:

Section of 1923 (AMNH loc. 147):

“Houldjin Gravels” beds— (“Early Oligocene”)
Yellow sands and gravels (10 ft).

“Irdin Manha” beds

White to grey arkosic concretionary sand-
stone and conglomerates, a thin local grey
clay (Main fossil layer) (30 ft).

Grey clayey sandstone, with some pink layers
(35 ft).

Barren red sandy clays (40 ft).

Grey clay (at base).

c. The list of the mammalian fauna:

Irdin Manha Formation (Early Late Eocene)
Order Rodentia Bowdich, 1821

Family Paramyidae Miller and Gidley,
1918

paramyid spp."””*®®’

Family Cocomyidae Dawson, Li, and Qi,
in press

Advenimus burkei Dawson, 1 964*®®’
Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Mongolonyx dolichognathus Szalay and
Gould, 1966*'®®’

Family Arctocyonidae Murray, 1 886
Paratriisodon gigas Chow, Li, and Chang,

19 7 3(148.(52)

Order Dinocerata Marsh, 1873

Family Uintatheriidae Flower, 1876
Gobiatherium mirificum Osborn and
Granger, 1932*'”®'

Order Perissodactyla Owen, 1 848
Family Brontotheriidae Marsh, 1873
Metatelmatheriurn cristatum Granger
and Gregory, 1943'®”

Protitan minor Granger and Gregory,

194 3(87’

?Protitan cingidatus Granger and Greg-
ory, 1943'®”

Upper (Late Eocene). Irdin Manha [77037]; only five meters in thickness;

Middle (Middle Eocene), Arshanto, with 35 mammalian species (see Section 2-2),
including 22 new ones field nos. [77027, 77034, 77036, 77039]; about 30 m,
Lower (Early Eocene). Bayan Ulan, with Mongololherium sp. (see 2-1); about
36 m.

Qi mentioned that only six sp>ecies were collected from the Irdin Manha Formation
and the others may be from the Middle zone. Until the study of the new collection
has been completed, it is very difficult to give accurate stratigraphic data for the fossils
listed here. Following the American authors, we tentatively refer all the specimens
collected from this area to the same horizon, the Irdin Manha Formation.

26

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Fig. 8.— Topographic map of the Shara Murun region of Inner Mongolia, showing camp sites (marked by an X) and collecting localities
of the Central Asiatic Expedition of the American Museum of Natural History. Contour interval is 100 feet.

Family Lophialetidae Radinsky, 1965
Breviodon sp.“^*’

Lophialetes expeditus Matthew and
Granger, 1925"2«>58'
cf Schlosseria magister Matthew and
Granger,

Family Deperetellidae Radinsky, 1965
cf. Teleolophus medius Matthew and
Granger, 1925"-'’"'^*'

Family Helaletidae Osborn, 1892
Helaleles fissus (Matthew and Granger,
1 925)<>-^>"5®>

?Helaletes fissus (Matthew and Granger,

1 925)0 26K1S8)

Helaletes sp."’®'
cf Hyrachyus sp."®®’

Family Rhinocerotidae Owen, 1845
Forstercooperia confluens (Wood,

19 6 3)0 59X206)

Family Chalicotheriidae Gill, 1872

Litolophus gobiensis (Colbert,

19 3 4)(65)0 56)

3C. Shara Murun Area®

^ As lo the East Mesa of Shara Murun Region, ihe typical section at Urtyn Obo was
published by Osborn (1929:5) (Fig. 9).

The fossil mammals of the Early Late Eocene described from East Mesa were few
and most forms were shared with North Mesa and Baron Sog Mesa.

Dr. Radinsky’s opinion is that:

“The main difficulty in correlating the strata exposed at East Mesa. Urtyn Obo,

(1) North Mesa; including Buckshot’s quarry.

Chimney Butte— the Ulan Shireh beds.

(2) Baron Sog Mesa— “the lower red beds or

Tukhum beds.”

Berkey, C. P., and F. K. Morris, 1927'‘'>

Berkey, C. P., W. Granger, and F. K. Morris,

1929(3)

Radinsky, L. B., 1964*'®^'

Chow Min-chen and A. K. Rozhdestvensky,

1960'”>

Qi Tao, 1974'"’"'

a. Location: West of Shara Murun River, Inner
Mongolia.

Coordinates; around 42°30'N; 1 1 1°00'E
(Fig. 8)

b. Stratigraphic Sequence (after Radinsky, 1964, p.
7-8):

and Nom Khong Shireh with the type Shara Murun and Ulan Gochu beds at Baron
Sog Mesa is that the lithology of these beds is too variable to allow correlation on
the basis of lithology alone.

“The main problem in working out the stratigraphic sequence east of the Shara
Murun River is the delimiting of the boundaries between the Ulan Shireh, Shara
Murun and Ulan Gochu beds. These apparently must be determined largely on
paleontology evidence, but the relevant collections have not yet been completely
studied, and the faunal successions are not well enough known to provide the nec-
essary information.

“It would thus appiear that the stratigraphic information for specimens collected
at East Mesa, Urtyn Obo and Nom KJiong Shireh cannot be completely trusted.
Specimens from beds called ’Shara Murun’ or ‘Ulan Gochu’ at those localities are
not necessarily the same age as the respective type fauna.” (Radinsky, 1965:9-10).

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

27

Fig. 9.— Section of Urtyn Obo. Baluchithere Camp. (Published in Osbom 1929:5, fig. 2, section 2; the explanation of the section following
Radinsky, 1964:9.)

1 . Baron Sog Mesa:

Baron Sog beds: white clays and sands ( 1 5-20
ft).

Ulan Gochu beds: red clay (50 ft).

Shara Murun beds: light varicolored clays, pre-
dominantly grey toward the top (over 200
ft).

Tukhum beds: red beds.

2. North Mesa:

Ulan Shireh beds: richly fossiliferous, multi-
colored, predominantly red clays (over 1 50
ft).

c. The list of the mammalian fauna:

Irdin Manha Formation (Early Late Eocene)
Order Lagomorpha Brandt, 1855
Family Leporidae Gray, 1821
Shamoiagus granger! Burke, 1941"°’
Order Rodentia Bowdich, 1821

Family Cocomyidae Dawson, Li, and Qi,
in press""’

Advenimus bohlini Dawson, 1964''’*’
cf. Advenimus sp."’*’

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875

?Harpagolestes orientalis Szalay and
Gould, 1966"'’*’

mesonychid gen. indet. (very large
form)"®*’

cf. Mesonyx sp."®*’

Order Pantodonta Cope, 1873

Family Coryphodontidae Marsh, 1876
Eudinoceras kholobolchiensis Osborn
and Granger, 1931"'"’

Family Pantolambdodontidae Granger and
Gregory, 1934

Pantolambdodon inermis Granger and
Gregory, 1934'**’

Pantolambdodon fortis Granger and
Gregory, 1934'**’

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873

Desmatotitan tukhumensis Granger and
Gregory, 1943'**’

Epimanteoceras formosus Granger and
Gregory, 1943'**’

Protitan bellus Granger and Gregory,
1943'**’

Microtitan mongoliensis (Osborn,
1925)'’**’'**’

Dolichorhinoides angustidens Granger
and Gregory, 1943'**’

Family Lophialetidae Radinsky, 1965
Breviodon acares Radinsky, 1965'’**’
cf. Breviodon acares Radinsky, 1965"**’
?Breviodon sp."**’

Lophialetes expeditus Matthew and
Granger, 1925'’*®’"**’

Lophialetes sp."**’

Rhodopagus pygmaeiis Radinsky,
1965"**’

Simplaletes ulanshierhensis Qi, 1980'"’“”

Family Deperetellidae Radinsky, 1965
?Teleolophus medius Matthew and
Granger, 1925"*®’"**’

Family Hyracodontidae Cope, 1879
?Triplopus proficiens (Matthew and
Granger, 1925)"*®’"*'*’

Teilhardia pretiosa Matthew and Grang-
er, 1926"**’"*'”

Family Amynodontidae Scott and Osbom,
1883

?Lushiamynodon sharamurenensis Xu,
1966'*'*’

Family Rhinocerotidae Owen, 1845
Eorstercooperia sp.' ' *■”

4. Shara Murun Formation (Late Late Eocene)

Shara Murun Area-Ula Usu

Berkey, C. P., and W. Granger, 1923'*’

Berkey, C. P., and F. K. Morris, 1927''"

Chow Min-chen and A. K. Rozhdestvensky,
I960'**’

28

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Radinsky, L. B„ 1964"5"'

Qi Tao, 1979"“’'

a. Location; Ula Usu, Baron Sog Mesa, west of the
Shara Murun River, Inner Mongolia.
Coordinates; around 42°30'N; 1 10°45'E

b. Stratigraphic Sequence (Baron Sog Mesa);

Baron Sog beds; light grey clays and sands (upper

white stratum) ( 1 5 ft).

Ulan Gochu beds; red clay (upper red beds) (50
ft).

Shara Murun beds; soft grey clay, with brown,
red, purple layers in the bottom part, main
fossil layer (over 200 ft).

Turkhun beds; hard red clay (Teilhardia pretiosa
beds).

c. The list of the mammalian fauna;

Shara Murun Formation (Late Late Eocene)
Order Lagomorpha Brandt, 1855
Family Leporidae Gray, 1821
Shamolagus medius Burke, l94l"o>"<”>
Gobiolagus tolmachovi Burke, 1941"“'
Order Rodentia Bowdich, 1821

Family Yuomyidae Dawson, Li, and Qi,
in press

Yuomys cavioides Li, 1975"°*'

Family indet. (Matthew and Granger, 1925;
4; not seen)"’“'

Order Creodonta Cope, 1875
Family Hyaenodontidae Leidy, 1869
Pterodon hyaenoides Matthew and
Granger, 1925"’“’

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
Rhinotitan kaiseni (Osborn, 1925)"°*"*’'
Rhinotitan mongoliensis (Osborn,
1 925)"°*"*’'

Rhinotitan andrewsi (Osborn,
1 925)"°*"*’'

Pachytitan ajax Granger and Gregory,

194 3(87,

Titanodectes minor Granger and Greg-
ory, 1943'*’’

Family Lophialetidae Radinsky, 1965
Rhodopagus? minimus (Matthew and
Granger, 1926)"°‘""°*’

Family Deperetellidae Radinsky, 1965
Deperetella cristata Matthew and
Granger, 1925"°“'"°*’

Family Hyracodontidae Cope, 1879
Triplopus? progressus (Matthew and
Granger, 1925)"°“'"°'”

Family Amynodontidae Scott and Osborn,
1883

Amynodon mongoliensis Osborn,
1936"“°"°'°’

Sianodon ulausuensis Xu, 1966'°'°’
Lushiamynodon sharamurenensis Xu,
1966'°'°’

Gigantamynodon promisus Xu, 1 966'° ' °’
Sianodon spp.'°'°’

Amynodontidae indet. '°'°'

Caenolophus promissus Matthew and
Granger, 1925"°“’"°*'

Caenolophus obliquus Matthew and
Granger, 1925"°“"'°*'

Family Rhinocerotidae Owen, 1845
Juxia sharamurenense Chow and Chiu,
1 964'“”"°^’

Family Chalicotheriidae Gill, 1872

Olsenia mira Matthew and Granger,
1925"°“"°°’"*°’"°*’

Order Artiodactyla Owen, 1848
Family Anthracotheriidae Gill, 1872
Ulausuodon parvus Hu, 1963'^°’

Family Hypertragulidae Cope, 1879
Archaeomeryx optatus Matthew and
Granger, 1925"°“’

II. Xin-jiang (Sinkiang) Region
5. Turpan (Turfan) Basin

Zhai Ren-jie, Zheng Jia-jian, and Tong Yong-
sheng, 1978'°“"

Zhai Ren-jie, 1978'°°”'°°*’

Zheng Jia-jian, 1978'°°°’

a. Location; Turpan and Shan-shan counties, Xin-
jiang Region.

Coordinates; 42°45'-43°12'N; 89°05'-9 1°36'E;
43°12'N; 91°36'E (Shi-san-jian-fang railway
station)

b. Stratigraphic Sequence;

Lian-kan Formation (Late Late Eocene)
Grey-red muddy sandstones, greyish-yellow,
white and red sandstones, orange-red sand-
stones and blue conglomerates [66014,
66020] (80 m).

Shi-san-jian-fang Formation (? Late Early Eocene
or Early Middle Eocene)

Yellow fresh limestones, greyish yellow coarse
sandstones, red muddy sandstones, light
red sandstones, and greyish red, white con-
glomerates [66019] (300 m).

Da-bu Formation (Early Eocene)

Purple-red mudstones with white calcareous
concretionary layers, grey-white mud-
stones, sandstones and conglomerates
[66015] (22 m).

c. The list of the mammalian fauna;

Lian-kan Formation (Late Late Eocene)

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873

?Rhinotitan sp.'°°°’

Family Lophialetidae Radinsky, 1965
Lophialetes expeditus Matthew and
Granger, 1925'°°°’"°°’
gen. et sp. indet. 1. 2. 3.'°°°’

Family Deperetellidae Radinsky, 1965
Teleolophus liankanensis Zheng, 1 978'°°°’
Family Amynodontidae Scott and Osborn,
1883

Amynodon mongoliensis Osborn,
1936'°°°’"“°’

Amynodon sp.'°°°’

Order Artiodactyla Owen, 1848

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

29

Table 3.— A comparison of the fossils from the Irdin Manha For-
mation and Shara Murun Formation.

Irdin Manha Fauna
(Early Late Eocene or
Late Middle Eocene)

Shara Murun Fauna
(Late Late Eocene)

Shamolagus grangeri

Shamolagus medius

paramyid spp.
Advenimus burkei
Advenimus bohlini
cf. Advenimus sp.

Rodentia indet.
Yuomys cavioides

Paracynohyaenodon morris
Propterodon irdinensis
Sarkastodon mongoliensis

Pterodon hyaenoides

Miacis invictus

Paratriisodon gigas
Hapalodectes serus
Andrewsarchus mongoliensis
Mongolonyx dolichognathus
Harpagolestes orientalis
cf, Mesonyx sp.

Pachyaena sp.

Mesonychidae indet.
Eudinoceras mongoliensis
Eudinoceras kholobolchiensis
Pantolambdodon inermis
Pantolambdodon fortis
Gobiatherium mirifkum

Metatelmatherium cristatum
Metatelmatherium parvum
Protitan grangeri
Protitan robustus
Protitan obliquidens
Protitan minor
?Protitan cingulatus
Protitan bellus
Microtitan mongoliensis
Gnathotitan berkeyi
Desmatotitan tukhumensis
Epimanteoceras formosus
Dolichorhinoides angustidens

Rhinotitan kaiseni
Rhinotitan mongoliensis
Rhinotitan andrewsi
Pachytitan ajax
Titanodectes minor

Litolophus gobiensis

Olsenia mira

Lophialetes expeditus
Lophialetes sp.

Breviodon ? minutus
Breviodon acares
cf Breviodon acares
?Breviodon sp.

?Rhodopagus pygmaeus
Simplaletes sujiensis
Simplaletes ulanshierhensis
cf Schlosseria magister

Rhodopagus? minimus

Teleolophus medius
TTeleolophus medius

Deperetella cristata

Helaletes mongoliensis
Helaletes fissus
?Helaletes fissus

Table 3.

— Continued.

Irdin Manha Fauna
(Early Late Eocene or
Late Middle Eocene)

Shara Murun Fauna
(Late Late Eocene)

Helaletes sp.
cf Hyrachyus sp.

Triplopus? prof dens
Eorstercooperia totadentata
Eorstercooperia confluens
Teilhardia pretiosa

Triplopus ? progressus
Juxia sharamurunense

?Lushiamynodon sharamure-
nensis

Amynodon mongoliensis
Sianodon ulausuensis
Sianodon spp.

Lushiamynodon sharamure-
nensis

Gigantamynodon promisus
Amynodontidae indet.
Caenolophus promissus
Caenolophus obliquus

Gobiohyus orientalis
Gobiohyus pressidens
Gobiohyus robustus
archaenodont indet.

Ulausuodon parvus

Mongoloryctes auctus

Archaeomeryx optatus

63 forms

24 forms

Family Anthracotheriidae Gill, 1872
Bothriodon

Family Flypertragulidae Cope, 1879
Xinjiangmeryx parvus Zheng, 1978<^^°>
Shi-san-jian-fang Formation (? Late Early Eocene
or Early Middle Eocene) [66019]

Order Anagalida Szalay and McKenna, 1 97 1
Family Eurymylidae Matthew, Granger,
and Simpson, 1929

Rhombomylus turpanensis Zhai, 1 978*^^*'
Order Notoungulata Roth, 1903
Family Arctostylopidae Schlosser, 1923
A natolostylops dubius Zhai, 1978'^’®’
Order Condylarthra Cope, 1881
Family Hyopsodontidae Lydekker, 1 889
Hyopsodus sp.'^®®’

Order Pantodonta Cope, 1873
Family Coryphodontidae Marsh, 1876
Coryphodon sp.*^®®’

Order Perissodactyla Owen, 1 848
Family Helaletidae Osborn, 1872
Heptodon tianshanensis Zhai, 1978'®®®’
Da-bu Formation (Early Eocene)

Order Pantodonta Cope, 1873
Family Coryphodontidae Marsh, 1876
Coryphodon dabuensis Zhai, 1978'®®®*
Order Dinocerata Marsh, 1873
Family Uintatheriidae Flower, 1876
Pyrodon xinjiangensis Zhai, 1978'®®®’

30

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

6. Jung-gur (Dzungar) Basin

Only four forms of Eocene mammals were dis-
covered in the Jung-gur Basin.

The stratigraphic and geographic distributions
of these fossils may be tentatively given as
following'’^-

Fossil

Locality

Horizon

Lophialetes cf

North area of

U-lun-gu For-

expeditus

the Basin

mation (Late

(around 46°N;
88°E)

Eocene)

Felidae

North of the

U-lun-gu For-

Manas Lake

mation (Late

(around 46°N;
87°E)

Eocene)

Eudinoceros

North area of

Hong-li-shan

sp.*'"*’

the Basin (Yi-

Formation

xi-bu-la-ke.

(Late Eocene or

south part of
Ulungur
depression,
around 46°N;
88°E)

earlier)

Bothriodon sp.

South area of

Lu-se (Green)

the Basin

Formation

(Manas River
area, around
44°N; 86“E)

(Late Eocene)

III. Shaan-xi (Shensi) Province

7. Lan-tian District

Chia Lan-po, Chang Yu-ping, Huang Wan-po,
Tang Ying-jun, Chi Hung-xiang, Yu Yii-zhu,
Ting Su-yin, and Huang Xue-shi, 1966*'*’
Wang Ban-yue, 1978"*’'’’

a. Location: Lin-tong County, Shaan-xi Province.
Coordinates: 34°24'N; 109°13'E (Lin-tong city)

b. Stratigraphic Sequence:

Hong-he Formation (Late Eocene)

Purplish red mudstone and sandy mudstone
interbedding with fine sandstone [65009,
65013] (200 m).

c. The list of the mammalian fauna:

Hong-he Formation (Late Eocene)

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873

Arctotitan honghoensis Wang, 1978"'’*’’
[65009]

Family Lophialetidae Radinsky, 1965
Breviodon sp. [65013]

Family Deperetellidae Radinsky, 1965
cf Deperetella sp."'"”

IV. Bei-jing (Peking) and He-bei (Hopei) Province

8. Chang-xin-dian (Changsintien) Locality

Young Chung-chien, 1934*^^*’

Chow Min-chen, 1953"*’’

Zhai Ren-jie, 1977'^*'"

a. Location: About 20 km southwest of the Bei-jing
(Peking) City.

Coordinates: 39°49'N; 1 16°14'E (Chang-xin-dian
town)

b. Stratigraphic Sequence: A thick series of conglom-
erates and red clays.

c. The list of the mammalian fauna:

Order Insectivora Bowdich, 1821
Family Erinaceidae Bonaparte, 1838
?Tupaiodon sp."*’’"^'”

Order Anagalida Szalay and McKenna, 1971
Family Eurymylidae Matthew, Granger,
and Simpson, 1929
Hypsimylus beijingensis Zhai, 1977*^*'”
Order Carnivora Bowdich, 1821
Family Miacidae Cope, 1880
Miacis sp.'^**”

Family Canidae Gray, 1821
gen. et sp. indet.*^**”

Order Pantodonta Cope, 1873

Family Coryphodontidae Marsh, 1876
?Eudinoceras sp.*-^'’’

Order Perissodactyla Owen, 1848
Family Hyracodontiade Cope, 1879
Imequincisoha sp.*^*"’* '*’'”

9. Ling-shan Locality

In 1961 a few fragmentary mammalian teeth
were found in the Ling-shan area (38°47'N;
114°38'E). Judging from a molar of an
amynodontid'**’*^'*’, the deposits yielding the fos-
sils may be correlated with the Yuan-chu For-
mation as Late Eocene in age.

V. Shan-dong (Shantung) Province

10. Wu-tu Basin

Li Chuan-kuei, 1962'"’*’

Chow Min-chen and Li Chuan-kuei, 1965**"

a. Location: Wu-tu coal mine, 10 km southeast of
Chang-le County, Shan-dong Province.
Coordinates: 36°42'N; 1 18°49'E (Chang-le city)

b. Stratigraphic Sequence:

Wu-tu Formation (Early Eocene)

Dark colored oil-shales, coals, and purple-red,
green mudstones and conglomerates
[62058].

c. The list of the mammalian fossils:

Wu-tu Formation (Early Eocene)

Order Perissodactyla Owen, 1 848
Family Isectolophidae Peterson, 1919
Homogalax wutuensis Chow and Li,
1965**""**’

11. Niu-shan Basin

Li Chuan-kuei, 1962"*’*’

Chow Min-chen and Li Chuan-kuei, 1965**"

a. Location: 8 km south of Lin-qu (Lin-chu) city.
Coordinates: 36°30'N; 1 18°32'E (Lin-qu city).

b. Stratigraphic Sequence:

Niu-shan Member of Wu-tu Formation (Early
Eocene)

Black oil shales, greyish green and brownish red
clay, mudstone and sandstone [62057].

c. The list of the mammalian fossils:

Niu-shan Member of Wu-tu Formation (Early
Eocene)

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

31

Order Perissodactyla Owen, 1848
Family Helaietidae Osborn, 1892
Heptodon niushanensis Chow and Li,

[965<5i)(i5«)

1 2. Xin-tai (Sintai) Basin

Tan, H. C, 1923"™'

Zdansky, 0„ 1930«3S)

Young Chung-chien and Bien Mei-nan, 1935'^'''’’
Li Chuan-kuei, 1962"°^'

a. Location: From north of the Men-yin city to the
north of Xin-tai city, extending NW-SE about 20
km, Shan-dong Province.

Coordinates: 35°54'N; 1 17°44'E (Xin-tai city).

b. Stratigraphic Sequence:

Guan-zhuang Formation (Middle Eocene)
(>1,500 m)

Upper: dark brown conglomerates and breccia.
Middle: greyish green mudstones and grey red
sandstone. Most fossils were collected from
this horizon.’

Lower: conglomerate sandstone intercalated
with marls.

— Unconformity —

Late Jurassic or Early Cretaceous Men-yin Series
yielding Euhelopus zdansky i

c. The list of the mammalian fauna:

Guan-zhuang Formation (Middle Eocene)

Order Rodentia Bowdich, 1821
gen. et sp. indet.'’^"’

Order Creodonta Cope, 1875
Family Hyaenodontidae Leidy, 1 869
?Thinocyon sichowensis Chow, 1975*'"’
Order Condylarthra Cope, 1881
Family ?Hyopsodontidae Lydekker, 1889
?Haplomylus sp.'’^^’

Order Tillodontia Marsh, 1 875
?Family Tillodontidae Marsh, 1875
Kuanchuanius shantunensis Chow,

196307)

Order Pantodonta Cope, 1873
Family Coryphodontidae Marsh, 1876
Coryphodon flerowi Chow, 1957'’''
Order Dinocerata Marsh, 1873
Family Uintatheriidae Flower, 1876
?cf Uintatherium sp.'^*’

Order Perissodactyla Owen, 1848
Family Palaeotheriidae Gill. 1872
Propalaeotherium sinensis Zdansky,
1930'™"

Family Equidae Gray, 1821
Hyracotheriinae, gen. et sp. indet. 1 . 2.'™^'
Heptaconodon ditbium Zdansky, 1930'™^'
Family ?Lophialetidae Radinsky, 1965
?Rhodopagus sp.'’™>'o®>

7 The Middle Eocene mammals of Xin-tai Basin were mostly collected by Tan (1922)
around a village of K.uan-chuang (perhaps the Chang-lu-kuan-chuang, 10 km SE of
Xin-tai city), except a chalicothere, Grangeria canina. which may have been discovered
from a younger horizon ("Ardyn Obo” or Sannoisian). Unfortunately, we cannot re-
examine where Tan’s localities were. All the materials collected by Li and his colleagues
(1960, 1962) were found in a small quarry, Xi-xi-zhou (Si-si-chou), 5 km WNW of
Xin-tai city (probably the same locality of Tan’s Hsi-kou).

Family Helaietidae Osborn, 1892
Hyrachyus spp.'™"

Family Eomoropidae Viret, 1958

Grangeria canina Zdansky, 1930'™""^^’
VI. Shan-xi (Shansi) Province
13. Yuan-chu Basin

Anderson, J. G., 1923"'

Young Chung-chien, 1937'’™'

Lee Yuen-yen, 1938'"’-"

Chow Min-chen, Li Chuan-kuei, and Chang Yu-
ping, 1973'^’’

Hartenberger, J.-L., J. Sudre, and M. Vianey-
Liaud, 1975'«"

a. Location: More than 20 Late Eocene mammalian
localities of Yuan-chu Basin mainly discovered
along both banks of the Yellow River from the
old city of Yuan-chu, Shan-xi Province, upward
to the village of Ren-cun (Jen-tsun), Mian-chi
County, He-nan Province.

Coordinates: 35°06'N; 1 1 l°53'E(old city ofYuan-
chii)

(Fig. 10)

b. Stratigraphic Sequence:

Yuan-chii Series

Bai-shui-cun (Pai-sui-tsun) Formation (Early Oli-
gocene)

Light blue and white mudstones, limestones
and lignites ( 1 5 m).

He-ti (Ho-ti) Formation (Late Late Eocene) (up
to 1,000 m)

Chai-li Member (“River section”):

Greyish white marls, light colored mudstones
and sandstones.

— Disconformity—

Ren-cun (Jen-tsun) Member:

Yellow, green, and varied colored clays,
shales, and sandstones, with very thick
conglomerates in the basal portion.

c. The list of the mammalian fauna:

Bai-shui-cun Formation (Early Oligocene)

Order Artiodactyla Owen, 1848
Family Anthracotheriidae Gill, 1 872
Brachyodus hui (Chow, 1958)*’'"’'’"'’'^’
He-ti Formation (Late Late Eocene)

Chai-li Member (“River section” and Loc. 1 of
Zdansky, Field no. 5301 of Chow):

Order Insectivora Bowdich, 1821
Family Leptictidae Gill, 1872
Ictopidium lechei Zdansky, 1930'™^''^’’
Order Primates Linnaeus, 1758
Family Omomyidae Trouessart, 1879
Hoanghonius stehlini Zdansky,

|930(235’1202’167)'52'

Order Rodentia Bowdich, 1 82 1

Family ?Ischyromyidae Alston, 1876
gen. et sp. indet. (or Ctenodactyloidea)"""
Family Cricetidae Rochebrune, 1883
Cricetodon schaubi Zdansky,

J 930(735’! ,52)'89’

Family Zapodidae Coues, 1875
?Plesiosminthus sp.

32

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Order Creodonta Cope, 1875
Family Hyaenodontidae Leidy, 1869
Hyaenodon yuanchuensis Young, 1937
(Loc. F 10)<22^>'52>

Order Carnivora Bowdich, 1821
Family Canidae Gray, 1821
Chailicyon crassidens Chow, 1975'“"'
Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
Rhinotitan mongoliensis (Osborn,
1 925)'-^'""^*"*''’

Family Amynodontidae Scott and Osborn,
1883

?Amynodon mongoliensis Osborn,

19 3 6(227)(M0,

?Amynodon sp. (=Cadurcodon ardynen-
sis of Young)'^^'""’'"

Sianodon sinensis (Zdansky,

Order Artiodactyla Owen, 1848
Family Anthracotheriidae Gill, 1872
Anthracokeryx sinensis (Zdansky,

1930)(52)(2I3)(235)

Anthracokeryx cf. sinensis (Zdansky,

1930)152X2 I 3M235)

artiodactyl indet. (=“Hoanghonius steh-
lini,” No. 2 of Wood and Chow)'“^"^^'
Ren-cun Member:

Order Primates Linnaeus, 1758
Family Omomyidae Trouessart, 1879
Hoanghonius stehlini Zdansky,

1930(235)(202)(67|,52) [53H_14]

Order Rodentia Bowdich, 1821

Family Yuomyidae Dawson, Li, and Qi,
in press

Yuomvs cavioides Li, 1975(52x108) 1-53 j 3_

14]

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Honanodon hebetis Chow, 1965'“""'’^’

[5314]

Order Tillodontia Marsh, 1875
Family Tillotheriidae Marsh, 1875
Adapidium huanghoense Young,

1937(227X79X37) [p]2]

Tillodontia gen. et sp. indet.'"’''^^’ [5313]

Order Perissodactyla Owen, 1 848

Family ?Lophialetidae Radinsky, 1965
?Rhodopagus sp.'^^sxiss)

Family Deperetellidae Radinsky, 1965

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

33

Deperetella depereti (Zdansky,

1930)<235KI58)(227, [p[2^ LOC.7]

Diplolophodon ( Deperetella) similis
Zdansky, [FI 2, F23]

Family Hyracodontidae Cope, 1879
Caenolophus cf. promissus Matthew and
Granger, 1925«”«'^»' [FI 2, F23]
Family Amynodontidae Scott and Osborn,
1883

Sianodon sinensis Chow and Xu, 1 965"’^'
Sianodon mienchiensis Chow and Xu,

1965'^2) [5314]

Amynodon mongoliensis Osborn,
■ 1936, 227, „40, [PI 2, P22]

Family Rhinocerotidae Owen, 1845
Prohyracodon cf. meridionalis Chow and
Xu, 1961'”""" [5312, 5314]

Family Eomoropidae Viret, 1958

Eomoropus quadridentatus Zdansky,
1930i235„9O)(i56) [Lqc. 7, Ec-lang-gou]
Granger ia ?major (Zdansky, 1930)'-^^'"^'’’
[Loc. 7]

Order Artiodactyla Owen, 1 848
Family Dichobunidae Gill, 1 872
?Dichobune sp.'^^’

Family Choeropotamidae Owen, 1845
Gobiohyus yuanchuensis Young,

19 3 7,227X52, [Pip P12]

Family Anthracotheriidae Gill, 1872
Anthracothema minima Xu, 1962*^'^**’*’
[5313]

Anthracokeryx sinensis (Zdansky,

1930)1235X2, 3, [53ii_i4]

Anthracosenex ambiguus Zdansky,
1930(235,(227, [Loc. 7, F12]

VII. He-nan (Honan) Province

14. Tan-tou Basin

For the location and the stratigraphic sequence of
the Basin see Section 1-6.

The list of the mammalian fauna of Tan-tou For-
mation (Early Eocene):

Order Dinocerata Marsh, 1873
Family Uintatheriidae Flower, 1876
Prodinoceratinae gen. et sp. indet."®'”
Order Pantodonta Cope, 1873

Family Archaeolambdidae Flerov, 1952
gen. et sp. indet."*'*'

15. Ji-yuan (Chiyuan) Basin

Liu Hsien-ting et al., 1963'"^’

Chow Min-chen, Li Chuan-kuei, and Chang Yu-
ping, 1973*”’

a. Location: Ji-yuan County, north bank of the Yel-
low River, He-nan Province.

Coordinates: 35°08'N; 1 12°35'E (Ji-yuan city)

b. Stratigraphic Sequence:

Ji-yuan Formation (new) (Late Eocene)

Red clays, brown-red sandstone, sandy clays,
and conglomerate (500 m).

— Unconformity—

Jurassic

c. The list of the mammalian fauna:

Ji-yuan Formation (Late Eocene)

Order Rodentia Bowdich, 1821

Family Yuomyidae Dawson, Li, and Qi,
in press

Yuomys cavioides Li, 1975(52)1108,

Order Perissodactyla Owen, 1 848

Family Amynodontidae Scott and Osborn,
1883

Sianodon chiyuanensis Chow and Xu,
1965'“’

Sianodon sinensis (Zdansky, 1 930)''>2k235,
Lushiamynodon obesus Chow and Xu,

1965(62)

1 6. Ling-bao Basin

Chow Min-chen, Li Chuan-kuei, and Chang Yu-
ping, 1973'^^’

Tong Yong-sheng and Wang Jing-wen, 1980"*'”

a. Location: 50 km north of Lu-shi city and south
bank of Yellow River.

Coordinates: 34'’34'N; 1 10°42'E (Ling-bao city)

b. Stratigraphic Sequence:

Hun-shui-he Formation (Late Eocene) (200 m)
Upper: Greyish green, reddish brown mud-
stones with intercalating sandstones.
Middle: Greyish brown sandy mudstones,
sandstones.

Lower: Reddish brown mudstones with inter-
calating sandstone and gravels.

Chuan-kou Formation (Late Middle Eocene or
Early Late Eocene)

Purplish red sandy mudstones and conglom-
erates.

Xiang-cheng Group (Paleocene to Eocene)
Purplish red silty mudstones, sandstones etc.
No fossils. (700 m)

Nan-chao Formation (Cretaceous)

Red clays, containing Macroblithes yaotunen-
sis.

c. The list of the mammalian fauna:

Hun-shui-he Formation (Late Eocene)

Order Perissodactyla Owen, 1 848
Family Brontotheriidae Marsh, 1873
gen. et sp. indet."*'”

Family Hyracodontidae Cope, 1879
Caenolophus sp."*'*’

Family Amynodontidae Scott and Osborn,
1883

Amynodontidae indet."*'”

Order Artiodactyla Owen, 1 848
Family Anthracotheriidae Gill, 1872
gen. et sp. indet."*'”

Family Hypertragulidae Cope, 1879
Archaeomeryx sp."*”

Chuan-kou Formation (Middle Eocene to Late
Eocene)

Order Perissodactyla Owen, 1 848
Family Helaletidae Osborn, 1892
Hyrachyus sp.

?Family Amynodontidae Scott and Osborn,
1883

?Sianodon sp."*”

34

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

17. Lu-shi Basin

Teilhard de Chardin, P., G. B. Barbour, and M.

N. Bien, 1935"”'

Lee Yue-yen, 1938"°-'

Li Chuan-kuei, 1957""“"

Chow Min-chen, Li Chuan-kuei, and Chang Yu-
ping, 1973’^-"

Tong Yong-sheng and Wang Jing-wen, 1980"*“"

a. Location: Lu-shi County, on the northern slope
of the east part of Tsin-ling Mountain. Length: 30
km. Width: 7-15 km.

Coordinates: 34°04'N; 1 10°02'E (Lu-shi city)
The main fossil locality, Men-chia-pu, or Field
no. 57202, is about 1.5 km southwest of the city
of Lu-shi. Almost all the fossils are from a patch
of sediments which is the filling of a sinkhole in
the limestone on an erosional surface antecedent
to the deposition of the Lu-shi Formation.

b. Stratigraphic Sequence:

Da-yu Formation (?01igocene)

Red conglomerates, light colored mudstones and
greenish shales; no mammalian fossils (800
m).

Chu-gou-yu Formation (Late Late Eocene)
Greyish green calcareous mudstone, sandy
mudstone with intercalating greyish yel-
low, coarse sandstone (700 m).

Lu-shi Formation (Early Late Eocene)

Dark red mudstones, brown, grey-white marls,
red conglomerates (400 m).

c. The list of the mammalian fauna [57202 Quarry]:
Lu-shi Formation (Early Late Eocene)

Order Primates Linnaeus, 1758
Family Anaptomorphidae Cope, 1883
Lushius qinlinensis Chow, 1961'^“"
Order Lagomorpha Brandt, 1855
Family Leporidae Gray, 1821
Lushilagus lohoensis Li, 1965""^'

Order Rodentia Bowdich, 1821

Family Cocomyidae Dawson, Li, and Qi,
in press

TsinUngomys youngi Li, 1963"“’

Order Creodonta Cope, 1875
Family Hyaenodontidae Leidy, 1869
Propterodon irdinensis Matthew and
Granger, 1925*“"’"-^'

Order Carnivora Bowdich, 1821
Family Miacidae Cope, 1880
Miacis lushiensis Chow, i975oi)(52)
Family Canidae Gray, 1821
Cynodictis sp.'^^'

Family Felidae Gray, 1821
cf Eusmilus sp.’”"^"

Order Condylarthra Cope, 1881

Family Arctocyonidae Murray, 1 866
Paratriisodon henanensis Chow, 1959‘^‘"
Paratriisodongigas Chow, Li, and Chang,

19 7 3(52)

Family Mesonychidae Cope, 1875

Honanodon macrodontus Chow,

1 qbSotDdss)

Lohoodon lushiensis (Chow,

1 965)<'‘0>I52)(168)

Order Taeniodonta Cope, 1876
Family Stylinodontidae Marsh, 1875
?Stylinodon sp.'^^’"''"

Order Pantodonta Cope, 1873

Family Coryphodontidae Marsh, 1876
Eudinocems sp.'“'

Order Perissodactyla Owen, 1 848
Family Brontotheriidae Marsh, 1873
?Microlitan sp. (or new genus)'^^'
Protitan grangeri (Osborn, 1925)'^^"*’''
Family Lophialetidae Radinsky, 1965
Breviodon miniitus (Matthew and
Granger, 1925)*^^'"^®'

Family Helaletidae Osborn, 1892
?Colodon sp.*^'^'

Family Deperetellidae Radinsky, 1965
Deperetella sp.‘”'

Family Hyracodontidae Cope, 1879
Caenolophus sp. (or Amynodonti-

dae)'52Ki59)

Family Amynodontidae Scott and Osborn,
1883

Lushiamynodon menchiapuensis Chow
and Xu, 1965''’"'<5'"

Sianodon honanensis Chow and Xu,

] 9^5)62)(52)

Family Rhinocerotidae Owen, 1 845
Prohyracodon sp.'”'

Eorstercooperia spp.'^-'

Family Eomoropidae Viret, 1958
Lunania youngi Chow, 1 957(^5)0 56h52)
Eomoropus sp.'^^'

Order Artiodactyla Owen, 1 848
Family Dichobunidae Gill, 1872
Dichobune sp.'^^'

Family Choeropotamidae Owen, 1845
Gobiohyus orientalis Matthew and
Granger, 1925'^'""^^'

Gobiohyus robustus Matthew and
Granger, 1925'^^'"^^'

Family Anthracotheriidae Gill, 1872
Anthracotherium spp.'”'

Tong and Wang (1980:23) reported that two
new Eocene mammalian horizons have been dis-
covered in recent years. One is beneath the main
fossils quarry [57202], yielding Lophialetes, Eudi-
noceras(l), mesonychid, Uintatheriinae, Gobio-
hyus and Breviodon etc. The others occurred in
the Chu-gou-yu Formation (Late Eocene), con-
taining Sciuravidae gen. et sp. nov., Yuomys sp.
nov., Palaeolaginae indet., Eomoropus major,
Breviodon sp. nov., Eorstercooperia sp., Archaeo-
meryx optatus etc.

18. Wu-cheng Basin

Gao Yu, 1976'”'

Wang Jing-wen, 1978"'""'
a. Location: Tong-bai County, He-nan Province.
Coordinates: 32°25'N; 1 13°30'E (Wu-cheng town)
(Fig. 11)

1983

LI AND TING-THE PALEOGENE MAMMALS OE CHINA

35

b. Dimension: About 250 km^.

c. Stratigraphic Sequence:

Wu-li-dui Formation (Late Eocene)

Thin greyish green mudstones interbedding with
brownish grey kerogen shales (550 m).

—Conformity—

Li-shi-gou Formation (Late Eocene)

Brownish yellow conglomeratic sandstones and
greyish green sandy mudstones (370 m).

—Conformity—

Mao-jia-po Formation (Late Eocene)

Greyish red breccia and sandy mudstone and
marls.

d. The list of the mammalian fauna:

Wu-li-dui Formation (Late Eocene)

Order Perissodacty la Owen, 1848
Family Hyracodontidae Cope, 1879
Imequincisoria mazhuangensis Wang,
1976"‘"*'

Imequincisoria /n/crads Wang, 1976"'*'^’
Imequincisona(l) sp.*”'”

Family Amynodontidae Scott and Osborn,
1883

Gigantamynodon sp.'-““’
cf. Lushiamynodon sp.'’*’

Sianodon sinensis (Zdansky, 1 930)'-‘“><’“'

Family Rhinocerotidae Owen, 1845
Juxia spp. nov.'^**'

Forstercooperiinae indet.'^*’

Li-shi-gou Formation (Late Eocene)

Order Rodentia Bowdich, 1821

Family Yuomyidae Dawson, Li, and Qi,
in press

Yuomys eleganes Wang, 1978'^“><’®'
Order Creodonta Cope, 1875

Family Hyaenodontidae Leidy, 1869
Hyaenodon sp.'^*’

Order Carnivora Bowdich, 1821
Carnivora indet.*’*’

Order Perissodactyla Owen, 1848

Family Lophialetidae Radinsky, 1965
?Breviodon sp.‘’*'

Lophialetidae gen. et sp. nov.*^*>

Family Deperetellidae Radinsky, 1965
Deperetella sp. nov.'^*>

Family Hyracodontidae Cope, 1879
Hyracodontidae gen. et sp. nov.'^*’
Family Amynodontidae Scott and Osborn,
1883

Lushiamynodon wuchengensis Wang,

1 97g(200)(78)

Amynodon mongoliensis Osborn,
1 936'-“'*^*>

Amynodontidae indet.'^*'

36

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Family Rhinocerotidae Owen, 1 845
Rhinocerotidae indet.'^®'

Pappacems sp.'

Forstercooperia sp. nov.'^*’

Family Eomoropidae Viret, 1958
Eomoropus sp.'^*'

Mao-jia-po Formation (Late Eocene)

Order Perissodactyla Owen, 1848

Family Deperetellidae Radinsky, 1965
Deperetella sp.*’*’

19. Xi-chuan (Sichuan) Basin

Teilhard de Chardin, 1930"’*’

Teilhard de Chardin, P., G. B. Barbour, and M.
N. Bien, 1935"’‘”

Chow Min-chen, Li Chuan-kuei, and Chang Yu-
ping, 1973'”’

Gao Yu, 1976*’*’

Xu Yu-xuan, Yan De-fa, Zhou Shi-quuan, Han
Shi-jing, and Zhang Yong-cai, 1979*”’’

a. Location: Southern slope of Tsinling Mountain,
about 150 km south of Lu-shi Basin. Xi-chuan
Basin (called also Li-quan-qiao [Li-kuan-chiao
Basin]) extends from east of He-tao-yuan, a small
village 45 km south of the Xi-chuan new city, He-
nan Province, to the west of Xi-jia-dian, 20 km
northwest of Jun-xian city, Hu-bei Province.
Coordinates: 32°45'N; 1 10°40'-1 1 1°20'E; 32*’46'N;

1 1 F27'E (Li-quan-quao town)

(Fig. 12)

b. Dimension: Length: more than 40 km (E-W);
Width: 10 km (N-S, at maximum).

c. Stratigraphic Sequence:

He-tao-yuan Formation (Early Late Eocene)
Greyish green mudstones with intercalating
marls, sandstones and conglomerates (800
m).

—Conformity—

Da-cang-fang Formation (Middle Eocene)
Greyish white, brownish yellow sandstones,
conglomerates and red sandy mudstone
(500-1,100 m).

—Conformity—

Yu-huang-ding Formation (Late Early Eocene or
Early Middle Eocene) Light red, white mud-
stone with intercalating purple-red or green
marls (400-900 m).

— Disconformity—

Hu-gang Formation (Cretaceous)

d. The list of the mammalian fauna:

He-tao-yuan Formation (Early Late Eocene)
Order Edentata Cuvier, 1798

Superfamily Megalonychoidea Simpson,
1931

Chungchienia sichuanica Chow, 1963'**’
Order Rodentia Bowdich, 1821
Family Cocomyidae Dawson, Li, and Qi,
in press

gen. et sp. indet. ("Sciuravus" sp.)"*’*”*’*"
Order Creodonta Cope, 1875
Family Oxyaenidae Cope, 1877
Prolaena parva Xu et al., 1979**-”’
Family Hyaenodontidae Leidy, 1869
T'Sinopa" sp.*’*’

T'Tritemnodon” sp.*’*’

Order Carnivora Bowdich, 1821
Family Miacidae Cope, 1880
Miacis lushiensis Chow, ) 975122 ik4I)

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

37

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
?Andrewsarchus sp.'''**'"^'’’

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
?Protitan sp.'^^"

Family Lophialetidae Radinsky, 1965
Lophialetes expeditus Matthew and
Granger, 1925*^^'”'“’

Lophialetes

Breviodon cf. minutus (Matthew and
Granger, 1925)'^’""^*'

Family Deperetellidae Radinsky, 1965
Teleolophus sichuanensis Xu et al.,

S979(22i)

Teleolophus cf. mediiis Matthew and
Granger, 1925'^^"

Family Helaletidae Osborn, 1892
?Colodon sp.‘”"^*’

Family Amynodontidae Scott and Osborn,
1883

Sianodon sp.'^^'"^''"

Family Rhinocerotidae Owen, 1845
Pwhyracodon sp.*^-'’

Da-cang-fang Formation (Middle Eocene)

Order Rodentia Bowdich, 1821

Family Cocomyidae Dawson, Li, and Qi,
in press

gen. et sp. indet.'^^"

Order Carnivora Bowdich, 1821
gen. et sp. indet.'^*’

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
gen. et sp. indet.*^^"

Order Pantodonta Cope, 1873

Family Coryphodontidae Marsh, 1876
gen. et sp. indet.'^-"

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
cf Palaeosyops sp.*^^'*

Family ?Lophialetidae Radinsky, 1965
gen. et sp. indet. (cf. Breviodon,
V5372y^^"

Family Amynodontidae Scott and Osborn,
1883

Euryodon minimus Xu et af, 1979'^^"
Yu-huang-ding Formation (Late Early Eocene or
Early Middle Eocene)

Order Anagalida Szalay and McKenna, 1971
Family Eurymylidae Matthew, Granger,
and Simpson, 1929
Rhombomylus sp.'--'*

Order Rodentia Bowdich, 1821

Family Cocomyidae Dawson, Li, and Qi,
in press

Advenimus hupeiensis Dawson, Li, and
Qi, in press’™’

Order Pantodonta Cope, 1873

Family Coryphodontidae Marsh, 1876
Asiocoryphodon conicus Xu, ) 976'22>>i2i6)
Asiocoryphodon lophodontus Xu,
1976(221,(216)

Coryphodon flerowi Chow,

1957(2211(24,(2161(1861

Manteodon cf youngi Xu, 1980''^^""’’'’
Order Dinocerata Marsh, 1873

Family Uintatheriidae Flower, 1876
?Gobiatherium sp.””’’

Order Perissodactyla Owen, 1848
Family Flelaletidae Osborn, 1892
cf Heptodon sp.'^^"

VIII. Hu-bei (Hupei) Province

20. Jun-xian Basin (see Section 2-19, Xi-chuan Basin)

21. Yi-chang (Ichang) District

Teilhard de Chardin, P., and Young Chung-chien,

193608”

Xu Yu-xuan, 1980’-'*’

a. Location: Yang-xi located on the south bank of
the Yang-tze River and 20 km SE of the Yidu city
(about 75 km SSE of the Yichang city), Hu-bei
Province.

Coordinates: 30°24'N; 1 1 1°26'E (Yidu city)
Mei-zi-xi: on the north bank of the Yang-tze River,
just opposite to the Yidu city of south bank.

b. Stratigraphic Sequence:

Dong-hu Formation (Late Early Eocene or Early
Middle Eocene)

Mainly red conglomerates, sandstones and
marls.

c. The list of the mammalian fauna:

Dong-hu Formation (Late Early Eocene or Early
Middle Eocene)

Order Pantodonta Cope, 1873
Family Coryphodontidae Marsh, 1876
Manteodon youngi Xu, 1980'-'*’ (con-
taining Eudinoceros cf kholobo-
chiensis of Teilhard de Chardin and
Young"*")

IX. An-hui (Anhui) Province

22. Lai-an Basin

Zhai Ren-jie, Bi Zhi-quo and Yu Zhen-jiang,

1976(240,

a. Location: About 15 km NE of Lai’-an city, Lai’-
an County, An-hui Province.

Coordinates: 32°27'N; ! 18°25'E (Lai’-an city)

b. Stratigraphic Sequence:

Huang-gang Formation (Pliocene)

Greyish yellow basalt.

— Unconformity —

Zhang-shan-ji Formation (?Early Eocene)

Brick-reddish thick sandy and calcareous mud-
stone with coarse sand and hne conglom-
erate (100 m).

— Disconformity—

Shun-shan-ji Formation (Paleocene)

Greyish red silt marls.

c. The list of the mammalian fauna:

Zhang-shan-ji Formation (?Early Eocene)

Order Anagalida Szalay and McKenna, 1 97 1
Family Eurymylidae Matthew, Granger,
and Simpson, 1929

Rhombomylus laianensis Zhai, Bi, and
Yu, 1976'-’"”

38

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

X. Jiang-xi (Kiangsi) Province

23. Chi-jiang (Chihkiang) Basin

a. Location, Dimension, and Stratigraphic Sequence:
See description in the Paleocene of Chi-jiang Basin
(Section 1-2).

b. The list of the mammalian fauna:

Ping-hu Formation (Early Eocene)

Order Dinocerata Marsh, 1873
Family Uintatheriidae Flower, 1876
Phenaceras lacustris Tong, 1979"*^’
[72027]

Ganatheriiim australis Tong, 1979“*^'
[72027]

24. Yuan-shui Basin

Chow Min-chen, 1959*””'’*’’

Chang Yu-ping and Tung Yung-sheng, 1963*’-’
Zheng Jia-jian, Tung Yung-sheng, and Chi Hung-
xiang, 1975*-”’

a. Location: About 30 km east of Xin-yu city, Jiang-
xi Province.

Coordinates: 27°52'N; 1 14°55'E (Xin-yu city)

b. Stratigraphic Sequence:

Lin-jiang Formation (?Oligocene)

Brownish red and black mudstone and shales
with intercalating marls (360^00 m).

— ?Disconformity —

Xin-yu Group (Eocene)

Upper Member: Purplish, brownish and grey-
ish mudstone and sandstone (500 m).
Ning-jia-shan Member (Early Eocene): Pur-
plish grey argillaceous sandstone and sandy
mudstone with intercalating greyish green
fine sandstone and calcareous nodules (600-
900 m).

— ?Disconformity —

Qing-feng-qiao Formation (?Late Cretaceous)
Purplish red thick conglomerate rock with argil-
laceous sandstone (>100 m).

c. The list of the mammalian fauna:

Ning-jia-shan Member (Early Eocene)

Order Carnivora Bowdich, 1821

Family Miacidae Cope, 1880
Xinyuictis tenuis Zheng, Tung, and Chi,
19 7 5(257) [72041]

Order Pantodonta Cope, 1873

Family Coryphodontidae Marsh, 1876
Cory’phodon ninchiashanensis Chow and
Tung, 1965*”’ [72041]

Order Dinocerata Marsh, 1873

Family Uintatheriidae Rower, 1876
?Probathyopsis sinyiiensis Chow and
Tung, 1962*’*” [72041]

Order Perissodactyla Owen, 1848
Family Helaletidae Osborn, 1872
?Heptodon sp.*-”’ [72041]

XL Hu-nan (Hunan) Province

25. Heng-yang Basin

Young Chung-chien, Bien Mei-nian, and Lee Yue-
yen, 1938*^’^’

Young Chung-chien, 1944*228)

Li Chuan-kuei, Chiu Chan-siang, Yan De-fa, and
Hsieh Shu-hua, 1979*"*”

a. Location: About 1 5 km SW of Heng-dong city,
Heng-dong County, Hu-nan Province.
Coordinates: 27°05'N; 1 12°57'E (Heng-dong city)

(Fig. 13)

b. The list of the mammalian fauna:

Ling-cha Formation (Early Eocene)

Order Insectivora Bowdich, 1821
Family indet.
gen. et sp. nov.*"*”

Order Anagalida Szalay and McKenna, 1 97 1
Family Eurymylidae Matthew, Granger,
and Simpson, 1 929

Matutinia nitidulus Li, Chiu, Yan, and
Hsien, 1979*"*” [76004]

Order Rodentia Bowdich, 1821

Family Paramyidae Miller and Gidley,
1918

Cocoinys lingchaensis (Li, Chiu, Yan, and
Hsien, 1979)*"*” [76003]

Order Pantodonta Cope, 1873

Family Coryphodontidae Marsh, 1876
Asiocoryphodon sp.'"*” [76003]

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

39

Fig. 14. — Map of the distribution of Lower Tertiary and the fossil localities of Bo-se Basin.

40

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Order Perissodactyla Owen, 1848
Family Equidae Gray, 1821
Propachynolophus hengyangensis
(Young, 1944)" [76003]

Mammalia, Order indet.

Family Didymoconidae Kretzoi, 1943
Hunanictis inexpectatus Li, Chiu, Yan,
and Hsien, 1979"'°' [76004]

?Paleocene*

Order Pantodonta Cope, 1873
Family Archaeolambdidae Flerov, 1952
Archaeolambda sp."'°' [76005'', 76005]
XII. Guang-xi (Kwangsi) Region
26. Bo-se Basin

Chow Min-chen, 1956*-°’

Tang Xin and Chow Min-chen, 1964"’"

Tang Ying-jun, You YU-zhu, Xu Qin-qi, Qiu Zhu-
ding, and Hu Yan-kun, 1975"'"’’

a. Location: Bo-se, Tian-dong and Tian-yang coun-
ties, Guang-xi Province.

Coordinates: 23°35'-23°50'N; 106°35'-107°10'E
(Eigs. 14, 15)

b. Dimensions: 90 km (NW-SE); 15 km (NE-SW,
at maximum) (800 km’).

c. Stratigraphic Sequence:

Gung-kang Formation (Early Oligocene)

Yellowish and greyish green mudstone

* Archaeolambda sp. (V5345, V5344) was found in the purplish red mudstone which
was underneath Ling-cha Formation and might be assigned to Paleocene.

interbedding with sandy mudstone and
sandstone [73092, 74067, 73075, 73984,
73988, 73989] (1,300-1,459 m).

—Conformity—

Na-duo Formation (Late Late Eocene)

Upper Coal Member: Greyish green mudstone
and sandy mudstone with intercalating coal
and sand-conglomerate rock.

Lower Coal Member: Greyish green silt sand-
stone with intercalating carbonaceous
mudstone, sandy marls and several coal
beds [73072, 73078, 73080, 7308 1 , 73086,
73088, 74067].

— Disconformity —

Dong-jun Formation (Early Late Eocene)

Upper part: Greyish yellow and white calcar-
eous mudstone and greyish red marls;
Lower part: Greyish white limestone marls and
rudaceous limestone; [74064, 74066,
74069] (20-50 m).

— Disconformity—

Liu-niu Eormation (?Eocene)

Brown red sandy mudstone, argillaceous sand-
stone interbedding with conglomerate rock
(60 m).

— U nconformity —

Mesozoic

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

41

d. The list of the mammalian fauna:

Na-duo Formation (Late Late Eocene)

Order Creodonta Cope, 1875
creodont indet.'^^*"’"

Order Condylarthra Cope, 1 88 1
Family Mesonychidae Cope, 1875
Guilestes acares Zheng and Chi, 1978'^^-^’
Guilestes cf. acares Zheng and Chi,

1 9 7 8(253)

cf. Harpagolestes sp.'-^’'

Family Phenacodontidae Cope, 1881
Eodesmatodon spanios Zheng and Chi,

1978«”)

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873

Metatelmatherium cf. browni Colbert,

1938(23)(I71)(66)

Family Deperetellidae Radinsky, 1965
Deperetella sp.i23)(m)d76)

Family Hyracodontidae Cope, 1879
Caenolophus sp."’*"’

Family Amynodontidae Scott and Osborn,
1883

Huananodon hui You, I977i224)(i76)
Paramynodon cf. birmanicus (Pilgrim
and Cotter, 1916)"’*'
Amynodontidae indet.'’^'"’"

Family Rhinocerotidae Owen, 1845
Guixia simplex You, 1977'’’‘""’'” [73081]
Family Eomoropidae Viret, 1958
Eomoropus cf. quadridentatus Zdansky,

1930(23)(171)(235)

Order Artiodactyla Owen, 1848
Family Entelodontidae Lydekker, 1883
Entelodontidae indet."’*’

Family Anthracotheriidae Gill, 1872
Anthracothema rubricae Pilgrim,

19 2 8(23)(154)(I45) [73Q81, 74067]

Anthracokeryx birmanicus Pilgrim and
Cotter, 1916"*'"'“"'’" [73078,
73088]

Anthracokeryx cf. barnbusae Pilgrim,

1928"54)(|45) [73080]

Anthracokeryx cf. birmanicus Pilgrim and
Cotter, 1916'“'"’"

Anthracokeryx sp.

Bothriodon chyelingensis Xu, 1 977121 iki76)
[73086]

Huananothema imparilica Tang,
1978"73)"76i [73086]

Heothema belUa Tang, i978"“i"2<>i
Heothema sp."’""’*' [73078, 73080]
Family Hypertragulidae Cope, 1879
Indomeryx cotteri Pilgrim, 1928"**'"'**'
[73080, 73081, 73086]

Indomeryx youjiangensis Qiu,
1978"**'" 76' [73086]

Indomeryx sp."**'"’*' [73086]
Notomeryx besensis Qiu, 1978"**'"’*'
[73072, 73086]

Family Tragulidae Milne Edwards, 1865
Tragulidae indet.'^*'"’"

Family Choeropotamidae Owen, 1845
Choeropotamidae gen. nov."’*'
Dong-jun Formation (Early Late Eocene)

Order Carnivora Bowdich, 1821
Family Felidae Gray, 1821
?Eusmilus sp.”*' [74064]

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Andrewsarchus crassum Ding, Zheng,

Zhang, and Tong, 1977'’*' [74069]
Order Pantodonta Cope, 1873
Family Coryphodontidae Marsh, 1876
Eudinoceras crassum Tong and Tang,

1977(186)

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
cf. Protitan sp.”*' [74069]

Family Deperetellidae Radinsky, 1965
Diplolophodon cf. similis Zdansky,

1930(’*'(23*'(158' [74066]

Teleolophus sp.'’*' [74066]

Family Amynodontidae Scott and Osborn,
1883

cf. Gigantamynodon sp.'’*' [74066]

Amynodon sp.'’*' [74066]

cf. Paramynodon sp.'’*' [74066]

Family Rhinocerotidae Owen, 1845
Prohyracodon sp.'’*' [74066]

?Ilianodon sp.'’*' [74064]
Eorstercooperia spp.'’*' [74066, 74069]
Order Artiodactyla Owen, 1848
Family Anthracotheriidae Gill, 1870
?Probrachyodus sp.'’*' [74066]

XIII. Yun-nan (Yunnan) Province
27. Lu-nan Basin

Young Chung-chien and Bien Mei-nian, 1939'““'
Chow Min-chen, 1957'’*', 1958'“'

Xu Yu-xuan and Chiu Chan-siang, 1962'*""
Zheng Jia-jian, Tang Ying-jun, Zhai Ren-jie, Ding
Su-yin, and Huang Xue-shi, 1978'***'

a. Location: Lu-nan County, Yun-nan Province.
Coordinates: 24°47'N; 103°16'E (Lu-nan city)

(Fig. 16)

b. Dimensions: 30 km long (N-S), 8 km wide (at
maximum).

e. Stratigraphic Sequence:

Xiao-tun Formation (Early Oligocene)

Brownish red argillaceous sandstone with inter-
calating sandy mudstone and greyish
medium-coarse sandstone (40-50 m).

— ?Conformity—

Lu-mei-yi Formation (Late Eocene)

Brownish red marls with intercalating greyish
white calcareous and sandy mudstone (483-
752 m).

— Unconformity—

Palaeozoic

d. The list of the mammalian fauna:

Lu-mei-yi Formation (Late Eocene)

42

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

LEGEND

Paleozoic

Eocene

Oligocene

^ Fault Line

• 73,053

0 1
I L

Geological Bounsdary

Geological Boundary
(Approx.)

Fossil Sites

2 3Km
J I

Fig. 16. — Map of the fossil localities of Lunan Basin.

( 1 ) Lu-mei-yi — Lu-nan area (?Early Late Eocene)
Order Creodonta Cope, 1875
Creodonta indet.'-'^’*^"^*

Order Carnivora Bowdich, 1821
Family Felidae Gray, 1821
Felidae indet.<“5><^“’'

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Ilonanodon sp.'‘^^“-‘‘’'

Order Tillodontia Marsh, 1875
Tillodontia indet.*-^^"-''^’

Order Perissodactyla Owen, 1848

Family Brontotheriidae Marsh, 1873

Protitan cf robustus Granger and Greg-
ory, 1943(755)(247)(87)

Brontotheriidae gen. et sp. indet.'-^^"-'*’’

Rhinotitan sp.i2'’'i255)(23)

Family Lophialetidae Radinsky, 1965
Breviodon sp.‘^”*

Lophialetes expeditus Matthew and
Granger, 1925«“>(> 26X247)

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

43

?Family Lophialetidae Radinsky, 1965

Rhodopagus pygmaeus Radinsky,

1965«55)(I58)

Rhodopagus sp.'“^'

Family Deperetellidae Radinsky, 1965
Deperetella sp.«3x2i')x247)

Teleolophus sp.'^'’’

Family Helaletidae Osborn, 1892

?HeIaletes mongoliensis Osborn,

Family Hyracodontidae Cope, 1879
Teilhardia pretiosa Matthew and
Granger, 1926'-”’"-’"’''^'

?Teilhardia sp/^ssxix?)

Caenolophus

Caenolophiis sp.'^'‘'“-55)(247>

Family Amynodontidae Scott and Osborn,
1883

Lushiamynodon menchiapuensis Chow
and Xu, 1962«>^'

Amynodon sp. (cf. sinensisy^'’’’'
Amynodon spp.'255)(63K247)

Amynodon hmanensis Chow, Xu, and
Zhen, 1964'‘'^«55x247)

Family Rhinocerotidae Owen, 1845
Prohyracodon sp.«5SM247)

Forstercooperia sp.*^^^"-'*^’

Family Eomoropidae Viret, 1958

Lunania youngi Chow, 1957123x2191

(255)(247M158)

Order Artiodactyla Owen, 1 848
Family Anthracotheriidae Gill, 1872
Anthracotheriidae gen. et sp.

Family Choeropotamidae Owen, 1845
Gobiohyus sp.'-^^"-'’"''

(2) An-yen-cun — Xiao-sha-he area (?Late Late
Eocene)

Order Creodonta Cope, 1875
Family Hyaenodontidae Leidy, 1869
Pterodon dahkoensis Chow, 1975”"
Order Carnivora Bowdich, 1821
Family Canidae Gray, 1821
Canidae indet.'-^'*^

Chailicyon crassidens Chow, 1975”"
Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
Brontotheriidae indet.'^

Rhinotitan quadridens Xu and Chiu,
1 962*^'^’'-“'"

Dianotitan hmanensis (Chow and Hu,

1958)12 19X255X44X49X27)

Brontotheriidae gen. et sp. indet.'^^^"^'’"
Family Lophialetidae Radinsky, 1965
Breviodon sahoensis Chow, Chang, and
Ting, 1974”-"'“5i

Family Deperetellidae Radinsky, 1965
Deperetella dienensis Chow, Chang, and
Ting, 1974”-"

Diplolophodon (Deperetella) cf. similis
Zdansky, 1939”'’«225)

Teleolophus cf. magnus Radinsky,

19 6 5(158X44)9

?Teleolophus sp.'^^^'

Teleolophus medius Matthew and Gran-
ger, 1956C20X2I91

Family Amynodontidae Scott and Osborn,
1883

Amynodon altidens Xu and Chiu,

1962(2(9X247)

Amynodon sp.(2'9i(247)
cf Metamynodon sp.*^' 9x242)
cf Paramynodon sp. (255x247)

Family Rhinocerotidae Owen, 1845
Prohyracodon progressa Chow and Xu,

1 96 1(159X219X61X44X247)

Prohyracodon meridionaleChow and Xu,

1 95 J (6 1X255X44X2 I 9X247)

Prohyracodon cf. orientale Koch,

1 897(‘59I(255X247)

Ilianodon hmanensis Chow and Xu,

J 95 ](6IX219X255)

Forstercooperia shiwopuensis Chow,
Chang, and Ting, 1974'''‘'i
Forstercooperia sp.*'''"

Juxia sp. (2^^x2421
""llndricotherium" sp. *255x247)
Indricotherium parvum Chow,

1 958*22X219)

Indricotherium cf parvum Chow, 1958”“"
Rhinocerotidae indet.*2' 9x247)

Family Eomoropidae Viret, 1958
Eomoropus ulterior Cho’w, 1962'2‘’X2I9X25S)
Eomoropus cf quadridentatus Zdansky,

1 9 3 0(36X235)

Order Artiodactyla Owen, 1848

Family Entelodontidae Lydekker, 1883
Eoentelodon yunnanense Chow,

1958*2(9X255X26)

Family Anthracotheriidae Gill, 1872
Prohrachyodus panchiaoensis Xu and
Chiu, 1962*2(5X219)

Brachyodus hui (Chov^, 1958)*2'3X22)
Anthracotheriidae indet.*2 '2x247)

28. Li-jiang Basin

Zhao Guo-guang, 19 6 5*249'

Zhang Yu-ping, You Yu-zhu, Ji Hong-xiang, and
Ding Su-yin, 19 7 8*242'

a. Location; Li-jiang Na-xi-zu Zi-zhi-xian, Y un-nan
Province.

Coordinates: 26°48'N; 100°16'E (Li-jiang city)

b. Stratigraphic Sequence:

Xiang-shan Formation (Late Eocene)

Purplish red coarse sandstone with intercalat-
ing greyish white calcareous mudstone
(150-200 m).

— Unconformity —

Li-jiang Formation (undetermined age)

c. The list of the mammalian fauna:

Xiang-shan Formation (Late Eocene)

9 The locality is uncertain.

44

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Order Creodonta Cope, 1875
Creodonta indet.'-"*^*'-^*’

Family Hyaenodontidae Leidy, 1869
Hyaenodontidae indet.*’"'’’

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875
Honanodon hebetis Chow, 1965«‘»7«255)(40)
Honanodon

Order Perissodactyla Owen, 1 848
Family Brontotheriidae Marsh, 1873
Brontotheriidae gen. et sp. indet.'^'*^'
Family Lophialetidae Radinsky, 1965
Breviodon sp. nov.'^'*^"^^^’

Schlosseria

Family Lophiodontidae Gill, 1872
?Lophiodon sp.*-‘*”'-^^’

Family Deperetellidae Radinsky, 1965
Deperetella sp.'-"’''”-^^*

Teleolophus sp.'^'*”'^^^’

Family Hyracodontidae Cope, 1879
Caenolophus sp.'’'*’'''^’^’

Family Amynodontidae Scott and Osborn,
1883

Amynodon

Family Rhinocerotidae Owen, 1 845
Prohyracodon sp.*^''’"^’^'

Family Eomoropidae Viret, 1958
Lunaniacf. youngiChow. I957i^3i<^‘»’iw55)
Eomoropus sp.'-‘*^>'^55)

Order Artiodactyla Owen, 1 848

Family Entelodontidae Lydekker, 1883
Eoentelodon sp. nov.<^^^“-''”

Family Anthracotheriidae Gill, 1872
Anthracokeryx sp.'-“^’'^^*'

Anthracothema

Family Hypertragulidae Cope, 1879
Hypertragulidae gen. et sp. indet.'-'*”

Section 3. Oligocene

I. Nei-mong-gol (Inner Mongolia) Region
1 . Shara Murun-Irdin Manha Area

lA. Early Oligocene: Ulan Gochu Formation and
Urtyn Obo Formation
Osborn, H. F., 1929"”'

Radinsky, L. B., 1964"^’’

Chow Min-chen and A. K. Rozhdestvensky,
I960'”’

Qi Tao, 1979"'”’

a. Location: Ulan Gochu Formation: The fossils
which have been published were mainly collected
at East Mesa, Twin Oboes, and Jhama Obo, but
also a few forms were found at Baron Sog Mesa
(Ulan Gochu, now called Ba-yan-obo).
Coordinates: Around 42°30'N; 1 1 1°30'E
Urtyn Obo Formation: Fossils collected from the

locality about 1 5 miles northeast of the East
Mesa.

b. Stratigraphic Sequence: Red clays; maximum 50
ft.

c. The list of the mammalian fauna:

Ulan Gochu Formation and Urtyn Obo Forma-
tion (Early Oligocene)

Order Anagalida Szalay and McKenna, 1 97 1
Family Anagalidae Simpson, 1931
Anagale gobiensis Simpson, I93li'62)(n3)
(Twin Oboes, Jhama Obo)

Order Lagomorpha Brandt, 1855
Family Leporidae Gray, 1921

Gobiolagus andrewsi Burke, 1941"°’
(Twin Oboes, Jhama Obo)
Gobiolagus{l) major Burke, 1941"°’
(Urtyn Obo)

Family Ochotonidae Thomas, 1897
Procaprolagus vetustus (Burke,
1941)(io’(264) (Xwin Oboes, Jhama
Obo)

Order Rodentia Bowdich, 1821

Family Ischyromyidae Alston, 1876
Hulgana ertnia Dawson, 1 968"’°’ (Jhama
Obo)

?Ischyromyidae indet."’°’ (Jhama Obo)
Family Cylindrodontidae Miller and Gid-
ley, 1918

Ardynomys sp."’°’ (Jhama Obo)

Order Condylarthra Cope, 1881
Family Mesonychidae Cope, 1875

Mongolestes hadrodens Szalay and
Gould, 1966"°®’ (Twin Oboes,
Jhama Obo)

cf. Plarpagolestes sp."°®’ (Twin Oboes)
Order Perissodactyla Owen, 1 848
Family Brontotheriidae Marsh, 1873
Embolotherium andrewsi Osborn,
1925<i39’(87’ (gast Mesa)
Embolotherium (?) grangeri Osborn,
1925(i39’(87’ (gast Mesa)
Embolotherium loucksii Osborn,
1929(87)(I39) (Baron Sog Mesa)
Metatitan primus Granger and Gregory,
1943'®°’ (Chimney Butte)

Metatitan progressus Granger and Greg-
ory, 1943'®°’ (Jhama Obo)
Parabrontops gobiensis (Osborn,
1925)'®°’"”’ (Urtyn Obo)
Titanodectes minor Granger and Greg-
ory, 1943'®°’ (Shara Murun)
Titanodectes ingens Granger and Greg-
ory, 1943'®°’ (Jhama Obo)

Family Deperetellidae Radinsky, 1965
Teleolophus magnus Radinsky, 1965"®®’
(Urtyn Obo)

Family Hyracodontidae Cope, 1879
Ardynia praecox Matthew and Granger,
1923(ii6)(i59) (Urtyn Obo)

Family Amynodontidae Scott and Osborn,
1883

Cadurcodon ardynensis Osborn,
1923(215)(i.34, (Urtyn Obo)
Cadurcodon sp.'°'®’

Family Rhinocerotidae Owen, 1845
Urtinotherium incisivum Chow and Chiu,
1964'46’ (Urtyn Obo, Jhama Obo)
IB. Middle-Late Oligocene: Houldjin Formation
Granger, W., and C. P. Berkey, 1922'®°’

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

45

Berkey, C. P., and F. K. Morris, 1927''*’

Chow Min-chen and A. K. Rozhdestvensky,
1960'“'

Qi Tao, 1979"‘""«

a. Location; Houldjin Locality is situated about 20
miles south-southeast of Iren Dabasu.
Coordinates: Around 43°30'N; 1 12°15'E

b. Stratigraphic Sequence: “yellow pebbly gravel
which forms the top of the escarpment” (15 ft).

c. The list of the mammalian fauna:

Houldjin Formation (Middle-Late Oligocene)

Order Pantodonta Cope, 1873
Family Coryphodontidae Marsh, 1876
Hypercoryphodon thomsoni Osborn and
Granger, 1932"''^' (Houldjin)

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
Hyotitan thomsoni Granger and Greg-
ory, 1943'®^’ (Camp Margetts)
Metatitan relictus Granger and Gregory,
194 3(87) (Camp Margetts)
Embolothehum ultimum Granger and
Gregory, 1943'*” (Baron Sog)
Family Deperetellidae Radinsky, 1965
?Deperetella sp."**' AMNH no. 81807
(Camp Margetts)

?Teleolophus sp."“' AMNH no. 81852
(Camp Margetts)

Family Amynodontidae Scott and Osborn,
1883

Cadurcodon sp."‘*'

Family Rhinocerotidae Owen, 1845
Indricotherium grangeri (Osborn,
19 2 3)(132)(84)'85) (Urtyn Obo, Jhama
Obo, Camp Margetts, Nom Khong
Obo)

Caenopus sp."‘” (Houldjin)

Family Chalicotheriidae Gill, 1872
Schizotherium sp.'"” (Nom Khong Obo)
Order Artiodactyla Owen, 1 848

Family Entelodontidae Lydekker, 1883
Entelodon dims Matthew and Granger,
1923'"” (Houldjin)

2. Deng-kou (Saint Jacques) Area

Teilhard de Chardin, P., 1926"’'”

Chow Min-chen and A. K. Rozhdestvensky,
1960'“'

a. Location: East bank of the Y ellow River, opposite
to the city of Deng-kou.

Coordinates: 40°18'N; 107°00'E

b. The list of the mammalian fauna (Middle Oli-
gocene)"

Order Insectivora Bowdich, 1821

Qi (1979) introduced a new name of the Oligocene in Shara Murun area as Nau-
gon-dai Formation. According to Qi’s point of view, the Lower Nau-gon-dai Formation
is overlying Urtyn Obo Formation, with Embolothehum andrewsi and Urtinothenum
mcislvum etc., more than 10 forms of fossils, and represents the Middle Oligocene in
age. The Upper Nau-gon-dai Formation correlates with the Houldjin Formation, con-
taining Indricotherium grangeri and Hypercoryphodon thomsoni. and is of Late Oli-
gocene age.

" During recent years two large collections of Oligocene mammals from different
horizons around the Saint Jacques area were collected by Wang B. Y. and Huang X.
S., respectively.

Family Erinaceidae Bonaparte, 1838
Amphechinus acridens (Matthew and
Granger, 1924)"’^'"“'

Order Lagomorpha Brandt, 1855
Family Leporidae Gray, 1 82 1

Ordolagus teilhardi (Burke,
1 94 1

Family Ochotonidae Thomas, 1897
Bohlinotona pusilla (Teilhard de Char-
din)"^^"'^^"""

Procaprolagus radicidens (Teilhard de
Chardin)""”"""'*"

Sinolagomys cf. major Bohlin,

1937(1311(1771(5)

Order Rodentia Bowdich, 1 82 1

Family Ctenodactylidae Zittel, 1893
Tataromys deflexus Teilhard de Char-
din, 1926"""''“’”

Tataromys plicidens Matthew and
Granger, 1923"""""*'

Karakoromys ?decessus Matthew and
Granger, 1923"""""*'

Family Cylindrodontidae Miller and Gid-
ley, 1918

Tsaganomys altaicus Matthew and
Granger, 1923"""""""“’“’'

Order Creodonta Cope, 1875
Family Hyaenodontidae Leidy, 1 869
?Hyaenodon sp."""'

Order Perissodactyla Owen, 1 848
Family Rhinocerotidae Owen, 1845
Indricotherium grangeri (Osborn,
1923)"""' (=Paraceratherium sp.
indet.)*"’'

Paraceratherium sp. (small form)""""""
?Aceratherium sp."""'

Family Chalicotheriidae Gill, 1872

Schizotherium avitum Matthew and
Granger, 1923""""'""

Order Artiodactyla Owen, 1848
Family Cervidae Gray, 1821
Eumeryx sp."""'

3. Hos Burd (Suhaitu) Basin

Chow Min-chen and A. K. Rozhdestvensky,
1960'“'

Qi Tao, 1975"“""

Jiang Yuan-ji, Wang Bao-liang, and Qi Tao,
1976'“’*'

a. Location; About 120 km NW of Alxa Zuoqi city
and 50 km west of Jartai Yanchi (Gi-lan-tai salt
lake).

Coordinates: Around 39°35'N; 104“’45'E

b. Dimensions: length: 80 km (NE-SW); width; 20
km (NW-SE).

c. Stratigraphic Sequence; Cha-gan-bu-la-ge For-
mation (Early Oligocene)

Upper: Brownish red and green sandy mudstone,
siltpelites with intercalated sandstones, and
gypsum. Main fossil beds [74096-74098],
(Thickness, more than 40 m.)

Middle; Red sandstones and conglomerates (70
m).

46

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Table A. — The list of the mammalian fauna of Hsanda Gol For-
mation (Middle Oligocene), Mongolia (after J. S. Mellett, 1968"^'",
revised).

Order Insectivora Bowdich, 1821

Family Erinaceidae Bonaparte, 1838
Amphechinus rectus (Matthew and Granger,

Amphechinus acridens (Matthew and Granger,

1 924)*'^'’"'^'’'

Tupaiodon morrisi Matthew and Granger, 1924"“*"“’
?Tupaiodon mimitus Matthew and Granger, 1934"“>"“*
Exallerix hsandagolensis McKenna and Holton,

1 957(130)11 14)

Order Lagomorpha Brandt, 1855

Family Ochotonidae Thomas, 1897
Desmatolagus gobiensis Matthew and Granger,

1 923"

Desmatolagus robustus Matthew and Granger,

1 923" I8>I130)(167)

?Ochotonolagus argyropuloi Gureev, 1960"“**-*’'""'’’*

Family Leporidae Gray, 1821

Ordolagus teilhardi (Burke, 1941)<i3D(io)(I30)

Procaprolagus vetustus (Burke, 1941)'“''hio)(130)

Agispelagus simplex Argyropulo, 1940***'’“*"“’"'"’

Order Rodentia Bowdich, 1821

Family Ischyromyidae Alston, 1876
Prosciurus lohiculus Matthew and Granger, 1923"“*"'**"'’^’
Prosciurus arboraptus Shevyreva, l97li‘’‘’n-76)

Family Cylindrodontidae Miller and Gidley, 1918
Pseudocylindrodon mongolicus Kowalski, 1974'**'”
Tsaganomys altaicus Matthew and Granger, 1923"
Cyclomylus lohensis Matthew and Granger, 1923"'”''**’*
Cyclomylus minutus Kowalski, 1974''*'**

Family ?Aplodontidae Trouessart, 1897
Selenomys mimicus Matthew and Granger, 1923""**''*'”

Family Ctenodactylidae Zittel, 1893

Tataromys dejlexus Teilhard de Chardin, 1926"“*''*'”"”'’
Tataromys gobiensis Kowalski, 1974'^'”

Tataromys plicidens Matthew and Granger, 1923'"**'*****
Tataromys sigmodon Matthew and Granger, 1923'"**'*'*'*
Tataromys cf. grangeri Bohlin, 1946'**'”'”

Karakoromys decessus Matthew and Granger, 1923'"**'*’'**
Karakoromys (f=Leptotataromys), cf. Karakoromys sp."“*

Family Zapodidae Coues, 1875
Plesiosminthus tangingoli (Bohlin, 1946)''***’'’'’

Family Cricetidae Rochebrune, 1883
Cricetops dormitor Matthew and Granger, 1923'"**'***'*
Cricetops cf. aeneus Shevyreva, 1965'*’***'’*'”

Eucricetodon asiaticus (Matthew and Granger,

1923)" '“*" ' '*

Family Rhizomyidae Miller and Gidley, 1918
Tachyoryctoides obrutschewt Bohlin, 1937<‘*‘»'5)
Tachyoryctoides pachygnathus Bohlin, 1937"“*'5)

Order Creodonta Cope, 1875

Family Hyaenodontidae Leidy, 1869
Hyaenodon pervagus Matthew and Granger, 1924'"'**
Hyaenodon aymardi Filhol, 1881"“*

Table 4. — Continued.

Hyaenodon ambiguus Martin, 1906"“*

Hyaenodon compressus Filhol, 1881"“’

Order Carnivora Bowdich, 1821
Family Miacidae Cope, 1880
Amphicticeps shackelfordi Matthew and Granger,
1924"“*'"'**

Family Canidae Gray, 1821
?Cynodictis elegans Matthew and Granger, 1924"-*'”'"'**
?Cynodictis constans (Matthew and Granger, 1924)"“*'"***
Amphicynodon teilhardi (Matthew and Granger,
1924)"“*'"”

Family Mustelidae Swainson, 1835
Palaeogale (=“Bunaelurus”) ulysses (Matthew and Grang-
er, 1924)"“*'"'**

Palaeogale parvula (Matthew and Granger, 1924)"“*'"'**
Plesictis Pomel, 1846; cf Plesictis sp."“*

Family Viverridae Gray, 1821
Palaeoprionodon gracilis Matthew and Granger,

1 924"“*" '^*

Family Felidae Gray, 1821
Nimravus Cope, 1879; cf Nimravus sp."-“*

Proailurus Filhol, 1879; cf Proailurus sp."“*

Order Perissodactyla Owen, 1 848
Family Hyracodontidae Cope, 1879

Indricotherium (=Baluchitherium) grangeri (Osborn,

1 923)"^^’"“*

Family Rhinocerotidae Owen, 1845
gen. et sp. indet."“*

Order Artiodactyla Owen, 1 848
Family Cervidae Gray, 1821
Eumeryx culminis Matthew and Granger, 1924"“*
Miomeryx Matthew and Granger, 1925"“’; cf Miomeryx

sp '130)

Family Gelocidae Schlosser, 1886
Pseudomeryx gobiensis T rofimov, 1957"“*

Family Bovidae Gray, 1821
Palaeohypsodontus asiaticus Trofimov, 1958"'*^’

Mammalia, Order indet.

Family Didymoconidae Kretzoi, 1943
Didymoconus (=“Tschelkaria”) colgatei Matthew and
Granger, 1924"“*'"'**

Didymoconus berkeyi Matthew and Granger, 1924"“*'"'**

Lower: Greyish white, light red conglomerates with
intercalated sandy mudstones (30 m).
d. The list of the mammalian fauna:

Cha-gan-bu-la-ge Formation (Early Oligocene)
Order Rodentia Bowdich, 1821
gen. et sp. indet.

Order Lagomorpha Brandt, 1855
gen. et sp. indet."'”’*

Order Condylarthra Cope, 188 1
Family Mesonychidae Cope, 1875
Harpagolestes alxaensis Qi, 1975"'”’*

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

47

Table 5. — The list of the mammalian fauna of Ardyn Obo
(=Ergeleen-Dzo) Formation (Early Oligocene), Mongolia
(according mainly to N. M. Janovskaja, E. N. Kurotchkin. and E.
V. Davjatkin, 7977'^®".

Order Insectivora Bowdich, 1821
Family Erinaceidae Bonaparte, 1838
?Amphechinus sp.

Family Leptictidae Gill, 1872
?Ictopidium sp.

Family Soricidae Gray, 1821
gen. et sp. indet.

Order Lagomorpha Brandt, 1855
Family Ochotonidae Thomas, 1 897
Desmatolagus robustus Matthew and Granger, 1923

Order Rodentia Bowdich, 1821
Family Cylindrodontidae Miller and Gidley, 1918
Ardynomys olseni Matthew and Granger, 1925
Ardynomys vinogradovi Shevyreva, 1972
Morosomys silenti Shevyreva, 1972
Family Cricetidae Rochebrune, 1883
Eucricetodon (Eumys) sp.

Order Creodonta Cope, 1875

Family Hyaenodontidae Leidy, 1869
Hyaenodon eminus Matthew and Granger, 1925
Pterodon mongoliensis (Dashzeveg, 1964)

Order Carnivora Bowdich, 1821
Family Canidae Gray, 1821
?Cynodictis sp.

Order Condylarthra Cope, 1881
Family and genus indet.

Order Perissodactyla Owen, 1848
Family Brontotheriidae Marsh, 1873
Embolotherium andrewsi Osborn, 1929
(Epimanteoceras amplus Janovskaja, 1976)
Epimanteoceras robustus (Granger and Gregory, 1943)
Metatitan relictus Granger and Gregory, 1943
Parabrontops gobiensis (Osborn, 1925)
Protembolotherium efremovi Janovskaja, 1954
Titanodectes ingens Granger and Gregory, 1943

Family Helaletidae Osborn, 1892

Colodon incept us Matthew and Granger, 1925
Colodon sp.

Family Hyracodontidae Cope, 1879
Ardynia praecox Matthew and Granger, 1923
gen. indet.

Family Amynodontidae Scott and Osborn, 1883
Amynodon mongoliensis Osborn, 1 936
Cadurcodon ardynensis (Osborn, 1924)

Caenolophus promissus Matthew and Granger, 1925
Gigantarnynodon cessator Gromova, 1954
Hypsa mynodon progressus Gromova, 1954
gen. indet.

Family Rhinocerotidae Owen, 1 845
Symphysorrhachis breviorostris Beliajeva, 1954

Family Chalicotheriidae Gill, 1872
Schizotherium avitum Matthew and Granger, 1925

Table 5. — Continued.

Order Artiodactyla Owen, 1 848

Family Entelodontidae Lydekker, 1883
Ergilobia gobiensis Trofimov, 1952
Archaeotherium sp.

Family Anthracotheriidae Gill, 1872
Brachyodus sp,
gen. indet.

Family Gelocidae Schlosser, 1886
Lophiomeryx angarae Matthew and Granger, 1925
Lophiomeryx gobiae Matthew and Granger, 1925
Miomeryx altaicus Matthew and Granger, 1925
Miomeryx sp.

Gobiomeryx dubius Trofimov, 1956

Family Cervidae Gray, 1821
Eumeryx sp.

Mammalia, Order indet.

Family Didymoconidae Kretzoi, 1943
Ardynictis furunculus Matthew and Granger, 1925

Order Perissodactyla Owen, 1 848
Family Brontotheriidae Marsh, 1873
Embolotherium grangeri Osborn,

Embolotheriinae indet."'’'’’

Family Amynodontidae Scott and Osborn,
1883

Paracadurcodon suhaituensis Xu,
1966«'5>

Amynodon alxaensis Qi, 1975"'’*’
Sianodon spp.'^'^’

Amynodontidae indet."

Perissodactyla indet."'’^’

Order Artiodactyla Owen, 1 848

Family Entelodontidae Lydekker, 1883
gen. et sp. indet.

?Family Cervidae Gray, 1821
gen. et sp. indet."'*'’’

II. Ning-xia (Ninghsia) Region
4. Ling-wu Basin

Young Chung-chien and Chow Min-chen, 1 956'-^^’
Hu Chang-kang, 1962*'*^’

a. Location: Qing-shui-ying (Ching-shui-ying), 40 km
ENE of the Ling-wu city.

Coordinates: 38°05'N; 106°20'E (Ling-wu city)

b. The list of the mammalian fauna:

Qing-shui-ying Formation' ' (Middle Oligocene)

Order Rodentia Bowdich, 1821

Family Cylindrodontidae Miller and Gid-
ley, 1918

Cyclomylus lohensis Matthew and
Granger, 1923'--’”"'”

Order Perissodactyla Owen, 1848
Family Rhinocerotidae Owen, 1845
Indricotherium grangeri (Osborn,

19 2 3)(23.1«132)

Family Chalicotheriidae Gill, 1 872
Schizotherium sp.''*'”

48

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Order Artiodactyla Owen, 1848

Family Entelodontidae Lydekker, 1883
Archaeotherium ordosius Young and
Chow, 1956'”^'

Family Cervidae Gray, 1821
“ Eumeryx"

5. Others

Two small Oligocene localities of Tong-xin and
Guyuan, yielding Indricotherium only, were
reported by Hu (1962:170). The fossils of these
localities have not yet been published.

III. Xin-jiang (Sinkiang) Region

6. Turpan (Turfan) Basin

Chow Min-chen and Xu yu-xuan, 1959'*°’

Zhai Ren-jie, Zheng Jia-jian, and Tong Yong-
sheng, 1978'^“"’

Zhai Ren-jie, 1978'-^'”

a. Location: East part of the Turpan Basin, along the
railway from Da-bu to Fei-yue.

Coordinates: Around 43°10'N; 91°35'E

(Fig. 6)

b. Stratigraphie Sequence:

Tao-shu-yuan-zi Group: Red and brownish yel-
low sandy clays, sandstones with intercalated
dark grey conglomerates [64080-64082] (700
m).

c. The list of the mammalian fauna:
Tao-shu-yuan-zi Group (Oligocene)

Lower part of the Tao-shu-yuan-zi Group (Early
Oligocene)

Order Perissodactyla Owen, 1848

Family Amynodontidae Scott and Osborn,
1883

Cadurcodon ardynense (Osborn,
1923)'--'’ [66018]

Upper part of the Tao-shu-yuan-zi Group
(Middle-Late Oligoeene)

Order Insectivora Bowdich, 1821
Family Erinaceidae Bonaparte, 1838
Amphechinus cf. rectus (Matthew and
Granger, 1924)'^^ [64081-14]
?Amphechinus sp.'^^” [64081]

Order Lagomorpha Brandt, 1855
Family Ochotonidae Thomas, 1897
Sinolagomys kansuensis Bohlin,
19 3 7(239)15) [64081-12]

Order Rodentia Bowdich, 1821

Family Ctenodactylidae Zittel, 1893
Tataromys cf. sigmodon Matthew and
Granger, 1923'"”«"«’ [64081-12]
Order Creodonta Cope, 1875
Family Hyaenodontidae Leidy, 1869
?Hyaenodon sp.'-°-’ [64082-4]

Order Perissodactyla Owen, 1 848
Family Rhinocerotidae Owen, 1845
Paraceratherium lipidus Xu and Wang,

1978'220)

Paraceratherium tienshanensis Chiu,
1 962"’°’"*’"^°''’

Dzungariotherium turfanensis Xu and
Wang, 1978'^“’

Indricotheriidae indet.'^^°’

?Aceratherium sp.'-^^-’ [64082-2]

Family Chalicotheriidae Gill, 1872
Schizotherium sp.'"”* [64080]

Order Artiodactyla Owen, 1 848
Family Anthracotheriidae Gill, 1872
gen. et sp. indet.*^^'” [64081]

Family Tragulidae Milne Edwards, 1864
?Tragulidae gen. et sp. indet.'^”-’ [64081]
Mammalia, Order indet.

Family Didymoconidae Kretzoi, 1943
Didymoconus berkeyi Matthew and
Granger, 1924'^°'”'"'” [64081]

7. Ha-mi Basin

Some molar teeth of Parabrontops sp. collected
from Ye-ma-quan, Ha-mi area, were described
by Hu Chang-kang in 1961"'". The age of the fossil
is most probably Early Oligocene.

8. Jung-gur (Dzungar) Basin

Only two forms of Oligoeene mammals were
reported from Jung-gur Basin:

1. Dzungariotheriutn orgosensis Chiu, 1973"*’
(Rhinocerotidae, Perissodactyla)

Locality: 20 km south of An-ji-hai bridge,
and east bank of Huo-er-gou-si (Orgos)
River.

Coordinates: Around 44°20'N; 85°25'E
Formation: He-se (Brown) Formation
(Middle-Late Oligocene); brown sands
and gravels.

2. Lophiomeryx sp. (Gelocidae, Artiodactyla),
Chiu, 1965"”

Locality: Hong-shan Farm south of Shi-he-
zi (Shih-ho-tzu) city.

Coordinates: Around 44°15'N; 86°00'E
Formation: He-se (Brown) Formation
(Middle-Late Oligocene).

IV. Gan-su (Kansu) Province

9. Taben buluk Area

Bohlin, B., 1942'*’

Bohlin, B., 1946'”

a. Location:

“Taben-buluk is a group of springs at the north-
ern foot of Anem-braruin-ola on the great caravan
road from Tunhuang to Sirtun .... The badlands
investigated by me extend chiefly towards the east
from this place, between it and the debouchment
of Tang-ho (Tangin-gol) from Nanshan. The area
from which the fossils were obtained does not
extend more than 20 kilometers eastwards from
Taben-buluk” (Bohlin, 1942:7). There are four
large ravines cut through this area, from west to
east: Taben buluk, Yindirte, Tieh-chiang-ku, and
Hsi-shui. The Yindirte is a most important fossil
locality in this area.

Tamu bulak, 60 km SSW of Dunhuang (Tun
huang) city, belongs to the Aksay Kazakzu auton-
omous county, Gansu Province.

Coordinates: Around 39°30'N; 94°35'E

b. Stratigraphic Sequence:

Taben buluk (Late or Latest Oligocene) (about
1,000 m)

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

49

Upp)er: Sandstones with intercalated heavy beds
of very coarse conglomerates.

Middle (main part of deposits): Clayey sedi-
ments with interbedded conglomerates.
Lower: Brick red clay and fine sandstones.

c. The list of the mammalian fauna:

Order Insectivora Bowdich, 1821
Family Erinaceidae Bonaparte, 1838
Amphechinus cf. rectus (Matthew and
Granger, 1924)'*'"“’

Amphechinus kansuensis (Bohlin, 1 942)'*’
Amphechinus minimus (Bohlin, 1942)'*’
?Erinaceidae indet.'*’

Family Soricidae Gray, 1821
Soricidae indet.'*’

Family Talpidae Gray, 1825
?Talpidae indet.'*’

Order Primates Linnaeus, 1758
Primate indet. (T. b. 557)'*’

Order Lagomorpha Brandt, 1855
Family Ochotonidae Thomas, 1897
Desmatolagus sp. (PShargaltensisY'’'
Sinolagomys kansuensis Bohlin, 1937'*"*’
Sinolagomys major Bohlin, 1937'*’'*’
Order Rodentia Bowdich, 1821
Family Sciuridae Gray, 1821
“Sciurus" sp."'’

Family Ctenodactylidae Zittel, 1893
Tataromys grangeri Bohlin, 1946'”
Tataromys sigmodon Matthew and
Granger, 1923'”'"®’

Tataromys cf plicidens Matthew and
Granger, 1923'^’'"*’

Yindirtemys woodi Bohlin, 1946'”
Family Cricetidae Rochebrune, 1883
cf Cricetodon sp.'”
aff. Eumys sp.'”

Family Rhizomyidae Miller and Gidley,
1918

Tachyoryctoides sp.‘”

Family Zapodidae Coues, 1875

Plesiosminthus asiaecentralis (Bohlin,
1946)'”

Plesiosminthus tangingoli (Bohlin,
1946)'”

Plesiosminthus parvulus (Bohlin, 1946)'^’
?Sicistinae sp. 1 and sp. 2'”

Order Carnivora Bowdich, 1821
gen. et sp. indet. 1 and 2'”

Order Perissodactyla Owen, 1848
Family Rhinocerotidae Owen, 1845
gen. et sp. indet. (small form)'”

Family Chalicotheriidae Gill, 1872
?Schizotherium sp.'”

Order Artiodactyla Owen, 1 848
Family Cervidae Gray, 1821
Eumeryx sp.'”

Cervulinae indet.'"’

Family Bovidae Gray, 1821
Bovinae indet.'”

Mammalia, Order indet.

Family Didymoconidae Kretzoi, 1943
?Didymoconus sp."”

There are several forms, such as Sayimys obli-
quidens^ohXin, \946, Kansupithecus, Proboscid-
ea sp., Cervulinae sp., and ?Aceratherium sp., col-
lected from the Taben buluk area, for which “we
must, however, count with the possibility that the
forms derive from Upper Miocene beds” (Bohlin,
1946:250).

10. Shargaltein Gol Area (Late Oligocene)

Bohlin, B., 1937'*’

Bohlin, B., 1942'*’

Bohlin, B., 1946'"’

a. Location:

The Shargaltein Gol (Shara River) is the south
upper course of Dang-he [Tong-he, also as Dun-
huang (Tun-huang) river]. The valley of the river
is situated between two mountains (NW-SE
direction): the north, Yeh-ma range, and the south
Shule nan-shan (Humboldt range). The fossils of
Shargaltein Gol are derived from two areas, the
Shih-chiang-tzu-ku and Wu-tao-ya-yu, lying about
10 km apart. The Shih-chiang-tzu-ku is located
at the north slope of the Humboldt range and near
the Ulan davan (Ulan pass). The Wu-tao-ya-yu
is 10 km northeast of the Shih-chiang-tzu-ku and
situated on the left bank of the Shara River. Boh-
lin (1953'”:9, Fig. 1 ) put the locality ofShih-chiang-
tzu-ku on the south bank of the Iqe-he (Yu-ke
River), Qing-hai Province, in his figure. That is
not correct compared with the description of
locality in the text (Bohlin, 1937:7). The localities
in the Shargaltein valley lie about 1 20 km east of
Taben buluk.

Coordinates: Around 39°N; 96°E

b. The list of the mammalian fauna:

Order Insectivora Bowdich, 1821
Family Erinaceidae Bonaparte, 1838
Amphechinus cf acridens (Matthew and
Granger, 1924)'*’"-'”

?Amphechinus sp.'*’

Erinaceidae, small species'*’

Insectivora indet.'*’

Order Lagomorpha Brandt, 1855
Family Ochotonidae Thomas, 1897
Desmatolagus shargaltensis Bohlin,
1937'*’'*’

?Desmatolagus parvidens Bohlin,
1937'*"*’

Desmatolagus sp.'*"*’

Desmatolagus sp., large form'*"*’
Sinolagomys kansuensis Bohlin, 1937'*"*’
Sinolagomys major Bohlin, 1937'*"*’
Sinolagomys gracilis Bohlin, 1942'*"*’
Order Rodentia Bowdich, 1821
Family Sciuridae Gray, 1821
gen. et sp. indet.'"’

Family Ctenodactylidae Zittel, 1893
Tataromys cf plicidens Matthew and
Granger, 1923'*""®’

Karakoromys cf decessus Matthew and
Granger, 1923'*""®’

50

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Leptotataromys gracilidens Bohlin,

1946(7)

Family Cylindrodontidae Miller and Gid-
ley, 1918

Tsaganomys altaicus Matthew and
Granger, 1923*^“"^'

Family Rhizomyidae Miller and Gidley,
1918

Tachyoryctoides obrutschewi Bohlin,

1937(5)

Tachyoryctoides intermedins Bohlin,

1937(5)

Tachyoryctoides pachygnathus Bohlin,
1937'5>

Family Zapodidae Coues, 1875
Sicistinae indet.'^’

Order Carnivora Bowdich, 1821
gen. et sp. indet.'^’

Order Perissodactyla Owen, 1848
Family Rhinocerotidae Owen, 1845
small rhinocerotid*”

Indricotherium sp.'^’

Order Artiodaetyla Owen, 1 848
Family Cervidae Gray, 1821
?Eumeryx sp.'^’

Cervulinae sp.'^’

Family Bovidae Gray, 1821
small hypselodont bovine'^’

Mammalia, Order indet.

Family Didymoconidae Kretzoi, 1943
Didymoconus sp.'^'

1 1 . Shih-ehr-ma-cheng Locality (?01igocene)

Bohlin, B., 1951«>

a. Location: Shih-ehr-ma-cheng, on the right bank
of the Po-yang-ho. Hui-hui-pu, 20 km ENE of
Yumen city (Lao-jun-maio oil field city); Gan-su
Province.

Coordinates: 39°50'N; 97°44'E (Yumen city)

b. Stratigraphic Sequence: Brick-red sandstone or fine
conglomerates laid down in very thick beds.

c. The list of the mammalian fauna:

Order Anagalida Szalay and McKenna, 1971
Family Anagalidae Matthew, Granger, and
Simpson, 1929

Anagalopsis kansuensis Bohlin, 1951'®*
Family Mimotonidae Li, 1977
Mimolagus rodens Bohlin, 1951'®’

V. Shaan-xi (Shensi) Province

12. Lan-tian District

Chia Lan-po, Chang Yu-ping, Huang Wan-po,
Tang Yin-jun, Chi Hung-siang, You Yii-zhu,
Ting Su-yin, and Huang Xue-shi"^’

Chang Yu-ping, Huang Wan-po, Tang Yin-jun,
Chi Hung-siang, You Yii-zhu, Tong Yung-
sheng. Ting Su-yin, Huang Xue-shi, and
Cheng Chia-chien, 1978'"’

Chow Min-chen, 1979'''-’

Six small localities of Early Oligocene age were
discovered around the Ba-he River from Xi-an
southeast to Lan-tian (coordinates around
34°15'N; 108°55'E). All these localities were
referred to the same formation: Bai-lu-yuan For-

mation, mainly white sandstones, about 400 m
thick. The fossils can be summarized as follows:

1. Tapiroidea indet.''*-’ (Perissodactyla), Yin-
po-cun, Hong-qin-bao, Lin-tong County
[65008],

2. Sianodon bahoensis Xu, 1965'^"*’'-"’ and
Sianodon sp."^"’ (Amynodontidae, Perisso-
dactyla), Mao-xi-cun, Xian city [63704],

3. Sianodon bahoensis Xu, 1965''*^* (Amyno-
dontidae, Perissodactyla), Xin-jie, Lantian
County [63705],

4. Lantianius xiehuensis Chow, 1964'^'”'®'”
(?Dichobunidae, Artiodaetyla), Kang-wan-
gou, Xie-hu, Lantian County [64017],

5. Palaeolaginae indet.'"”’ (Leporidae, Lago-
morpha) and Artiodaetyla indet. Gao-po,
Lantian County [64005],

6. Amynodon sp.*''^’ (Amynodontidae, Peris-
sodactyla) and Brontotheriidae indet. ''*^’
(Perissodactyla), Gao-wan-gou, Wei-nan
County [64017],

VI. Shan-xi (Shansi) Province

13. Yuan-chu Basin

Only a fragmentary lower jaw of Brachyodus
hui (Chow, 1958) was found from the Early Oli-
gocene of Bai-shui-cun, Yuan-chu County (see also
in Section 2- 1 3).

VII. Guang-xi (Kwangsi) Region

14. Bo-se Basin

a. Location, Dimensions, and Stratigraphic Se-
quence: See the description of the Eocene of Bo-
se Basin, Guang-xi Province (Section 2-26).

b. The list of the mammalian fauna:

Gung-kang Formation (Early Oligoeene)

Order Carnivora Bowdich, 1821
Family Canidae Gray, 1821
Pachycynodon sp.'"^’

Order Perissodactyla Owen, 1848
Family Rhinocerotidae Owen, 1845
?Forstercooperia sp.'"*’

Order Artiodaetyla Owen, 1 848
Family Anthracotheriidae Gill, 1870
Anthracokeryx gungkangensis Qiu,
1977'"“’ [73079, 74067]
Anthracokeryx kwangsiensis Qiu,
1977'"“’ [73075]

Anthracokeryx spp.'"“’ [73089]
Heothema media Tang, 1978'"®’ [73088]
Heothema angusticalxia Tang, 1978'"®’
[73084]

15. Yong-le Basin

Tang Yin-jun, You Yii-zhu, Xu Qin-qi, Qiu Zhu-
ding, and Hu Yan-kun, 1974"®*’

a. Location: About 20 km NW of Bo-se city, Bo-se
County, Guang-xi Province.

Coordinates: 23°54'N; 106°37'E (Bo-se city)

b. Stratigraphic Sequence:

Gung-kang Formation (Early Oligocene)

Greyish yellow mudstone interbedding with
yello'wish brown sandstone and a few coarse
sandstones [773088, 73091, 74072A] (650
m).

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

51

c. The list of the mammalian fauna:

Gung-kang Formation (Early Oligocene)

Order Perissodactyla Owen, ! 848
Family Chalicotheriidae Gill, 1872

Schizotherium nabanensis Zhang,
1976««>

Schizotherium sp.

Family Amynodontidae Scott and Osborn,

1883

Huananodon hypsodonta You, 1977'^-“"
Family Rhinocerotidae Owen, 1845
Guixia youjiangensis You, 1977<^-'*>
[73091]

Order Artiodactyla Owen, 1 848

Family Anthracotheriidae Gill, 1872
Bothriodon tientongensis Xu, 1977'-'"
[773088]

Anthracokeryx spp."^'*’ [73091]
Heothema chengbiensis Tang, 1978"^"’
[74072A, 73091]

VIII. Gui-zhou (Kweichow) Province

In 1979 some excellent mammalian fossils,
mainly perissodactyls, were collected from the
western part (near Shui-cheng-Pan Xian) of Gui-
zhou Province. The materials have not been pub-
lished yet, so it is hard to determine the age of
the fauna, which may be Oligocene"’'", or even
Eocene.

IX. Yun-nan Province

16. Qu-jing (Chu-ching) Basin

Young Chung-chien and Bien Mei-nien, 1939'-'"’
Chow Min-chen, 1957'^"'^^’

Xu Yu-xuan, 1961'^""

Zhang Yu-ping, You Yu-zhu, Ji Hong-xiang, and
Ding Su-yin, 1978'^""

a. Location: Qu-jing County, Yun-nan Province.
Coordinates: 25°36'N; 103°49'E (Qu-jing city)

b. Stratigraphic Sequence:

Cai-jia-chong Formation (Early Oligocene)

Greyish white and greyish green marls with
intercalating thin greyish green mudstone
(480 m).

c. The list of the mammalian fauna:

Cai-jia-chong Formation (Early Oligocene)

Order Perissodactyla Owen, 1 848
Family Brontotheriidae Marsh, 1873
Brontotheriidae gen. et sp.
indet.'^

Family Hyracodontidae Cope, 1879
Caenolophus sp.""^"”'"'"'*"

Family Amynodontidae Scott and Osborn,
1883

Cadurcodon ardynensis (Osborn,

1923)'2I2"174)

Cadurcodon sp.""^""'"'’
Gigantamynodon giganteus Xu,

195 1(212,(247,

Gigantamynodon cf. giganteus Xu,

1961'I74,'247,

Gigantamynodon sp."^'"'"'"’
cf. Metamynodon sp.< "'“><2 12x247,

Family Rhinocerotidae Owen, 1845
Indricotherium qujingensis Tang,

1978(174,

Indricotherium sp.'2' 2x247,

Order Artiodactyla Owen, 1848

Family Anthracotheriidae Gill, 1872
Bothriodon chowi Xu, 1961'^""'-'*"
?Anthracotheriidae indet."^""

Family Hypertragulidae Cope, 1879
cf. Miomeryx sp.<2' 2x247,

17. Lu-nan Basin

a. Location, Dimensions, and Stratigraphic Se-
quence: See the description of the Eocene of Lu-
nan Basin (Section 2-27).

b. The list of the mammalian fauna:

Xiao-tun Formation (Early Oligocene)

Order Perissodactyla Owen, 1848
Family Hyracodontidae Cope, 1879
Hyracodontidae gen. et sp. indet.'2”’
Family Amynodontidae Scott and Osborn,
1883

cf. Gigantamynodon giganteus Xu,

1 95 1(212,(255,(247)

Order Artiodactyla Owen, 1848

Family Anthracotheriidae Gill, 1872
Bothriodon sp. '27x247,

Artiodactyla indet.'2^^’

18. Luo-ping Basin

Chow Min-chen and Xu Yu-xuan, 1959"’“'

Chiu Chan-siang, 1962'""

a. Location: Luo-ping and Shi-zong counties, Yun-
nan Province.

Coordinates: 24°58'N; 104°20'E (Luo-ping city)-
24°51'N; 103°59'E (Shi-zong city).

b. Stratigraphic Sequence: Clay and lignite beds.

c. The list of the mammalian fauna:

Order Perissodactyla Owen, 1 848
Family Rhinocerotidae Owen, 1845
Indricotherium sp.'^°’

Indricotherium intermedium Chiu,
1962"w

Indricotheriinae indet.'""

52

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

CHAPTER 2. -THE SYSTEMATIC AND STRATIGRAPHIC DISTRIBUTION TABLE

OF THE CHINESE PALEOGENE MAMMALS

Forms

Paleocene

Eocene

Oligocene

E.-M. L. E. M. L. E. M. L.

Geographical

distribution

Multituberculata
Taeniolabidae
Prionessus lucifer M. et G.

Sphenopsalis nobilis M. G. et S.

Lambdopsalidae
Lambdopsalis bulla C. et Q.

Multituberculata indet.

Edentata

Emanodontidae
Ernanodon antelios D.

Megalonychoidea
Chungchienia sichuanica C.

Insectivora
Deltatheridiidae
Sarcodon pygmaeus M. et G.

Leptictidae
Ictopidium lechei Z.

Erinaceidae

Amphechinus acridens (M. et G.)

Amphechinus cf. acridens (M. et G.)

Amphechinus kansuensis (B.)

Amphechinus minimus (B.)

Amphechinus cf. rectus (M. et G.)

?Amphechinus sp.

?Erinaceidae indet.

Erinaceidae, small species
?Tupaiodon sp.

Soricidae indet.

?Talpidae indet.

Family indet.

Hyracolestes ermineus M. et G.

Insectivora gen. et sp. indet.

Insectivora gen. et sp. nov.

Insectivora gen. et sp. nov. [""Sinosinopa
sinensis Q.”]

?Pantolestes sp.

Primates

Adapidae

Petrolemur brevirostre T.

Anaptomorphidae
Lushius qinlinensis C.

Omomyidae
Hoanghonius stehlini Z.

Primates indet.

Anagalida
Anagalidae
Anagale gobiensis S.

Anagalopsis kansuensis B.

Anaptogale wanghoensis X. +

+ + 1-9,'^ 2-1

-1- 1-9

+ 1-9

+ 1-8

-t- 1-1

+ 2-19

4- 1-9

+ 2-13

4- 3-1, 3-2

3-10
4- 3-10

4- 3-10

4- 4- 3-6, 3-9

4- 3-6

-H 3-9

4- 3-10

4- 2-8

4- 3-9

4- 3-9

4- 1-4

4- 4- 1-2, 3-10

2-25

4- 2-2

2- 3A

4- 1-1

4- 2-17

-h 2-13

4- 3-9

3- 1

4-? 3-11

1-4

The numbers (for example. 1-9) represent, first, that of the Section (or Epoch) and second, that of the basin or area in Chapter 1.

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

53

CHAPTER 2.- Continued.

Forms

Paleocene
E.-M. L.

Eocene
E. M,

L.

Oligocene

E. M L.

Geographical

distribution

Chianshania jianghuaiensis X.

+

1-4

Diacronus anhuiensis X.

+

1-4

Diacronus wanghuensis X.

+

1-4

Hsiuannania maguensis X.

+

1-5

Hsiuannania minor D. et Z.

+

1-2

Hsiuannania tabiensis X.

+

1-4

Hsiuannania sp.

+

1-4

Huaiyangale chianshanensis X.

+

1-4

cf. Huaiyangale leura D. et T.

+

1-1

Huaiyangale sp.

+

1-4

Linnania lofoensis C. et al.

+

1-1

Stenanagale xiangensis W.

+

1-3

Wanogale hodungensis X.

+

1-4

Pseudictopidae

Allictops inserrata Q.

+

1-4

Anictops tabiepedis Q.

+

1-4

Anictops aff. tabiepedis Q.

+

1-4

Cartictops canina D. et T.

+

1-4

Haltictops meilingensis D. et T.

+

1-1

Haltictops mirabilis D. et T.

+

1-1

Paranictops majuscula Q.

+

1-4

Paranictops sp.

+

1-4

Pseudictops chaii T.

+

1-8

cf. Pseudictops tenuis D. et Z.

+

1-2

Pseudictops lophiodon M., G. et S.

+

+

1-9, 2-1

Pseudictopidae gen. et sp. indet.

+

1-6

Eurymylidae

Heomys orientalis L.

+

1-4

Heomys sp.

+

1-4

Hypsimylus beijingensis Z.

+

2-8

Matutinia nitidulus L. et al.

+

2-25

Rhombomylus laianensis Z. et al.

+

2-22

Rhombomylus turpanensis Z.

+

2-5

Rhombomylus sp.

+

2-20

Eurymylidae gen. et sp. indet.

+

1-8

Eurymyloidea indet.

+

1-4

Mimotonidae

Mimolagus rodens B.

+?

3-1 1

Mimotona borealis C. et Q.

+

1-9

Mimotona robusta L.

+

1-4

Mimotona wana L.

+

+

1-4

Mimotona sp.

+

1-4

Zalambdalestidae

Anchilestes impolitus Q. et L.

+

1-4

Lagomorpha

Leporidae

Gobiolagus andrewsi B.

+

3-1

Gobiolagus (?) major B.

+

3-1

Gobiolagus tohnaclwvi B.

+

2-4

Lushilagus lohoensis L.

+

2-17

Ordolagus teilhardi (B.)

+

3-2

Shamolagus granger! B.

+

2-3C

Shamolagus medius B.

+

2-4

Palaeolaginae indet.

+

3-12

54

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

CHAPTER 2. — Continued.

Paleocene

Eocene

Oligocene

Geographical

distnbution

Forms

E.-M. L.

E.

M.

L.

E.

M. L.

Ochotonidae

Bohlinotona pusilla (T.)

+

3-2

?Desmatolagus pamdens B.

+

3-10

Desmatolagus shargaltensis B.

+

3-10

Desmatolagus sp. (?shargaltensis)

+

3-9

Desmatolagus spp.

+

3-10

Procaprolagus radicidens (T.)

+

3-2

Procaprolagus vetustus (B.)

+

3-1

Sinolagomys gracilis B.

+

3-10

Sinolagomys kansuensis B.

+

+

3-6, 3-9, 3-10

Sinolagomys major B.

+

3-9, 3-10

Sinolagomys cf. major B.

+

3-2

Lagomorpha indet.

+

3-3

odentia

Ischyromyidae

Hulgana ertnia D.

+

3-1

paramyid spp.

+

2-3A

paramyid gen. et sp. nov.

medius Q.”]

+

2-2

?Ischyromyidae indet.

+

+

2-13, 3-1

Cylindrodontidae

Ardynomys sp.

+

3-1

Cyclomylus lohensis M. et G.

+

3-4

Tsaganomys altaicus M. et G.

+

+

3-2, 3-10

Sciuravidae indet.

+

2-17

Cocomyidae

Advenimus bohlini D.

+

2-3C

Advenimus burkei D.

+

2-3B

Advenimus hupeiensis D. et al.

+

2-20

cf. Advenimus sp.

+

2-3C

Cocomys lingchaensis (L. et al.)

+

2-25

Cocomyidae indet. {"Sciuravus sp.”)

+

2-19

Tamquammys wilsoni D. L. et Q.

+

2-2

Tsinlingomys youngi L.

+

2-17

Cocomyidae gen. et sp. indet.

+

2-19

Yuomyidae

Yuomys cavioides L.

+

2-13, 2-15, 2-4

Yuomys eleganes W.

+

2-18

Ctenodactylidae

Karakoromys? decessus M. et G.

+

3-2, 3-10

Leptotataromys gracilidens B.

+

3-10

Tataromys dejlexus T.

+

3-2

Tataromys grangeri B.

+

3-9

Tataromys plicidens M. et G.

+

3-2

Tataromys cf plicidens M. et G.

+

3-9, 3-10

Tataromys sigmodon M. et G.

+

3-9

Tataromys cf sigmodon M. et G.

+

3-6

Yindirtemys woodi B.

+

3-10

Sciuridae

""Sciurus” sp.

+

3-9

gen. et sp. indet.

+

3-10

Cricetidae

Cricetodon schaubi Z.

+

2-13

cf Cricetodon sp.

+

3-9

aff. Eumys sp.

+

3-9

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

55

CHAPTER 2. -Continued.

Paleocene

Eocene

Oligocene

Geographical

distribution

Forms

E.-M.

L.

E. M

L.

E.

M.

L.

Zapodidae

Piesiosminthus asiaecentralis (B.)

+

3-9

Plesiosminthus parvulus (B.)

+

3-9

Piesiosminthus tangingoli (B.)

+

3-9

Plesiosminthus sp.

+

2-13

?Sicistinae spp.

+

3-9

Sicistinae indet.

+

3-10

Rhizomyidae

Tachyoryctoides intermedius B.

+

3-10

T achyoryctoides obrutschewi B.

+

3-10

Tachyoryctoides pachygnathus B.

+

3-10

T achyoryctoides sp.

+

3-9

?Rodentia indet.

+

1-9

Rodentia indet.

+

2-12

Rodentia indet.

+

3-3

Creodonta

Hyaenodontidae

Hyaenodon yuanchuensis Y.

+

2-13

Hyaenodon sp.

+

2-18

?Hyaenodon sp.

+

3-2, 3-6

Propterodon irdinensis M. et G.

+

2-3A, 2-17

Paracynohyaenodon morrisi M. et G.

+

2-3A

Pterodon dahkoensis C.

+

2-27

Pterodon hyaenoides M. et G.

+

2-4

T'Sinopa” sp.

+

2-19

?Thinocyon sichowensis C.

+

2-12

?“ Tritemnodon” sp.

+

2-19

Hyaenodontidae indet.

+

2-28

Oxyaenidae

Prolaena parva X. et al.

+

2-19

Sarkastodon mongoliensis G.

+

2-3A

Creodonta indet.

+

2-26, 2-27, 2-28

Carnivora

Miacidae

Miacis invictus M. et G.

+

2-3A

Miacis lushiensis C.

+

2-17, 2-19

Miacis sp.

+

2-8

Pappictidops acies W.

+

1-1

Pappictidops obtusus W.

+

1-1

Pappictidops orientalis Q. et L.

+

1-4

Xinyuictis tenuis Z. et al.

+

2-24

Canidae

Chailicyon crassidens C.

+

2-13, 2-27

Cynodictis sp.

+

2-17

Pachycynodon sp.

+

3-14

Canidae gen. et sp. indet.

+

2-8, 2-27

Felidae

cf. Eusmilus sp.

+

2-17

?Eusmilus sp.

+

2-26

Felidae indet.

+

2-6, 2-27

Carnivora gen. et sp. indet.

+

2-2

Carnivora gen. et sp. indet.

+

2-18, 2-19

Carnivora gen. et spp. indet.

+

3-9, 3-10

Condylarthra

Arctocyonidae

Paratriisodon gigas C. et al.

+

2-3B, 2-17

Paratriisodon henanensis C.

+

2-17

56

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

CHAPTER 2. — Continued.

Paleocene

Eocene

Oligocene

Geographical

distribution

Forms

E.-M. L.

E.

M.

L.

E. M. L.

Hyopsodontidae

Decoredon elongetus X.

+

1-4

?Haplomylus sp.

+

2-12

Hyopsodus sp.

+

2-5

Palasiodon siurenensis C. et al.

+

1-1

Yuodon protoselenoides C. et al.

+

1-1

Hyopsodontidae indet.

+

1-2

Mesonychidae

Andrewsarchus crassum D. et al.

+

2-26

Andrewsarchus mongoliensis O.

+

2-3A

?Andrewsarchus sp.

+

2-19

Dissacus magushanensis Y. et T.

+

1-5

?Dissacus rotundus W.

+

1-3

?Dissacus sp.

+

1-9

Dissacusium shanghoensis C. et al.

+

1-1

Guilestes acares Z. et C.

+

2-26

Guilestes cf. acares Z. et C.

+

2-26

Hapalodectes serus M. et G.

+

2-3A

?Hapalodectes sp.

+

1-2

Harpagolestes alxaensis Q.

+

3-3

?Harpagolestes orientalis S. et G.

+

2-3C

cf. Harpagolestes sp.

+

+

2-26, 3-1

Honanodon hebetis C.

2-13, 2-28

Honanodon macrodontus C.

+

2-17

Honanodon sp.

+

2-27, 2-28

Hukoutherium ambigum C. et al.

+

1-1

Jiangxia chaotoensis Z. et al.

+

1-2

Lohoodon lushiensis (C.)

+

2-17

Mesonyx sp.

+

2-3A

cf Mesonyx sp.
?Mesonyx sp. nov. [“M.

+

2-3C

obtusidens Q.”]

+

2-2

Mongolestes hadrodens S. et G.

+

3-1

Mongolonyx dolichognathus S. et G.
Mongolonyx sp. nov. [“M.

+

2-3B

prominentis Q.”]

+

2-2

Pachyaena sp.

+

+

2-1, 2-3 A

?Pachyaena sp.

+

1-1

Plagiocristodon serratus C. et Q.

+

+

1-9, 2-1

Yantanglestes conenxus (Y. et T.)

+

1-4

Yantanglestes datangensis (W.)

+

1-1

Yantanglestes feiganensis (C. et al.)

+

1-1

Mesonychidae gen. et spp. indet.

+ +

1-6, 1-7, 1-8

Mesonychidae gen. et sp. indet.

+

2-19

Periptychidae

?Ectoconus sp.

+

1-1

Pseudanisonchus antelios Z. et al.

+

1-2

Phenacodontidae

Eodesmatodon spanios Z. et C.

+

2-26

Notoungulata

Arctostylopidae

Allostylops periconotus Z.

+

1-2

Anatolostylops dubius Z.

+

2-5

Arctostylopidae gen. et sp. nov.

+

1-1

Asiostylops spanios Z.

+

1-2

Palaeostylops iturus M. et G.

+

+

1-9, 2-1

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

57

CHAPTER 2. — Continued.

Paleocene

Eocene

Oligocene

Geographical

distnbution

Forms

E.-M.

L.

E.

M.

L.

E. M. L.

Palaeostylops macrodon M., G. et S.

+

+

1-9, 2-1

Sinostylops progressus T. et Y.

+

1-5

Sinostylops promissus T. et Y.

+

1-4

Taeniodonta

Stylinodontidae

?Stylinodon sp.

+

2-17

Tillodontia

Esthonychidae

Lofochaius brachyodus C. et al.

+

1-1

Meiostylodon zaoshiensis W.

+

1-3

Kuanchuanius shantunensis C.

+

2-12

Adapidium huanghoense Y.

+

2-13

Family indet.

Dysnoetodon minuta Z.

+

1-1

Tillodontia gen. et sp. indet.
Tillodontia gen. et sp. nov.

+

2-13, 2-27

[''Ulanius chowi Q.”]

+

2-2

Pantodonta

Archaeolambidae

Archaeolambda dayuensis T.

+

1-2

Archaeolambda cf. planicanina F.

+

1-2, 1-8

Archaeolambda tabiensis H.

+

1-4

Archaeolambda yangtzeensis H.

+

1-5

Archaeolambda sp.

+

+

1-2, 2-25

Nanlingilambda chijiangensis T.

+

1-2

Archaeolambdidae indet.

+

1-2

Archaeolambdidae gen. et spp. nov.

+

1-1

Archaeolambdidae indet.

+

2-14

Bemalambdidae

Bemalambda crassa C. et al.

+

1-1

Bemalambda nanhsiungensis C. et al.

+

1-1, 1-3

Bemalambda pachyoesteus C. et al.

+

1-1

Bemalambda shizikouensis W. et D.

+

1-2

Bemalambda spp.

+

1-1, 1-4

Hypsilolambda chalingensis W.

+

1-3

Hypsilolambda impensa W.

+

1-3

Hypsilolambda spp.

+

1-3

Bemalambdidae indet.

+

1-3, 1-4, 1-6, 1-7

Coryphodontidae

Asiocoryphodon conicus X.

+

2-19

Asiocoryphodon lophodontus X.

+

2-19

Asiocoryphodon sp.

+

2-25

Coryphodon dabuensis Z.

+

2-5

Coryphodon flerowi C.

+

+

2-19, 2-12

Coryphodon ninchiashanensis C. et T.

+

2-24

Coryphodon sp.

+

2-5

Eudinoceras crassum T. et T.

+

2-26

Eudinoceras kholobolchiensis O. et G.

+

2-3C

Eudinoceras mongoliensis O.

+

2-3A

Eudinoceras sp.

+

2-6, 2-8, 2-17

Hypercoryphodon thomsoni O. et G.

+

3-1

Manteodon youngi X.

+

2-21

Manteodon cf. youngi X.
Corphodontidae gen. et sp. nov.

+

2-19

Metacoryphodon luminis Q.”]

+

2-2

58

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 21

CHAPTER 2. -Continued.

Paleocene

Eocene

Oligocene

Geographical

distribution

Forms

E.-M. L.

E

M.

L.

E. M. L.

Coryphodontidae gen. et sp. nov.

["?Metacoryphodon minor Q.”]

+

2-2

Coryphodontidae gen. et sp. indet.

+

2-19

Harpyodidae

Harpyodus decorus W.

+

1-2

Harpyodus euros Q. et L.

+

1-4

Pantolambdodontidae

Dilambda speciosa T.

+

1-8

Pantolarnbdodon fortis G. et G.

+

2-3C

Pantolambdodon inermis G. et G.

+

2-3C

?Pantolambdodon sp. nov. [“P. minor” Q.]

+

2-2

Pastoraiodontidae

Altilambda pactus C. et W.

+ +

1-1, 1-4

Altilambda tenuis C. et W.

+

1-4

Convallisodon convexus C. et Q.

+

1-9

Convallisodon haliutensis C. et Q.

+

1-9

Pastoralodon lacustris C. et Q.

+

1-9

?Pastoralodon lacustris C. et Q.

+

2-1

Pastoraiodontidae indet.

+

1-6

Phenacolophidae

Ganolophus lanikenensis Z.

+

1-2

Minchenella ("Conolophus”) grandis Z.

+

1-1

Tienshanilophus lianmuqinensis T.

+

1-8

Tienshanilophus shengjinkouensis T.

+

1-8

Tienshanilophus subashiensis T.

+

1-8

Yuelophus validus Z.

+

1-1

Phenacolophidae gen. et sp. nov.

+

1-1

Dinocerata

Uintatheriidae

Ganatherium australis T.

+

2-23

Gobiatherium mirificum O. et G.

+

+

2-2, 2-3B

Gobiatherium sp. nov. [“G. major Q.”]

+

2-2

Gobiatherium sp. nov. [“G.

monolabotum Q.”]

+

2-2

?Gobiatherium sp.

+

2-19

Houyanotherium primigenum T.

+

1-8

Houyanotherium simplum T.

+

1-8

Jiaoluotherium turfanense (C.)

+

1-8

Mongolotherium efremovi F.

+

2-1

Phenaceras lacustris T.

+

2-23

?Probathyopsis sinyuensis C. et T.

+

2-24

Prodinoceras diconicus T.

+

1-8

Pyrodon xinjiangensis Z.

+

2-5

Pyrodon sp.

+

2-1

cf. Uintatherium sp.

+

2-12

Prodinoceratinae indet.

+

2-14

Uintatheriinae indet.

+

2-17

Perissodactyla

Brontotheriidae

Arctotitan honghoensis W.

+

2-7

Desmatotitan tukhumensis G. et G.

+

2-3C

Desmatotitan sp.

+

2-2

Dianotitan lunanensis (C. et H.)

+

2-27

Dolichorhinoides angustidens G. et G.

+

2-3C

Embolotherium andrewsi O.

+

3-1

Embolotherium grangeri O.

+

3-3

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

59

CHAPTER 2.-Contimied.

Forms

Paleocene

Eocene

Oligoccne

Geographical

(distribution

E.-M.

L.

E.

M.

L.

E.

M. L.

Embolotherium (?) grangeri O.

+

3-1

Embolotherium loucksii O.

+

3-1

Embolotherium ultimum G. et G.

+

3-1

Epimanteoceras formosus G. et G.

+

2-3C

Gnathotitan berkeyi (O.)

+

2-3A

Hyotitan thomsoni G. et G.

+

3-1

?Larnbdotherium sp.

+

2-1

Metatelmatherium cf. browni C.

+

2-26

Metatelmatherium cristatum G. et G.

+

2-3B

Metatelmatherium parvum G. et G.

+

2-3A

Metatitan primus G. et G.

+

3-1

Metatitan progressus G. et G.

+

3-1

Metatitan relictus G. et G.

+

3-1

Microtitan mongoliensis (O.)

+

2-3A, 2-3C

Microtitan sp. nov. elongatus Q.”]

+

2-2

?Microtitan sp. (or new genus)

+

2-17

Pachytitan ajax G. et G.

+

2-4

cf. Palaeosyops sp.

+

2-19

Parabrontops gobiensis (O.)

+

3-1

Parabrontops sp.

+

3-7

Protitan bellus G. et G.

+

2-3C

?Protitan cingulatus G. et G.

+

2-3B

Protitan grangeri (O.)

+

2-3A, 2-17

Protitan minor G. et G.

+

2-3B

Protitan obliquidens G. et G.

+

2-3A

Protitan robustus G. et G.

+

2-3A

Protitan cf robustus G. et G.

+

2-27

?Protitan sp.

+

2-19

cf Protitan sp.

+

2-26

Rhinotitan andrewsi (O.)

+

2-4

Rhinotitan kaiseni (O.)

+

2-4

Rhinotitan mongoliensis (O.)

+

2-4, 2-13

Rhinotitan quadridens X. et C.

+

2-27

Rhinotitan sp.

+

2-27

?Rhinotitan sp.

+

2-5

Titanodectes ingens G. et G.

+

3-1

Titanodectes minor G. et G.

+

+

2-4, 3-1

Brontotheriidae indet.

+

2-16, 2-27, 2-28

Brontotheriidae indet.

+

3-12, 3-16

Embolotheriinae indet.

+

3-3

Palaeotheriidae

Propalaeotherium sinensis Z.

+

2-12

Equidae

Heptaconodon dubium Z.

+

2-12

Propachynoloph us hengyangensis (Y.)

+

2-25

Hyracotheriinae indet.

+

2-12

Isectolophidae

Homogalax wutuensis C. et L.

+

2-10

Lophialetidae

Breviodon acares R.

+

2-3C

cf Breviodon acares R.

+

2-3C

Breviodon minutus (M. et G.)

+

2-17

Breviodon cf minutus (M. et G.)

+

2-19

Breviodon ?minutus (M. et G.)

+

+

2-2, 2-3A

Breviodon sahoensis C. et af

+

2-27

Breviodon sp. nov.

+

2-28

60

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

CHAPTER 2. -Continued.

Paleocene

Eocene

Oligocene

Geographical

distribution

Forms

E.-M. L.

E.

M.

L.

E.

M. L.

Breviodon sp.

+

2-3B, 2-7,

2-27

?Breviodon sp.

+

2-3C, 2-18

Lophialetes expeditus M. et G.

+

+

2-2, 2-3A,B,C,

2-5, 2-19,
2-27

Lophialetes cf. expeditus M. et G.

+

2-6

Lophialetes sp.

+

2-3C, 2-19

?Lophialetes sp.

+

2-3A

Rhodopagus ?minimus (M. et G.)

+

2-4

Rhodopagus pygmaeus R.

+

2-3C, 2-27

?Rhodopagus pygmaeus R.

+

2-3A

Rhodopagus sp.

+

+

2-12, 2-13, 2-27

Schlosseria magister M. et G.

+

2-2

cf. Schlosseria magister M. et G.

+

2-3B

Schlosseria sp. nov. [“S. dimera Q.”]

+

2-2

Schlosseria sp. nov. [“S', masculus Q.”]

+

2-2

Schlosseria sp.

+

2-28

Simplaletes sujiensis Q.

+

2-3A

Simplaletes ulanshierhensis Q.

+

2-3C

Lophialetidae indet.

+

2-5, 2-28, 2-19

Lophiodontidae

?Lophiodon sp.

+

2-28

Deperetellidae

Deperetella cristata M. et G.

+

2-4

Deperetella depereti (Z.)

+

2-13

Deperetella dienensis C. et al.

+

2-27

Deperetella sp. nov.

+

2-18

Deperetella sp.

+

2-17, 2-18, 2-26,

2-27, 2-28

cf Deperetella sp.

+

+

2-7, 3-1

Diplolophodon {Deperetella) similis Z.

+

2-13

Diplolophodon {Deperetella) cf similis Z.

+

2-26, 2-27

Teleolophus liankanensis Z.

+

2-5

Teleolophus rnagnus R.

+

3-1

Teleolophus cf rnagnus R.

+

2-27

Teleolophus medius M. et G.

+

2-3A, 2-27

cf Teleolophus medius M. et G.

+

2-3B,C

Teleolophus sichuanensis X. et af

+

2-19

Teleolophus sp. nov. [“T. primarius Q.”]

+

2-2

Teleolophus sp. nov. [“T. rectis Q."\

+

2-2

Teleolophus sp.

+

+

2-26, 2-27, 2-28,
3-1

Helaletidae

Colodon cf inceptus M. et G.

+

2-2

?Colodon sp.

+

2-17, 2-19

Helaletes fissus (M. et G.)

+

2-3B

?Helaletes fissus (M. et G.)

+

2-3B

Helaletes mongoliensis (O.)

+

2-3A

?Helaletes mongoliensis (O.)

+

2-27

Helaletes sp.

+

2-3B

Heptodon niushanensis C. et L.

+

2-11

Heptodon tianshanensis Z.

+

2-5

?Heptodon sp.

+

2-1, 2-24

cf Heptodon sp.

+

2-19

Hyrachyus sp. nov. [“//. crista Q.”]

+

2-2

Hyrachyus sp. nov. [“//. medius Q.”]

+

2-2

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

61

CHAPTER 2. — Continued.

Paleocene

Eocene

Oligocene

Geographical

distnbution

Forms

E.-M. L.

E.

M.

L.

E.

M. L.

Hyrachyus sp. nov. [“?//. minor Q.”]

+

2-2

Hyrachyus sp. nov. [“//. neimongoliensis Q.”]

+

2-2

Hyrachyus spp.

+

+

2-12, 2-3B,

2-16

Tapiroidea gen. et sp. nov. ["'Euryletes
magniis Q.”j

Tapiroidea gen. et sp. nov. ['"Euryletes

+

2-2

medius Q.”]

Tapiroidea gen. et sp. nov. Euryletes

+

2-2

minimus Q.”]

+

2-2

Tapiroidea indet.
Hyracodontidae

+

3-12

Ardynia praecox M. et G.

+

3-1

Caenolophus medius C.

+

2-27

Caenolophus obliquus M. et G.

+

2-4

Caenolophus promissus M. et G.

+

2-4

Caenolophus spp.

+

+

+

2-16, 2-17, 2-26,

2- 27, 2-28,

3- 16

Imequincisoria mazhuangensis W.

+

2-18

Imequincisoria micracis W.

+

2-18

Imequincisoria (?) sp.

+

2-18

Imequincisoria sp.

+

2-8

Teilhardia pretiosa M. et G.

+

2-3C, 2-27

TTeilhardia sp.

+

2-27

Triplopus? proficiens (M. et G.)

+

2-3A, 2-3C

Triplopus? progressus (M. et G.)

+

2-4

Hyracodontidae gen. et sp. nov.

+

2-18

Hyracodontidae indet.

+

3-17

Amynodontidae

Amynodon altidens X. et C.

+

2-27

Amynodon alxaensis Q.

+

3-3

Amynodon lunanensis C. et al.

+

2-27

Amynodon m.ongoliensis O.

+

2-4, 2-5, 2-13,

2-18

Amynodon spp.

+

+

2-5, 2-13, 2-26,

2- 27, 2-28,

3- 12

Cadurcodon ardynensis O.

+

3-1, 3-6, 3-16

Cadurcodon sp.

+

3-1, 3-16

Euryodon minimus X. et al.

+

2-19

Gigantamynodon giganieus X.

+

3-16

Gigantamynodon cf. giganteus X.

+

3-16, 3-17

Gigantamynodon promisus X.

+

2-4

Gigantamynodon sp.

+

+

2-18, 3-16

cf. Gigantamynodon sp.

+

2-26

Huananodon hui Y.

+

2-26

Huananodon hypsodonta Y.

+

3-15

Lushiamynodon menchiapuensis C. et X.

+

2-17, 2-27

Lushiamynodon obesus C. et X.

+

2-15

Lushiamynodon sharamurenensis X.

+

2-4

? Lushiamynodon sharamurenensis X.

+

2-3C

Lushiamynodon wuchengensis W.

+

2-18

cf Lushiamynodon sp.

+

2-18

cf Metamynodon sp.

+

+

2-27, 3-16

Paracadurcodon suhaituensis X.

+

3-3

Paramynodon cf birmanicus (P. et C.)

+

2-26

62

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 21

CHAPTER 2. — Continued.

Paleocene

Eocene

Oligocene

Geographical

distribution

Forms

E.-M.

L.

E. M. L.

E.

M.

L.

cf. Pararnynodon sp.

+

2-26, 2-27

Sianodon hahoensis X.

+

3-12

Sianodon chiyuanensis C. et X.

+

2-15

Sianodon honanensis C. et X.

+

2-17

Sianodon mienchiensis C. et X.

+

2-13

Sianodon sinensis (Z.)

+

2-13, 2-15, 2-18

Sianodon ulausuensis X.

+

2-4

Sianodon spp.

+

+

2-4, 2-16, 2-19,

3-3, 3-12

Amynodontidae indet.

+

+

2-4, 2-16, 2-18,

2-26, 3-3

Rhinocerotidae

?Aceratherium spp.

+

3-2, 3-6, 3-9

Caenopus sp.

+

3-1

Dzungariotherium orgosensis C.

+

3-7

Dzungahotherium turfanensis X. et W.

+

3-6

Forstercooperia conjluens (W.)

+

2-3B

Forstercooperia shiwopuensis C. et al.

+

2-27

Forstercooperia totadentata (W.)

+

2-3A

Forstercooperia sp. nov. (“T. elongata Q.”)

+

2-2

Forstercooperia sp. nov. grandis Q?”)

+

2-2

Forstercooperia sp. nov.

+

2-18

Forstercooperia spp.

+

2-3C, 2-17,

2-26, 2-27

?Forstercooperia sp.

+

3-14

Forstercooperiinae indet.

+

2-18

Guixia simplex Y.

+

2-26

Guixia youjiangensis Y.

+

3-15

llianodon lunanensis C. et X.

+

2-27

?Ilianodon sp.

+

2-26

Indricotherium grangeri (O.)

+

3-1, 3-2, 3-4

Indricotherium intermedium C.

+

3-18

Indricotherium qujingensis T.

+

3-16

Indricotherium parvum C.

+

2-27

Indricotherium cf. parvum C.

+

2-27

“?Indricotherium" sp.

+

2-27

Indricotherium spp.

+

+

3-5, 3-16, 3-18,

3-10

Juxia sharamurenense C. et C.

+

2-4

Juxia spp. nov.

+

2-18

Juxia sp.

+

2-27

Pappaceras sp.

+

2-18

Paraceratherium lipidus X. et W.

+

3-6

Paraceratherium tienshanensis C.

+

3-6

Paraceratherium sp. (small form)

+

3-2

Prohvracodon meridionale C. et X.

+

2-27

Prohyracodon cf meridionale C. et X.

+

2-13

Prohyracodon progressa C. et X.

+

2-27

Prohyracodon cf orientale K.
Prohyracodon sp. (“Caenolophus cf

+

2-27

promissus")

+

2-13

Prohvracodon spp.

+

2-17, 2-19, 2-26,

2-27, 2-28

Urtinotherium incisivum C. et C.

H-

3-1

Urtinotherium sp. nov. [“?U. minor Q.”]

+

2-2

Indricotheriinae indet.

+

+

3-18, 3-6

small rhinocerotid

+

3-10

Rhinocerotidae indet.

+

2-18, 2-27

Rhinocerotidae indet.

+

3-9

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

63

CHAPTER 2. -Continued.

Forms

Paleocenc

Eocene

Oligocene

Geographical

distribution

E.-M.

L.

E. M. L.

E.

M. L.

Eomoropidae

Eomoropus quadridentatus Z.

+

2-13

Eomoropus cf. quadridentatus Z.

+

2-26, 2-27

Eomoropus ulterior C.

+

2-27

Eomoropus sp.

+

2-17, 2-18, 2-28

Grangeria canina Z.

?+

2-12

Grangeria? major (Z.)

+

2-13, 2-17

Lunania youngi C.

+

2-17, 2-27

Lunania cf. youngi C.

+

2-28

Chahcotheriidae

Litolophus gobiensis (C.)

+

2-3B

Olsenia mira M. et G.

+

2-4

Schizotherium avitum M. et G.

+

3-2

Schizotherium nabanensis Z.

+

3-15

Schizotherium sp.

+

+

3-1, 3-4, 3-6,

3-15

?Schizotherium sp.

+

3-9

Perissodactyla indet.

+

3-3

Artiodactyla

Anthracotheriidae

Anthracokeryx cf bambusae P.

+

2-26

Anthracokeryx birmanicus P. et C.

+

2-26

Anthracokeryx cf birmanicus P. et C.

+

2-26

Anthracokeryx gungkangensis Q.

+

3-14

Anthracokeryx kwangsiensis Q.

+

3-14

Anthracokeryx cf moriturus P.

+

2-26

Anthracokeryx sinensis (Z.)

+

2-13

Anthracokeryx cf sinensis (Z.)

+

2-13

Anthracokeryx sp.

+

2-26, 2-28

Anthracokeryx spp.

+

3-14, 3-15

Anthracosenex ambiguus Z.

+

2-13

Anthracothema minima X.

+

2-13

Anthracothema rubricae P.

+

2-26

Anthracothema sp.

+

2-28

Anthracotherium spp.

+

2-17

Bothriodon chowi X.

+

3-16

Bothriodon chyelingensis X.

+

2-26

Bothriodon tientongensis X.

+

3-15

Bothriodon sp.

+

+

2-5, 2-6, 3-17

Brachyodus hui (C.)

+

+

2-27, 3-13

Heothema angusticalxia T.

+

3-14

Heothema bellia T.

+

2-26

Heothema chengbiensis T.

+

3-15

Heothema media T.

+

3-14

Heothema sp.

+

2-26

Huananotherna imparilica T.

+

2-26

Probrachyodus panchiaoensis X. et C.

+

2-27

?Probrachyodus sp. nov.

+

2-26

Ulausuodon parvus H.

+

2-4

Anthracotheriidae indet.

+

+

+

2-15, 2-26, 2-27,

3-1, 3-6

?Anthracotheriidae indet.

+

3-16

Dichobunidae

Dichobune spp.

+

2-13, 2-17

Lantianius xiehuensis C.

+

3-12

Choeropotamidae

Gobiohyus orientalis M. et G.

+

2-3A, 2-17

64

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

CHAPTER 2. — Continued.

Paleocene

Eocene

Oligocene

Geographical

distribution

Forms

E.-M.

L.

E. M. L.

E.

M.

L.

Gobiohyus pressidens M. et G.

+

2-3A

Gobiohyus robustus M. et G.

+

2-3A, 2-17

Gobiohyus yuanchuensis Y.

+

2-13

Gobiohyus sp.

+

2-27

Choeropotamidae indet.

+

2-26

Entelodontidae

Archaeotherium ordosius Y. et C.

+

3-1, 3-4

Entelodon dims M. et G.

+

3-1

Eoentelodon yunnanense C.

+

2-27

Eoentelodon sp. nov.

+

2-28

achaenodont indet.

+

2-3A

Entelodontidae indet.

+

+

2-26, 3-3

Hypertragulidae

Archaeomeryx opt at us M. et G.

+

2-4, 2-16, 2-17

Archaeomeryx sp.

+

2-3A

Indomeryx cotteri P.

+

2-26

Indomeryx youjiangensis Q.

+

2-26

Indomeryx sp.

+

2-26

cf. Miomeryx sp.

+

3-16

Notomeryx besensis Q.

+

2-26

Xinjiangmery’x parvus Z.

+

2-5

Hypertragulidae indet.

+

2-28

Gelocidae

Lophiomerys sp.

+

3-8

Tragulidae

Tragulidae indet.

+

2-26

?Tragulidae indet.

+

3-6

Cervidae

Eumeryx spp.

+

+

3-2, 3-4, 3-9,

3-10

Cervulinae indet.

+

3-9, 3-10

Cervidae indet.

+

3-3

Bovidae

small hypselodont bovine

+

3-10

Bovinae indet.

artiodactylid indet. (= ^^Hoanghonius

+

3-9

stehlinC no. 2 of W. et C.)

+

2-13

Artiodactyla indet.

+

3-12, 3-17

Mammalia, Order indet.

Didymoconidae

Archaeoryctes notialis Z.

+

1-2

Didymoconus berkeyi M. et G.

+

3-6

?Didymoconus sp.

+

+

3-9, 3-10

Hunanictis inexpectatus L. et al.

2-25

Mongoloryctes auctus (M. et G.)

+

2-3A

Zeuctherium niteles T. et Y.

+

1-4

Mammalia, Order et Family indet.

Obtususdon hanhuaensis X.

+

+

1-4

Wanotherium xuanchengensis T. et Y.

+

1-5

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

65

CHAPTER 3. -THE INDEX

Systematic Index

Form

Distribution

Page

Aceratherium

?sp

. . . . 3-2,'3 3-6,

... 45, 48,

3-9

49

achaenodont

indet

. . . . 2-3A

25

Adapidium

huanghoense Y

. . . . 2-13

32

Advenimus

bohlini D

.... 2-3C

27

burkei D

. . . . 2-3B

25

hupeiensis D. et al

. . . . 2-20

37

cf. sp

. . . . 2-3C

27

Agispelagus

simplex A

.... HG

46

Allictops

i riser rat a Q

. . . . 1-4

17

Allostylops

periconotus Z

. . . . 1-2

16

Altanius

orlovi D. et M

.... NB

22

Altilambda

pactus C. et W

. . . . 1-1, 1-4

.... 13, 19

tenuis et W.

1-4

19

Amphechinus

acridens (M. et G.)

. HG, 3-2

.... 46, 45

cf. acridens (M. et G.) . . . .

. . . . 3-10

49

kansuensis (B.)

. . . . 3-9

49

minimus (B.)

. . . . 3-9

49

rectus (M. et G.)

.... HG

46

cf rectus (M. et G.)

. . . . 3-6, 3-9

.... 48, 49

?sp

. . . . AO, 3-6,

... 47, 48,

3-10

49

Amphicticeps

shackelfordi M. et G HG 46

Amphicynodon

teilhardi (M. et G.) HG 46

Amynodon

altidens X. et C 2-27 43

alxaensis Q 3-3 47

lunanensis C. et al 2-27 43

mongoliensis 0 2-4, 2-5, 2-13 28, 33,

2-18, AO 35,47

?mongoliensis 0 2-13 32

spp 2-5, 2-13, 2-26, . . 28, 32,

2-27,2-28,3-12 41,43,

44. 50

Amynodontidae

indet 2-4. 2-16, 2-18, .. . 28, 33

2-26,3-3 35,41,47

Form

Distribution

Page

Anagale

gobiensis S

. . . . 3-1

44

Anagalopsis

kansuensis B

. . . , 3-11

50

Anaptogale

wanghoensis X

. . . . 1-4

19

Anatolostylops

dubius Z

.... 2-5

29

Anchilestes

impolitus Q. et L

. . . . 1-4

19

Andrewsarchus

crassum D. et al

.... 2-26

41

mongoliensis O

.... 2-3A

24

?sp

. . . . 2-19

37

Anictops

tabiepedis Q

. . . . 1-4

19

aff tabiepedis Q

. . . . 1-4

19

Anthracokeryx

cf bambusae P

.... 2-26

41

birmanicus P. et C

.... 2-26

41

cf birmanicus P. et C. . . .

.... 2-26

41

gungkangensis Q

. . . . 3-14

50

kwangsiensis Q

. . . . 3-14

50

sinensis (Z.)

. . . . 2-13

.... 32,33

cf sinensis (Z.)

. . . . 2-13

32

Sp

. . . . 2-26, 2-28

. . . . 41, 44

spp

. . . . 3-14, 3-15

. . . . 50, 51

Anthracosenex

ambiguus Z

. . . . 2-13

33

Anthracothema

minima X

. . . . 2-13

33

rubricae P

.... 2-26

41

sp

.... 2-28

44

Anthracotherium

spp

. . . . 2-17

34

Anthracotheriidae

gen. et sp. indet

. . . . 2-16, 2-27, . . . .

. . . 33, 43,

AO, 3-6

47, 48

?indet

. . . . 3-16

51

Archaeolambda

dayuensis T

. . . . 1-2

16

planicanina F

. . . . NB

22

cf planicanina F

. . . . 1-2, 1-8

. . . . 16, 21

tabiensis H

.... 1-4

17

vangtzeensis H

. . . . 1-5

20

sp

. . . . 1-2, 2-25

. ... 16,40

Archaeolambdidae

gen. et spp. nov

. . . . 1-1

13

gen. et sp. indet

. . . . 1-2, 2-14

. . .. 17,33

The numbers (for example, 3-2) represent, first, that of the Section (or Epoch) and second, that of the basin or area m Chapter 1. Abbreviations mean; GA: Gashato; NB;
Naran Bulak; HG: Hsanda Gol; AO: Ardyn Obo.

66

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Systematic Index-

— Continued.

Form

Distribution

Page

Archaeomeryx

optatus M. et G

2-4, 2-17

28, 34

sp

2-3A, 2-16 . . .

25, 33

Archaeoryctes

notialis Z.

1-2

16

Archaeotherium

ordosius Y. et C

3-4

48

sp

AO

47

Arctostylopidae

gen. et sp. nov

1-1

13

Arctotitan

honghoensis W

2-7

30

Ardynia

praecox M. et G

3-1, AO

44, 47

Ardynictis

funmculus M. et G

AO

47

Ardynomys

olseni M. et G

AO

47

vinogradovi S

AO

47

sp.

3-1

44

Artiodactyla

indct

3-12, 3-17. . . .

50, 51

artiodactylid

indet. (="Hoanghonius

2-13

32

stehlini,” no. 2 of
W. et C.)

A siocoryphodo n

conicus X

2-19

37

lophodontus X

2-19

37

sp

38

Asiostylops

spanios Z

1-2

16

Bemalambda

rrn^.^n C et al.

1-1

14

nanhsiungensis C. et al

1-1, 1-3

14, 17

pachvoesteus C. et al

1-1

14

shizikouensis W. et D

1-2

17

Sp.

1-1, 1-4

14, 20

Bemalambdidae

indet

1-3, 1-4, 1-6, 1

-7 . . 17, 20

Bohlinotona

pusilla (T.)

3-2

45

Bothhodon

chowi X

3-16

51

chvelingensis X

2-26

41

tientongensis X

3-15

51

sp

2-5, 2-6, 3-17.

. 29,30,51

Bovinae

indet

3-9

49

Systematic Index— Continued.

Form

Distribution

Page

Brachyodus

hui (C.)

. . . . 2-13, 2-27,

31,43,

3-13

50

sp

.... AO

47

Breviodon

acares R

. . . . 2-3C

27

cf. acares R

.... 2-3C

27

minutus (M. et G.)

. . . . 2-17

34

cf. minutus (M. et G.) . . . .

. . . . 2-19

37

?minutus (M. et G.)

.... 2-2, 2-3A

... 24, 25

sahoensis C. et al. . . .

i.~n

43

Sp. nov

.... 2-17, 2-28

... 34, 44

sp

. . . . 2-3B, 2-7,

26, 30,

2-27

42

?sp

. . . . 2-3C, 2-18

. . . 27, 35

Brontotheriidae

gen. et sp. indet

. . . . 2-16, 2-27,

33, 43,

2-28, 3-12,

44, 50,

3-16

51

Cadurcodon

ardvnense (O.)

. . . . 3-1, AO,

44, 47,

3-6, 3-16

48, 51

sp

. . . . 3-1, 3-16

44, 45, 51

Caenolophus

medius C

. . . . 2-27

43

obUquus M. et G

. . . . 2-4

28

promissus M. et G

. . . . 2-4, AO

... 28, 47

sp

. . . . 2-16, 2-17,

. . 33, 34,

2-26, 2-27,

41, 43

2-28, 3-16

44, 51

Caenopus

Sp

. . . . 3-1

45

Canidae

gen. et sp. indet

. . . . 2-8, 2-27

... 30, 43

Carnivora

gen. et sp. indet

. . . . 2-2, 2-18, 2-19, .

. . 23, 35,

3-9, 3-10

37, 49, 50

Cartictops

canina D. et T

. . . . 1-4

19

Cervidae

gen. et sp. indet

. . . . 3-3

47

Cervulinae

sp

. . . . 3-9, 3-10

... 49, 50

indet

. . . . 3-9

49

ChaiUcyon

crassidens C

. . . . 2-13, 2-27

... 32, 43

Chianshania

gianghuaiensis X

. . . . 1-4

19

Choeropotamidae

indet

.... 2-26

41

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

67

Systematic \nAex— Continued.

Systematic XnAe's.— Continued.

Form

Dislnbuiion

Page

Chungchienia

sichuanica C

. . . 2-19

36

Cocomys

lingchaensis (L. et al.)

. . . 2-25

38

Cocomyidae

indet. (“ Sciuravus sp.”) ....

. . . 2-19

36

Colodon

inceptus M. et G

. . . AO

47

cf. inceptus M. et G

. . . 2-2

23

?sp

. . . 2-17, 2-19

34, 37,

AO

47

Convallisodon

convexus C. et Q

. . . 1-9

22

haliutensis C. et Q

. . . 1-9

22

Coryphodon

dabuensis Z

. . . 2-5

29

flerowi C

. . . 2-12, 2-19

. . . 31, 37

ninchiashanensis C. et T. . .

. . . 2-24

38

sp

. . . 2-5

29

Coryphodontidae
gen. et sp. nov. [“Meta-

coryphodon minor Q.”] . .

. . . 2-2

23

gen. et sp. nov. [“Meta-

coryphodon luminis Q.”]

. . . 2-2

23

gen. et sp. indet

. . . 2-19

37

Creodonta

indet

. . . 2-26, 2-27,

41.42,

2-28

44

Cricetodon

schaubi Z

. . . 2-13

31

cf sp

. . . 3-9

49

Cr ice tops

cf aeneus S

. . . HG

46

dormitor M. et G

. . . HG

46

Cyclomylus

lohensis M. et G

. . . HG, 3-4

... 46, 47

minutus K

. . HG

46

Cynodictis

?constans (M. et G.)

. . HG

46

?elegans M. et G

. . HG

46

?sp

. . . 2-17, AO

... 34, 47

Decoredon

elongetus X

. , . 1-4

19

Deperetella

cristata M. et G

. . . 2-4

28

depereti (Z.)

. . . 2-13

33

dienensis C. et al

1.11

43

sitnilis (Z.)

. . . 2-13

33

cf similis (Z.)

. . . 2-26, 2-27

. . . 41, 43

Sp. nov.

. , . 2-18

35

sp

. . . 2-3C, 2-7, 2-17,.

. . 30, 34,

2-18, 2-26,

36, 41,

2-27, 2-28

43, 44, 45

Form Distribution Page

Desmatolagus

gobiensis M. et G

?parvidens B

robust us M. et G

shargaltensis B

spp

sp. (?shargaltensis)

Desmatotitan
lukhumensis G. et G. . .
sp

Diacronus

anhuiensis X

wanghuensis X

Dianotitan

lunanensis (C. et H.) . . . .
Dichobune

spp

Didymoconus

berkeyi M. et G

(="Tschelkaria”) col- . . .

gatei M. et G.

?sp

Dilambda

speciosa T

Diplolophodon (Deperetella)
cf. similis Z

Dissacus

magushanensis Y. et T. .

?rotundus W

?sp

Dissacusium

shanghoensis C. et al. . . .

Dolichorhinoides
angustidens G. et G

Dysnoetodon

niinuta Z

Dzungariotherium

orgosensis C

turfanensis X. et W

Ectoconus

?sp

Embolotherium

andrewsi O

grange ri O

?grangeri O

loucksii O

ultimum G. et G

Embolotheriinae

indet

Enteiodon

dirus M. et G

HG 46

3-10 49

HG, AO 46, 47

3-10 49

3-10 49

3-9 49

2-3C 27

2-2 24

1-4 19

1- 4 19

2- 27 43

2- 13,2-17 33,34

HG, 3-6 46, 48

HG 46

3- 9, 3-10 49, 50

1-8 21

2-26,2-27 41,43

1-5 20

1-3 17

1-9, GA 21,22

1- 1 13

2- 3C 27

1-1 14

3- 8 48

3-6 48

1-1 14

3-1, AO 44,47

3-3 47

3-1 44

3-1 44

3-1 45

3-3 47

3-1 45

68

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Systematic InAe^— Continued.

Systematic \nAtx— Continued.

Form

Distribution

Page

Entelodontidae

indet

. . 2-26, 3-3

41,47

Eodesmatodon

spanios Z. et C

. . 2-26

41

Eoentelodon

Yunnanense C

. . 2-27

43

sp

. . 2-28

44

Eomoropus

major Z

. . 2-17

34

quadridentatus Z

. . 2-13

33

cf. quadridentatus Z

. . 2-26, 2-27

... 41, 43

ulterior C

. . 2-27

43

sp

. . 2-17, 2-18, . . .

. . . 34, 36,

2-28

44

Epimanteoceras

amplus J

. . AO

47

formosus G. et G

. . 2-3C

27

robustus (M. et G.)

. . AO

47

Ergilobia

gobiensis T

. . AO

47

Erinaceidae

small species

. . 3-10

49

'.’indet

. . 3-9

49

Ernanodon

antelios D

1-1

13

Eucricetodon

asiaticus (M. et G.)

. . IIG

46

sp

. . AO

47

Eudinoceras

crassum T. et T

. . 2-26

41

kholobolchiensis O. et G. ...

. . 2-3C

27

cf kholobolchiensis O. et G.

. . 2-21

37

mongoliensis O

. . 2-3A

24

sp

. . 2-6, 2-8, 2-17. .

.... 30, 34

Eumeryx

culminis M. et G

. HG

46

sp

. AO, 3-2, 3-4. . .

47, 45, 48,

3-9, 3-10

49, 50

Eumys

aff sp

. 3-9

49

?Eurymylidae

gen. et sp. indet

1-8

21

Eurymyloidea

indet

. . 1-4

17

Form

Distribution

Page

Exallerix

hsandagolensis M. et H

. HG

46

Felidae

indet

. 2-6, 2-27

. . . 30, 42

Eorstercooperia

confJuens (W.)

. . 2-3B

26

shiwopuensis C. et al

. . 2-27

43

totadentata (W.)

. . 2-3A

25

sp. nov. [“E’. elongata"] ....

. . 2-2

24

sp. nov. [“F. grandis”]

. . 2-2

24

Sp. nov

. . 2-18

36

spp

. . 2-3C, 2-17,

27, 34,

2-26, 2-27

41, 43

?sp

. . 3-14

50

Forstercooperiinae

intipt

2-18

35

Ganatherium

australis H

. . 2-23

38

Ganolophus

lanikenensis Z

. . 1-2

16

Gigantamynodon

cessator G

. . AO

47

giganteus X

. . 3-16

51

cf giganteus X

. . 3-16, 3-17

51

promisus X

. . 2-4

28

spp

. . 2-18, 2-26,

35,41,

3-16

51

Gnathotitan

berkevi (O.)

. . 2-3A

25

Gobiatherium

mirificum O. et G

. . 2-2, 2-3B

. . . 23, 25

sp. nov. [“G. major Q.”] ...

. . 2-2

23

sp. nov. [“G. monolabotum .

. . 2-2

23

Q.”]

?sp

. 2-19

37

Gobiohyus

orient alis M. et G

. . 2-3A, 2-17

... 25, 34

pressidens M. et G

. . 2-3A

25

robustus M. et G

. . 2-3A, 2-17

. . . 25, 34

yuanchuensis Y

. . 2-13

33

sp

. . 2-27

43

Gobiolagus

andrewsi B

. 3-1

44

?major B

. 3-1

44

tolmachovi B

. 2-4

28

Eurymylus

laticeps

GA, NB

22

Euryodon

minimus X. et al

2-19

37

Gobiometyx

dubius T AO 47

Gomphos

elkema S GA 22

Eusmilus

cf. sp 2-17 34

?sp 2-26 41

Grangeria

canina Z 2-12 31

?major (Z.) 2-13 33

1983

LI AND TING-THE PALEOGENE MAMMALS OE CHINA

69

Systematic InAes.— Continued.

Systematic XnAe-x.— Continued.

Form

Dislribulion

Page

Guilestes

acares Z. et C

2-26

41

cf. acares Z. et C

2-26

41

Guixia

simplex Y

2-26

41

youjiangensis Y

3-15

51

Haltictops

nipiliriQpn^i^ V) Pt T

1-1

13

mirnhiUs D. ef T. .

1-1

13

Hapalodectes

serus M. et G

2-3A

24

?sp

1-2

16

Haplornylus

?sp

2-12

31

Harpagolestes

alxaensis Q

3-3

46

?orientalis S. et G

2-3C

27

cf. sp

2-26, 3-1

. . . . 41, 44

Harpyodus

decorus W.

1-2

16

euros Q. et L

1-4

19

Helaletes

ftssus (M. et G.)

2-3B

26

?fissus (M. et G.)

2-3B

26

mongoliensis (O.)

2-3A, 2-27 ....

. ... 25,43

sp

2-3B

26

Heomys

orientalis L

1-4

17

sp

1-4

19

Heothema

angusticalxia T

3-14

50

bellia T

2-26

41

chengbiensis T

3-15

51

media T

3-14

50

sp

2-26

41

Heptaconodon

dubium Z

l.\l

31

Heptodon

niiishanensis C. et L

2-11

31

tianshanensis Z

2-5

29

cf sp

2-19

37

?sp 2-1, 2-24 22, 38

Hoanghonius

stehlini Z

2-13

.... 31, 32

Homogalax

wutuensis C. et L

2-10

30

Honanodon

hebetis C

2-13, 2-28

. . . . 32, 44

macrodontus C

. . 2-17

34

sp

1.11 1-19.

. . . . 42, 44

Form

Dislribulion

Page

Ilouyanotheriiim

primigenum T

. . . 1-8

21

simplum T. .

1-8

21

Hsiuannania

maguensts X

. . . 1-5

20

minor D. et Z

. . . 1-2 .

16

tabiensis X

. . . 1-4

17

sp

. . . 1-4

17

Hiiaiyangale

chianshanensis X

.1-4

19

cf leura D. et T

. . . I-I

13

sp

. . . 1-4

19

Huananodon

hut Y

. . . 2-26

41

hvpsodonta Y

. . . 3-15

51

Hiiananothema

imparilica T

. . . 2-26

41

Hukoiitherium

ambigum C. et al

1-1

13

Hidgana

ertnia D

. . . 3-1

44

Hunanictis

inexpectatus L. et al

. . . 2-25

40

Hyaenodon

ambigiius M

. . . HG

46

avmardi F

. . . HG

46

compressus F

HG

46

eminus M. et G

. . . AO

47

pervagus M. et G

. . HG

46

yuanchuensis Y

. . . 2-13

32

sp

. . . 2-18

35

?sp

. . . 3-2, 3-6

45, 48

Hyaenodontidae

indet

. . . 2-28

44

Hyopsodits

sp

. . . 2-5

29

Hyopsodontidae

indet. ...

. . . 1-2

16

Hyotitan

thomsoni G. et G

. . . 3-1

45

Hypercoryphodon

thomsoni O. et G

. . . 3-1

45

Hypertragulidae

gen. et sp. indet

. . . 2-28

44

Hypsamynodon

progressus G

. . . AO

47

hypselodont bovine (small)

. . . 3-10

50

Ilypsilolambda

chalingensis W

. . . 1-3

17

impensa W

. . . 1-3

17

spp

. . . 1-3

17

70

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 21

Systematic \nAe\— Continued.

Form Distribution Page

Hypsimylus

beijingensis Z

Hyrachyus

sp. nov. [“//. crista'] . .
sp. nov. [“//. niedius Q.”]
sp. nov. [“//. minor Q.”]
sp. nov. [“//. neimongol-
iensis Q.”]

spp

Hyracodontidae

gen. et sp. indet

gen. et sp. nov

Ilyracolestes

erm incus M. et G

Hyracotheriinae
gen. et sp. indet

Iciopidium

lechei Z

?sp

Ilianodon

lunanensis C. et X

?sp

Imequincisona

mazhuangensis W

micracis W

sp

?sp

Indomeryx

cotteri P

youjiangensis Q

sp

Indricotheriuin
gran gen O

intermedium C

parvum C

cf. parvum C

qujingensis T

sp

?sp

Indricotheriidae
indet

Insectivora

gen. et sp. nov

gen. et sp. indet

2-8 30

2-2 23

2-2 23

2-2 23

2-2 23

2-3B, 2-12 26, 31,

2- 16 33

AO, 3-17 47, 51

2-18 35

1- 4, GA 17,22

2- 12 31

2-13 31

AO 47

2-27 43

2-26 41

2-18 35

2-18 35

2-8 30

2-18 35

2-26 41

2-26 41

2- 26 41

3- 1, HG, 3-2 45,46,

3- 4 47

3-18 51

2-27 43

2- 27 43

3- 16 51

3-5, 3-10, 48, 50,

3-16,3-18 51

2- 27 43

3- 6, 3-18 48, 51

2-2 23

1-2,2-25 16,38,

3-10 49

Ischyromyidae

?indet 2-13, 3-1 31, 44

Jiangxi a

chaotoensis Z. ei a\ 1-2 16

Systematic Index— Continued.

Form

Distribution

Page

Jiaoluotherium

turfanense (C.)

. . . 1-8

. . . . 21

Juxia

sharamurenense C. et C. . .

. . . 2-4

.... 28

spp

. . . 2-18, 2-27

.. 35,43

Kansupithecus

. . . 3-9 (Mio.)

.... 49

Karakoromys

decessus M. et G

... HG

.... 46

cf. decessus M. et G

. . . 3-10

.... 49

?decessus M. et G

3-2

.... 45

(?=Leptotataromvs) sp

HG

.... 46

Khashanagale

zofiae S. et M

. . . GA

.... 22

?sp. nov

. . . GA

.... 22

Kuanchuanius
shantunensis C

i.\i

. . . . 31

Lagomorpha

indet

... 3-3 ..

.... 46

Lambdopsalis

bulla C. et Q

. . . 1-9

. . . . 21

Lambdotherium

“^sp

. . . 2-1

.... 22

Lantianius

xiehuensis C

. , . 3-12

.... 50

Leptotataromys
gracilidens B

. . . 3-10

.... 50

Linnania

Infnpn^i^ C et a!

. . 1-1

13

Litolophus

gobiensis (C.)

. . . 2-3B

.... 26

Lqfochaius

hrnrhyoHiJS C et al

, . . 1-1

14

Lohoodon

lushiensis (C.)

. . . 2-17

.... 34

Lophialetes

expeditus M. et G

. . . 2-2, 2-3A,B,C, . .

24, 25,

2-5, 2-19, 2-27

26, 27,

cf expeditus M. et G

... 2-6

28, 37, 42
30

spp

. . . 2-3A, 2-3C, ....

25, 27,

2-17, 2-19

34, 37

Lophialetidae

gen. et sp. indet

. . . 2-5, 2-18

. . 28, 35

?gen. et sp. indet. (cf

. . . 2-19

. . . . 37

Breviodon)

Lophtodon

?sp

. . . 2-28

.... 44

Lophiomeryx

angarae M. et G

. . . AO

.... 47

1983

LI AND TING-THE PALEOGENE MAMMALS OE CHINA

71

Systematic Index— Systematic Index—

Form

Dislnbulion

Page

gobiae M. et G

... AO

47

Sp

... 3-8

48

Lunania

voungi C

. . . 2-17, 2-27. . . .

34, 43

cf voungi C

. . . 2-28

44

Lushiamynodon

menchiapuensis C. et X. . . .

. . . 2-17, 2-27. . . .

34, 43

obesus C. et X

. . . 2-15

33

sharamurenensis X

... 2-4

28

?sharamurenensis X

. . . 2-3C

27

wuchengensis W

. . . 2-18

35

cf sp

. . . 2-18

35

Lushilagus

lohoensis L

. . . 2-17

34

Lushius

qinlinensis C

... 2-17

34

Manteodon

voungi X

. . . 2-21

37

cf voungi X

. . . 2-19

37

Matutinia

nitidulus L. et al

. . . 2-25

38

Meiostylodon

zaoshiensis W

. . . 1-3

17

Mesonychidae

gen. et sp. indet

. . . 1-6, 1-7, 1-8, .

. .. 20,21,

2-3A, 2-19

24, 37

mesonychid

. . . 2-17

34

Mesonyx

sp

cf. sp

?sp. nov. [“M obtusidens

Q-”]

Metamynodon

cf. sp 2-27, 3-16 43, 51

Metatelmathehum

cf browni C 2-26 41

chstatum G. et G 2-3B 25

parvum G. et G 2-3A 25

Metatitan

primus G. et G 3-1 44

progressus G. e\ G 3-1 44

relictus G. et G 3-1, AO 45, 47

Miacis

invictus M. et G 2-3A 24

lushiensis C 2-17,2-19 34,36

sp 2-8 30

Microtitan

mongoliensis (O.) 2-3A,C 25, 27

sp. nov. 2-2 24

elongatus Q.”]

Form

Dislnbulion

Page

sp

1-1

24

?sp. (or new genus)

.... 2-17

34

Mimolagus

rodens B

. . . 3-11

.... 50

Mimotona

borealis C. et Q

. . . 1-9

. . . . 21

robusta L

... 1-4

17

wana L

.. 1-4

17, 19

sp

... 1-4

. . . . 19

Minchenella (Conolophus)
grandis Z

. . . 1-1

13

Miomeryx

ahaicus M. et G

. AO

.... 47

sp

. . HG, AO,

46, 47,

3-16

51

Mongolestes

hadrodens S. et G

. . , 3-1

44

Mongolonyx

dolichognathus S. et G

. . . 2-3B

. . . , 25

sp. nov. [“A/

1.1

. . . . 23

prominentis'"]

Mongoloryctes
auctus (M. et G.)

. . . 2-3A

. . . . 25

Mongolotherium

efremovi F

. . . 2-1. NB

11

plantigradum F

. . . NB

11

Morosomys

silenti S

... AO

.... 47

Multituberculata
indet

. . . 1-8

20

Nanlingilambda

chijiangensis T

. . . 1-2

. . 16

Nimravus

cf sp

. . HG

.... 46

Notomeryx

besensis Q

. . . 2-26

. . . . 41

Obtususdon

hanhuaensis X

. . . 1-4

. . 17, 19

Ochotonolagus
?arg\>ropuloi G

. . HG

.... 46

Olsenia

mira M. et G

... 2-4

.... 28

Ordolagus

teilhardi (B.)

. . HG, 3-2

. . 46, 45

Pachyaena

sp

... NB, 2-1, 2-3A . . .

22, 24

?sp

. . 1-1

13

Pachycynodon
sp

. . . 3-14

. . . . 50

2-3A 24

2-3C 27

2-2 23

72

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Systematic InAe'x.— Continued.

Form

Distribution

Page

Pachytitan

aja.x G. et G

. . . 2-4

. . . . 28

Palaeogale

parvula (M. et G.)

(=" Bimaelurus") ulvsses . . .
(M. et G.)

. . . HG

. . HG

. . . . 46

. . . . 46

Palaeohypsodontus
asiaticiis T

. . . HG

. . . . 46

Palaeolaginae

indet

. . . 3-12

. . . . 50

Palaeophonodon
gracilis M. et G

. . . HG

... 46

Palaeostylops

iturus M. et G

tnacrodon M., G. et S

. . . 1-9, GA, NB, 2-1 .

. . 1-9, GA, NB, 2-1 .

21, 22
t~>

Palaeosyops

cf. sp 2-19 37

Palasiodon

siurenensis C. ex a\ 1-1 13

Pantolambdodon

fortis G. et G

. . 2-3C

27

inennis G. et G

. . 2-3C

27

?sp. nov. [“f. minor Q.”] , . .

23

Panto testes

?sp

.. 2-3A

25

Pappaceras

sp

. . 2-18

36

Pappictidops

acies W

. 1-1 . .

13

obtusus W

. . 1-1

13

orientalis Q. et L

. . 1-4

19

Parabrontops

gobiensis (O.)

. . 3-1, AO

. . . . 44, 47

sp

. . 3-7

48

Paracadiircodon

suhaituensis X

. . 3-3

47

Paraceratherimn

lipidus X. et W

. . 3-6

48

tienshanensis C

. . 3-6

48

sp. (small form)

. . 3-2 . .

45

Paracynohyaenodon

morrisi M. et G

.. 2-3A

24

paramyid

gen. et sp. nov. ['\4siamys . .

':>-7

23

mediiis Q.”]

spp

. . 2-3A,B

.... 24, 25

Paramynodon

cf. binnanicus (P. et C.) . . . .

. . 2-26

41

sp

. . . . 41, 43

Systematic Index—

Form

Distribution

Page

Paranictops

majuscida Q

1-4

19

sp

1-4

19

Paratriisodon

gigas C. et al

2-3B, 2-17 . . .

. . . . 25, 34

henanensis C

2-17

34

Pastoralodon

lacustris C, et Q

1-9

77

?lacustris C. et Q

2-1

22

Pastoralodontidae

indet

.... 1-6

20

Perissodactyla

indet

.... 3-3

47

Petrolemur

brevirostre T

.... 1-1

13

Phenaceras

lacustris T

.... 2-23

38

Phenacolophidae

gen. et sp. nov

.. . 1-1

13

Plienacolophus

fallax M. et G

.... GA

Plagiocristodon

serratus C. et Q

1-9, 2-1

21 22

cf. Plesictis

sp

HG

46

Plesiosmintlius

asiaecentralis (B.)

.... 3-9

49

parvuius (B.)

.... 3-9

49

tangingoli (B.)

... HG, 3-9

. . . . 46, 49

sp

. . . . 2-13

31

Praolestes

nanus M. et G

.... GA

46

Primates

indet

.... 3-9

49

Prionessus

lucifer M. et G

. . . . 1-9, GA, 2-1 . .

. . . . 21, 22

cf Proailurus

sp

.... HG

46

?Prohathyopsis

sinviiensis C. et T

. . . . 2-24

38

Proboscidea

sp

. . . . 3-9 (Mio.)

49

Prohrachyodus

panchiaoensis X. et C. ...

7-')7

43

?sp. nov

. . . . 2-26

41

Procaprolagus

radicidens (T.)

. . . . 3-2

45

vet us t us (B.)

. . . . 3-1, HG

... 44, 46

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

73

Systematic InAe's.— Continued.

Form

Distribution

Page

Prodinoceras

diconicus T

. . . . 1-8

21

martvr M. et G

. . . . GA

22

Prodinoceratinae

gen. et sp. indet

. . , 2-14

33

Prohyracodon

meridionalis C. et X

43

cf. meridionalis C. et X. . . .

. . . 2-13

33

cf. orientale K

7-71

43

progressa C. et X

. . . 2-27

43

sp. ("Caenolophus cf.

. . . 2-13

33

promissus")

spp

. . . 2-17, 2-19, 2-26,
2-27, 2-28

. . 34, 37,
41, 43, 44

Prolaena

parva X. et al

. . . 2-19

36

Propachynolophiis

hengyangensis (Y.)

. . . 2-25

40

Propalaeotherium

sinensis Z

. . . 2-12

31

Propterodon

irdinensis M. et G

. . . 2-3A, 2-17

. . . 24, 34

Prosciurus

arboraptus S

... HG

46

lohiculus M. et G

... HG

46

Protembolotherium

efremovi J

... AO

47

Protitan

bellus G. et G

. . . 2-3C

27

?cingulatus G. et G

. . . 2-3B

25

granger i (O.)

. . . 2-3A, 2-17

. . . 25, 34

minor G. et G

. . . 2-3B

25

obliquidens G. et G

. . . 2-3A

25

robustus G. et G

. . . 2-3A

25

cf robustus G. et G

. . . 2-27

42

?spp

. . . 2-19, 2-26

. .. 37,41

Pseuda nisonchus

nntpUns 7 et al

\-7

16

Pseudictops

chaii T

... 1-8

21

lophiodon M., G. et S

. . . 1-9, GA, NB, 2-1

21,
46, 22

cf tenuis D. et Z

... 1-2

16

Pseudictopidae

indet

. . . . 1-6

20

Pseudocylindrodon

mongolicus K HG 46

Pseudomeryx

gobiensis T HG 46

Pterodon

dakhoensis C 2-27 43

Systematic Index— Continued.

Form

Distribution

Page

hvaenoides M. et G

2-4

28

mongoliensis (D.)

AO

47

Pyrodon

xinjiangensis Z

2-5

29

Sp

2-1 . .

22

Rhinocerotidae

indet

2-18, 2-27, . . . ,

. . 36, 43,

HG, 3-9

46, 49

small rhinocerotid

3-10

50

Rhinotitan

andrewsi (O.)

2-4

28

kaiseni (O.)

2-4

28

mongoliensis (O.)

2-4, 2-13

.... 28, 32

quadridens X. et C

2-27

43

spp

2-5, 2-27

.... 28, 42

Rhodopagus

?minimus (M. et G.) ....

2-4

28

pvgmaeus R

2-3C, 2-27 . . . .

.... 27,43

?pvgmaeus R

2-3A

25

sp

2-12, 2-13

. . 31,32,

2-27

43

Rhombomylus

laianensis Z. et al

2-22

37

turpanensis Z

2-5

29

sp

2-20

37

Rodentia

indet

1-9, 2-12, 3-3. .

21, 31, 46

Sarcodon

pvgmaeus M. et G

1-9, GA

.... 21, 46

Sarkastodon

mongoliensis G

2-3A

24

Sayimys

obliquidens B

3-9 (Mio.)

49

Schizotherium

avitum M. et G

AO, 3-2

.... 47, 45

nabanensis Z

3-15

51

spp

3-1, 3-4, 3-6, . .

45, 47, 48,

3-9, 3-15

49, 51

Schlosseria

magister M. et G

77 17

cf magister M. et G

2-3B

26

sp. nov. [“X

2-2

23

dimera Q.”]

sp. nov. [“5

2-2

23

masculus Q.”]

sp

2-28

44

“Sciuravus”

sp

2-19

36

Sciuravidae

gen. et sp. nov

2-17

34

Sciuridae

gen. et sp. indet

3-10

49

74

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Systematic InAe's.— Continued. Systematic Index— Continued.

Form

Distribution

Page

'^Sciurus"

sp

3-9

49

Selenomys

mimicus M. et G

IIG

46

Shamolagus

gr anger i B

2-3C

27

medius B

2-4

28

Sianodon

hahoensis X

3-12

50

chivuanensis C. et X

2-15

33

honanensis C. et X

2-17

34

mienchiensis C. et X

2-13

33

sinensis (Z.)

2-13, 2-15, ... .

.. 32,33,

2-18

35

ulausuensis X

2-4

28

spp

2-4, 2-16,

.. 28,33,

2- 19, 3-3,

3- 12

37, 47, 50

Sicistinae

• spp

.... 3-9

49

indet

. . . . 3-10

50

Simplaletes

sujiensis Q

.... 2-3A

25

ulanshierhensis Q

. . . . 2-3C

27

Sinolagomys

gracilis B

. . . . 3-10

49

kansuensis B

. . . . 3-6, 3-9, 3-10. .

. . . . 48, 49

major B

. . . . 3-9, 3-10

49

cf. major B

.... 3-2

45

“Sinopd"

?sp

. . . . 2-19

36

Sinostylops

progressus T. et Y

. . . . 1-5

20

promissus T. et Y

.... 1-4

17

Soricidae

indet

.... AO, 3-9

. . . . 47, 49

SphenopsaUs

nohilis M., G. et S

. ... 1-9, GA

. . . . 21, 46

Stenanagale

xiangensis W

. . . . 1-3

17

Stylinodon

?sp

.... 2-17

34

Symphysorrhachis

brevirostris B

.... AO

47

Tachyoryctoides

intermedius B

3-10

50

obrutschewi B

. ... HG, 3-10

. . . . 46, 50

pachvgnaihus B

.... HG, 3-10

. . . . 46, 50

sp

.... 3-9

49

?Talpidae

indet

.... 3-9

49

Form

Distribution

Page

Tamquammys

wilsoni D., L. et Q

23

Tapiroidea

gen. et sp. nov

23

["'Euryletes medius Q.”]

gen. et sp. nov

1.1

23

Euryletes magnus Q.”]

gen. et sp. nov

1.1

23

Y" Euryletes minimus Q.”]

indet

. . 3-12

50

Tataromys

deflexus T

. . 3-2, HG

... 45, 46

gobiensis K

. . HG

46

grange ri B

. . 3-9

49

cf. grangeri B

. . HG

46

plicidens M. et G

. . HG, 3-2

... 46, 45

cf plicidens M. et G

. . 3-9, 3-10

49

sigmodon M. et G

. . HG, 3-9

... 46, 49

cf sigmodon M. et G

. . 3-6

48

Teilhardia

pretiosa M. et G

. . 2-3C, 2-27

, . . 27, 43

?sp

. . 2-27

43

Teleolophus

liankanensis Z

. . 2-5

28

magnus R.

. . 3-1

44

cf magnus R

1-11

43

medius M. et G

. . 2-3A, 2-27 ....

. ... 25,43

cf medius M. et G

. . 2-3B, 2-3C, . . .

26, 27,

2-19

37

?medius M. et G

. . 2-3C

27

sichuanensis X. et al

. . 2-19

37

sp. nov. [“Y.

. . 2-2

24

rectis Q.”]

sp. nov. Y'T.

. . 2-2

24

primarius Q.”]

sp

. . 2-26, 2-27

41,43,

2-28, 3-1

44, 45

Thinocyon

?sichowensis C

l.M

31

Tienshanilophus

lianmuqinensis T

.. 1-8

21

shengjinkouensis T

. . 1-8

21

1-8

21

Tillodontia

gen. et sp. nov

1-1

23

Y^Ulanius chowi Q.”]

gen. et sp. indet

. . 2-13, 2-27

. . . . 32, 42

Titanodectes

ingens G. et G

. . 3-1, AO

. . . . 44, 47

minor G. et G

. . 2-4, 3-1

. . . . 28, 44

Tragulidae

indet

. . 2-26

41

?indet

. . 3-6

48

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

75

Systematic IxiAc's.— Continued.

Form

Distribution

Page

Thphpus

proficiens (M. et G.)

. . 2-3A, 2-3C ....

. ... 25,27

?progressus (M. et G.)

. . 2-4

28

Tritenmodon

?sp

. . 2-19

36

Tsaganomys

altaicus M. et G

. . HG, 3-2,

46, 45, 50

Tsinlingomys

voimgi L

3-10

. . 2-17

34

Tupaiodon

?minutus M. et G

. . HG

46

moirisi M. et C

. . HG

46

?sp

. . 2-8

46

?cf. Uintalherium
sp

. . 2-12

31

Uintatheriinae

indet

. . 2-17

34

Ulaimiodon

parvus H

. . 2-4

28

Urtinotherium

incisivum C. et C

. . 3-1

44

sp. nov. {'"?U. minor Q.”] . .

. . 2-2

24

Wanogale

hodungensis X

, . 1-4

19

Wanotherium

xuanchengensis T. et Y

, . 1-5

20

Xinjiangmeryx

parvus Z

, . 2-5

29

Xinyuictis

tenuis Z. et al

. 2-24

38

Yantangiestes

iLestes) conenxus (Y. et T.) . .

, 1-4

20

(Lestes) datangensis (W.) . . . .

. 1-1

13

(Dissacus) feiganensis

. 1-1

13

(C. et al.)

Yindirtemys

woodi B

. 3-9

49

Yue tophus

validus Z

. 1-1 . ...

13

Yuodon

protoselenoides C. et al

. 1-1

13

Yuomys

cavioides L

. . 2-4, 2-13, 2-15 .

28, 32,

eleganes W. . . .

. . 2-18

33

35

Sp. nov

. . 2-17

34

Zeuctherium

niteles T. et Y

. . 1-4

19

76

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Stratigraphic Formation Index

Pin-yin

English

Age

Page

—

Ardyn Obo Formation

Early Oligocene

47

—

Arshanto Formation

Middle Eocene

22

Bai-lu-yuan Formation

Pai-lu-yuan Formation

Early Oligocene

50

Bai-shui-cun Formation

Pai-shui-tsun Formation

Early Oligocene

31,50

-

Bayan Ulan Formation

Early Oligocene

22

Cai-jia-chong Formation

Tsai-chia-chung Formation

Early Oligocene

51

Cha-gan-bu-la-ge Formation

Cha-gan-bu-la-ge Formation

Early Oligocene

45

Chi-jiang Formation

Chih-kiang Formation

Late Paleocene

15

Chang-xin-dian Formation

Chang-sin-tien Formation

Late Eocene

30

Chuan-kou Formation

Chuan-kou Formation

Middle Eocene

33

Chu-gou-yu Formation

Chu-kou-yu Formation

Late Eocene

34

Da-bu Formation

Ta-pu Formation

Early Eocene

28

Da-cang-fang Formation

Ta-tsang-fang Formation

Middle Eocene

36

Da-tang Member

Ta-tang Member

Late Paleocene

13

Da-yu Formation

Ta-yu Formation

?01igocene

34

Da-zhang Formation

Ta-chang Formation

Late Paleocene

20

Dai-jia-ping Formation

Te-chia-ping Formation

Cretaceous

17

Dan-xia Formation

Tan-hsia Formation

?Eocene

13

Dong-hu Formation

Tung-hu Formation

Early Eocene

37

Dong-jun Formation

Tung-chun Formation

Late Eocene

40

Dou-mu Formation

Tou-mu Formation

Late Paleocene

17

Gao-yu-gou Formation

Kou-yu-kou Formation

Middle Paleocene

20

—

Gashato Formation

Late Paleocene

22

Gong-kang Formation

Kung-kang Formation

Early Oligocene

40, 50

Guan-zhuang Formation

Kuan-chuang Formation

Middle Eocene

31

He-se Formation

Ho-se (Brown) Formation

Middle Oligocene

48

He-tao-yuan Formation

Ho-tao-yuan Formation

Late Eocene

36

He-ti Formation

Ho-ti Formation

Late Eocene

31

Hong-he Formation

Hung-ho Formation

Late Eocene

30

Hong-li-shan Formation

Hung-li-shan Formation

Late Eocene

30

—

Houldjin Formation

Middle Oligocene

44

—

Hsanda Gol Formation

Middle Oligocene

46

Hu-gang Formation

Hu-kang Formation

Cretaceous

36

Fluang-gang Formation

Huang-kang Formation

Miocene

37

Hun-shui-he Formation

Hun-shui-ho Formation

Late Eocene

33

—

Irdin Manha Formation

Late Eocene

24

Ji-yuan Formation

Chi-yuan Formation

Late Eocene

33

Lan-ni-keng Member

Lan-ni-keng Member

Late Paleocene

15

Li-jiang Formation

Li-kiang Formation

?Eocene

43

Li-shi-gou Formation

Li-shih-kou Formation

Late Eocene

35

Lian-kan Formation

Lien-kan Formation

Late Eocene

28

Lin-jiang Formation

Lin-kiang Formation

?01igocene

38

Ling-cha Formation

Ling-cha Formation

Early Eocene

38

Ling-shan locality

Ling-shan locality

Late Eocene

30

Liu-niu Formation

Liu-niu Formation

? Eocene

40

Lu-mei-yi Formation

Lu-mei-yi Formation

Late Eocene

41

Lu-se Formation

Lu-se (Green) Formation

Late Eocene

30

Lu-shi Formation

Lu-shi Formation

Late Eocene

34

Mao-jia-po Formation

Mao-chia-po Formation

Late Eocene

35

Meng-yin Series

Meng-yin Series

Cretaceous

31

Na-duo Formation

Na-duo Formation

Late Eocene

40

Nan-chao Formation

Nan-chao Formation

Cretaceous

33

Nan-xiong Group

Nan-hsiung Group

Cretaceous

13

Nao-mu-gen Formation

No-mo-gen Formation

Late Paleocene

77

—

Naran Bulak Formation

Early Eocene

Nau-gon-dai Formation

Nau-gon-dai Formation

Middle Oligocene

45

Ning-jia-shan Member

Ning-shia-shan Member

Early Eocene

38

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

77

Stratigraphic Formation Index — Continued.

Pin-yin English Age Page

Niu-shan Member Niu-shan Member Early Eocene 30

Nong-shan Formation Nung-shan Formation Late Paleocene 13

Ping-hu Formation Ping-hu Formation Early Eocene 15,38

Qing-feng-qiao Formation Ching-feng-chiao Formation Cretaceous 38

Qing-shui-ying Formation Ching-shui-ying Formation Middle Oligocene 47

Qiu-ba Formation Chiu-pa Formation Cretaceous 20

Ren-cun Member Jen-tsun Member Late Eocene 31

San-sheng-kung Saint Jacques Middle Oligocene 45

— Shara Murun Formation Late Eocene 27

Shang-hu Formation Shang-hu Formation Early-Middle Paleocene 13

— Shar-gal-tein Late Oligocene 49

Shi-san-jian-fang Formation Shih-san-chien-fang Formation Early Eocene 28

Shi-zi-kou Formation Shih-tze-kou Formation Early-Middle Paleocene 15

Shih-chiang-tzu-ku Shih-chiang-tzu-ku Late Oligocene 49

Shih-ehr-ma-cheng Shih-ehr-ma-cheng ?Early Oligocene 50

Shuang-ta-si Group Shuang-ta-ssu Group Late Paleocene 20

Shun-shan-ji Formation Shun-shan-chi Formation Paleocene 37

Su-ba-shi Formation Su-pa-shih Formation Cretaceous 20

— Tabenbuluk Late Oligocene 48

Tai-zi-cun Formation Tai-tzu-tsun Formation Late Paleocene 20

Tan-tou Formation Tan-tou Formation Early Eocene 20, 33

Tao-shu-yuan-zi Group Tao-shu-yuan-tze Group Oligocene 48

— Ulan Gochu Formation Early Oligocene 44

U-lun-gur Formation Wu-lun-ku Formation Late Eocene 30

— Urtyn Obo Formation Early Oligocene 44

Wang-he Formation Wang-ho Formation Cretaceous 17

Wang-hu-dun Formation Wang-hu-tun Formation Early-Middle Paleocene 17

Wang-wu Member Wang-wu Member Late Paleocene 15

Wu-li-dui Formation Wu-li-tui Formation Late Eocene 35

Wu-tao-ya-yu Wu-tao-ya-yu Late-Middle Oligocene 49

Wu-tu Formation Wu-tu Formation Early Eocene 30

Xiang-cheng Group Hsiang-cheng Group Paleocene-Eocene 33

Xiang-shan Formation Hsiang-shan Formation Late Eocene 43

Xiao-tun Formation Hsiao-tun Formation Early Oligocene 41,51

Xin-yu Group Sin-yu Group Eocene 38

Xuan-nan Formation Hsuan-nan Formation Cretaceous 20

— Yindirte Late Oligocene 48

Yu-huang-ding Formation Yu-huang- ting Formation Early Eocene 36

Yuan-qu Series (Group) Yuan-chu Series (Group) Late Eocene-Oligocene 31

Zhai-li Member Chai-li Member Late Eocene 31

Zhang-shan-ji Formation Chang-shan-chi Formation Early Eocene 37

Zao-shi Formation Zhao-shih Formation Middle Paleocene 17

Zhu-gui-keng Member Chu-kuei-keng Member Late Paleocene 13

78

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

CHAPTER 4. -THE BIBLIOGRAPHY

1. Anderson, J- G. 1923. Essays on the Cenozoic of North
China. Mem. Geol. Surv. China, ser. A, no. 3 (both in
English and Chinese).

2. Berkey, C. P., AND W. Granger. 1923. Later sediments
of the desert basins of Central Mongolia. Amer. Mus. No-
vitates, 77:1-16.

3. Berkey, C. P., W. Granger, and F. K. Morris. 1929.
Additional new formations in the late sediments of Mon-
golia. Amer. Mus. Novitates, 385:1-12.

4. Berkey, C. P., AND F. K. Morris. 1927. Geology of Mon-
golia. Amer. Mus. Nat. Hist., 475 pp.

5. Bohlin, B. 1937. Oberoligozane Saugetiere aus dem Shar-
galtein-Tal (Western Kansu). Pal. Sin., n.s., C, 3:1-66.

6. . 1942. The fossil mammals from the Tertiary

deposit of Taben-buluk, West Kansu. Pt. I, Insectivora and
Lagomorpha. Pal. Sin., n.s., C, 8a: 1-1 13.

7. . 1946. The fossil mammals from the Tertiary

deposit of Taben-buluk, West Kansu. Pt. II, Simpliciden-
tata, . . . Pal. Sin., n.s., C, 8b: 1-259.

8. . 1951. Some mammalian remains from Shih-ehr-

ma-cheng, Hui-hui-pu Area, Western Kansu. Sino-Swedish
Expedition Pubis., 35:1-46.

9. . 1953. Fossil reptiles from Mongolia and Kansu.

Sino-Swedish Expedition Pubis., 37:1-1 13.

10. Burke, J. J. 1941. New fossil Leporidae from Mongolia.
Amer. Mus. Novitates, 1117:1-23.

11. Chang Yu-ping, Huang Wan-po, Tang Ying-jun, Chi
Hung-giang, You Yu-zhu, Tung Yung-sheng, Ting
Su-YiN, Huang Xue-shin, and Cheng Chia-chien. 1978.
The Cenozoic of Lantian Area, Shensi. Mem. Inst. Vert.
Pal. Paleoanthrop., ser. A, no. 14: 1-64. Science Press, Peking.
(In Chinese.)

12. Chang Yu-ping, AND Tung Yung-sheng. 1963. On the
age of the Redbeds of Yuanshi Basin, Kiangsi. Vert.
PalAsiat., 7(2): 177-181. (In Chinese, with English sum-
mary.)

13. . 1963. Subdivision of “Redbeds” of Nanhsiung

Basin, Kwangtung. Vert. PalAsiat., 7(3):249-262. (In
Chinese, with English summary.)

14. Cheng Chia-chien, Tang Ying-jun, Chiu Chan-siang,
AND Yeh Hsiang-kuei. 1973. Notes on the Upper Cre-
taceous-Lower Tertiary of the Nan-hsiung Basin, N.
Kwangtung. Vert. PalAsiat., 1 1 ( 1 ): 1 8-30. (In Chinese, with
English summary.)

15. Chia Lan-po, Chang Yu-ping, Huang Wan-po, Tang
Ying-jun, Chi Hung-giang, You Yu-zhu, Ting Su-yin,
AND Huang Xue-shi. 1966. The Cenozoic of Lantian,
Shensi. Pp. 1-31, in Monograph of Field Conference of
Lantian Cenozoic (IVPP ed.). Science Press, Peking. (In
Chinese.)

16. Chiu Chan-siang. 1962. Giant rhinoceros from Loping,
Yunnan, and discussion on the taxonomic characters of
I ndricotherium grangeri. Vert. PalAsiat., 6(1):58-71. (In
Chinese, with English summary.)

17. . 1965. First discovery of Lop/t/owcryx in China.

Vert. PalAsiat., 9(4):395-397. (In Chinese, with English
summary.)

18. . 1973. A new genus of giant rhinoceros from Oli-

gocene of Dzungaria, Sinkiang. Vert. PalAsiat., 1 1(2): 182-
191. (In Chinese, with English summary.)

19. Chow Min-chen. 1953. Note on the age of Changsintien
Gravels. Acta Pal. Sin., l(4):201-205. (In Chinese.)

20. . 1956. The new formation of Eocene in South

China. Di-Zhi-Zhi-Shi, 4:19-20. (In Chinese.)

21. . 1957. Mammalian fauna and correlation of Ter-

tiary and Early Pleistocene of South China. Ke-Xue-tong-
bao, 13:394-399. (In Chinese.)

22. . 1957. Notes on some mammalian fossils from

the Late Cenozoic of Sinkiang. Vert. PalAsiat., 1(1):33^1.
(In English, with Chinese summary.)

23. . 1957. On some Eocene and Oligocene mammals

from Kwangsi and Yunnan. Vert. PalAsiat., l(3):201-204.
(In English, with Chinese summary.)

24. . 1957. A new Coryp/iotfon from Sintai, Shantung.

Vert. PalAsiat., l(4):301-304. (In English, with Chinese
summary.)

25. . 1958. Mammalian faunas and correlation of

Tertiary and Early Pleistocene of South China. J. Paleontol.
Soc. India, Birbal Sahni Memorial, 3: 1 23-130. (In English.)

26. . 1958. fsocntc/oifow— a new primitive entelodont

from the Eocene of Lunan, Yunnan. Vert. PalAsiat., 2(1):
30-36. (In English, with Chinese summary.)

27. . 1958. Some Oligocene mammals from Lunan,

Yunnan. Vert. PalAsiat., 2(4):263-268. (In Chinese, with
English summary.)

28. . 1958. New material of Tertiary mammals from

Sinkiang. Vert. PalAsiat., 2(4):289-294. (In Chinese, with
English summary.)

29. . 1959. A new arctocyonid from the Upper Eocene

of Lushih, Honan. Vert. PalAsiat., 3(3):133-138. (In En-
glish, with Chinese summary.)

30. . 1959. The discovery of the Eocene vertebrate

fossils from Sinyu, Kiangsi. Palaeovert. et Palaeoanthrop.,
l(2):79-80. (In Chinese.)

3 1 . . 1959. A record of the earliest sabre-toothed cats

from the Eocene of Lushih, Honan. Science Record, 2(10):
347-349. Science Press, Peking. (In English.)

32. . 1960. Discovery of the Paleocene mammal in

T urfan Basin and a summary of Cenozoic mammalian hori-
zons of Sinkiang. Acta Pal. Sin., 8(2): 155-158. (In Chinese,
with English summary.)

33. . 1960. Prodinoceras and a summary of mam-

malian fossils of Sinkiang. Vert. PalAsiat., 4(2):99-102. (In
English.)

34. . 1961. A new tarsioid primate from the Lushih

Eocene, Honan. Vert. PalAsiat., 5(1): 1-5. (In Chinese, with
English summary.)

35. (Min). 1961. The discovery of Eocene mam-

malian fossils of Chu-yang, Hopei. Vert. PalAsiat., 5(3):286.
(In Chinese.)

36. . 1962. A new species of primitive chalicothere

from the Tertiary of Lunan, Yunnan. Vert. PalAsiat., 6(3):
219-224. (In Chinese, with English summary.)

37. . 1963. Tillodont materials from Eocene of Shan-

tung and Honan. Vert. PalAsiat., 7(2):97-104. (In Chinese,
with English summary.)

38. . 1963. A xenarthran-like mammal from the Eocene

of Honan. Scientia Sinica, 12(1 2): 1889-1 893. (In English.)

39. . 1964. A lemuroid Primates from the Eocene of

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

79

Laotian, Shensi. Vert. PalAsiat., 8(3):257-262. (In Chinese
and English.)

40. . 1965. Mesonychids from the Eocene of Honan.

Vert. PalAsiat., 9(3):286-29 1 . (In Chinese, with English
summary.)

41. . 1975. Some carnivores from the Eocene of China.

Vert. PalAsiat., 13(3): 165-1 68. (In Chinese, with English
summary.)

42. . 1979. Tertiary mammalian faunas of the Laotian

District, Shensi. Prof Papers Strat. Pal., 7:98-108. Geol.
Publ. House, Peking. (In Chinese.)

43. Chow Min-chen, Chang Yu-ping, Wang Ban-yue, and
Ting Su-YiN. 1973. New mammalian genera and species
from the Paleocene of Nanhsiung, N. Kwantung. Vert.
PalAsiat., il(l):31-35. (In Chinese, with English sum-
mary.)

44. Chow Min-chen, Chang Yu-ping, and Ting Su-yin.
1974. Some Early Tertiary Perissodactyla from Lunan
Basin, E. Yunnan. Vert. PalAsiat., 12(4):262-278. (In
Chinese, with English summary.)

45. Chow Min-chen AND Chiu Chan-siang. 1962. The Ter-
tiary and Lower Quaternary of E. Yunnan. 9th Annual
Conf & 2nd Natl. Cong. Pal. Soc. China, Abstracts of Papers,
p. 62. Peking. (In Chinese.)

46. . 1963. New genus of giant rhinoceros from Oli-

gocene of Inner Mongolia. Vert. PalAsiat., 7(3):230-239.
(In Chinese, with English summary.)

47. . 1964. An Eocene giant rhinoceros. Vert. Pal-

Asiat., 8(3):264-268. (In Chinese, with English summary.)

48. Chow Min-chen, and Hu Chang-kang. 1956. The
occurrence of the Paleogene mammal in Sinkiang. Acta Pal.
Sin., 4(2):239-242. (In Chinese, with English summary.)

49. Chow Min-chen, AND Hu Cheng-zhi. 1959. A new species
of Parabrontops from the Oligocene of Lunan, Yunnan.
Acta Pal. Sin., 7(2):85-88. (In Chinese, with English sum-
mary.)

50. Chow Min-chen, AND Li Chuan-kuei. 1963. A fossil of
Homogalax from the Eocene of Shantung. Scientia Sinica,
12(9): 141 1-1412. (In English.)

51. . 1965. Homogalax and Heptodon of ShanXung.

Vert. PalAsiat., 9(1): 15-21. (In Chinese, with English sum-
mary.)

52. Chow Min-chen, Li Chuan-kuei, and Chang Yu-ping.
1973. Late Eocene mammalian faunas of Honan and Shansi
with notes on some vertebrate fossils collected therefrom.
Vert. PalAsiat., i 1(2):165-181. (In Chinese, with English
summary.)

53. Chow Min-chen and Qi Tao. 1978. Paleocene mam-
malian fossils from Nomogen Formation of Inner Mon-
golia. Vert. PalAsiat., 16(2):77-85. (In Chinese, with English
summary.)

54. Chow Min-chen, Qi Tao, AND Li Yung. 1976. Paleocene
stratigraphy and faunal characters of Mammalian fossils on
Nomogen Commune, Si-zi-wang-qi, Nei Mongol. Vert.
PalAsiat., 14(4):228-233. (In Chinese, with English sum-
mary.)

55. Chow Min-chen and A. K. Rozhdestvensky. 1960.
Exploration in Inner Mongolia. Vert. PalAsiat., 4(1): 1-10.
(In English.)

56. Chow Min-chen and Tung Yung-sheng. 1962. Notes
on some new uintathere materials of China. Vert. PalAsiat.,
6(4):368-374. (In Chinese, with English summary.)

57. . 1965. A new coryphodont from the Eocene of

Sinyu, Kiangsi. Vert. PalAsiat., 9(1):1 14-121. (In Chinese,
with English summary.)

58. Chow Min-chen AND Wang Ban-yue. 1978. A new pan-
todont genus from the Paleocene of South China. Vert.
PalAsiat., 16(2):86-90. (In Chinese, with English sum-
mary.)

59. . 1979. Relationship between the pantodonts and

tillodonts and classification of the Order Pantodonta. Vert.
PalAsiat., 17(l):37-48. (In Chinese, with English sum-
mary.)

60. Chow Min-chen and Xu Yu-xuan. 1959. Indhcothe-
rium from Hami Basin, Sinkiang. Vert. PalAsiat., 3(2):93-
98. (In English.)

61. . 1961. New primitive true rhinoceroses from the

Eocene of Iliang, Yunnan. Vert. PalAsiat., 5(4):29 1-305.
(In Chinese, with English summary.)

62. . 1965. Amynodonts from the Upper Eocene of

Honan and Shansi. Vert. PalAsiat., 9(2): 190-204. (In
Chinese, with English summary.)

63. Chow Min-chen, Xu Yu-xuan, and Zhen Shuo-nan.
1964. Amynodon from the Eocene of Lunan, Yunnan. Vert.
PalAsiat., 8(4):355-361. (In Chinese, with English sum-
mary.)

64. Chow Min-chen AND Zheng Jia-jian. 1980. The mam-
mal-bearing Early Tertiary horizons of China. PaleoBios
(Berkeley), 32:1-7.

65. Colbert, E. H. 1934. Chalicotheres from Mongolia and
China in the American Museum. Bull. Amer. Mus. Nat.
Hist., 67(8):353-387.

66. . 1938. Fossil mammals from Burma in the Amer-

ican Museum of Natural History. Bull. Amer. Mus. Nat.
Hist., 74(6):255^36.

67. Dashzeveg, D., and M. C. McKenna. 1977. Tarsoid
primate from the Early Tertiary of the Mongolian People’s
Republic. Acta Pal. Polonica, 22(2):1 19-137.

68. Dawson, M. R. 1964. Late Eocene rodents (Mammalia)
from Inner Mongolia. Amer. Mus. Novitates, 2191:1-15.

69. . 1968. Oligocene rodents (Mammalia) from East

Mesa, Inner Mongolia. Amer. Mus. Novitates, 2324:1-12.

70. Dawson, M. R., Li Chuan-kuei, and Qi Tao. [In press].
Eocene ctenodactyloid rodents (Mammalia) of Eastern and
Central Asia. Special Paper, Carnegie Museum of Natural
History.

71. Defense Mapping Agency, Washington, D.C. 1979.
Gazetteer of the People’s Republic of China (Pinyin to Wade-
giles, Wade-giles to Pinyan). Washington, D.C., 919 pp.

72. Ding Su-yin. 1979. A new edentate from the Paleocene
of Guang-dong. Vert. PalAsiat., 17(l):57-64. (In Chinese,
with English summary.)

73. Ding Su-yin and Tong Yong-sheng. 1979. Some Paleo-
cene anagalids from Nanxiong, Guangdong. Vert. PalAsiat.,
17(2): 137-145. (In Chinese.)

74. Ding Su-yin and Zhang Yu-ping. 1979. The insectivore
and anagalids of Chijiang Basin, Jiangxi. Pp. 354-359, in
The Mesozoic and Cenozoic Red Beds of South China,
Science Press, Beijing, 432 pp. (In Chinese.)

75. Ding Su-yin, Zheng Jia-jian, Zhang Yu-ping, and Tong
Yong-sheng. 1977. The age and characteristic of the
Liuniu and the Dongjun faunas, Baise Basin of Guangxi.
Vert. PalAsiat., 15(1):35^5. (In Chinese, with English
summary.)

80

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

76. Flerov, K. K. 1957. A new coryphodont from Mongolia,
and on evolution and distribution of pantodonts. Vert.
PalAsiat., 1(2);73-81.

77. Gao Hung-hsiang. 1975. Paleocene mammal-bearing
beds of Chaling Basin, Hunan. Vert. PalAsiat., 13(2):89-
95. (In Chinese, with English summary.)

78. Gao Yu. 1976. Eocene vertebrate localities and horizons
of Wucheng and Xichuan basins, Henan. Vert. PalAsiat.,
14(l):26-34. (In Chinese.)

79. Gazin, C. L. 1953. The Tillodontia. Smithsonian Misc.
Coll., 121(10):226.

80. Gingerich, P. D. 1976. Systematic position of the alleged
primate Lantianius xiehuensis Chow, 1 964, from the Eocene
of China. J. Mamm., 57(1):194-198.

81. Granger, W. 1938. A giant oxyaenid from the Upper
Eocene of Mongolia. Amer. Mus. Novitates, 969:1-5.

82. Granger, W., and C. P. Berkey. 1922. Discovery of
Cretaceous and older Tertiary strata in Mongolia. Amer.
Mus. Novitates, 42:1-7.

83. Granger, W., and W. K. Gregory. 1934. An apparently
new family of amblypod mammals from Mongolia. Amer.
Mus. Novitates, 720:1-8.

84. . 1935. A revised restoration of the skeleton of

Baliichithehum. gigantic fossil rhinoceros of Central Asia.
Amer. Mus. Novitates, 787:1-3.

85. . 1936. Further notes on the gigantic extinct rhi-

noceros, Bahichitherium, from the Oligocene of Mongolia.
Bull. Amer. Mus. Nat. Hist., 72(1): 1-73.

86. . 1938. A new titanothere genus from the Upper

Eocene of Mongolia and North America. Bull. Amer. Mus.
Nat. Hist., 74(6):435^36 (Appendum).

87. . 1943. A revision of the Mongolian titanotheres.

Bull. Amer. Mus. Nat. Hist., 80(10):349-389.

88. Hartenberger, J.-L. 1980. Donnees et Hypotheses sur
la Radiation Initiale des Rongeurs. Palaeovertebrata, Mem.
Jubil. Hommage a Rene Lavocat, Montpellier, pp. 285-
302.

89. Hartenberger, J.-L., J. Sudre, and M. Vianey-Liaud.
1975. Les Mammiferes de L’Eocene Superieur de Chine
(Gisement de River Section); Leur Place dans L’Historie
des Faunas Eurasiatiques. 3e Reunion Annuelle des Sci-
ences de la Terre, p. 186.

90. Hu Chang-kang. 1959. On some Tertiary chalicotheres
of North China. Palaeovertebrata et Palaeoanthropologia,
1(3): 125-1 32. (In Chinese.)

9 1 . . 1961. The occurrence of Parabrontops in Hami,

Siankiang. Vert. PalAsiat., 5(l):41-42. (In Chinese, with
English summary.)

92. . 1962. Cenozoic mammalian fossil localities in

Kansu and Ningshia. Vert. PalAsiat., 6(2):162-172. (In
Chinese, with English summary.)

93. . 1963. A new Eocene anthracothere. Vert.

PalAsiat., 7(4):310-317. (In Chinese, with English sum-
mary.)

94. . 1964. Archaeotheriurn ordosius from Oligocene

Inner Mongolia. Vert. PalAsiat., 8(3):3 1 2-3 1 5. (In Chinese,
with English summary.)

95. Huang XuE-SHi. 1977. zlrc/iaco/awWa fossils from Anhui.
Vert. PalAsiat., 1 5(4):249-260. (In Chinese.)

96. . 1978. Paleocene Pantodonta of Anhui. Vert.

PalAsiat., 16(4):275-281. (In Chinese.)

97. Ideker, j., and Yan De-fa. 1980. Lestes (Mammalia) a

junior homonym of Lestes (Zygoptera). Vert. PalAsiat.,
18(2): 138-1 40. (In Chinese and English.)

98. Jiang Yuan-ji, Wang Bao-liang, and Qi Tag. 1976.
Stratigraphy of the Early Oligocene Chaganbulage Forma-
tion, Haosibuerdu Basin, Ningxia. Vert. PalAsiat., 14(1):
35-41. (In Chinese.)

99. Kowalsk.1, K. 1974. Middle Oligocene rodents from
Mongolia. Pal. Polonica, 30:147-178.

100. Kjelan-Jaworowska, Z. 1969. Archaeolambdidae Fle-
rov (Pantodonta) from the Paleocene of the Nemegt Basin,
Gobi Desert. Pal. Polonica, 19:133-140.

101. Kretzoi, M. 1942. Auslandische Saugetierefossilien der
Ungarischen Museen. Foldtani Kozloni, 72(1 -3): 140-1 48.

102. LeeYuen-yen. 1938. Some new fossil localities in Eastern
Tsinling. Bull. Geol. Soc. China, 18(3-4):227-239. (In
English.)

103. . 1938. The Early Tertiary deposits of the Y uanchii

Basin on the Honan-Shansi border. Bull. Geol. Soc. China,
18(3-4):24 1-257. (In English.)

104. Li Chuan-kuei. 1957. The discovery of Paleogene mam-
malian fossils of Lushi, Honan. Vert. PalAsiat., 1(3):265-
266. (In Chinese and English.)

105. . 1962. Some notes on Cenozoic vertebrate hori-

zons of Shantung. 32nd Ann. Conf. Geol. Soc. China,
Abstr. Papers, 1:26. (In Chinese.)

106. . 1963. Paramyid and sciuravids from North China.

Vert. PalAsiat., 7(2): 15 1-1 60. (In Chinese, with English
summary.)

107. . 1965. Eocene leporids of North China. Vert.

PalAsiat., 9(l):23-36. (In Chinese and English.)

1 08. . 1975. Yuomys, a new ischyromyoid rodent genus

from the Upper Eocene of North China. Vert. PalAsiat.,
13(l):58-70. (In Chinese, with English summary.)

109. . 1977. Paleocene eurymyloids (Anagalida, Mam-

malia) ofQianshan, Anhui. Vert. PalAsiat., 15(2): 103-1 18.
(In Chinese, with English summary.)

no. Li Chuan-kuei, Chiu Chan-siang, Yan De-fa, and Hsieh
Shu-hua. 1979. Notes on some Early Eocene mammalian
fossils of Hengtung, Hunan. Vert. PalAsiat., 17(!):71-82.
(In Chinese, with English summary.)

111. Lindsay, E. H. 1978. Eucricetodon asiaticus {MdiWht'N
and Granger), an Oligocene rodent (Cricetidae) from Mon-
golia. J. Paleontol., 52(3):590-595.

112. Liu Hsien-ting ET AL., 1963. The field investigations of
the Mesozoic and Cenozoic of Chi-yuan, Honan. Unpub-
lished manuscript.

1 13. McKenna, M. C. 1963. New evidence against tupaioid
affinities of the mammalian Family Anagalidae. Amer. Mus.
Novitates, 2158:1-16.

114. McKenna, M. C., AND C. P. Holton. 1967. Anewinsec-
tivore from the Oligocene of Mongolia and a new subfamily
of hedgehogs. Amer. Mus. Novitates, 231 1:1-12.

115. Matthew, W. D., AND W. Granger. 1923. The fauna of
the Houldjin Gravels. Amer. Mus. Novitates, 97:1-6.

1 16. . 1923. The fauna of the Ardyn Obo Formation.

Amer. Mus. Novitates, 98:1-5.

117. . 1923. New Bathyergidae from the Oligocene of

Mongolia. Amer. Mus. Novitates, 101:1-5.

1 18. . 1923. Nine new rodents from the Oligocene of

Mongolia. Amer. Mus. Novitates, 102:1-10.

1 19. . 1924. New Carnivora from the Tertiary of Mon-

golia. Amer. Mus. Novitates, 104:1-9.

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

120. . 1924. New Insectivora and ruminants from the

Tertiary of Mongolia. Amer. Mus. Novitates, 105:1-7.

121. . 1925. Fauna and correlation of the Gashato For-

mation of Mongolia. Amer. Mus. Novitates, 189:1-12.

1 22. . 1 925. New creodonts and rodents from the Ardyn

Obo Formation of Mongolia. Amer. Mus. Novitates, 193:
1-7.

123. . 1925. New ungulates from the Ardyn Obo For-

mation of Mongolia. Amer. Mus. Novitates, 195:1-12.

124. . 1925. New mammals from the Shara Murun

Eocene of Mongolia. Amer. Mus. Novitates, 196:1-1 1.

125. . 1925. New mammals from the Irdin Manha

Eocene of Mongolia. Amer. Mus. Novitates, 198:1-10.

126. . 1925. The smaller perissodactyls of the Irdin

Manha Formation, Eocene of Mongolia. Amer. Mus. No-
vitates, 199:1-9.

127. . 1926. Two new perissodactyls from the Arshanto

Eocene of Mongolia. Amer. Mus. Novitates, 208:1-5.

128. Matthew, W.D., W. Granger, AND G.G. Simpson. 1928.
Paleocene multituberculates from Mongolia. Amer. Mus.
Novitates, 331:1^.

129. . 1929. Additions to the fauna of the Gashato

Formation of Mongolia. Amer. Mus. Novitates, 376:1-12.

130. Mellett, J. S. 1968. The Oligocene Hsanda Gol For-
mation, Mongolia: a revised faunal list. Amer. Mus. Nov-
itates, 2318:1-16.

131. Muizon, Christian DE. 1977. Revision des Lagomorphes
des Couches a Baluchitherium (Oligocene Superieur) de
San-tao-lo (Ordos, Chine). Bull. Mus. Natl. Hist. Nat. (Paris),
3e ser., no. 488 (Sci. Terre 65):265-294.

132. Osborn, H. F. 1923. Baluchitherium grangeri, a giant
hornless rhinoceros from Mongolia. Amer. Mus. Novitates,
78:1-15.

133. . 1923. Titanotheres and lophiodonts in Mongo-

lia. Amer. Mus. Novitates, 91:1-5.

134. . 1923. Cadurcotherium from Mongolia. Amer.

Mus. Novitates, 92:1-2.

135. . 1924. £Mr/inocera5, Upper Eocene amblypod of

Mongolia. Amer. Mus. Novitates, 145:1-5.

136. . 1924. TnJrf’wrarc/tws, giant mesonychid of Mon-

golia. Amer. Mus. Novitates, 146:1-5.

137. . 1924. Cadurcotherium ardynense. Oligocene,

Mongolia. Amer. Mus. Novitates, 147:1^.

138. . 1925. Upper Eocene and Lower Oligocene titano-

theres of Mongolia. Amer. Mus. Novitates, 202:1-12.

139. . 1929. Embolotherium, gen. nov., of the Ulan

Gochu, Mongolia. Amer. Mus. Novitates, 353:1-20.

140. . 1936. Amynodon mongoliensis irom \hc \]x>x>c:r

Eocene of Mongolia. Amer. Mus. Novitates, 859:1-9.

141. Osborn, H. F., and W. Granger. 1931. Coryphodonts
of Mongolia, Eudinoceras mongoliensis Osborn, E. kho-
lobolchiensis sp. nov. Amer. Mus. Novitates, 459:1-13.

142. . 1932. Coryphodonts and uintatheres from the

Mongolian Expedition of 1930. Amer. Mus. Novitates, 552:
1-16.

143. Pei Wen-chung, Chow Min-chen, and Cheng Chia-
CHiEN. 1963. Cenozoic of China. Edited by the Committee
of Stratigraphy of China. Science Press, Peking, 3 1 pp. (In
Chinese.)

144. Pen Shi-ling. 1975. Cenozoic vertebrate localities and
horizons of Dzungaria Basin, Sinkiang. Vert. PalAsiat., 13(3):
185-189. (In Chinese.)

145. Pilgrim, G. E. 1928. Artiodactyla of the Eocene of Burma.
Pal. Indica, n.s., 8:1-39.

146. Qi Tao. 1975. An Early Oligocene mammalian fauna of
Ningxia. Vert. PalAsiat., 1 3(4):2 1 7-224. (In Chinese, with
English summary.)

147. . 1979. A general account of the Early Tertiary

mammalian faunas of Shara Murun area. Inner Mongolia.
A separated paper of 2nd Congress of Stratigraphy, China,
Beijing, pp. 1-9. (In Chinese, with English summary.)

148. . 1980. Irdin Manha Upper Eocene and its mam-
malian fauna at Huhebolhe Cliff in Central Inner Mongolia.

Vert. PalAsiat., 18(l):28-32. (In Chinese, with English
summary.)

149. . 1980. A new Eocene lophialetid genus of Inner

Mongolia. Vert. PalAsiat., 1 8(3):2 1 3-2 1 9. (In Chinese, with
English summary.)

150. Qiu Zhan-xiang. 1977. New genera of Pseudictopidae
(Anagalida, Mammalia) from Middle-Upper Paleocene of
Qianshan, Anhui. Acta Pal. Sin., 1 6( 1 ): 1 28- 1 48. (In Chinese,
with English summary.)

151. Qiu Zhan-xiang AND Li Chuan-kuei. 1972. New finds
of Paleocene mammalian fossil from Anhui. “Fossil,” p.
24. (In Chinese.)

152. . 1977. Miscellaneous mammalian fossils from the

Paleocene of Qianshan Basin, Anhui. Vert. PalAsiat., 1 5(2):
94-102. (In Chinese.)

153. Qiu Zhan-xiang, Li Chuan-kuei, Huang Xue-shi, Tang
Ying-jun, Xu Qin-qi, Van De-fa, andZhang Hong. 1977.
Continental Paleocene stratigraphy of Qianshan and Xuan-
cheng Basin, Anhui. Vert. PalAsiat., 15(2):85-93. (In
Chinese.)

154. Qiu Zhu-ding. 1977. Note on the new species of Tn/Zira-
cokeryx from Guangxi. Vert. PalAsiat., 15(l):54-58. (In
Chinese.)

155. . 1978. Late Eocene hypertragulids of Baise Basin,

Kwangsi. Vert. PalAsiat., 16(1):7-12. (In Chinese, with
English summary.)

1 56. Radinsky, L. B. 1 964. Paleomoropus, a new Early Eocene
chalicothere (Mammalia, Perissodactyla), and a revision of
Eocene chalicotheres. Amer. Mus. Novitates, 2179:1-28.

157. . 1964. Notes on Eocene and Oligocene fossil

localities in Inner Mongolia. Amer. Mus. Novitates, 2180:
1-11.

158. . 1965. Early Tertiary Tapiroidea of Asia. Bull.

Amer. Mus. Nat. Hist., 129(2): 18 1-263.

159. . 1967. A review of the rhinocerotoid family Hyra-

codontidae (Perissodactyla). Bull. Amer. Mus. Nat. Hist.,
136(1):1^5.

160. Romer, A. S. 1966. Vertebrate Paleontology. Univ. Chi-
cago Press (3rd ed.), 468 pp.

161. Russell, D. E. 1967. Le Paleocene Continental D’Ame-
rique du Nord. Mem. Mus. Natl. Hist. Nat., n.s. C, Tom
16(2): 1-99.

161a. bHiKAMA, I. 1941. Vertebrate fossils of the Chinese con-
tinent. Contrib. Inst. Geol. Pal., Tohoku Imp. Univ., 36:
1-103.

162. Simpson, G. G. 1931. A new insectivore from the Oli-
gocene, Ulan Gochu horizon, of Mongolia. Amer. Mus.
Novitates, 505:1-22.

163. . 1945. The principles of classification and a clas-

sification of mammals. Bull. Amer. Mus. Nat. Hist., 85:1-
350.

82

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

164. South China “Red beds” Research Group, IVPP. 1977.
Paleocene vertebrate horizons and mammalian faunas of
South China. Scientia Sinica, 20(5):665-678. (In English.)

165. SuLiMSKi, A. 1969. Paleocene genus Pseudictops Mat-
thew, Granger and Simpson, 1 929 (Mammalia) and its revi-
sion. Pal. Polonica, 19:101-129.

166. Sych, L. 1971. Mixodontia, a new Order of Mammalia
from the Paleocene of Mongolia. Pal. Polonica, 25:147-
158.

167. . 1975. Lagomorpha from the Oligocene of Mon-

golia. Pal. Polonica, 33:183-200.

168. SzALAY, F. S., AND S. J. GouLD. 1966. Asiatic Mesony-
chidae (Mammalia, Condylarthra). Bull. Amer. Mus. Nat.
Hist., 132(2): 1-1 73.

169. SzALAY, F. S., AND M. C. McKenna. 1971. Beginning of
the Age of Mammals in Asia: the Late Paleocene Gashato
Fauna, Mongolia. Bull. Amer. Mus. Nat. Hist., 144(4):271-
317.

170. Tan, H.C. 1923. New research on the Mesozoic and Early
Tertiary geology in Shantung. Bull. Geol. Surv. China, no.
5(2). (In Chinese and English.)

171. Tang Xin and Chow Min-chen. 1964. A review of ver-
tebrate-bearing Lower Tertiary of South China. Vert.
PalAsiat., 8(2):1 19-133. (In Chinese, with English sum-
mary.)

172. Tang H. AND Chow M. C. 1965. The vertebrate-bearing
Early Tertiary of South China: a review. Intemat. Geol.
Review, 7(8): 1338-1 352.

173. Tang Ying-jun. 1978. Two new genera of Anthraco-
theriidae from Kwangsi. Vert. PalAsiat., 16(1): 13-21. (In
Chinese.)

174. . 1978. New materials of Oligocene mammalian

fossils from Qujing Basin, Yunnan. Prof Papers of Strat.
Pal., 7:75-79. Geol. Publ. House, Peking. (In Chinese.)

175. Tang Ying-jun AND Yan De-fa. 1976. Notes on some
mammalian fossils from the Paleocene of Qianshan and
Xuancheng, Anhui. Vert. PalAsiat., 14(2):91-99. (In
Chinese.)

176. Tang Ying-jun, You Yu-zhu, Xu Qin-qi, Qiu Zhu-ding,
AND Hu Yan-kun. 1974. The Lower Tertiary of the Baise
and Yungle basins, Kwangsi. Vert. PalAsiat., 12(4):279-
292. (In Chinese, with English summary.)

177. Teilhard de Chardin, P. 1926. Mammiferes Tertiaires
de Chine et de Mongolie. Ann. Pal., 15:1-51 (Paris).

178. . 1930. On the occurrence of a Mongolian Eocene

Perissodactyla in the Red Sandstone of Hsichuan, S. W.
Honan. Bull. Geol. Soc. China, 8:331-332.

179. Teilhard de Chardin, P., G. B. Barbour, and M. N.
Bien. 1935. A geological reconnaissance across Easting
Tsinling. Bull. Geol. Surv. China, no. 25.

180. Teilhard DE Chardin, P., AND P. Leroy. 1942. Chinese
fossil mammals— a complete bibliography, analysed, tab-
ulated, annotated and indexed. Inst. Geo. -Biol., Pekin, 8:
1-142.

181. Teilhard de Chardin, P., and C. C. Young. 1936. A
Mongolian Amblypoda in the Red Beds of Ichang (Hupei).
Bull. Geol. Soc. China, 1 5(2):2 1 7-224.

182. Tong Yong-sheng. 1978. Late Paleocene mammals of
Turfan Basin, Sinkiang. Mem. Inst. Vert. Pal. Paleoan-
throp., 13:81-101. (In Chinese.)

183. . 1979. A Late Paleocene primate from South

China. Vert. PalAsiat., 17(l):65-70. (In Chinese, with
English summary.)

184. . 1979. The new materials of archaeolambds from

South Jiangxi. Pp. 377-38 1 , in The Mesozoic and Cenozoic
Red Beds of South China, Science Press, Beijing, 432 pp.
(In Chinese.)

185. . 1979. Some Eocene uintathere materials of Chi-

jiang Basin, Jiangxi. Pp. 395-399, in The Mesozoic and
Cenozoic Red Beds of South China, Science Press, Beijing,
432 pp. (In Chinese.)

186. Tong Yong-sheng and Tang Ying-jun. 1977. A new
species of Eudinoceras. Vert. PalAsiat., 15(2): 139-1 42. (In
Chinese.)

187. Tong Yong-sheng and Wang Jing-wen. 1979. Some
new materials of the Paleocene and Eocene mammals from
W. Henan. Vert. PalAsiat., 17(2): 176. (In Chinese.)

188. . 1979. New observation on the Lower Tertiary of

West Henan. A separated paper of 2nd Congress of Stra-
tigraphy, China, Beijing. (In Chinese, with English sum-
mary.)

189. . 1980. Subdivision of the Upper Cretaceous and

Lower Tertiary of the Tantou Basin, the Lushi Basin and
the Lingbao Basin of W. Henan. Vert. PalAsiat., 18(1):21-
27. (In Chinese, with English summary.)

190. Tong Yong-sheng, Zhang Yu-ping, Wang Ban-yue, and
Ding Su-yin. 1976. The Lower Tertiary of the Nanxiong
and Chijiang basins. Vert. PalAsiat., 14(1): 16-25. (In
Chinese.)

191. Tong Yong-sheng, Zhang Yu-ping, Zheng Jia-jian,
Wang Ban-yue, and Ding Su-yin. 1979. The discussion
of the Lower Tertiary strata and mammalian fauna of Chi-
jiang Basin, Jiangxi. Pp. 400^06, in The Mesozoic and
Cenozoic Red Beds of South China. Science Press, Beijing,
432 pp. (In Chinese.)

192. Trofimov, B. 1958. New Bovidae from the Oligocene of
Central Asia. Vert. PalAsiat., 2:244-247.

192a. Van Valen, L. 1966. Deltatheridia, a new Order of
mammals. Bull. Amer. Mus. Nat. Hist., 132(1):1-126.

193. Wang Ban-yue. 1975. Paleocene mammals of Chaling
Basin, Hunan. Veit. PalAsiat., 1 3(3): 1 54-162. (In Chinese.)

194. . 1976. Late Paleocene mesonychid from Nan-

xiong Basin, Guangdong. Vert. PalAsiat., 14(4):259-262.
(In Chinese, with English summary.)

195. . 1978. Two new miacids from Paleocene of

Nanhsiung, Kwangtung. Vert. PalAsiat., 16(2):91-96. (In
Chinese, with English summary.)

196. . 1978. Perissodactyla from the Late Eocene of

Lantian, Shensi. Prof Papers Strat. Pal., 7:1 18-121. Geol.
Publ. House, Peking. (In Chinese, with English summary.)

197. . 1979. A new species of Harpyodus and its sys-

tematic position. Pp. 366-372, in The Mesozoic and Ceno-
zoic Red Beds of South China, Science Press, Beijing, 432
pp. (In Chinese.)

198. Wang Ban-yue and Ding Su-yin. 1979. Some bema-
lambds from Chijiang Basin, Jiangxi. Pp. 351-353, in The
Mesozoic and Cenozoic Red Beds of South China, Science
Press, Beijing, 432 pp. (In Chinese.)

1 99. Wang Jing-wen. 1976. A new genus of Forstercooperi-
inae from the Late Eocene ofTongbo, Henan. Vert. PalAsiat.,
14(2):104-111. (In Chinese.)

200. . 1978. Fossil Amynodontidae and Ischyromyidae

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

83

of Tongbo, Henan. Vert. PalAsiat., 16(l):22-29. (In
Chinese.)

201. Wang Tze-yi AND Hu Chang-kang. 1963. AnOligocene
mammalian horizon in the Yuan-chii Basin, South Shansi.
Vert. PalAsiat., 7(4):357-360. (In Chinese, with English
summary.)

202. Woo Ju-KANG AND Chow Min-chen. 1957. New mate-
rials of the earliest primate known in China— Hoanghonius
stehlini. Vert. PalAsiat., l(4):267-272. (In English, with
Chinese summary.)

203. Wood, A. E. 1942. Notes on the Paleocene lagomorph,
Eurymylus. Amer. Mus. Novitates, 1 162:1-7.

204. . 1970. The Early Oligocene rodent Ardynomys

(Family Cylindrodontidae) from Mongolia and Montana.
Amer. Mus. Novitates, 2418:1-18.

205. Wood, H. E. 1938. Cooperia totadentata, a rem.arkable
rhinoceros from the Eocene ofMongolia. Amer. Mus. Novi-
tates, 1012:1-20.

206. . 1963. A primitive rhinoceros from the Late

Eocene of Mongolia. Amer. Mus. Novitates, 2146:1-1 1.

207. Wood, H. E. et al. 1941. Nomenclature and correlation
of the North American continental Tertiary. Bull. Geol.
Soc. Amer., 52: 1—48.

208. Xu Qin-qi. 1976. New materials of Anagalidae from the
Paleocene of Anhui. (A). Vert. PalAsiat., 14(3):174-184.
(In Chinese.)

209. . 1976. New materials of Anagalidae from the

Paleocene of Anhui. (B). Vert. PalAsiat., 14(4):242-25 1 .
(In Chinese.)

210. . 1977. Two new genera of old Ungulata from the

Paleocene of Qianshan Basin, Anhui. Vert. PalAsiat., 1 5(2):
1 19-125. (In Chinese.)

21 1. . 1977. New materials of Bothriodon from Bose

Basin of Guangxi. Vert. PalAsiat., 1 5(3):202-206. (In
Chinese.)

212. Xu Yu-xuan. 1961. Some Oligocene mammals from
Chuching, Yunnan. Vert. PalAsiat., 5(4):3 15-329. (In
Chinese, with English summary.)

213. . 1962. Some new anthracotheres from Shansi and

Yunnan. Vert. PalAsiat., 6(3):232-250. (In Chinese, with
English summary.)

214. . 1965. A new genus of amynodont from the Eocene

ofLantian, Shensi. Vert. PalAsiat., 9( 1 ):83-88. (In Chinese,
with English summary.)

215. . 1966. Amynodonts of Inner Mongolia. Vert.

PalAsiat., 10(2):123-190. (In Chinese, with English sum-
mary.)

216. . 1976. Some new forms of Coryphodontidae from

the Eocene of Sichuan, Honan. Vert. PalAsiat., 14(3):185-
193. (In Chinese, with English summary.)

217. . 1979. Amynodonts from the Lower Oligocene

ofLantian and Sian, Shensi. Prof. Papers Strat. Pal., 7: 109-
120. Geol. Publ. House, Peking. (In Chinese.)

218. . 1980. New material of fossil Manteodon yoimgi

from Yi-chang, Hubei. Vert. PalAsiat., 18(4):296-298. (In
Chinese.)

219. Xu-YU-xuAN AND Chiu Chan-siang. 1962. Early Ter-
tiary mammalian fossils from Lunan, Yunnan. Vert.
PalAsiat., 6(4):3 13-332. (In Chinese, with English sum-
mary.)

220. Xu Yu-xuan and Wang Jing-wen. 1978. New materials

of giant rhinoceros. Mem. Inst. Vert. Pal. Paleoanthrop., 1 3:
132-140. (In Chinese.)

221. Xu Yu-xuan, Yan De-fa, Zhou Shi-quan, Han Shi-jing,
and Zhang Yong-cai. 1979. The subdivision of the Red
Beds of South China. Pp. 416—432, in The Mesozoic and
Cenozoic Red Beds of South China, Science Press, Beijing,
432 pp. (In Chinese.)

222. XuE XiANG-xi. 1978. The discovery of the Paleocene
mammalian fossils from Luo-nan, Shaanxi. Vert. PalAsiat.,
16(4):287. (In Chinese.)

223. Yan De-fa AND Tang Ying-jun. 1976. Mesonychids from
the Paleocene of Anhui. Vert. PalAsiat., 14(4):252-258. (In
Chinese.)

224. You Yu-zhu. 1977. Note on the new genus of Early Ter-
tiary Rhinocerotidae from Bose, Guangxi. Vert. PalAsiat.,
15(l):46-53. (In Chinese.)

225. Young Chung-chien. 1932. On some fossil mammals
from Yunnan. Bull. Geol. Soc. China, 1 1:283-393. (In
English.)

226. . 1934. A review of the Early Tertiary formations

of China. Bull. Geol. Soc. China, 13(3):469-503. (In English.)

227. . 1937. An Early Tertiary vertebrate fauna from

Yuanchii. Bull. Geol. Soc. China, 17(3-4):413— 438. (In
English.)

228. . 1944. Note on the first Eocene mammals from

South China. Amer. Mus. Novitates, 1268:1-3.

229. Young Chung-chien and Bien Mei-nian. 1935. Ceno-
zoie geology of the Wenho-Ssushui District of Central Shan-
tung. Bull. Geol. Soc. China, 14:221-246. (In English.)

230. . 1939. The Cenozoic geology of Lunan. Di-Zhi-

Lun-Ping, 4(3-4): 165-172. (In Chinese.)

231. . 1940. New horizons of Tertiary mammals in

southern China. Proc. 6th Pacific Science Congress, 2:531-
534.

232. Young Chung-chien, Bien Mei-nian, and Lee Yuen-yen.
1938. “Red Beds” of Hunan. Bull. Geol. Soc. China, 1 8(3-
4):259-300. (In English.)

233. Young Chung-chien and Chow Min-chen. 1956. Some
Oligocene mammals from Lingwu, North Kansu. Acta Pal.
Sinica, 4(4):447-460. (In Chinese, with English summary.)

234. . 1962. Some reptilian fossils from the “Red Beds”

of Kwangtung and Chekiang. Vert. PalAsiat., 6(2): 1 30-137.
(In Chinese, with English summary.)

235. Zdansky, O. 1930. Die AltertiSren Saugetiere Chinas nebst
Stratigraphischen Bemerkungen. Pal. Sin., 6(2): 1-87. (In
German, with Chinese summary.)

236. Zha! Ren-jie. 1977. Supplementary remarks on the age
of Changxindian Formation. Vert. PalAsiat., 1 5(3): 173-
176. (In Chinese, with English summary.)

237. . 1978. Two new Early Eocene mammals from

Sinkiang. Mem. Inst. Vert. Pal. Paleoanthrop., 13:102-106.
(In Chinese.)

238. . 1978. More fossil evidences favouring an Early

Eocene connection between Asia and Neoarctic. Mem. Inst.
Vert. Pal. Paleoanthrop., 13:107-1 15. (In Chinese.)

239. . 1978. Late Oligocene mammals from the Tao-

shuyuanzi Formation of Eastern Turfan Basin. Mem. Inst.
Vert. Pal. Paleoanthrop., 13:126-131. (In Chinese.)

240. Zhai Ren-jie, Bi Zhi-guo, and Yu Zhen-jiang. 1976.
Stratigraphy of Eocene Zhangshanji Formation with note

84

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

on a new species of eurymylid mammal. Vert. PalAsiat.,
14(2): 100-103. (In Chinese, with English summary.)

241. Zhai Ren-jie, Zheng Jia-jian, and Tong Yong-sheng.
1978. Stratigraphy of the mammal-bearing Tertiary of the
Turfan Basin, Sinkiang. Mem. Inst. Vert. Pal. Paleoan-
throp., 13:68-81. (In Chinese.)

242. Zhang Yu-ping. 1976. The Early Tertiary chalicotheres
of Bose and Yungle basins, Guangxi. Vert. PalAsiat., 14(2):
128-130. (In Chinese, with English summary.)

243. . 1978. Two new genera of Condylarthra phena-

colophids from the Paleocene of Nanxiong Basin, Guang-
dong. Vert. PalAsiat., 16(4):267-274. (In Chinese, with
English summary.)

244. . 1979. A new genus of phenacolophids. Pp. 373-

376, in The Mesozoic and Cenozoic Red Beds of South
China, Science Press, Beijing, 432 pp. (In Chinese.)

245. . 1980. A new tillodont-like mammal from the

Paleocene of Nanxiong Basin, Guangdong. Vert. PalAsiat.,
18(2):126-130. (In Chinese, with English summary.)

246. . 1980. Minchenelta, new name for Conolophus

Zhang, 1978. Vert. PalAsiat., 18(3):257. (In Chinese, with
English summary.)

247. Zhang Yu-ping, You YO-zhu, Ji Hong-xiang, and Ding
Su-YiN. 1978. Cenozoic stratigraphy of Yunnan. Prof.
Papers. Strat. Pal., 7:1-21. Geol. Publ. House, Beijing. (In
Chinese, with English summary.)

248. Zhang Yu-ping, Zheng Jia-jian, AND Ding Su-YiN. 1979.
The Paleocene condylarths of Jiangxi. Pp. 382-386, in The
Mesozoic and Cenozoic Red Beds of South China, Science
Press, Beijing, 432 pp. (In Chinese.)

249. Zhao Guo-guang. 1965. The preliminary review of the
Cenozoic strata and geological structure of Da-li and Li-
jiang Area, Northwest Yunnan. Di-Zhi-Lun-Ping, 23(5):
345-355. (In Chinese.)

250. Zheng Jia-jian. 1978. Description of some Late Eocene
mammals from Lian-kan Formation of Turfan Basin, Sin-
kiang. Mem. Inst. Vert. Pal. Paleoanthrop., 13:116-125.
(In Chinese.)

251. . 1979. A new genus of Didymoconidae from the

Paleocene of Jiangxi. Pp. 360-365, in The Mesozoic and
Cenozoic Red Beds of South China, Science Press, Beijing,
432 pp. (In Chinese.)

252. . 1979. The Paleocene notoungulates of Jiangxi.

Pp. 387-394, in The Mesozoic and Cenozoic Red Beds of
South China, Science Press, Beijing, 432 pp. (In Chinese.)

253. Zheng Jia-jian AND Chi Hung-xiang. 1978. Some of the
Latest Eocene Condylarthra mammals from Guangsi, South
China. Vert. PalAsiat., 16(2):97-102. (In Chinese, with
English summary.)

254. Zheng Jia-jian AND QiuZhan-xiang. 1979. Adiscussion
of Cretaceous and Lower Tertiary continental strata of South
China. Pp. 1-78, in The Mesozoic and Cenozoic Red Beds
of South China, Science Press, Beijing, 432 pp. (In Chinese.)

255. Zheng Jia-jian, Tang Ying-jun, Zhai Ren-jie, Ding
Su-YiN, AND Huang XuE-SHi. 1978. Early Tertiary strata
of Lunan Basin, Yunnan. Prof. Papers Strat. Pal., 7:22-29.
Geol. Publ. House, Beijing. (In Chinese.)

256. Zheng Jia-jian, Tong Yong-sheng, Ji Hong-xiang, and
Zhang Fa. 1973. The subdivision of the “Red Beds” of
Chijiang Basin, Jiangxi. Vert. PalAsiat., 1 1(2):206-21 1. (In
Chinese.)

257. Zheng Jia-jian, Tong Yong-sheng, and Ji Hong-xiang.

1975. Discovery of Miacidae (Carnivora) in Yuanshui
Basin, Kiangsi Province. Vert. PalAsiat., 1 3(2):96-104. (In
Chinese.)

258. Zhou Ming-zhen (Minchen Chow). 1979. Vertebrate
paleontology in China (1949-1979). Vert. PalAsiat., 17(4):
263-276. (In Chinese.)

259. Zhou Ming-zhen, Zhang Yu-ping, Wang Ban-yue, and
Ding Su-yin. 1977. Mammalian fauna from the Paleo-
cene of Nanxiong Basin, Guangdong. Pal. Sin., n.s., C., 20:
1-100. (In Chinese, with English summary.)

260. Aprnponyjio, A. M. 1940. OOsop HaxojjoK TpeTHHHbix
FpHsyHOB Ha TeppHTopmt CCCP h Cmcjkhbix OBJiacTCH
Ashh. HPMPOJJA, 12:74-82. JlcHHurpaa.

(Argyropulo, a. I. 1940. A survey of the findings of Ro-
dentia (Tertiary) on the territory of USSR and of the con-
tiguous regions of Asia. Priroda, 12:74-82, Leningrad.
(In Russian.))

261. BejiacBa, E. H., B. A. TpocJjHMOB, h B. K). PeiueMOB.
1974. OcHOBHbie 3xaHbi SaojiioitHH MjieKOHaTaKjmnx B
HoagHCM Mesosoe-HajieoreHe U, IJseHTpajibHOH Abhh.
CoBMecT. CoBCT. -Monro. Haji. 3Kcnefl., Tpygbi Bbin 1:
Oayna h BHOTpaxnrpactjHR Mesosoa h KaHHOsoa Monrojinn.
exp. 19-45. (MocKBa).

(Beliajeva, E. L, B. a. Trofimov, and V. Yu. Reshe-
xov. 1974. General stages in evolution of Late Mesozoic and
Early Teritary mammalian fauna in Central Asia. The
Joint Soviet-Mongolian Paleontological Expedition, Trans.
Vol. 1: Mesozoic and Cenozoic faunas and biostratigraphy
of Mongolia, pp. 19^5) (Moscow). (In Russian, with short
English summary.))

262. TpoMOBa, B. H. 1952. O HpnMHXHBHbix XnmnnKax h3
Hajieorena Monrojinn h Kasaxexana. Tpyjjbi HUH AH
CCCP, 41:51-77.

(Gromova, V. I. 1952. On the primitive carnivores from
the Paleogene of Mongolia and Kazakhstan. Trans. Pal.
Inst. Acad. Sci. USSR, 41:51-77. (In Russian.))

263. . 1952. HpHMnxHBHbie Tannpoo6pa3Hbie h3 Haji-

eorena MoHrojinti. Tpyjtbi HHH AH CCCP, 41:99-119.

. 1952. Primitive tapirids from the Paleogene of

Mongolia. Trans. Pal. Inst. Acad. Sci. USSR, 41:99-119.
(In Russian.))

264. FypecB, A. A. 1960 . 3afliteo6pa3Hbie OjiHroitena Monro-
jiHH H Kaaaxcxana. Tpygbi Haji. Hhcx. AH CCCP, 77:5-34.

(Gureev, A.A. 1960. Oligocene Lagomorpha of Mongo-
lia and Kazakhstan. Trans. Pal. Inst. Acad. Sci. USSR,
77:5-34. (In Russian.))

265. JJamaeBer, jj. 1976. HoBbie MeaoHnxHjtbi (Condylarthra,
Mesonychidae) m3 HajieoreHa MoHrojiHH. CoBMeex. Co-
Bex. -Mohxo. Haji. 3Kcnejt. Tpyjjbi 3:14-31.

(Dashzeveg, D. 1976. New mesonychids (Condylarthra,
Mesonychidae) from the Early Paleogene of Mongolia. The
Joint Soviet-Mongolian Paleontological Expedition, Trans.
Vol. 3:14-31. (In Russian.))

266. . 1977. Byrjt HaiipaMjtax Monroji Apjj Vjic flaxb

Hyopsodus (Leidy, 1870, Mammalia, Condylarthra) — hhh
Ahxhbi OjijtBop. CoBMeex. CoBex. -Monro, Haji. 3Kcneg.
Tpyjtbi 4: fbayna, Ojiopa n BnocxpaxHrpacJjHa Meaoaoji
n KaHH0305i Monrojinn. exp. 7-13.

. 1977. On the first occurrence of Hyopsodus

Leidy, 1870 (Mammalia, Condylarthra) in Mongolian
People’s Republic. The Joint Soviet-Mongolian Paleonto-
logical Expedition, Trans. Vol. 4:7-13. (In Russian.))

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

85

267. PeiuexoB, B. K). 1975 . 063op PaHHexpeTMUHbix TaHHpo-
o6pa3Hbix MoHro.iHH m CCCP. Cobmccx. CoBex.-MoHro.
naji. 3KcncA. TpyAti 2: McKonaemaH OayHa h fpjiopa
MoHro.xnH. exp. 19-53.

(Reshexov, V. Yu. 1975. Review of Early Tertiary Tapi-
roidea of Mongolia and USSR. The Joint Soviet-Mongolian
Paleontological Expedition, Trans. Vol. 2: Eossil fauna and
flora of Mongolia, pp. 19-53. (In Russian.))

268. . 1976. O nepBoii Haxo/tKe po/ta Coryphodoit Owen,

1845 (Mammalia, Pantodonta) b UajieoreHe MonrojibCKOH
HapoflHOH Pecny6.nHKH. CoBMeex. CoBex. -Monro. Uaji.
3Kcne)(., Tpyflbi 3, exp. 9-13.

( . 1976. On the first occurrence of Corypliodon

Owen, 1845 (Mammalia, Pantodonta) in Paleogene of Mon-
golian People’s Republic. The Joint Soviet-Mongolian Pa-
leontological Expedition, Trans. Vol. 3:9-13. (In Russian.))

269. TpocjjHMOB, B. A. 1952. O Potta Pseudiclops — Cboc-
o6pa3HOM HaceKOMOflttnoM H3 Hn>KHexpexHrHbix Oxjio-
Hcennit Monrojinn. Tpy^bi Tlaji. Hn-xa AH CCCP, 41:7-12.

(Trofimov, B. A. 1952. On the genus Pseudictops — a
peculiar insectivore from the Lower Tertiary deposits of
Mongolia. Trans. Pal. Inst. Acad. Sci. USSR, 41:7-12.
(In Russian.))

270. . 1975. HoBbie /lannhie O Buginhacitar Kielan-

Jaworowska et Sochava, 1969 (Mammalia, Multituberculata)
H3 MonrojiHH. Cobmccx. Cobcx. -Monro. Haji. 3Kcnefl.,
Tpy^bi 2:7-13.

( . 1975. New data on Buginhaatar Kielan-Jawor-

owska (Mammalia, Multituberculata) from Mongolia. The
Joint Soviet-Mongolian Paleontological Expedition, Trans.
Vol. 2:7-13. (In Russian.))

271. 0.iiepoB, K. K. 1952. HoBbie Dinocerata h3 Monrojinn.
floKJi. AH CCCP, 86(5): 1029-1032.

(Elerov, K. K. 1952. New Dinocerata from Mongolia.
Dokl. Akad. Nauk. USSR, 86(5): 1029-1032. (In Russian.))

272. . 1952. HanxoAonxbi (Pantodonta) CooOpannue

MonrojibCKoit HajieonxonorHHecKOH 3KcnejtHLtnen Ana-
.iteMHH HayK CCCP. Tpy^bi Haji. Hn-Ta AH, CCCP, 41:
43-50.

( . 1952. Pantodonts (Pantodonta) collected by the

Mongolian Paleontological Expedition of the Academy of
Science of the USSR. Trans. Pal. Inst. Acad. Sci. USSR,
41:43-50. (In Russian.))

273. . 1957. JInHOitepaxbi MonrojiHH. Tpyflbi Hn-Ta AH,

CCCP, 67, exp. 1-82.

( . 1957. Dinocerata of Mongolia. Trans. Pal. Inst.

Acad. Sci. USSR, 67:1-82. (In Russian.))

274. OjiepoB, K. K. n Jl,. flamxeBer. 1974. Hobbih HpeAcxa-
BMxejib Archaeolambda (Mammalia, Pantodonta) h3 Ha-
.xeorepa MonrojibCKon HapoflHOH PecnyOjiHKH. Cobmccx.
Cobcx. -Monro. Haji. 3KcncA., TpbiAti, 1:60-63.

(Elerov, K. K., and D. Dashzeveg. 1974. New spe-
cies of Archaeolambda (Mammalia, Pantodonta) from Early

Tertiary deposits in Mongolian People's Republic. The Joint
Soviet-Mongolian Paleontological Expedition, Trans. Vol.
1:60-63. (In Russian.))

275. lIIeBbipeBa, H. C. 1965. HoBbie 0.miroitenoBbie Xombkh
CCCP H Monrojinn. Haji. >Kypnaji, 1:105-114.

(Shevyreva, N. S. 1965. New Oligocene hamsters of
the USSR and Mongolia. Pal. Zhumal, 1:105-114. (In Rus-
sian.))

276. . 1971. HoBbie CpcAneojinroiienoBbie Tpbixynbi

Kaxaxcxanan Monrojinn. TpyAbi Haji. MH-CT AH CCCP,
130:70-86.

( . 1971. New rodents from the Middle Oligocene

of Kazakhstan and Mongolia. Trans. Pal. Inst., Acad. Sci.
USSR, 130:70-86. (In Russian.))

277. . 1976. riajieorenoBbie Tpbiaynbi Axnn. (Ce-

MencxBa Paramyidae, Sciuravidae, Ischyromyidae, Cylin-
drodontidae). TpyAbi Haji. HH-CT CCCP, 158:1-113.

( . 1976. Paleogene rodents of Asia. — Eamilies

Paramyidae, Sciuravidae, Ischyromyidae, Cylindrodon-
tidae. Trans. Pal. Inst. Acad. Sci. USSR, 158:1-113. (In
Russian.))

278. UIcBbipeBa, H. C., B. M. HxuKBaAxe n B. M. >Kerajuio.
1975. HoBbie Jlanubie O Oayne HoxBonounbix Meexona-
xojKAcnnn Taiuaxo MonrobCKan HapoAnaa PeenyOnuKa.
CooGme AKaA- HayK. Tpyarin CCCP, 77(l):225-228.

(Shevyreva, N. S., V. M. Chkhikvadze, and V. I.
Zhegallo. 1975. New data on the vertebrate fauna of
the Gashato Eormation (Mongolian People's Republic),
Bull. Acad. Sci. Georgian SSR, 77( l):225-228. (In Russian,
with English summary.))

279. HnoBCKaa, H. M. 1975. HpuMuxuBnaa <t>opMa Bponxo-
repna n3 SotienoBbix OxjiojKennn Monrojinn. Cobmccx
CoBex-Monro. Haji. Dkciica-, TpyAU, 2:14-18.

(Janovskaja, N. M. 1975. Primitive form of brontotheres
from Eocene deposits in Mongolia. The Joint Soviet-Mon-
golian Paleontological Expedition, Trans. Vol. 2:14-18. (In
Russian.))

280. . 1976. MomojibiH Epimanteoceras ampins sp. nov.

(Mammalia, Perissodactyla, Brontotheriidae). Cobmccx.
Cobcx. -Monro. Haji. Dkchca., TpyAbi, 3:38^6.

( . 1976. Epimanteoceras ampins sp. nov. (Mamma-

lia, Perissodactyla, Brontotheriidae) from Mongolia. The
Joint Soviet-Mongolian Paleontological Expedition, Trans.
Vol. 3:38-46. (In Russian.))

281. RnoBCKan, H. M., E. H. KypouKun n E. B. J],eB}ixKnn.
1977. MecxonaxojKAenne Dprnjmn-JJxocxpaxoxnri Hn»c-
nero Ojinroitena b lOro-Bocxonnon Monrojinn. Cobmccx.
Cobcx. -Monro. Haji. 3KcncA., TpyAbi, 4:14-33.

(Janovskaja, N. M., E. N. Kuroxchkin, and E. V.
Devjaxkin. 1977. Ergeleen-Dzo locality — the stratotype
of Lower Oligocene in southeast Mongolia. The Joint So-
viet-Mongolian Paleontological Expedition, Trans. Vol. 4:
14-33. (In Russian.))

86

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Appendix l.—A comparative list of the Chinese authors’ names in English and Pinyin.

Pinyin

English Chinese

Chen Guan-fang

W-Mfs

Ding Su-yin

Ting Su-yin T

Hu Chang-kang

Hu Chang-kang

Hu Cheng-zhi

- - -

Huang Wan-po

Huang Wan-po MJj'hli

Huang Xue-sh i (h)

- - -

Ji Hong-xiang

Chi Hung-giang

Jia Lan-po

Chia Lan-po

Li Chuan-kui

Li Chuan-kuei

Liu Dong-sheng

Liu Tung- sen

Liu Xian -ting

Liu Hsian -ting

Pei Wen-zhong

Pei Wen-chung

Qi Tao

Chi Tao Tf

Qiu Zhan-xiang

Chiu Chan-siang

Qiu Zhu-ding

Chiu Chu-ting EPWJffI

Tang Xin

Tang Hs i n ft

Tang Ying-jun

Tong Yong-sheng

Tung Yung-sheng

Wang Ban-yue

- - -

Wang Jing-wen

- - -

Wu Ru-kang

Woo Ju-kang

Wu Wen-yu

- - -

Xu Q,in-qi

Xu Yu-xuan

Hsu Yu-hsiuan

Xue Xiang-xi

Hsieh Hsiang-hsu

Yan De-fa

WMlk

Yang Zhong-jian

Young Chung-chien

Ye Xiang-kui

Yeh Hsiang-kuei

You Yu-zhu

- - -

Zha i Ren- j i e

Chai Jen-chieh

Zhang Yu-ping

Chang Yu-ping

Zhen Shou-nan

Zheng J i a-j i an

Cheng Chia-chien

Zhou Ming-zhen

Chow Min-chen

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

87

Appendix 11.

— A comparative table of the localities in

"Conventional English" or Wade-giles and Piny

in.

Pinyin

Wade-gi les

"Conventional English"

Chinese

A i - 1 i -ge-mi ao

Ai -1 i -ke-mi ao

-

Al-xa Zuo-qi

A-la-shan Tso-chi

-

An-hui (sheng)

An-hui (sheng)

Anhwei (province)

An- j i -ha i

An-ch i -ha i

Ans i ha i

An- ren-cun

An- jen-tsun

An jen tsun

-

-

Arshan to

RtLl^lr

Ba-he

Pa-ho

Paho

M

Ba i - 1 u-yuan

Pa i - 1 u-yuan

Pa i 1 uy uan

Ba i -shu i -cun

Pa i -shu i - ts un

Pa i shu i tsun

Ban-q i ao

Pan-ch ' i ao

Ranch i ao

ft #

-

-

Baron Sog Mesa

Bayan Gol

Pa-yen Kao-le

Bayan Gol (Teng kow)

-

-

Bayen Ulan

Bei - j ing

Pe i -ch i ng

Pek 1 ng

it M

Bo-se

Pai -se

Ba i se

H fe

-

-

Camp Hargetts

Ca i - j i a-chon g

Tsai -ch i a-chung

Tsa i ch i achung

Cha-gan-bu- 1 a-ge

-

-

Cha-1 ing

Ch ' a- 1 ing

Chal ing

Chang-xi n -d i an

Chang-hs i n- t i en

Changs i n t i en

Ch i - j i ang

Ch ' i h-ch i ang

Ch i hk i ang

ft a:

-

-

Chimney Butte

fflSlLLi

Chu-gou-yu

Chu-kou-yu

Chukouyu

Chuan-kou

Chuan-kou

Chuankou

Jll p

Da-bu

Ta-pu

Tapu

A ^

Da-cang-fang

Ta-tsang-fang

Tatsangfang

Da- j i an

Ta-ch i en

Tach i en

Da-ke

Ta-ko

Dahko

A oj

Da- tang

Ta-t ' ang

Tatang

A ii

Da-ye-ma-ban

Ta-yeh-ma-pan

Tayehmapan

Da-yu

Ta-yu

Tayu

A IM

Da-zhang

Ta-chang

Tachang

A m-

Dan-xi a

Tan-hs i a

Tanya

)3 B

88

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Appendix II. — Continued.

Pinyin

Wade-giles

"Conventional English"

Chinese

Dan g-he

Tang-ho

Tongho

iej

Deng-fang

Teng-fang

Tengfang

)(|5

Deng-kou

Teng-kou

Tengkow

m □

D i ng-ch i - 1 i ng

T i ng-ch i h- 1 i ng

T i ngch i h 1 i ng

Don g-hu

Tung-hu

Tunghu

Dong- j un

T ung-chun

Tungchun

Dou-mu

Tou-mu

Toumu

IS ®

Dun-huan g

Tun-huang

T unhuang

» 'M

-

-

East Mesa

Fe i -y ue

Fe i -yueh

Fe i yueh

A K

Gao-po

Kao-po

Kaopo

-

-

Gashato

Gong-kang

Kun g-k ' ang

Kungkang

m

Gu-yuan

Ku-y uan

Kuyuan

@ IS

Guan-zhuang

Kuan-chuang

Kuanchuang

li

Guang-dong (sheng)

Kuang-tung (sheng)

Kwantung (province)

Guang-xi (Zhuang-zu

Kuang-hsi (Chuang-tze

Kwangs i (Chuang

r- m

Zi -zh i -qu)

Tzu-ch i h-ch ' u)

Autonomous Region)

Gui-zhou (sheng)

Kuei-chow (sheng)

Kweichow (province)

Ha-mi

Ha-mi

Hami

Han-hua-wu

Han-hua-wu

Hanhuawu

He-bei (sheng)

Ho-pei (sheng)

Hopei (province)

He-nan (sheng)

Ho-nan (sheng)

Honan (province)

He-tao-y uan

Ho- tao-yuan

Hetaoyuan

He-t i

Ho-t i

Hot i

/SI ifi

Heng-dong

Heng-tung

Hengtung

Heng-yan g

Heng-yang

Hengyang

ilr PH

Hon g-he

Hung-ho

Hungho

a isj

Hong-q i n-bao

Hung-ch i n-pao

Hungch i npao

Hos Burd

Hao-ssu-pu-ehr-tu

-

-

-

Hou 1 d j i n

-

-

Hsanda Gol

illii/SJ

Hu-bei (sheng)

Hu-pei (sheng)

Hupei (province)

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

89

Appendix 11.

— Continued.

Piny in

Wade-gi les

"Conventional English"

Chinese

Hu-he-bol -he

-

-

Hu-nan (sheng)

Hu-nan (sheng)

Hunan (province)

Hu i -hu i -pu

Hu i -hu i -pu

Hu i hu i pu

Hun-shu i -he

Hun-shu i -ho

Hunshui ho

Huo-yan-shan

Huo-yan-shan

Huoyanshan

E ren-hot

E rh- 1 i en-hao- 1 ' e

1 ren Dabasu

-

-

1 rd i n Manha

Jar-ta i

Ch i -1 an-tai

Ch i 1 an ta i (salt 1 ake )

J i -yuan

Ch i -yuan

T s i yuan

J i ang-xi (sheng)

Chiang-hsi (sheng)

Kiangsi (province)

-

-

Jhamo Obo

J un-x i an

Chun-hs i en

Kunhs i en

Jung-gar

Chun-ko-erh

Dzungar

/-illicit

Kang-wan-gou

Kang-wan-kou

Kangwankou

La i -an

La i -an

La i an

3ft ^

Lan-t i an

Lan-t i en

Lan t i en

m

L i -guan-q i ao

L i -kuan-ch i ao

L i kuanch i ao

L i - j i a- 1 ao-wu

L i -ch i a- 1 ao-wu

L i ch i ai aowu

L i -j i ang

L i -ch i ang

L i ki ang

ffi il

L i -mu-pi ng

L i -mu-p ’ i ng

L i mup i ng

L i -sh i -gou

L i -sh i h-kou

Li sh i hkou

$±'i^

L i an-kan

L i en-k ' an

L i enkan

iS ^

L i an-mu-x i n

L i en-mu-hs i n

L i enmus i n

L i n-qu

L i n-ch ' u

L i nchu

ig m

L i n- j i ang

L i n-ch i ang

L i nk i ang

US fl;

L i n-tong

L i n-t ‘ ung

L i n t ung

is m

L i ng-bao

L ing-pao

L ingpao

L i ng-cha

L i ng-ch ' a

L ingcha

L i ng-shan

L ing-shan

L i ngshan

M. dj

L i ng-wu

L i ng-wu

L i ngwu

M iEti

L i u-n i u

L i u-n i u

L i un i u

A nil

Lu-me i -y i

Lu-me i -y i

Lume i y i

E&Hb

Lu-nan

Lu-nan

Lunan

90

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Appendix H. — Continued.

Piny in

Wade-giies

"Conventional English"

Chinese

Lu-sh i

Lu-sh i h

Lush i

/s R

Luan-chuan

Luan-ch ' uan

Luanchuan

m jii

Luo- fu-zha i

Lo- f u-cha i

Lofocha i

Luo-nan

Lo-nan

Lonan

Luo-p i ng

Lo-p‘ ing

Lopi ng

Ma-gu-shan

Ma-ku-shan

Makushan

ftfeiLl

Ma-na-s

Ma-na-ssu

Manas

Mao-j i a-po

Mao-ch i a-po

Maoch i apo

Mao-xi -cun

Mao-hs i - tsun

Maos i tsun

Me i -z i -xi

Me i -tsu-hs i

Me i tzes i

tSTM

Meng- j i a-po

Meng-ch i a-po

Mench i apu

Meng-y i n

Meng-yin

Mengy i n

m m

Mi an-ch i

Mi en-ch i h

M i ench i h

M ite

Na-duo

Na-tu

Nado

m is

Na- 1 ou

Na- lou

Na 1 ou

IP «

Nan-kang

Nan-kan g

Nan kang

Nan-xiong

Nan-hs i ung

Nanhs i ung

m m

Nao-mu-gen

-

Nomogen

-

-

Naran Bulak

Ne i Mong-gol

Ne i -meng-ku

1 nne r Mongol i a

(Zi -zh i -qu)

(Tzu-ch i h-chu)

Nemegt

N i ng- j i a-shan

N i ng-ch i a-shan

N ingch i ashan

T^iii

Ning-xia (Hui-zu

N i ng-hs i a (Hui -tsu

N i ngs i a (Hui Autonomous

T S

Zi -zhi -qu)

Tzu-ch i h-ch ' u)

Region)

Oiling f&E)

N i u-shan

N i u-shan

N i ushan

^ Lij

-

-

Nom Khong Obo (Shireh)

-

-

North Mesa

dtS±i!l

Nong-shan

Nung-shan

Nunshan

■nk til

P i ng-hu

P ' i ng-hu

P i nghu

Qi -ke-ta i

Ch i -ko-ta i

Ch i keta i

Qi an-shan

Ch ' ien-shan

Chienshan, Tsienshan

m li]

Qi n- 1 i ng

Ch ' in-1 ing

Ts i n 1 ing (moun tain)

1983

LI AND TING-THE PALEOGENE MAMMALS OF CHINA

91

Appendix II. — Continued.

Pinyin

Wade-gi les

"Conventional English"

Chi nese

Qing-shu i -y i ng

Ch ' i ng-shui -y i ng

Ch i ngshu lying

Qu-j ing

Ch ' il-ch i ng

Chuch ing

ft m

Qu-yang

Ch ' u-yang

Chuyang

ft P0

Ren-cun

Jen-ts * un

Jentsun

San-sheng-gong

San-sheng-kung

Saint Jacques, Santancho,

Santaolo

Shan-dong (sheng)

Shan-tung (sheng)

Shantung (province)

Shan-shan

Shan-shan

Shanshan

§15 ^

Shan-xi (sheng)

Shan-hsi (sheng)

Shansi (province)

Shaan-xi (sheng)

Shaan-hsi (sheng)

Shensi (province)

Shang-hu

Shang-hu

Shanghu

± m

Shang-xi u-ren

Shang-hs i u-j en

Shangs i ujen

-

-

Sha rgal te i n

Sheng- j i n-kou

Sheng-ch i n-kou

Shengch i nkou

Sh i -e r-ma-cheng

-

Sh i h-eh r-ma-cheng

+~3,r

Sh i -men

Sh i h-men

Sh i hmen

5 n

Sh i -san- j i an-fang

Shih-san-chi en- fang

Sh i sanch i en fang

+Hra];^

Sh i -zi -kou

Sh i h- tzu-k ' ou

Sh i tzekou

Sh i -zong

Sh i h-tsung

Shihtsung, Shichong

!/f ^

-

-

Shi hch i angtzuku

Shuang-ta-s i

Shuang-t ' a-ssu

Shuangtassu

Si-zi-wang Qi

Ssu-tzu-wang Ch ’ i

-

Su-ba-sh i

-

-

Su- j i -deng-en- j i (Mesa)

-

-

Su-ha i t

Su-hai -tu

Suha i tu

Ta i -z i -cun

T ' a i -tzu-t ‘ sun

Tai tzetsun

Tan-tou

T‘an-t 'ou

Tan tou

if ^

Tao-shu-yuan-zi

T ' ao-shu-yuan-tzu

Taoshuyuan tze

Tamu Bulak

-

Taben Buluk

Ti an-dong

T ‘ i en-tung

Ti en tung

ffl

Ti an-yang

T ' i en-yang

Ti enyang

H P0

Tong-ba i

Tung-pai

Tungpeh

Ton g-x i n

T ung-hs i n

Tungs i n

'C'

92

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 21

Appendix II. — Continued.

Pinyin

Wade-gi les

"Conventional English"

Chinese

T u rpan

T ' u- 1 u-p ' an

Turfan

B±#§

-

-

Tukhum

-

-

Twin Oboes

xx^lLi

-

-

Ulan Usu

-

-

Ulan Gochu

-

-

Ulan Shi reh

U 1 ungu r

Wu- 1 un-ku

U1 ungur

-

-

Urtyn Obo

Wan-1 ie

Wan - 1 i eh

Wan 1 i eh

m n

Wang-da-wu

Wang-ta-wu

Wangtawu

-a AM

Wan g-he

Wang-ho

Wangho

E /E

Wang-hu-dun

Wang-hu-tun

Wanghutun

We i -nan

We i -nan

We i nan

if 1^1

Wu-cheng

Wu-ch 'eng

Wucheng

Wu- 1 i -du i

Wu- 1 i -tu i

Wul i tu i

EMlft

Wu-tu

Wu-tu

Wu-tu

E m

-

-

Wutaoyayu

Eitaiits

Xi -an

Hs i -an

S i an

X i -chuan

Hs i -ch ' uan

S i chuan

m ill

X i - j i a-d i an

Hs i -ch i a- t i en

S i ch i atien

X i -X i -zhou

Hs i -hs i -chou

S i s i chow

Xar-Mo non

-

Shara Murun

Xi a-meng-tang

Hs i a-meng- tang

S i amengtang

Xi a-xi u-ren

Hs i a-hs i u- jen

S i as i u jen

TfcC

X i ang-shan

Hsian g-shan

S i angshan

^ E

X i ao-sha-he

Hs i ao-sha-ho

S i aoshaho

X i ao- tun

Hs i ao- tun

S i aotun

d' E

X i e-hu

Hs ieh-hu

Hs iehhu

X i n-cun- ] i

Hsin-tsun-1 i

S i n ts un 1 i

Xin-jiang (Weiwu'er

Hsin-chiang (Wei-wu-erh

Sinkiang (Uygur Autonomous

if a

Z i -zh i -qu)

Tzu-ch i h-Ch ' u)

Region)

X i n- j ie

Hs i n-ch i eh

S i nch i eh

if St

X i n-ta i

Hs i n-t ' a i

S i n ta i

if S

1983

LI AND TING- THE PALEOGENE MAMMALS OF CHINA

93

Appendix II. — Continued.

Pinyin

Wade-gi les

"Conventional Engl i sh"

Chinese

Xi n-yu

Hs i n-yil

S i nyu

ft

Xuan-cheng

Hsuan-cheng

Hsuancheng

a ^

Yang-x i

Yang-hs i

Yangs i

n m

Yang-xi ao-wu

Yang-hs i ao-wu

Yangs i aowu

Ye-ma-quan

Yeh-ma-chuan

Yehmachuan

Y i -chang

1 -ch ' ang

1 chang

S e

Yi-du

1 -tu

Itu

Y i -xi -bu- ] a-ke

- _ -

- - -

Yong-le

Yung-1 e

You-shan

Yu-shan

Yu-huang-di ng

Yu-huang-t i ng

Yu-men

Yu-men

Yuan-qu

Yuan-chil

Yun-nan (sheng)

Yun-nan (sheng)

Zao-sh i

Tsao-sh i h

Zeng-b i - 1 i ng

T seng-p i - ] i ng

Zeng-de-ao

Tseng-te-ao

Zha i - 1 i

Cha i - 1 i

Zhang- j i a-wu

Chang-ch i a-wu

Zhang-shan-j i

Chang-shan-ch i

Zhu-j i

Chu-ch i

Yindi rte

H if.

Yunglo

^ *

Youshan

ill

Yuhuangt i ng

Yumen

i n

Yuanchu

S ft

Yunnan (province)

Zhaosh i h

$ Hr

Changp i 1 i ng

Tsenteao

Cha i 1 i

m M

Changch i awu

Changshanch i

Chuch i

^ m

Appendix III. —Map of Chinese Paleogene Mammalian Fossil Localities.

70”

Paleocene

1. Nan-xiong (Nanhsiung): Shang-hu Fm., Nong-shan (Nungshan) Fm.

2. Chi-jiang (Chihkiang): Shi-zi-kou (Shihtzekou) Fm., Chi-jiang Fm.

3. Cha-ling: Zao-shi (Zhaoshih) Fm.

4. Qian-shan (Chienshan): Wang-hu-dun (Wanghutun) Fm., Dou-mu

(Toumu) Fm.

5. Xuan-cheng (Hsuancheng): Shuang-ta-si (Shuangtassu) Gr.

6. Tan-tou: Gao-yu-gou (Kaoyukou) Fm., Da-zhang (Ta-

chang) Fm.

7. Luo- nan: Shi-men.

8. Turpan (Turfan): Tai-zi-cun (Taitzetsun) Fm.

9. Nao-mu-gen (Nomogen): Nomogen Fm.

Eocene

10. Bayan Ulan: Bayan Ulan Fm.

1 1. Arshanto and Irdin Manha: Arshanto Fm., Irdin Manha Fm.

12. Camp Margetts: Irdin Manha Fm.

13. Shara Murun: Irdin Manha Fm.

14. Ula Usu: Shara Murun Fm.

15. Turpan: Da-bu (Tapu) Fm.

16. Turpan: Shi-san-jian-fang (Shihsanchienfang) Fm.

17. Turpan: Lian-kan (Lienkan) Fm.

18. Jung-gur (Dzungar): U-lun-gu Fm., Hong-li-shan Fm.

19. Jung-gur: Lu-se (Green) Fm.

20. Lan-tian: Hong-he (Hungho) Fm.

2 1 . Beijing (Peking): Chang-xin-dian (Changsintien) Fm.

22. Qu-yang (Chuyang): Ling-shan.

23. Wu-tu: Wu-tu Fm.

24. Niu-shan: Wu-tu Fm.

25. Xin-tai (Hsintai): Guan-zhuang (Kuanchuang) Fm.

26. Yuan-chu: He-ti (Hoti) Fm.

27. Tan-tou: Tan-tou Fm.

28. Ji-yuan (Chiyuan): Ji-yuan Fm.

29. Ltng-bao (Lingpao): Chuan-kou Fm., Hun-shui-he (Hunshuiho)

Fm.

30. Lu-shi (Lushih): Lu-shi Fm., Chu-gou-yu (Chukouyu) Fm.

31. Wu-cheng: Mao-jia-po (Maochiapo) Fm., Li-shi-gou (Li-

shihkou) Fm., Wu-li-dui (Wulitui) Fm.

32. Xi-chuan (Sichuan): Yu-huang-ding (Yuhuangting) Fm., Da-cang-

fang (Tatsangfang) Fm., He-tao-yuan (Hotaoyuan) Fm.

33. Jun-xian (Chunhsien): Yu-huang-ding Fm., Da-cang-fang Fm.

34. Yi-chang (Ichang): Dong-hu (Tunghu) Fm.

35. Lai-an: Zhang-shan-ji (Changshanchi) Fm.

36. Chi-jiang: Ping-hu Fm.

37. Yuan-shui: Xin-yu (Sinyu) Gr.

38. Heng-yang: Ling-cha Fm.

39. Bo-se (Baise): Dong-jun (Tungchun) Fm., Na-duo Fm.

40. Lu-nan: Lu-mei-yi Fm.

41. Li-jiang (Likiang): Xiang-shan (Hsiangshan) Fm.

Oligocene

42. Houldjin: Houldjin Fm., Nao-gon-dai Fm. (in part).

43. Artyn Obo and Ulan Gochu: Artyn Obo Fm., Ulan Gochu Fm.

44. Deng-kou (Tengkow): Saint Jacques.

45. Hos-burd: Cha-gan-bu-la-ge Fm., Suhaitu.

46. Ling-wu: Qing-shui-ying (Chingshuiying) Fm.

47. Tong-xin (Tungsin) and Gu-yuan (Kuyuan).

48. Turpan: Tao-shu-yuan-zi (Taoshuyuantze) Fm.

49. Ha-mi: Ye-ma-quan (Yehmachuan).

50. Jung-gur: He-se (Brown) Fm.

51. Taben-buluk: Yinderte.

52. Shargaltein: Shih-chiang-tzu-ku and Wu-tao-ya-yu.

53. Shi-er-ma-cheng.

54. Lan-tian: Bai-lu-yuan (Pailuyuan) Fm.

55. Yuan-chu: Bai-shui-cun (Paishuitsun) Fm.

56. Bo-se: Gong-kang (Kung-kang) Fm.

57. Yong-le (Yunglo): Gong-kang Fm.

58. Gui-zhou (Kueichow): Oligocene or Eocene

59. Qu-jing (Chuching): Cai-jia-chong (Tsaichiachung) Fm.

60. Lu-nan: Xiao-tun (Hsiaotun) Fm.

61. Luo-ping.

Appendix III —Map of Chinese Paleogene Mammalian Fossil Localities.

Paleocene

1. Nan>xiong (Nanhsiung): Shang-hu Fm., Nong-shan (Nungshan) Fm.

2. Chi-jiang (Chihkiang): Shi-zi-kou (Shihtzekou) Fm., Chi>jiang Fm.

3. Cha-ling: Zao-shi (Zhaoshih) Fm.

4. Qian>shan (Chienshan): Wang-hu*dun (Wanghutun) Fm.. Dou-mu

(Toumu) Fm.

5. Xuan-cheog (Hsuancheng): Shuang-ta>si (Shuangtassu) Gr.

6. Tan-lou: Gao-yu-gou (Kaoyukou) Fm., Da-zhang (Ta-

chang) Fm.

7. Luo-nan; Shi-men.

8. Turpan (Turfan): Tai-zi-cun (Taitzelsun) Fm.

9. Nao-mu-gen (Nomogen): Nomogen Fm.

Eocene

10. Bayan Ulan: Bayan Ulan Fm.

1 1. Arshanto and Irdin Manha: Arshanto Fm., Irdin Manha Fm.

12. Camp Margeus: Irdin Manha Fm.

13. Shara Murun: Irdin Manha Fm.

14. Ula Usu; Sham Murun Fm.

15. Turpan: Da-bu (Tapu) Fm.

16. Turpan: Shi-san-jian-fang (Shihsanchienfang) Fm.

17. Turpan: Lian-kan (Lienkan) Fm.

18. Jung-gur (Dzungar): U-lun-gu Fm., Hong-li-shan Fm.

19. Jung-gur Lu-se (Green) Fm.

20. Lan-lian: Hong-he (Hungho) Fm.

21. Beijing (Peking); Chang-xin-dian (Changsintien) Fm.

22. Qu-yang (Chuyang): Ling-shan.

23. Wu-lu: Wu-tu Fm.

24. Niu-shan: Wu-tu Fm.

25. Xin-lai (Hsintai): Guan-zhuang (Kuanchuang) Fm.

26. Yuan-chu: He-li (Hoti) Fm.

27. Tan-lou: Tan-tou Fm.

28. Ji-yuan (Chiyuan): Ji-yuan Fm.

29. Ling-bao (Lingpao); Chuan-kou Fm., Hun-shui-he (Hunshuiho)

Fm.

30. Lu-shi (Lushih); Lu-shi Fm., Chu-gou-yu (Chukouyu) Fm.

31. Wu-cheng: Mao-jia-po (Maochiapo) Fm.. Li-shi-gou (Li-

shihkou) Fm.. Wu-li-dui (Wulitui) Fm.

32. Xi-chuan (Sichuan): Yu-huang-ding (Yuhuangiing) Fm.. Da-cang-

fang (Tatsangfang) Fm., Hc-iao-yuan (Hotaoyuan) Fm.

33. Jun-xian (Chunhsien): Yu-huang-ding Fm.. Da-cang-fang Fm.

34. Yi-chang (Ichang): Dong-hu (Tunghu) Fm.

35. Lai-an: Zhang-shan-ji (Cbangshanchi) Fm.

36. Chi-jiang: Ping-hu Fm.

37. Yuan-shui: Xin-yu (Sinyu) Gr.

38. Heng-yang: Ling-eha Fm.

39. Bo-se (Baise): Dong-jun (Tungchun) Fm., Na-duo Fm.

40. Lu-nan: Lu-mei-yi Fm.

41. Li-jiang (Likiang): Xiang-shan (Hsiangshan) Fm.

Oligocene

42. Houidjin; Houldjin Fm., Nao-gon-dai Fm. (in part).

43. Artyn Obo and Ulan Gochu: Artyn Obo Fm., Ulan Gochu Fm.

44. Deng-kou (Tengkow): Saint Jacques.

45. Hos-burd; Cha-gan-bu-la-ge Fm., Suhaitu.

46. Ling-wu; Qing-shui-ying (Chingshuiying) Fm.

47. Tong-xin (Tungsin) and Gu-yuan (Kuyuan).

48. Turpan: Tao-shu-yuan-zi (Taoshuyuanize) Fm.

49. Ha-mi: Yc-ma-quan (Ychmachuan).

50. Jung-gur: He-se (Brown) Fm.

51. Taben-buluk: Yinderte.

52. Shargaltein: Shih-chiang-tzu-ku and Wu-tao-ya-yu.

53. Shi-er-ma-cheng.

54. Lan-tian; Bai-lu-yuan (Pailuyuan) Fm.

55. Yuan-chu; Bai-sbui-cun (Paishuitsun) Fm.

56. Bo-se: Gong-kang (Kung-kang) Fm.

57. Yong-le (Yunglo): Gong-kang Fm.

58. Gui-zhou (Kueichow): Oligocene or Eocene

59. Qu-jing (Chuching): (Z!ai-jia-chong (Tsaichiachung) Fm.

60. Lu-nan: Xiao-tun (Hsiaotun) Fm.

61. Luo-ping.

'C

•(

(

1

s

Appendix \(in Pinyin).

LARGE M

XINJIANG

GANSU

NINGXIA

SHANXI

SHAANXI

HEBEI

SHANDONG

HENAN

HUBEI

ANHUI

Indricotl

Taben buluk
Shargaltlen

Taoshuyuanzi Fm.

Qingshuiying Fm.

Cadurc

?Shiehrmacheng

Bailuyuan Fm.
Baishuicun Fm.

Depere

Lophia

Schlosi

Llankan Fm.

Honghe Fm.
Heti Fm.

Changxindian Fm.

Jlyuan Fm.
Lishigou Fm.

Hetaoyuan Fm.

Guanzhuang Fm.

Dacangfang Fm.

Hepto

Dabu Fm.

Wutu Fm.

Archaeo

Taizlcun Fm.

Doumu Fm.

Bemala

Appendix IV.— ,-l lenialive correlation of the Chinese Paleogene Formations coniaming mammalian fossils (in English).

Appendix IV.- I icntauvc correlation of the Chinese Paleogerte Formations containing mammalian fossils (in Pinyinl.

1

)

\

1

\

i

I

HUNAN

JIANGXI

GUANGDONG

GUANGXI

YUNNAN

NORTH AMERICA

EUROPE

Whitneyan

Chattian

Orellan

Stampian

Gongkang Fm.

Caijiachong Fm.

Chadronian

Sannorsian

Naduo Fm.
Dongjun Fm.

Lumeiyi Fm.

Duchesnean

Uintan

Ludian

Bridgerian

Lutetian

Lingcha Fm.

Xinyu Gr.

Wasatchian

Cuisian

Chijiang Fm.

Nongshan Fm.

Tiffanian

Thanetian

Zaoshi Fm.

Torrejonian

Puercan

Copies of the following Bulletins of Carnegie Museum of Natural History may be obtained at the priees
listed from the Publieations Seeretary, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pittsburgh,
Pennsylvania 15213.

1. Krishtalka, L. 1976. Early Tertiary Adapisorieidae and Erinaceidae (Mammalia, Insectivora) of North

Ameriea. 40 pp., 13 figs $2.50

2. Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the
late Pleistoeene paleoecology of the central Appalachian Mountains of Virginia. 88 pp., 21 figs. . . .

$12.00

3. Wetzel, R. M. 1977. The Chacoan peccary, Catagonus (Rusconi). 36 pp., 10 figs

$6.00

4. Coombs, M. C. 1978. Reevaluation of early Miocene North American Moropus (Perissodactyla,

Chalicotheriidae, Schizotheriinae). 62 pp., 28 figs $5.00

5. Clench, M. H., and R. C. Leberman. 1978. Weights of 1 5 1 species of Pennsylvania birds analyzed

by month, age, and sex. 87 pp $5.00

6. Schlitter, D. A. (ed.). 1978. Ecology and taxonomy of African small mammals. 214 pp., 48 figs.

$15.00

7. Raikow, R. J. 1978. Appendicular myology and relationships of the New World nine-primaried

oscines (Aves: Passeriformes). 43 pp., 10 figs $3.50

8. Berman, D. S, and J. S. McIntosh. 1978. Skull and relationships of the Upper Jurassic sauropod

Apatosaurus (Reptilia, Saurischia). 35 pp., 1 1 figs $3.00

9. Setoguchi, T. 1978. Paleontology and geology of the Badwater Creek area, central Wyoming. Part

16. The Cedar Ridge local fauna (Late Oligocene). 61 pp., 30 figs $4.50

10. Williams, D. E. 1978. Systematics and ecogeographic variation of the Apache pocket mouse (Roden-

tia: Heteromyidae). 57 pp., 23 figs $4.00

11. Guilday, J. E., H. W. Hamilton, E. Anderson, and P. W. Parmalee. 1978. The Baker Bluff Cave

deposit, Tennessee, and the Late Pleistocene faunal gradient. 67 pp., 16 figs $5.00

12. Swanepoel, P., and H. H. Genoways. 1978. Revision of the Antillean bats of the genus Brachyphylla

(Mammalia: Phyllostomatidae). 53 pp., 17 figs $4.00

13. Schwartz, J. H., and H. B. Rollins (eds.). 1979. Models and methodologies in evolutionary theory.

105 pp., 36 figs $6.00

14. Engesser, B. 1979. Relationships of some insectivores and rodents from the Miocene of North

America and Europe. 68 pp., 12 figs., 20 plates $5.00

15. Bentz, G. D. 1979. The appendicular myology and phylogenetic relationships of the Ploceidae

and Estrildidae (Aves: Passeriformes). 25 pp., 5 figs $2.00

16. Mares, M. A. 1980. Convergent evolution among desert rodents: a global perspective. 51 pp., 25

figs $3.50

17. Lacher, T. E., Jr. 1981. The comparative social behavior of Kerodon rupestris and Galea spixii and

the evolution of behavior in the Caviidae. 71 pp., 40 figs $6.00

1 8. McIntosh, J. S. 1981. Annotated catalogue of the dinosaurs (Reptilia, Archosauria) in the collections

of the Carnegie Museum of Natural History. 67 pp., 22 figs $6.00

19. Bristow, C. R., and J. J. Parodiz. 1982. The stratigraphical paleontology of the Tertiary non-marine

sediments of Ecuador. 52 pp., 24 figs $5.00

20. Williams, S. L. 1982. Phalli of Recent genera and species of the family Geomyidae (Mammalia:

Rodentia). 62 pp., 30 figs $5.00

BULLETIN

of CARNEGIE MUSEUM OF NATURAL HISTORY

THE D/FF/C/L/5 COMPLEX OF SPHAERODACTYLUS
(SAURIA, GEKKONIDAE) OF HISPANIOLA

ALBERT SCHWARTZ AKp RICHARD THOMAS

NUMBER 22

PITTSBURGH, 1983

BULLETIN

of CARNEGIE MUSEUM OF NATURAL HISTORY

THE DIFFICILIS COMPLEX OF SPHAERODACTYLUS
(SAURIA, GEKKONIDAE) OF HISPANIOLA

PART 1. SPHAERODACTYLUS DIFFICILIS, S. CLENCHI, AND S. LAZELLI

ALBERT SCHWARTZ

Research Associate, Section of Amphibians and Reptiles
(address: Miami-Dade Community College, North Campus,

Miami, Florida 33167)

PART 2. SPHAERODACTYLUS SAVAGEI, S. COCHRAN AE, S. DARLINGTONI,
S. ARMSTRONGI, S. STREPTOPHORUS , AND CONCLUSIONS

RICHARD THOMAS

Biology Department, University of Puerto Rico, Rio Piedras, Puerto Rico 00931

ALBERT SCHWARTZ

NUMBER 22

PITTSBURGH, 1983

BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY
Number 22, pages 1-60, 15 figures

Issued 1 April 1983

Price $7.00 a copy

Mary R. Dawson, Acting Director

Editorial Staff: Hugh H. Genoways, Editor; Duane A. Schlitter,
Associate Editor; Stephen L. Williams, Associate Editor;
Mary Ann Schmidt, Technical Assistant.

© 1983 by the Trustees of Carnegie Institute, all rights reserved.

CARNEGIE MUSEUM OF NATURAL HISTORY, 4400 FORBES AVENUE
PITTSBURGH, PENNSYLVANIA 15213

CONTENTS

Part 1 5

Abstract 5

Introduction 5

Acknowledgments 6

Methods 6

Systematic Accounts 7

Sphaerodactyius difficilis 7

S. d. difficilis 7

S. d. lycauges 1 1

S. d. euopter 1 3

S. d. typMopous 14

S. d. peratus 17

S. d. diolenius 19

S. d. anthracornus 22

Sphaerodactyius clenchi 24

S. c. clenchi 25

S. c. apocoptiis 28

Sphaerodactyius lazelli 29

Literature Cited 30

Part 2 31

Abstract 31

Introduction 31

Acknowledgments 31

Systematic Accounts 31

Sphaerodactyius savagei 31

S. s. savagei 32

S. s. jiumilloensis 35

Sphaerodactyius cochranae 38

Sphaerodactyius darlingtoni 39

S. d. darlingtoni 40

S. d. noblei 42

S. d. bobilini 43

S. d. mekistus 45

Sphaerodactyius armstrongi 46

S. a. armstrongi 46

S. a. hypsinephes 49

Sphaerodactyius streptophorus sphenophanes 50

Discussion 54

Ecological Notes on the difficilis Complex 55

Geographic Relationships of the Species 58

Relationships with Other Sphaerodactyius 59

Literature Cited 60

ii

¥

i

t£.

-£i

PART 1. SPHAERODACTYLVS DIFFICILIS, S. CLFNCHI,

AND S. LAZELLI

Albert Schwartz

ABSTRACT

Three Hispaniolan species of Sphaerodactylus of the difficilis
complex are discussed in detail. Sphaerodactylus difficilis is shown
to be composed of seven subspecies, and S. clenchi of two. Of
these two species, S. difficilis is widely distributed in the Repub-
lica Dominicana but is restricted to northern Haiti and one lo-

cality in central Haiti, whereas S. clenchi occurs only in the
extreme northeastern Republica Dominicana. The holotype of
S. lazelli from northern Haiti is still the only known specimen.
Aside from meristic data, information on ecology, altitudinal
distribution, and sympatry-allopatry are also given.

INTRODUCTION

Throughout the Bahama Islands (including the
Turks Bank), Cuba, Isla de la Juventud, Puerto Rico
and its associated offshore islands (and including
Isla Mona, Isla Monito, and Isla Desecheo), the Vir-
gin Islands, and the northern Lesser Antilles, occurs
a group of moderately large geckos of the genus
Sphaerodactylus which characteristically have a
prominent dark scapular patch or blotch with an
associated pair of ocelli. This group of forms has its
greatest diversity in part on Puerto Rico (where all
species are members of this assemblage; Thomas
and Schwartz, 1966a) and in part on the centrally
located island of Hispaniola. Shreve (1968) appro-
priately used the trinomial name of the first-de-
scribed member {S. notatus Baird) to refer to this
group of lizards.

There have been known only two Cuban (and Isla
de la Juventud) species of the group— 5. notatus and
5. bromeliarum Peters and Schwartz. Sphaerodac-
tylus notatus itself occurs as well on the Great and
Little Bahama banks, on the North American main-
land in southeastern Florida, along the Florida Keys,
and on the Swan Islands (Schwartz, 1966). The
southern Bahamas (Great and Little Inagua) and the
Turks Islands have the related S. inaguae Noble and
KJingel and S. underwoodi Schwartz. In Puerto Rico
and the Virgin Islands (as well as the Lesser Antilles
as far south as St. Barthelemy) occur seven species
of this group— V. macrolepis Gunther, V. roosevelti
Grant, V. klauberi Grant, S. gaigeae Grant, V.
nicholsi Grant, S. parthenopion Thomas, S. beattyi
Grant. Other forms are V. levinsi Heatwole from
Isla Desecheo, S. monensis Meerwarth from Isla
Mona, and S. micropithecus Schwartz from Isla
Monito. From this brief geographical review, it is
apparent that the notatus group is widely distributed
throughout the Greater Antilles.

On Hispaniola, the notatus group has achieved a
diversification greater than it has on Puerto Rico.
We are certain that the roster of the Hispaniolan
species is as yet incomplete. Barbour (1914:265)
named S. difficilis as the first Hispaniolan member
of the group; the description of this species was based
on four specimens from the Republica Dominicana
(two from La Vega, two from Puerto Plata); even
in this short series, Barbour noted that there were
several variable characters, and he was uncertain
that all four specimens represented the same species.
Noble and Hassler (1933) named S. armstrongi on
the basis of two specimens from near Paraiso in the
Sierra de Baoruco, Republica Dominicana, on the
Peninsula de Barahona, and S. altavelensis from Isla
Alto Velo south of Isla Beata which is in turn off
the southernmost point of this same peninsula. From
the southern shore of the Bahia de Samana, Ruibal
( 1 946) named S. cochranae on the basis of two spec-
imens. Most of these species were considered mem-
bers of the difficilis complex on Hispaniola by Shreve
(1968).

The greatest step forward in our knowledge of
Hispaniolan difficilis complex members is the re-
view by Shreve (1968). Impressed with the differ-
ences in accumulated specimens at the Museum of
Comparative Zoology at Harvard University, Shreve
undertook a review of the complex on Hispaniola.
He examined 295 specimens from both Haiti and
the Republica Dominicana, and named randi, sav-
agei, and juanilloensis as subspecies of V. notatus
(with which Shreve considered S. difficilis conspe-
cific; difficilis was regarded as a recognizable sub-
species of 5. notatus); S. lazelli from Cap-Haitien
in northern Haiti, based on a single specimen; S.
darlingtoni from Pico Diego de Ocampo, Republica
Dominicana, from two specimens; S. noblei from

5

6

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 22

Los Bracitos, Republica Dominicana, with a distri-
bution on the Peninsula de Samana, the eastern Cor-
dillera Septentrional, and the southern shore of the
Bahia de Samana; S. clenchi from the Peninsula de
Samana; and S. brevirostratus as an abundant and
widely distributed species in both Haiti and the Re-
publica Dominicana. Shreve (1968) recognized two
subspecies of S. brevirostratus (b. brevirostratus and
b. enriquilloensis).

My own interest, along with that of Richard
Thomas, in this group of Hispaniolan sphaerodac-
tyls began with a visit to Haiti in 1962. Since that
time, both of us have spent considerable time be-
tween 1963 and 1979 collecting in Haiti and the
Republica Dominicana. The result of these expe-
ditions is that we have good representations of these

geckos from throughout much of Hispaniola, and
certainly a far better coverage than did Shreve. We
have examined over 3,000 specimens, most of which
we have ourselves collected. In two papers (Schwartz
and Thomas, 1977; Thomas and Schwartz, 1977)
we have named five new species of the difficilis com-
plex {S. ocoae, S. zygaena, S. streptophorus, S. cry-
phius, S. nycteropus). Schwartz (1977) reviewed the
status of N. randi, regarding it as a species distinct
from S. difficilis and naming two new subspecies,
methorius and strahrni. We also now have in press
(Thomas and Schwartz, manuscript) a paper clari-
fying the relationships of S. brevirostratus and S.
altavelensis and naming a new species from north-
western Haiti.

ACKNOWLEDGMENTS

Our collecting in Hispaniola between 1 968 and 1971 was under
the sponsorship of National Science Foundation grants GB-7977
and B-023603. Much of our material is deposited in the Albert
Schwartz Field Series (ASFS). Many of the Museum of Com-
parative Zoology (MCZ) specimens were collected under Na-
tional Science Foundation grant GB-6944 to Ernest E. Williams.

I wish to acknowledge the assistance in the field of Patricia A.
Adams, Robert K. Bobilin, Donald W. Buden, Jeffrey R. Buffet,
David A. Daniels, David C. Duval, Danny C. Fowler, Eugene
D. Graham, Jr., Ronald F. Klinikowski, Mark D. Lavrich, David
C. Leber, John K. Lewis, John C. Lucio, James W. Norton,
Dennis R. Paulson, S. Craig Rhodes, James A. Rodgers, Jr., Bruce
R. Sheplan, William W. Sommer, James B. Strong, and T. Mark
Thurmond; without their capable assistance, I would not have
such a large quantity of specimens available to me. In addition,
C. Rhea Warren has from time to time given me specimens from
Haiti and He de la Tortue. My visit to Isla Saona in late 1968
would have been impossible without the organizing ability of

Sixto J. Inchaustegui. I have been the guest of the Alcoa Explo-
ration Company at Cabo Rojo on the Peninsula de Barahona;
my stays there were made pleasant by the cooperation of Patrick
N. Hughson and the late Ruth Hamor. In addition to material
in the ASFS, I have borrowed specimens from American Mu-
seum of Natural History (AMNH), British Museum (Natural
History) (BMNH), Museum of Comparative Zoology (MCZ), and
National Museum of Natural History (USNM); for the courtesies
extended in these loans I am grateful to Richard G. Zweifel, Alice
G. C. Grandison, George W. Foley, Ernest E. Williams, the late
James A. Peters, and George R. Zug. I have also studied speci-
mens in the collection of Lewis D. Ober (LDO). Holotypes and
paratypes have been deposited in the above collections, as well
as in the collections of the Carnegie Museum of Natural History
(CM) and the Museum of Zoology at Louisiana State University
(LSUMZ). The dorsal view illustrations are the work of Dr. Leber;
I am once more in his debt.

METHODS

I have taken the measurements and counts used by myself and
Thomas previously in our studies of Antillean geckos (that is,
Thomas and Schwartz, 1966a, on Puerto Rican Sphaerodacty-
lus). These data include; 1 ) measurement of snout-vent length in
millimeters; 2) dorsal scales between axilla and groin; 3) ventral
scales between axilla and groin; 4) scales around body at mid-
body; 5) number of supralabials on both sides to below center of
eye; 6) number of intemasal scales; 7) number of fourth toe
lamellae; 8) the presence of, and amount of, keeling on the gular,
chest, and abdominal scales; and 9) the length and width (=lateral
extent) of the escutcheon in males. Counts of dorsal, ventral, and
midbody scales, supralabials, and fourth toe scales were taken
on specimens with snout-vent lengths above 25 mm in larger
taxa. The number of intemasals and the extent of ventral keeling
were determined on many specimens regardless of size. Exces-
sively low counts of fourth toe lamellae are usually due to the

fact that this count was begun distad to the small plantar scales,
with the first scale that was transversely entire; in many specimens
with very low lamellar counts, this is due to the presence of one
or more basal digital scales that are longitudinally divided and
thus not included in the lamellar count. Likewise, abnormally
low counts of length or breadth of escutcheon are due to lack of
complete development of this group of specialized ventral scales
in males. However, mean and modal differences in lamellar counts
are of little importance in this complex, and the species differ-
ences (where they exist) in escutcheon size are so obvious that
they are not obscured by the abnormally low counts in some
specimens.

In the taxonomic treatment I have followed a temporarily con-
servative course regarding the status of “S. difficilis.'' Rather than
regarding this taxon as a subspecies of 5". notalus, I consider it a
species distinct from that form. Clarification of the status of these

1983

SCH^NhKTZ-SPHAERODACTYLUS SYSTEM ATICS I

7

two species will be attempted in a later paper. The present paper
discusses the variation and systematics of three species— 5. dif-
ficilis, S. clenchi, and S. lazelli.

I have not here, nor will we later, include a discussion of S'.
samanensis Cochran or S. callocricus Schwartz as part of this
complex, because they are so set apart in morphology and color
pattern that I doubt their close relationship with the geckos dis-
cussed here. S. samanensis was included in the difficilis group by
Shreve (1968) and S. callocricus was described subsequently. In
S. samanensis the dorsal body scales are very strongly keeled
and weakly imbricate; the snout is narrow and pointed; the color
pattern is boldly cross-banded on the body; and the head pattern

is also banded in part but is not easily associated with any of the
difficilis complex head patterns. In fact, the pattern, particularly
that of the head, is interestingly similar to that of Sphaerodactylus
cinereus Wagler (see Graham and Schwartz, 1 978, for usage) from
Hispaniola and Cuba, a form that Thomas and Schwartz ( 1 966b),
with some reservation, included in the S. nigropunctatus {=S.
decoratus) complex. I am not adamant about their lack of affin-
ities with the difficilis group; I only assert that samanensis and
callocricus are peripheral to the forms discussed herein and else-
where in this series of papers. Perhaps when their variation in
color pattern is more completely known, we will have better
insight into their relationships.

SYSTEMATIC ACCOUNTS

Sphaerodactylus difficilis Barbour

Sphaerodactylus difficilis Barbour, 1914, Mem. Mus. Comp. Tool.,

44:265.

Definition.— \ species of Sphaerodactylus with
large, acute, strongly keeled, flattened, imbricate
dorsal scales, axilla to groin 22 to 40; no area of
middorsal granules or granular scales; dorsal body
scales with four to seven hair-bearing organs, each
with a single hair, around apex. Dorsal scales of tail
keeled, acute, imbricate, and flat-lying; ventral scales
of tail smooth, rounded, enlarged (often only slight-
ly) midventrally; gular scales usually smooth, but
occasionally weakly to strongly keeled; chest scales
smooth; ventral scales rounded, imbricate, axilla to
groin 23 to 37, smooth; scales around midbody 37
to 60; intemasals 0 to 3 (mode 1); upper labials to
mid-eye 3 (rarely 4); escutcheon with a broad and
compact central area and extensions onto thighs to
near underside of knee (3-8 by 8-30).

Color pattern sexually dichromatic and variable
among the subspecies. Adult males dorsally pinkish
gray or gray to tan or brown, usually with scattered
dark brown scales giving a coarsely salt-and-pepper
effect, trilineate head pattern obsolescent or (usu-
ally) absent, the head (and chin and throat) often
covered by small to large dark brown to black dots,
the throats with yellow to orange ground color; dark
scapular patch usually absent, and if present, usually
very restricted and diffusely edged; a pair of small
pale ocelli present or absent (by population); venter
variable, from gray to flesh. Females with same dor-
sal pattern as males, although distinctly lineate in
some populations, but head with a prominent tri-
lineate pattern, and pattern brown on a buffy to tan
ground color, and head scales dotted; scapular patch
variable (by population) from relatively large, brown
to black, with an associated pair of pale (white to
bufly or gray) ocelli, to small with a single included

ocellus, or both patch and ocelli entirely absent;
ventral color as in males. Iris color yellow, tan, or
brown. Juveniles with a more intense female pat-
tern.

Distribution. — \n Haiti, from the Presqu’ile du
Nord Quest (Bombardopolis, Mole St. Nicholas) east
along the northern Haitian coast to Terrier Rouge
and inland as far as Grande Riviere du Nord, Don-
don, Ennery, and Terre Sonnain near Gonaives. In
the Republica Dominicana, from Monte Cristi in
the northwest, east to the base of the Peninsula de
Samana (Sanchez); central and eastern portions of
the Republica Dominicana (Santiago and La Vega
provinces east to central La Altagracia Province, but
absent from most coastal regions in La Altagracia
Province), thence westward both along the southern
coast and inland to San Juan and Azua provinces,
and along the eastern coast of the Peninsula de Ba-
rahona to Enriquillo and into the uplands of the Si-
erra de Baoruco (La Lanza). A single record from
Hinche, Departement de I’Artibonite, in central
Haiti; Cayos Siete Hermanns (Muertos, Monte Chi-
co, Monte Grande) off the northern Dominican coast,
and Isla Pascal in the Bahia de Samana (Shreve,
1968:14); He de la Tortue. Altitudinal distribution
from sea level to 2,000 ft (610 m) on the northern
slopes of the Cordillera Central between La Vega
and Jarabacoa, and on the southern slopes of the
Cordillera Septentrional in the vicinity of La
Cumbre, and to 2,400 ft (732 m) in the Sierra de
Baoruco (Fig. 1).

Sphaerodactylus difficilis difficilis Barbour

Sphaerodactylus difficilis QaTbour, 1914, Mem. Mus. Comp. Zool.,

44(2):265.

Type-locality. — ?>2Lnimgo de la Vega, La Vega
Province, Republica Dominicana.

Holotype. -MCZ 7834.

8

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Fig. 1. — Map of Hispaniola, showing distributions of Sphaerodactylus difficilis (triangles) and S. lazelli (hexagon). Subspecies of S.
difficilis are named on the map.

1983

SCHWARTZ- SPHAERODACTYLUS SYSTEM ATICS I

9

Fig. 2. — Dorsal views of the subspecies of Sphaerodactylus difficilis. as follow: A) difficilis (male, ASFS V14170); B) difficilis (female,
ASFS Vi 788); C) lycauges (holotype female, CM 52251); D) euopter (holotype female, CM 54142); E) typhlopous (holotype female,
USNM 166966); F) peratus (holotype female, CM 52264); G) diolenius (holotype female, USNM 166967); FI) anthracomus (holotype
female, CM 52279),

Definition.— A subspecies of S. difficilis charac-
terized by a combination of high number of dorsal
scales (25-34) between axilla and groin, moderate
number (42-55) of midbody scales, female shoulder
pattern consisting of a small indistinct dark scapular
spot and a single median pale ocellus, and males
without patch or ocellus (Figs. 2A and 2B).

Distribution. — Rex^\xh\ica. Dominicana, from the
type-locality south onto the northern slopes of the
Cordillera Central (between La Vega and Jaraba-
coa), west to near Santiago, north onto the southern
slopes of the Cordillera Septentrional as far as the
pass across these mountains at La Cumbre, and east
to Los Bracitos, in La Vega, Santiago, Espaillat,

10

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Duarte (and presumably Salcedo), and extreme
southern Puerto Plata provinces.

Variation.— The. series of 65 S. d. difficilis has the
following counts and measurements (means in pa-
rentheses): largest male (ASFS VI 8 180) 34 mm
snout-vent length, largest female (ASFS VI 8 197)

33 mm; dorsal scales between axilla and groin 25-

34 (29.2); ventral scales between axilla and groin
24-37 (30.2); midbody scales 42-55 (48.1); supra-
labials to mid-eye 3/3 (50 individuals), 4/4 (1); in-
ternasals 0 (7 individuals), 1 (50), 2 (8); fourth toe
lamellae 9-13 (1 1.0; mode 10); gular scales usually
smooth (35 individuals) but at times keeled (19) or
partially keeled (11); escutcheon 4-6 (5.4) by 12-27
(21.9).

Males are grayish pink to brown above, regularly
heavily flecked with darker brown to give a salt-
and-pepper effect. The heads are dull yellow, and
the throats are yellow to orange, often heavily dotted
or almost marbled with dark brown. The upper sur-
face of the head may be either plain or rather heavily
covered by brown dots or spots, these dots arranged
in a more or less lineate pattern following the ar-
rangement of the dark head stripes of the females.
The scapular patch is almost always absent but is
barely discernible in young males (snout-vent length
28 mm: ASFS VI 8 185) which still retain remnants
of the juvenile (=female) pattern; an ocellus, prom-
inent in females, is absent in males.

Females are colored dorsally like males and are
likewise salt-and-pepper in aspect, except that there
is a tendency for the dorsal flecking to be arranged
in longitudinal lines, giving a vaguely lineate aspect.
The head pattern is trilineate, with a single median
dark brown line from the snout onto the shoulders
(where it becomes darker brown to black and forms
the restricted scapular patch) and a pair of lines from
the snout across the lores and through the eyes onto
the shoulders, where they fade and become lost in
the dorsal body pattern. There is regularly also a
moderately sharp ventrolateral (postauricular) line
from the ear opening along the neck and above the
forelimb insertion, but this line is less sharply de-
fined than the primary head lines noted above. The
scapular patch is small and restricted, usually dark
brown to black, and contains a single median pale
ocellus which varies from white to pale gray or buffy.
The scapular patch is at times diffuse (ASFS V 1 6 1 03)
and in some females it is only barely indicated (ASFS
V18197). The throat in females is usually immac-
ulate, but there may be a few tiny scattered brown
flecks or even some more or less diagonally oriented
dark gray to brownish lines on the sides of the throat.

The ventral color in both sexes is gray, flesh, or
yellow-flesh.

Juveniles resemble females in pattern and color
details, except that the juvenile pattern is more in-
tense than is that of the females.

The iris color appears to be quite variable. It was
recorded as golden in the series from 4 km NW La
Vega, tan to grayish at 4 km N La Vega, and brown
at 14.4 km E La Vega.

Mertens (1939:42-43) reported a female from
Jarabacoa and two males, two females and six sub-
adults and juveniles from Moca. Females in the lat-
ter series had “einem einzigen weissen Fleck in der
Mitte” of the scapular patch, whereas the female
from Jarabacoa apparently lacked any shoulder
markings. Presumably this latter female represents
a member of that sex where the scapular patch and
ocellus were extremely reduced.

The series from Los Bracitos, Duarte Province,
is assigned to S. d. difficilis on the basis of female
and juvenile patterns. These old specimens are pres-
ently somewhat discolored and have been dissected
ventrally, so that reliable ventral and midbody counts
are presently impossible. Counts from this series
have not been included in the computations.

Remarks.— The holotype of S. d. difficilis (MCZ
7834) is an adult male with a snout-vent length of
about 28 mm. The specimen is much shriveled, and
consequently taking scale counts is difficult and sub-
ject to inaccuracies. However, the dorsal scales be-
tween axilla and groin number about 27, and mid-
body scales about 45; there are 3/3 supralabials to
the eye center, 1 intemasal, and 1 1 fourth toe la-
mellae. The escutcheon is 6 scales long. All these
counts fall within the parameters of the population
herein regarded as Y. d. difficilis. Although the spec-
imen is now patternless, Barbour (1914:266) de-
scribed the specimen as “light gray-brown, with
many darker spots scattered irregularly over the
whole dorsal area; some cover a single scale; others
are composed of several dark scales juxtaposed.”
This pattern style, along with the smooth gular and
chest scales, occurs in males of many populations
of V. difficilis, including the nominate subspecies.
The three paratypes (MCZ 7835, from the type-lo-
cality; MCZ 5444 — two specimens— from Puerto
Plata; Barbour, 1921:274) cannot now be located in
the Harvard collections. The former is S. d. difficilis-,
the latter two presumably pertain to the northern
Dominican subspecies.

S. d. difficilis is common in the region about the
type-locality. We have taken specimens in shaded
and moist cacao groves, running freely in the late

1983

^Cim ARTZ- S PH AERO DACTYLUS SYSTEM ATICS I

1 1

afternoon in the deep ground litter of cacao leaves,
as well as in palm frond trash piles associated with
cacao groves. Two geckos from east of Santiago were
taken from under dead agave plants in xeric scrub,
whereas the lizard from north of La Cumbre was
taken from the thatch roof of the kitchen of an oc-
cupied native hut, and two other specimens were
collected on the porch of a hotel between Jarabacoa
and La Vega. S. d. difficilis is ecologically tolerant
of both xeric and mesic situations and of both ed-
ificarian and natural situations as well. The distri-
bution of the subspecies ranges from about 300 ft
(10 m) at the type-locality to 2,000 ft (610 m) in
both the Cordillera Central and Cordillera Septen-
trional. The elevation at Los Bracitos may be even
higher.

S. d. difficilis occurs sympatrically with S. dar-
lingtoni (Tenares, Cruce de Pimentel) in the eastern
portion of the Valie de Cibao and probably else-
where. The scale counts of these two species are
quite comparable but S. darlingtoni in the area of
sympatry is a much smaller lizard (maximally sized
males and females with respective snout-vent lengths
of 25 mm and 27 mm, in contrast to 34 mm and
33 mm in S. d. difficilis). In addition, S. darlingtoni
is patterned quite differently than S. d. difficilis and
is very dark (brown) dorsally.

Specimens examined. — Dominicana; Santiago

Province, 7 km E Santiago (ASFS V2929-30); 4 km S La Cumbre,
1,700 ft (519 m) (ASFS VI 8093-97, V18175-206). Puerto Plata
Province. 1 km N La Cumbre, 2,000 ft (610 m) (ASFS V18101).
La Vega Province, La Vega (MCZ 7834— holotype); 4 km NW
La Vega (ASFS V1786-91); 3 km NW La Vega (ASFS V1783-
85); 2 km NW La Vega (ASFS V16103, ASFS V16111); 4 km
N La Vega (ASFS V4 179-81); 2 km S La Vega (ASFS V4166-
67); 14.4 km E La Vega (ASFS V4213); 12 km NE Jarabacoa,
2,000 ft (610 m) (ASFS VI 8389, ASFS VI 8433). Espaiilat Prov-
ince, 8 km N Moca (ASFS V4331). Duarte Province, 5 km NW
San Francisco de Macoris (ASFS V 1 4 1 69); 4 km NW San Fran-
cisco de Macoris (ASFS V14170); 3 km S San Francisco de Ma-
coris (ASFS V2960); Los Bracitos (AMNH 45201-02, AMNH
45204, AMNH 45206-07, AMNH 45209-12, AMNH 45214-
1 5, AMNH 452 1 7-20); 7.5 mi(l 2.0 km) NW Cruce de Pimentel,
400 ft (122 m) (ASFS V33480); 6.4 mi (10.2 km) SE Tenares
(ASFS V33494-98, ASFS V33539-44).

Sphaerodactylus difficilis lycauges, new subspecies

Holotype. — CM 5225 1 , an adult female, from Cap-
Haitien, Departement du Nord, Haiti, one of a series
collected by natives for Richard Thomas on 7-8
April 1966. Original number ASFS V10128.

Paratypes.-IAUMZ 21903-07, LSUMZ 21909-14, MCZ
1 19359-66, CM 52252-59, same data as holotype; USNM
167258, same locality as holotype, E. Cyphale, 6 April 1966;

USNM 1 67259-65, ASFS V 1 0063-88, same locality as holotype,
natives, 7 April 1966; ASFS V10231-34, USNM 167266-72,
same locality as holotype, natives, 9 April 1966; MCZ 63179-
214, Bombardopolis, Dept, du Nord Quest, Haiti, A. S. Rand
and J. D. Lazell, 22 July 1960; MCZ 63219-21, Port-de-Paix,
Dept, du Nord Quest, Haiti, A. S. Rand and J. D. Lazell, 16 July
1960; ASFS V10235, Anse de la Riviere Salee, 10 km (airline)
NW Port Margot, Dept, du Nord, Haiti, E. Cyphale, 10 April
1966; MCZ 9365-66, Grande Riviere du Nord, Dept, du Nord,
Haiti, W. M. Mann, 1912; ASFS V10026, 6 km (est.) SW Li-
monade, Dept, du Nord, Haiti, E. Cyphale, 5 April 1966; ASFS
V10167, 1 mi (1.6 km) E Terrier Rouge, Dept, du Nord, Haiti,
R. Thomas, 8 April 1966.

Associated specimens. — Hmtv. Dept, du Nord Quest, Mole St.
Nicholas (ASFS V49584-85); Ballade, 5.5 mi (8.8 km) S Port-
de-Paix, 100 ft (31 m) (ASFS V49902-48); 5.3 mi (8.5 km) SE
Port-de-Paix (ASFS V46974-79); 0.8 mi (1.3 km) NW Ballade
(ASFS V47070-77); Ballade, 12.3 mi (19.7 km) NW Bassin Bleu
(ASFS V46954-81, ASFS V46989-7000); Deux Gar9ons, 7.7 mi
(12.3 km) NW Bassin Bleu, 200 ft (61 m) (ASFS V46949-53);
1.1 mi (1.8 km) SE Bassin Bleu, 400 ft (122 m) (ASFS V46948).
Dept, du Nord, 1 mi (1.6 km) E Cormier Plage (ASFS V47424,
ASFS V47427); Plage Rival, just N Cap-HaYtien (ASFS V39273);
Carrefour La Mori (ASFS V50263-64, ASFS V50283-87); 2.2
mi (3.5 km) S Plaisance, 1,100 ft (336 m) (ASFS V40 172-76,
ASFS V45913-15, ASFS V50280); 2.3 mi (3.7 km) S Plaisance,
1,100 ft (336 m) (ASFS V40347); 0.2 mi (0.3 km) W Gaubert,
850 ft (259 m) (ASFS V47431-33, ASFS V47458-93); 1.2 mi
(2.1 km) NE Dondon, 1,400 ft (427 m) (ASFS V38604-06); Bois
Neuf, 4. 5-5.4 mi (7. 2-8. 6 km) SE Dondon, 1,300 ft (397 m)
(ASFS V47844, ASFS V47852-56, ASFS V4794 1 , ASFS V48298-
486); 8.2 mi (13.1 km) E Terrier Rouge (ASFS V39025). Dipt,
de I'Artibonite, Ennery, 1 ,000 ft (305 m) (ASFS V40 152-61, ASFS
V40313-21, ASFS V45848-50, ASFS V45919-29, ASFS
V4772 1-47, ASFS V47804-21); 1 .2 mi ( 1 .9 km) W Ennery, 1,100
ft (336 m) (ASFS V40 186-2 10, ASFS V40330, ASFS V44900-
27, ASFS V459 1 8, ASFS V47698-7 1 8, ASFS V47748-96, ASFS
V50140); Terre Sonnain, 1 mi (1.6 km) N Les Poteaux, 400 ft
(122 m) (ASFS V40241^2, ASFS V40300-01, ASFS V46405-
07); 1 1 mi (17.6 km) N Carrefour Joffre, 600 ft (183 m) (ASFS
V40419, ASFS V40433, ASFS V44847-51); 14.5 mi (23.2 km)
N Carrefour Joffre (ASFS V40427).

Definition.— A subspecies of S. difficilis charac-
terized by a combination of low number of dorsal
scales (22-33) between axilla and groin, low number
(37-50) of midbody scales, female shoulder pattern
consisting of a relatively large and prominent dark
brown to black scapular patch and a pair of pale
ocelli (Fig. 2C), and males usually without both patch
and ocelli, although both are at times indicated.

Distribution. — Haiti, from Bombardo-
polis and Mole St. Nicholas on the Presqu’ile du
Nord Quest in the west, east to the vicinity of Ter-
rier Rouge (probably as far as the Riviere Massacre),
and inland to the vicinities of Grand Riviere du
Nord, Dondon, Limonade, Ennery and Terre Son-
nain.

Description of holotype.— An adult female with a
snout-vent length of 31 mm, tail length 25 mm.

12

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

distal half regenerated; dorsal scales axilla to groin
29, ventral scales axilla to groin 31, midbody scales
49, supralabials to mid-eye 3/3, 1 internasal, fourth
toe lamellae 1 1 , gular, chest, and ventral scales
smooth.

Dorsal ground color dark brown, back streaked
with dark brown and yellowish, with a rather defi-
nite longitudinally striped pattern; head pattern
trilineate, the median line very narrow on the snout,
expanding behind the eyes where it is abruptly con-
stricted, then once more expanded and constricted
to give a more or less subcircular figure, then once
more expanded and leading to the scapular patch
which is black and large and with a pair of cream
ocelli at its posterior comers; lateral head lines across
lores, through eye, and onto neck as far posteriorly
as the scapular patch, where they disappear; ven-
trolateral lines from ear over shoulder to groin,
thereby giving a lineate appearance to the lower sides;
tail longitudinally marked with a diffuse series of
dark longitudinal dashes; venter gray, but heavily
overlaid with dark brown scale edges to give a squa-
mate appearance; iris brown.

Variation.— Ont hundred nineteen S. d. lycauges
have the following measurements and counts: larg-
est males (ASPS VI 0232, ASPS VI 0046, MCZ
63179) 32 mm snout-vent length, largest female
(MCZ 63191) 34 mm; dorsal scales between axilla
and groin 22-33 (26.4); ventral scales between axilla
and groin 24-35 (28.7); midbody scales 37-50 (43.4);
supralabials to mid-eye 3/3 (88 individuals), 2/4(1),
% (1), 4/4 (3); internasals 0 (9), 1 (94), 2(16); fourth
toe lamellae 8-13 (10.5; mode 1 1); gular scales usu-
ally smooth (100 individuals) but rarely keeled (4)
or partially keeled (1); escutcheon 4-8 (6.1) by lb-
24 (20.2).

Males are gray to gray-brown dorsally and have
the back covered with widely spaced dark scales in
most cases, although exceptional males lack this fea-
ture (MCZ 63179). The scapular patch and ocelli
are usually lacking, but a few males (MCZ 63181,
MCZ 63195) still have the ocelli vaguely indicated
as a retention of a portion of the juvenile pattern.
The head ground color may be faintly orange, dark
yellow-gray or orange-gray, but in other (less ma-
ture?) individuals, the head is unicolor with the dor-
sum. The venter is yellowish brown to gray, and
frequently the dark ventral scale edges give a squa-
mate appearance to the belly. The throats are vari-
able, from more or less immaculate but with a gray-
ish suffusion to heavily and densely dotted with dark
brown to black; the degree of throat spotting is not
necessarily correlated with the amount of dorsal ce-

phalic spotting or dotting, because some males (ASPS
V 1 0048) with heavily spotted heads have the throats
unmarked; the opposite is also true.

Pemales in general resemble the holotype, and
there is a strong tendency for the dorsal dark scales
to be aligned into a series of dorsal longitudinal
lines. The scapular blotch is dark and relatively well
developed for the species, and the two included ocel-
li are cream and often rather large and not dot-like.
Some large females have both dorsal and head pat-
terns very obscured; even in such females, the scap-
ular patch and ocelli are still evident. The ventral
ground color is gray to yellowish, but the heavily
pigmented scale edges give a dark squamate ap-
pearance to the venter. In some females, the throat
is yellowish, and there are rarely any markings other
than, at the extreme, a grayish suffusion or a few
tiny scattered dark gray flecks. In both sexes the iris
is brown.

Juveniles show the intensification of the female
pattern; the dorsum is distinctly lined longitudinal-
ly, and the scapular patch and ocelli are conspicu-
ous.

Comparisons. — In all body scale counts, lycauges
averages lower than nominate difficilis\ the differ-
ences in midbody means of 48.1 ± 0.7 (twice stan-
dard error of mean) in difficilis and 43.4 ± 0.6 in
lycauges are statistically significant. The two sub-
species are easily distinguished also by the lineate
pattern and large scapular patch with ocelli in ly-
cauges in contrast to the non-lineate pattern and
reduction of the patch and a single ocellus in diffi-
cilis. Males of the two forms are less easily distin-
guished chromatically, but male lycauges appear to
be less densely flecked with dark scales dorsally and
generally have the heads with more dorsal dotting
than do difficilis. The squamate ventral appearance
of both sexes of lycauges is not a common feature
of difficilis.

Despite the much longer series of lycauges than
difficilis, many less specimens of the former have
any keeling on the throat (five of 1 05 lycauges versus
30 of 65 difficilis). The two subspecies are about the
same size, although male lycauges do not appear to
reach so large a size as do male difficilis (32 versus
34), and female lycauges are slightly larger than fe-
male difficilis (34 versus 33). Modally, lycauges has
1 1 fourth toe lamellae, whereas difficilis has 10.

Remarks. — Sphaerodactylus d. lycauges is quite
common at Cap-Haitien and Bombardopolis, but it
appears to be rare in non-urban situations. Because
most of the Cap-Haitien specimens were native col-
lected, we have no precise ecological data on them.

1983

SCHWAKTZ-SPHAERODACTYLUS SYSTEMATICS I

13

but presumably most were taken within the city
itself. Elsewhere, in edificarian situations, speci-
mens have been taken under trash in cafeieres (Bal-
lade), under a fallen thatch roof (Deux Gar9ons),
and in a semi-xeric cafei^re-cacaoiire (Carrefour La
Mort), as well as under scattered leaf litter on a xeric
hillside (Cormier Plage). I have the impression that
this subspecies prefers edificarian situations and
elsewhere is found most commonly in mesic rather
than xeric areas, although it does not shun the latter
completely.

The distribution of Y. d. lycauges extends from
the Presqu’ile du Nord Quest east along the northern
Haitian littoral to between Terrier Rouge and Ouan-
aminthe, and thence inland (in the central portion
of this range) to Dondon, Ennery, and Terre Sonnain
and above Carrefour Joffre near Gonaives. How the
sphaerodactyls have reached the valley of the Ri-
viere d’Ennery and onto the southern slopes of the
Montagnes de Terre Neuve remains uncertain; we
have never collected the species at higher elevations
(1,000 m) in the area near Carrefour Marmelade so
that its route of entry to these interior areas is not
directly across the Chaine de Marmelade. Maxi-
mum elevation (397 m) is in the vicinity of Dondon,
but the specimens from Bombardopolis show that
in the west these geckos ascend to elevations of
slightly less than 500 m on the Plateau de Bombar-
dopolis in the Montagnes du Nord Quest. The east-
ern limit of S. d. lycauges appears to be the Riviere
Massacre, but in the Monte Cristi region in the Re-
publica Dominicana, on the east side of the Riviere
Massacre, there seems to be some genetic effect of
lycauges (see later discussions). Aside from the very
short series from Hinche in central Haiti, S. difficilis
remains unknown from elsewhere in Haiti.

The long series from Bombardopolis may repre-
sent still another subspecies of S. difficilis. Although
there appear to be no scale differences between the
Bombardopolis and more eastern samples, the
Bombardopolis females to a large extent lack a dor-
sal lineate pattern and seem (as preserved) paler in
general than the much darker lots from the type-
locality and elsewhere (including Port-de-Paix and
Mole St. Nicholas) on this coast.

Elymology.— The name lycauges is from the Greek meaning
“at the gray twilight,” a reference to the time of activity of these
geckos.

Sphaerodactylm difficilis euopter, new subspecies

Holotype. — Cyi 54142, an adult female, from the
vicinity of Palmiste, He de la Tortue, Haiti, one of

a series collected by natives for C. Rhea Warren on
15-17 August 1 970. Qriginal number ASPS V20223.

Paratypes.-AS¥S V20213-16, ASFS V20222, ASFS V20224-
27, ASFS V20237, CM 54143^7, MCZ 125643^9, USNM
194019-25, same data as holotype.

Definition.— subspecies of S. difficilis charac-
terized by a combination of moderate number of
dorsal scales (24-31) between axilla and groin, low
number (41-50) of midbody scales, both male and
female shoulder patterns consisting of a large black
scapular patch with two included white ocelli (Fig.
2D), and females with a dorsal pattern of four lon-
gitudinal orange lines.

Distribution. — \\q de la Tortue, Haiti.

Description of holotype.— adult female with a
snout-vent length of 29 mm, tail length 19 mm;
dorsal scales axilla to groin 28, midbody scales 46,
supralabials to mid-eye 3/3, 1 intemasal, fourth toe
lamellae 1 1, gular, chest, and ventral scales smooth.

Dorsal ground color tan with four longitudinal
orange lines in life, of which the lateralmost on each
side begins at the eye and extends almost to the
groin, the two dorsalmost begin abruptly behind the
black scapular patch, the lines fairly conspicuously
outlined with dark brown to black; head pattern
trilineate but somewhat blurred, the pale interspaces
between the lines with dark brown stippling; scap-
ular patch large, black, roughly triangular with its
apex pointed anteriorly and including two white
ocelli, the lateral comers of the triangle abutting
upon the lateralmost of the orange longitudinal lines,
and the entire shoulder figure in a clear tan area;
tail vaguely lined dorsally with tan and brown; ven-
ter yellow-gray, with some dark brown scale-edging
posteriorly to give a squamate appearance; throat
concolor with venter and with some vague scattered
grayish flecks; iris yellow.

Variation. — ThirXy S. d. euopter have the follow-
ing measurements and counts: largest males (ASFS
V20215, ASFS V20216) 27 mm snout-vent length,
largest female (CM 54142 — holotype) 29 mm; dor-
sal scales between axilla and groin 24-31 (27.8);
ventral scales between axilla and groin 23-3 1 (26.7);
midbody scales 41-50 (45.3); supralabials to mid-
eye 3/3 (23 individuals) or 4/4 ( 1 ); internasals 1 (29)
or 2 (1); fourth toe lamellae 9-14 (10.9; mode 1 1);
throat scales almost always smooth (one specimen
has the throat scales partially keeled); escutcheon
4-6 (4.7) by 10-24 (17.0).

Males are tan to (rarely) brown dorsally with a
combination of isolated dark brown to black scales
and vague longitudinal buffy to tan lines. The heads

14

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

and throats are orange; the tops of the heads are
variably spotted with black, and the throats are
heavily spotted with black; the degree of throat spot-
ting is variable and correlated with the degree of
dorsal head spotting. The venter is dark gray and
has dark brown scale edges to give a squamate ap-
pearance. All (nine) males have a conspicuous black
scapular patch with two included white ocelli; the
patch is basically triangular with its apex pointed
anteriorly, but it tends to be somewhat cordate in
shape (indented along its posterior edge) in some
specimens (MCZ 125649), and it may be somewhat
reduced in intensity but nonetheless quite obvious
(MCZ 125643).

Females have the dorsum tan to rich brown with
four longitudinal orange lines, outlined with dark
brown to black. The head pattern is trilineate but
in most adult females is somewhat blurred with ad-
ditional dark brown stippling in the tan to buIFy-tan
interspaces between the three head lines. The scap-
ular patch is large, black, and has two included white
ocelli. Although the patch is usually triangular, it
may be cordate as in the males. The pale head lines
may terminate anterior to the black patch and join
transversely with each other by means of a pale bar
(ASFS V20225). The preocular portions of all dark
head lines are especially diffuse, and the median line
is obscure or even absent in most specimens. The
ventral color varies between yellow-grey to flesh,
and the dark scale edges give a squamate appearance
to the belly. The iris in both sexes is yellow.

Juveniles are patterned like females, but the pat-
tern is more vivid, and the dorsal longitudinal lines
are especially distinct (but less bright) than in the
adults.

Comparisons.— ThQTQ is no difficulty in separat-
ing euopter from nominate difficilis, because in the
latter subspecies females have a small black scapular
patch with only a single ocellus, and both ocelli and
patch are absent in male difficilis. The differences
between adjacent lycauges on the Haitian mainland
and euopter on Tortue are less obvious but never-
theless are quite striking. Male lycauges lack a scap-
ular patch and ocelli, whereas these features occur
in all euopter. Although female lycauges tend toward
a lineate dorsal pattern, it is never so discrete and
striking as in female euopter. The heads of male
lycauges are faintly orange, dark yellow-grey, or or-
ange-gray, whereas those of male euopter are much
brighter orange. The two subspecies also differ in
maximum size; male lycauges reach a size of 32 mm,
females 34 mm, in contrast to 27 mm and 29 mm

in male and female euopter. In general, lycauges is
a larger and huskier lizard than is euopter. In ad-
dition, only in euopter, of all the subspecies of 5.
difficilis, do both sexes have a well developed scap-
ular patch and ocelli, and the name euopter is an
allusion to this condition.

Remarks.— S. d. appears to be fairly com-

mon on He de la Tortue, at least in the Palmiste
area. Warren stated that this region is mesic, but
because the specimens were taken by natives, we
have no concrete data on the habitats involved.

There is little doubt that euopter is a local insular
derivative of lycauges. The two subspecies resemble
each other in the female lineate pattern (although
this is much less well expressed in the latter sub-
species) and the distinctly squamate appearance of
the venter because of the darkly pigmented scale
edges. However, euopter is quite distinctive when
compared directly with lycauges.

Sphaerodactylus difficilis typhlopous,
new subspecies

//otoype. — USNM 166966, an adult female, from
3 km NE Sosiia, Puerto Plata Province, Republica
Dominicana, one of a series collected by Richard
Thomas on 1 5 October 1 963. Original number ASFS
V1660.

Paratypes.- ASFS V 1 654-56, ASFS V 1 66 1-63, ASFS V 1 666-
69, CM 52260-63, USNM 167273, same data as holotype; USNM
167274, same locality as holotype, hatched from egg taken 15
October 1963 by R. Thomas; USNM 167276, Sosua, Puerto Plata
Province, Republica Dominicana, A. Schwartz, 1 9 October 1963;
ASFS V1688, 8 km E Imbert, 1,100 ft (336 m), Puerto Plata
Province, Republica Dominicana, R. Thomas, 17 October 1963;
ASFS V32 188-208, Hojas Anchas, 9.0 mi (14.4 km) NE Alta-
mira, ca. 800 ft (244 m), Puerto Plata Province, Republica Do-
minicana, D. C. Fowler, A. Schwartz, B. R. Sheplan, 27 October
1971; ASFS V33734— 67, 3.2 mi (5. 1 km) SE Caspar Hernandez,
Espaillat Province, Republica Dominicana, D. C. Fowler, B. R.
Sheplan, 10 November 1971; ASFS V33769-80, 3.0 mi (4.8 km)
NW Caspar Hernandez, Espaillat Province, Republica Domini-
cana, D. C. Fowler, B. R. Sheplan, 10 November 1971.

Associated specimens. — Republics. Dominicana: Monte Cristi
Province, 1 km S Palo Verde (ASFS VI 339-46, ASFS V 17640-
44); 4 km SE Monte Cristi (ASFS VI 7632); Cana, 14.4 mi (23.0
km) NW Mao, 300 ft (92 m) (ASFS V33253). Cayos Siete Her-
manos, Isla Monte Crande (ASFS V 17705-09); Isla Monte Chico
(ASFS VI 769 1-94); Isla Muertos (USNM 76718-23). Vaherde
Province. 5.9 mi (9.4 km) N Cruce de Cuayacanes (ASFS V32023).
Santiago Rodriguez Province, 1 .8 mi (2.9 km) W Los Quemados,
500 ft (153 m) (ASFS V32069, ASFS V33476); 3.3 mi (5.3 km)
W Los Quemados, 850 ft (259 m) (ASFS V33474-75). Santiago
Province, 3.4 mi (5.4 km) SE Los Montones, Rio Bao, 1,600 ft
(488 m) (ASFS V3378 1); Rio Bao, 5 km SE Los Montones Abajo,
610 m (ASFS V4 100 1-03).

1983

SCn^N A^WIZ-SPHAERODACTYLUS SYSTEMATICS 1

15

Definition.— \ subspecies of S. difificilis charac-
terized by a combination of high number of dorsal
scales (25-34) between axilla and groin, moderate
number of midbody scales (41-53), female shoulder
pattern of scapular blotch and ocelli absent or vague-
ly indicated (Fig. 2E), and males without indication
of scapular patch or ocelli.

Distribution. — K&'pnhhca. Dominicana, from the
vicinity of Monte Cristi in the west, east as far as
Caspar Hernandez, and inland to the vicinity of Los
Quemados and Los Montones in Santiago Rodri-
guez and Santiago provinces; specimens from the
western Valle de Cibao and the northern slopes of
the interior Cordillera Central intergradient between
typhlopous, dijficilis and probably lycauges (see dis-
cussion); also the Cayos Siete Hermanos off the Do-
minican coast at Monte Cristi.

Description of holotype.— An adult female with a
snout-vent length of 31 mm, tail length 23 mm, tail
almost completely regenerated; dorsal scales axilla
to groin 25, ventral scales axilla to groin 26, mid-
body scales 48, supralabials to mid-eye 3/3, 1 in-
temasal, fourth toe lamellae 13, gular, chest, and
ventral scales smooth.

Dorsal ground color medium brown with scat-
tered darker brown and buffy dots giving a flecked
appearance to the back. Three cephalic lines more
or less diffuse, the median line very obscure and on
the snout hardly differentiable from the tan head
ground color; median line reduced on neck to two
or three small, isolated, dark, more or less subcir-
cular areas followed by two small blackish markings
in the area of the scapular patch; ocelli and patch
absent; ventrolateral neck stripe present and as bold
as lateral head stripe; throat gray in life, venter gray-
ish.

Variation. — The series of 85 S. d. typhlopous has
the following measurements and counts: largest
males (ASPS V32 1 90, ASPS V32 1 92) 33 mm snout-
vent length, largest female (ASPS V33747) 32 mm;
dorsal scales between axilla and groin 25-34 (30.3);
ventral scales between axilla and groin 24-36 (30.0);
midbody scales 41-53 (46.8); supralabials to mid-
eye 3/3 (60 individuals); intemasals 0(1), 1 (79), 2
(4), 3(1); fourth toe lamellae 9-15 (12.1; mode 1 1);
throat scales almost always smooth (70 individuals)
but rarely keeled (10) or partially keeled (5); es-
cutcheon 3-6 (4.7) by 18-27 (21.8).

Males have the dorsum pinkish gray, tan, grayish
brown, or brown, and, as full adults, have the back
covered with a rather coarse marbling which ap-
parently represents an expansion and fusion of a

more orthodox salt-and-pepper pattern. The dark
body pattern often extends onto the upper surface
of the head, which may also be dotted with pale gray
to buffy dots. The head is dull yellowish to bright
yellow, and the dark head markings may be so ex-
tensive as to leave only a reticulum of the yellowish
cephalic ground color exposed. The throats are bright
orange, yellow-orange, gray, yellow-gray, or gray,
and are either immaculate or have a dark brown
pattern composed of a reticulum or isolated spots.
The venter is grayish brown to yellowish gray. Only
one adult male has an indication of a scapular pat-
tern; in this specimen (ASPS V3 2 1 9 1 ) there is a very
small black scapular patch and two white ocelli which
are peripheral to the patch remnants.

Females are in general like the holotype in color
and pattern, but females from Hojas Anchas have
a small black scapular patch and two white ocelli.
One female and one juvenile from Caspar Hernan-
dez have one tiny white ocellus but no dark scapular
patch. Likewise, females from the western and
southern portions of the range of typhlopous (Monte
Cristi, Palo Verde, Los Quemados) have one or two
small ocelli and a restricted black or dark scapular
patch. These specimens are discussed below. The
females are grayish to brown dorsally, and full adults
have the back variously marbled with dark brown
and tan with buffy or orange spots, so that the entire
effect is a coarse salt-and-pepper condition. The three
head lines are present, but the median line shows a
strong tendency to be reduced on the snout, or at
least very obscured on that portion of the head, due
to the deposition of inter-stripe pigment. Throats
are pale yellow to gray, usually immaculate, but in
some specimens the throat is heavily clouded with
gray or may even have a few to many dark brown
to black flecks or spots. The venters are grayish to
white and not obviously squamate. The iris is yel-
low.

Juveniles show an intensification of the female
pattern. A hatchling from Sosua (ASPS V2031) is
dark brown with scatted pale ocelli over the back
and sides, and of two juveniles (ASPS V32207-08)
from Hojas Anchas, one has a small dark scapular
patch and two ocelli, whereas the other has neither
of these features indicated. Even in these very young
juveniles (snout-vent lengths 15 to 18 mm), there
is a tendeney for the head lines to be obscure as in
adult females.

Comparisons.— S. d. typhlopous females are
readily differentiated from female dijficilis, lycauges,
and euopter, in that these three subspecies have

16

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

scapular patches with one or two ocelli, whereas no
scapular pattern occurs except under very redueed
conditions in typhlopous. Likewise, male typhlopous
differ from male euopter in that they lack a scapular
patch and ocelli, as well as being much more coarse-
ly patterned dorsally. Male typhlopous resemble male
lycauges and difficilis, but typhlopous males are more
coarsely patterned dorsally and have (as full adults)
more heavily spotted heads with a bright inter-spot
reticulum. Female typhlopous and euopter are also
distinguishable by the presenee in the latter of four
longitudinal orange lines.

In all body scale counts, typhlopous averages be-
tween difficilis and lycauges. In midbody scale eounts,
the mean of typhlopous (46.8 ± 1.5) is signihcantly
different from lycauges (43.4 ± 0.6), but not from
euopter (45.3 ± 1.1) or difficilis (48.1 ± 0.7). The
mean (12.1) of fourth toe lamellae is greater than
the means of difficilis, lycauges, and euopter, al-
though the mode of 1 1 in typhlopous is identical
with the modes in lycauges and euopter; difficilis has
a mode of 10.

Remarks. — S. d. typhlopous occupies the north-
ern Dominican lowlands north of the Cordillera
Septentrional (on whose southern slopes occurs S.
d. difficilis). Present material does not demonstrate
intergradation between these two subspecies in
southern Puerto Plata Province, but almost certain-
ly intergrades will be encountered in this region be-
cause typhlopous occurs (farther to the west at Im-
bert and Hojas Anchas in the northern foothills of
this range) at elevations of 1,100 ft and about 800
ft (336 and 244 m) and very likely ascends to the
northern slopes of the Septentrional to intergrade
with S. d. difficilis.

Elsewhere, I interpret specimens from at least the
western Valle de Cibao and the northern flank of
the Cordillera Central as being intergradient be-
tween difficilis, typhlopous, and lycauges. Through-
out mueh of this area of presumed intergradation,
there are only single specimens from most localities.
The exception is south of Palo Verde, whenee there
are 13 lizards. A few individuals from the Monte
Cristi-Palo Verde area have a small scapular patch
and one (sometimes two) tiny gray ocelli. A female
from north of Cruce de Guayacanes has a reduced
black scapular patch and two peripheral ocelli. Two
females from Los Quemados have a much reduced
dark brown scapular patch with either one or two
tiny gray ocelli. However, a female from Cana has
neither patch nor ocelli. All in all, we suggest that
these specimens show both influence of difficilis
(small pateh, one oeellus) and lycauges (large patch,

two ocelli) upon an otherwise typhlopous popula-
tion. The western portion of the Cordillera Septen-
trional is low and arid (in contrast to the mesic
forested uplands of this range in its eastern and cen-
tral portions), and S. d. typhlopous has crossed the
mountains in this western region. 5. d. difficilis oc-
curs in the extreme eastern portion of the xeric Valle
de Cibao in the vicinity of Santiago. Probably this
subspecies occurs farther to the west in this valley
and finally meets typhlopous at its western extrem-
ity. However, the presenee of a reduced scapular
patch and two ocelli is not characteristic of either
typhlopous or difficilis-, it is characteristic of the more
western lycauges, which is known from very elose
(45 km at Terrier Rouge) to the Dominieo-Haitian
border where intergradient specimens occur. The
major geographie barrier between lycauges and the
intergradient speeimens is the Riviere Massacre, and
it seems likely that there has been some genetic ex-
change across this river between lycauges and typh-
lopous.

The topotypic series of S. d. typhlopous was se-
eured in seagrapes (Coccoloba) and almonds (Ter-
minalia) along the coast, and a single juvenile was
found in a hotel bungalow at Sosua. The lizard from
Imbert was taken under palm trash on a mesic hill-
slope, and geckos were secured at Palo Verde in dry
logs, under termitaria, and in piles of dry wood,
sticks, leaves, and palm slats in riverine woods and
adjacent to a banana grove. The lizard from Monte
Cristi was taken in old palm trash at a garbage dump
in Acac/a-cactus lowlands. The long series from Ho-
jas Anehas was taken from within a staek of very
old logs in the sheltered yard of an abandoned sugar
mill, and those from Gaspar Hernandez were se-
eured in coastal Cocos trash.

Perhaps the most unusual situation for S. difficilis
is shown by the specimens from Los Quemados.
One of these (ASFS V32069) was taken from under
bark and a woody bracket-fungus 6 ft (1 .8 m) up on
a leguminous tree in rather mesic riverine woods in
otherwise very xeric country. Another lizard was
seen 12 ft (3.7 m) up in a similar tree at the same
locality. Both trees were alive but riddled with ter-
mites and had cavities filled with termitaria. A third
lizard was taken from a hollow deeayed tree limb
at this locality. Of two lizards (ASFS V3 3474-75)
from this same region, one was taken under the bark
of a standing dead tree in a shaded but not mesic
Acacia ravine, and the second was secured in trash
at the base of a tree. The occurrence of S. difficilis
in any situation above the ground surface is unusual,
and it is remarkable that three lizards were seen or

1983

KKYZ- SPHAERODACTYLUS SYSTEMATICS 1

17

taken in this situation in the Los Quemados area.
The generally xeric habitats of the western Cibao
may restrict S. difficilis to more mesic situations in
the valley and may even have compelled the lizards
to becoming partially arboreal in this region. The
specimens from Los Montones, which lies at an el-
evation of 1 ,600-2,000 ft (488-6 1 0 m), were native-
collected; the area there is mesic deciduous canopy
forest, quite different from areas occupied by S. d.
typhlopous in the floor of the Valle de Cibao.

S. difficilis from the Cayos Siete Hermanos (15
specimens from three islets) we group tentatively
with typhlopous. These lizards do not differ in scale
counts from our series of typhlopous, and females
lack scapular patches and ocelli as does that sub-
species. There is, however, a strong tendency for the
females and juveniles to be lineate dorsally (ASFS
VI 7694, ASFS VI 7708, USNM 76718). The loca-
tion of the Siete Hermanos, off the northern coast
near the Bahia de Manzanillo, and thus close to the
ranges of both lycauges to the west and typhlopous
to the east, and the presence in this region of the
mouths of the Riviere Massacre (which separates
the ranges of lycauges and typhlopous) and the Rio
Yaque del Norte, suggests that these islands have
been colonized fortuitously by individuals of both
subspecies. Thus the lineate pattern of female and
juvenile Siete Hermanos S. difficilis may be the re-
sult of advent of lycauges upon an otherwise more
or less typhlopous population.

S. d. typhlopous is known to be sympatric with S.
darlingtoni at one locality (north of Cruce de Guay-
acanes) in the western (but very mesic) portion of
the Cordillera Septentrional; at this locality, S. dar-
lingtoni far outnumbers V. difficilis. Presumably the
two species occur sympatrically elsewhere in the
central and western Cordillera Septentrional, but S.
darlingtoni has not as yet been collected at La
Cumbre, where S. difficilis is extremely abundant.
The two species are easily distinguished in this re-
gion, since S. d. darlingtoni has a greater number
of midbody scales (48-59) than does S. d. typhlo-
pous (41-53) and is a much smaller lizard (males
25, females 29 snout-vent length) than is S. d. typh-
lopous (males 33, females 32 snout-vent length).

Etymology.— The name typhlopous from the Greek for “step-
ping in blindness,” an allusion to the usual absence of ocelli in
this subspecies.

Sphaerodactylus difficilis peratus, new subspecies

Hoiotype. — CM 52264, an adult female, from 5
km NW Los Yayales, Maria Trinidad Sanchez Prov-
ince, Republica Dominicana, one of a series col-

lected by Albert Schwartz and Richard Thomas on
30 October 1963. Original number ASFS VI 923.

Paratypes.- ASFS y 1919-22, ASFS VI 924-29, same data as
hoiotype; CM 52265-70, USNM 167277-81, same locality as
hoiotype, A. Schwartz, 18 July 1968; 1.4 mi (2.2 km) SE Los
Yayales, Maria Trinidad Sanchez Province, Republica Domini-
cana, A. Schwartz, 24 November 1971; MCZ 1 19368, 6 km SE
Nagua, Maria Trinidad Sanchez Province, Republica Domini-
cana, R. Thomas, 26 October 1963; MCZ 1 19367, 3.3 km S
Cabrera, Maria Trinidad Sanchez Province, Republica Domini-
cana, R. Thomas, 29 November 1964; LSUMZ 21915-19, 4.8
km S Cabrera, Maria Trinidad Sanchez Province, Republica Do-
minicana, D. W. Buden, R. Thomas, 29 November 1964; ASFS
V33704-31, 2.1 mi (3.4 km) NE Rio San Juan, Maria Trinidad
Sanchez Province, Republica Dominicana, D. C. Fowler, B. R.
Sheplan, 10 November 1971; ASFS VI 6069, 4 km N Azucey,
Maria Trinidad Sanchez Province, Republica Dominicana, A.
Schwartz, 23 December 1968; ASFS V34161, 1.0 mi (1.6 km) S
Cano Abajo, Maria Trinidad Santhez Province, Republica Do-
minicana, native collector, 24 November 1971; ASFS V34239-
59, 1.0 mi (1.6 km) S Cano Abajo, Maria Trinidad Sanchez
Province, Republica Dominicana, native collectors, 26 Novem-
ber 1971; MCZ 43395-96, Sanchez, Samana Province, Republica
Dominicana, P. J. Darlington, July 1938.

Definition. — A subspecies of S. difficilis charac-
terized by a combination of very high number of
dorsal scales (28-40) between axilla and groin, very
high number of midbody scales (46-60), female
shoulder pattern of scapular blotch and ocelli absent
(Fig. 2F), and males without indication of scapular
patch and ocelli.

Distribution. — Republica Dominicana, the north-
eastern coast and inland in mesic situations, from
Rio San Juan on the north to Azucey on the south,
and east (apparently) to Sanchez on the base of the
Peninsula de Samana.

Description of hoiotype. — An adult female with a
snout-vent length of 28 mm, tail length 23 mm,
almost completely regenerated; dorsal scales axilla
to groin 38, ventral scales axilla to groin 29, mid-
body scales 58, supralabials to mid-eye 4/4, inter-
nasals 1, fourth toe lamellae 13, gular scales keeled,
chest and ventral scales smooth.

Dorsal ground color dark brown, heavily flecked
with black to dark brown, the flecking more or less
aligned into a series of about seven longitudinal
stripes, all of which are narrow and relatively in-
distinct; head pattern trilineate but median line ab-
sent on snout, thrice constricted behind eyes, and
disappearing abruptly before reaching scapular area
which is occupied by a series of tiny black longi-
tudinal flecks which may be remnants of the scap-
ular patch; ocelli absent; snout heavily strippled with
brown; venter and throat cream, throat flecked with
numerous tiny brown flecks.

18

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Variation. — The series of 83 S. d. peratus has the
following measurements and counts: largest male
(ASFS V34240) 33 mm snout-vent length, largest
female (ASFS V33721) 33 mm; dorsal scales be-
tween axilla and groin 28-40 (34.0); ventral scales
between axilla and groin 26-35 (29.4); midbody
scales 46-60 (5 1 .8); supralabials to mid-eye 3/3 (43
individuals), 2/2 (1), 3/4 (5), 4/4 (3), 4/5 (1); inter-
nasals 0 (2), 1 (81); fourth toe lamellae 10-14(12.4;
mode 13); gular scales smooth (41 individuals),
weakly keeled (1 5) or strongly keeled (25), chest and
ventral scales smooth; escutcheon 3-8 (5.0) by 8-
30 (23.2).

Males are brownish gray to tan or dark brown,
regularly heavily spotted dorsally with dark brown;
the heads are yellowish with dark brown spots, often
with additional white, cream, or pale yellow frost-
ing. One male also had some scattered yellow scales
between the brown spots on the back. Throats are
yellow to orange-yellow, usually immaculate; a few
males have the throat marked with discrete dark
brown dots or marbling. The ventral color varies
from grayish brown to tannish gray or yellowish
gray. There is no indication of the scapular patch
or ocelli.

Females are much less heavily marked than males,
and the dorsum is salt-and-pepper with the dark
scales often arranged into more or less distinct lon-
gitudinal lines. The head pattern is distinct behind
the eyes but is often obscured on the snout. In some
individuals (ASFS VI 924, ASFS V4257, both adults)
the median head line continues down the midline
of the back, and a young individual (ASFS V1925)
also shows this condition very clearly. The dorsum
is light gray to dark brown, and the dorsal markings
are thus variably visible against a dark ground. The
scapular patch and ocelli are absent. The venters are
light gray, and the throats are cream, pale yellow,
or yellow, with a few tiny scattered brown flecks.
The iris is yellow to brown or gray-brown in both
sexes.

Juveniles are patterned like females, but all pat-
tern elements are more distinct; as noted above, the
lineate female pattern is expressed in some juveniles
(ASFS V14103, ASFS VI 925) even more clearly
than in the adults.

Comparisons. — \n lacking a scapular patch and
ocelli in females, S. d. peratus differs from difficilis,
lycauges, and euopter, all of which have these pat-
tern elements expressed. From typhlopous, which
also usually lacks the patch and ocelli, peratus differs
in much higher midbody scale counts (means 51.8

± 0.8 in peratus, 46.8 ± 1.5 in typhlopous). In fact,
the higher mean in Peratus midbody counts is sig-
nificantly different from those of all other subspecies
of S. difficilis.

The mean of dorsal scale counts in peratus is also
higher than those of all other subspecies. S. d. per-
atus also has a very high percentage of specimens
with keeled gular scales (including individuals with
at least some keeling); the percentage in peratus (49%)
is higher than that in difficilis (47%). Male N. d.
peratus are distinguishable from males of the other
subspecies in being heavily spotted dorsally.

Remarks.— k.X the type-locality and Nagua, N. d.
peratus was collected in palm trash along a sandy
beach, where the geckos were extremely abundant.
Elsewhere, lizards were taken under trash in Theo-
broma groves (Azucey, San Francisco de Macoris),
and under Cocos trash (San Francisco de Macoris,
Cano Abajo).

Shreve (1968) described S. clenchi as a species
distinct from N. difficilis', the former occurs on the
Peninsula de Samana. S. clenchi, surely a S. difficilis
derivative, differs from S. difficilis in that the former
lacks sexual dichromatism and has very small dorsal
scales (and thus has very high body scale counts).
It is interesting that S. d. peratus, that subspecies of
V. difficilis which occurs closest to S. clenchi (and
even apparently sympatrically with it at Sanchez at
the base of the Peninsula de Samana), has the small-
est scales and thus the highest body scale counts of
all the subspecies of S. difficilis.

Intergradation between peratus and typhlopous
remains as yet unknown. The two subspecies ap-
proach each other rather closely at Sosua and Caspar
Hernandez, a distance of about 35 km airline, yet
there is no especial tendency for specimens from
either of these localities to have higher or lower scale
counts than lizards from other localities well within
the respective ranges of either subspecies. Likewise,
material is still lacking between the Nagua-Los Yay-
ales- Azucey area on one hand and Los Bracitos on
the other— a distance of about 38 km airline. It
should be recalled that S. d. difficilis occurs at Los
Bracitos, a locality in the extreme eastern uplands
of the Cordillera Septentrional, whereas in this re-
gion, peratus is known only from lowland and more
or less coastal localities.

S. d. peratus is broadly sympatric with S. dar-
lingtoni. The local subspecies of the latter (S. d.
noblei) has dorsal, ventral, and midbody seale counts
reaching less high extremes than does peratus, but
the ranges in all these counts overlap broadly. How-

1983

scnw ARTZ- SPHAERODACTYLUS SYSTEM ATICS 1

19

ever, S. d. noblei is a much smaller lizard (snout-
vent lengths in males to 25 mm, in females to 27
mm) than is S. d. peratus (males and females both
to 33 mm). S. d. noblei is a dark brown lizard and
has a scapular pattern, whereas S. d. peratus is gen-
erally paler and lacks a scapular pattern.

Etymology.— The name peratus is from the Greek for “that
which may be passed over” in allusion to the pattern similarities
between this subspecies and 5”. d. typhlopous.

Sphaemductylus difficilis diolenius, new subspecies

— USNM 166967, an adult female, from
2 mi (3.2 km) SE San Cristobal, San Cristobal Prov-
ince, Republica Dominicana, one of a series col-
lected by Albert Schwartz and Richard Thomas on
15 June 1963. Original number ASPS V7777.

Paratypes.—ASFS V7776, ASFS V7778-80, same data as ho-
lotype; ASFS V34971-72, La Romana, La Romana Province,

Republica Dominicana, 17 November 1971, D. C. Fowler, B. R.
Sheplan; ASFS V28901-06, 1.8 mi (2.9 km) S Boca del Soco,
San Pedro de Macoris Province, Republica Dominicana, 16 July
1971, D. C. Fowler, A. Schwartz; USNM 167282-88, Villas del

Mar, 5 km E Guayacanes, San Pedro de Macoris Province, Re-
publica Dominicana, 5 June 1969, J. A. Rodgers, Jr., A. Schwartz;
LSUMZ 2 1 920-3 ! , 1 km W Guayacanes, San Pedro de Macoris
Province, Republica Dominica, 2 August 1968, J. K. Lewis, A.
Schwartz; MCZ 1 19369-70, 8 km E Boca Chica, San Pedro de
Macoris Province, Republica Dominicana, 21 July 1964, R.
Thomas; AMNH 49990-92, 31 mi (50.6 km) E Santo Domingo,
San Pedro de Macoris Province, Republica Dominicana, 1 Au-
gust 1935, W. G. Hassler; CM 52271-73, 7 mi (11.2 km) E Boca
Chica, San Pedro de Macoris Province, Republica Dominicana,
14 June 1963, D. C. Leber; MCZ 119371-72, EDO 7-5219-23,
Aeropuerto Punta Caucedo (=Aeropuerto Intemacional de las
Americas), Distrito Nacional, Republica Dominicana, 9 July ! 968,
R. K. Bobilin, J. K. Lewds, J. B. Ober, L. D. Ober, R. A. Ober,
A. Schwartz; ASFS V598, 5.5 km NE Guerra, Distrito Nacional,
Republica Dominicana, 22 August 1963, D.C. Leber; CM 52274-
77, 6.7 km S Bayaguana, Distrito Nacional, Republica Domini-
cana, 22 August 1963, D. C. Leber, R. Thomas; ASFS X7743-
53, 9.8 mi (1 5.7 km) E Santo Domingo, Distrito Nacional, Re-
publica Dominicana, 14 June 1963, R. F. Klinikowski, D. C.
Leber, R. Thomas; USNM 167289-93, Santo Domingo (Hotel
Jaragua), Distrito Nacional, Republica Dominicana, 1 3 June 1 963,
D. C. Leber, R. Thomas; CM 52278, Santo Domingo, old airport,
Distrito Nacional, Republica Dominicana, 1 7 July 1963, R.
Thomas; ASFS V2472, 8.5 km W Santo Domingo, Distrito Na-
cional, Republica Dominicana, 20 June 1964, R. Thomas; ASFS
V28487-90, 9.2 km W Rio Ozama, west of Santo Domingo,
Distrito Nacional, Republica Dominicana, 23 June 1971, D. C.
Fowler, A. Schwartz; MCZ 119373, 17 km NW Santo Domingo,
Distrito Nacional, Republica Dominicana, 24 July 1964, R.
Thomas; ASFS V4149, 17 km NW Santo Domingo, Distrito
Nacional, Republica Dominicana, hatched from egg between 1-
7 September 1964.

Associated specimens.— Dominicana: La Altagra-
cia Province, 1 2 km E Otra Banda (ASFS V17618); Higiiey (ASFS
V21832); 3.2 mi (5.1 km) W Higiiey (ASFS V759-62). El Seibo

Province. 7 km W El Cuey (ASFS V 1 7586-609); 1.1 mi (1.8 km)
W Miches (ASFS V28785); 3.3 mi (5.3 km) SW Miches (ASFS
X7900-01); 3 km N El Valle (ASFS V3159); 2.6 km N Hato
Mayor (ASFS V35291-92); 3.5 mi (5.6 km) S Sabana de la Mar
(ASFS X7938-39); Sabana de la Mar (ASFS V3 123-24); Cuevas
de Cano Hondo (ASFS V35285-90); north side of sheltering
peninsula of Bahia de San Lorenzo, approximately 2 km E of tip
(ASFS V3 152-56); 20.2 mi (32.3 km) NW, 3.4 mi (5.4 km) N
La Vacama, Playa de Guaco {ASFS V29393^14). San Pedro de
Macoris Province, 1 km W San Pedro de Macoris, west side of
Rio Magua (ASFS V41041-51). San Cristobal Province, 7 km
SE Yamasa, 122 m (ASFS V40926-28); 3.4 mi (5.4 km) W Sa-
bana Grande de Palenque (ASFS X7 178-79); 2 km E Juan Baron
(ASFS V28519-20). La Vega Province, 1.5 km W Jayaco, 183
m (ASFS V40862). Peravia Province, 4.8 mi (7.7 km) S Ban!
(ASFS V29441-5 1). Azua Province. 9.7 mi {15.5 km) E Azua
(ASFS X8067-94, ASFS VI 9335-51, ASFS V21 102-04); 2.7 mi
(4.3 km) W Azua (ASFS V3 1039-50); Monte Rio (ASFS V21 164);
4 km W, 6 km N Azua (ASFS V2 1168-71); Barreras (ASFS
V2 1272-358, ASFS V31033, ASFS V3 1069-1 75); 12.4 mi (19.8
km) SE Guanito, 900 ft (275 m) (ASFS V31338-41). San Juan
Province, San Juan (ASFS V402); 15 km SE San Juan (ASFS
V41 1); 15 km E San Juan (ASFS V21545^8). Sdnehez Ramirez
Province, 1 km SE La Mata (ASFS V18579, ASFS V33667).
Province unknown, Isla Pascal, Bahia de Samana (AMNH 41979-
85).

Definition.— A subspecies of S. difficilis charac-
terized by a combination of moderate number of
dorsal scales (23-35) between axilla and groin, mod-
erate number of midbody scales (40-53), female
shoulder pattern consisting of a pair of pale ocelli
and without a scapular patch (at best represented
by a diffuse dark bar between ocelli), and males
without indication of scapular patch but ocelli usu-
ally present (Fig. 2G).

Distribution. —Republica Dominicana, from the
Valle de San Juan (San Juan) and the Llanos de Azua
(Barreras) in the west, east to La Romana Province
(La Romana) along the coast, and inland as far as
La Vega Province (Jayaco) and Sanchez Ramirez
Province (La Mata), thence east to El Seibo Province
(Hato Mayor and north to Sabana de la Mar), east
along the coast of the Bahia de Samana as far as the
vicinity of Laguna Redonda and thence south to La
Altagracia Province (Otra Banda, Higiiey); presum-
ably also the subspecies at Hinche, Dept, de I’Ar-
tibonite, Haiti.

Description of holotype. — An adult female with a
snout-vent length of 30 mm, tail length 24 mm;
dorsal scales axilla to groin 3 1 , ventral scales axilla
to groin 27, midbody scales 44, supralabials to mid-
eye 3/3, 1 intemasal, fourth toe lamellae 1 1, gular,
chest, and ventral scales smooth.

Dorsal ground color medium tan with gray scales
arranged in a series of about five very broken Ion-

20

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

gitudinal lines and with scattered dull orange dorsal
ocelli. Three cephalic lines clear, the median one
broken in the interorbital region; scapular patch re-
stricted to a small bar between the two white ocelli;
ventrolateral postauricular stripe present and mod-
erately prominent; venter gray in life, throat with
scattered dark brown flecks.

Variation. — ¥o\xr hundred forty-six S. d. diolen-
ius have the following measurements and counts:
largest males (ASFS V3 1069, ASFS V3 1 338) 32 mm
snout-vent length, largest females (ASFS V21275,
ASFS V31094) 33 mm; dorsal scales between axilla
and groin 23-35 (28.0); ventral scales between axilla
and groin 24-37 (29.8); midbody scales 40-53 (46.6);
supralabials to mid-eye 3/3 (272 individuals), 3/4
(9), 4/4 (3); internasals 1 (400), 0 (18), 2 (22), 3 (3);
fourth toe lamellae 7-15 (10.9; mode 1 1); gular scales
usually smooth (417 individuals) but occasionally
keeled (7) or slightly keeled (14); escutcheon 3-8
(5.5) by 8-26 (17.8).

Because of its broad distribution, S. d. diolenius
is somewhat more variable both in dorsal ground
color and pattern than are any of the other subspe-
cies. Males have been recorded as tan, sandy, yel-
lowish tan, brown, dark brown, gray or even black
dorsally; the upper surface of the head is often or-
ange to grayish yellow, and may or may not have
dark brown spots or dots overlying the brighter
ground color. The dorsum is heavily spotted, the
dark brown spots arranged in a lineate pattern, al-
though this seems to be a condition more prevalent
in fully adult males. In some smaller males, the
dorsum is salt-and-pepper. A pair of pale yellow,
bufly, or white ocelli is very regularly present, at
times darkly outlined and rarely connected by a very
narrow and diffuse darker area which is a remnant
of the juvenile pattern. The throat varies from gray-
ish to yellow or orange, usually in the case of the
latter brighter colors corresponding in intensity to
the color of the head. Throat markings vary from
scattered brown flecks or dots to a more complete
dark reticulum; the degree of throat markings is cor-
related with the amount of dark head markings. The
ventral color was recorded as grayish brown, grayish
yellow, yellowish brown, pinkish, light gray, dull
gray-orange, gray, or flesh. Some males have a fairly
prominent pair of orange ocelli at the base of the
tail, and the tail is usually yellowish above and be-
low, often in strong contrast to the more sombre
tones of the dorsum and venter.

Females have been recorded as tan, dark brown,
wood brown, or yellowish tan; no females were re-

ported as gray dorsally. The pattern of the back
consists of dark brown scales arranged in a more or
less lineate pattern. The usual head pattern is boldly
trilineate, and the median line is seldom complete
anterior to the eyes. The scapular patch is absent or
at best restricted to a dark bar lying between the
ocelli which are pale yellow, white, or pale bufly.
Females often have scattered orange spots or bufly
flecking on the dorsum, and there may be a pair of
yellow ocelli at the beginning of the tail. The ventral
color was recorded as brown, dark brown, gray, yel-
lowish brown, pale yellow-gray, gray-orange, flesh,
or yellow-gray. The upper surface of the head and
the throat are not orange or yellow but are unicolor
with the dorsum and venter respectively. The throat
pattern consists of grayish to brownish flecking or
marbling; some females, especially subadults, lack
any prominent throat markings. The iris color is
regularly yellow or yellow-brown in both sexes.

Juveniles are colored and patterned like females;
the tail is orange dorsally and ventrally. In hatch-
lings, the tip the tail is white, followed by a broad
black band.

Comparisons.— S. d. diolenius has a moderate
number of midbody scales (mean 46.6 ± 0.3) and
in this character differs significantly from the sub-
species difficilis, lycauges, and peratus. The number
of dorsal scales between the axilla and groin in di-
olenius averages less than those in difficilis, typhlo-
pous, and peratus, and greater than those in lycauges
and euopter. Female diolenius, with a pair of ocelli
and (at best) a small scapular patch or with the patch
absent, differ from female difficilis, typhlopous, and
peratus, all of which have either one ocellus or no
ocelli, and the patch either well expressed or absent.
From euopter, female diolenius differ in having a
much more reduced scapular patch and in lacking
the distinct longitudinal orange lines. Female di-
olenius most closely resemble female lycauges in
northern Haiti, but lycauges has less midbody scales
(means 43.4 ± 0.6 versus 46.6 ± 0.3), and the ranges
of the two subspecies are widely separated by (for
the most part) the interior Dominican Cordillera
Central, although diolenius possibly extends into the
Plateau Central in Haiti (see Remarks). Even if the
Plateau Central is occupied by diolenius, this sub-
species and lycauges are probably separated by the
Massif du Nord. Male diolenius have a pair of pale
ocelli and thus are easily differentiated from males
of all other subspecies thus far described, with the
exception of euopter. Males of the latter subspecies
have a prominent black scapular patch with ocelli.

1983

SCH^N KKTZ- SPIIAERODACTYLUS SYSTEM ATICS I

21

whereas the patch is never present in diolenius. It
should also be noted that male lycauges occasionally
have ocelli present, much as in the regular fashion
of male diolenius.

Remarks. — S. d. diolenius is the most widely dis-
tributed of the subspecies, occupying most of eastern
and central-southern Repiiblica Dominicana. Shreve
( 1 968) noted the occurrence of “Y notatus difficilis"
at Hinche in central Haiti. I have examined the four
specimens (BMNH 1948.1. 4. 27-.30) upon which
Shreve based his record; the series consists of one
adult and one subadult female, and one adult and
one subadult male. Shreve (1968:5), quoting field
notes on this series, stated that the upper surfaces
were fawn brown, prominently spotted with jet black.
There were two white ocelli which lay lateral to a
pair of jet black spots. The two males now lack
scapular patches and ocelli (and thus dilfer from
most male diolenius) and have the dorsum with scat-
tered darker brownish dots. The larger male lacks
head pattern or dotting, and the smaller has some
faint brownish dotting on the head. Both are very
drab lizards. The larger female has the typical tri-
lineate diolenius head pattern, with the body heavily
dotted with dark brown, and with two very prom-
inent ventrolateral lines which extend to the fore-
limb insertion. The black scapular patch is reduced
to an elongate bar (which is nevertheless conspic-
uous) with a pair of pale ocelli. The smaller female
resembles the larger female closely in scapular pat-
tern, but the dorsum is more lineate than dotted.
There is nothing distinctive about the body scale
counts, except that the midbody scales for the small-
er female are 39, slightly less than the low count of
40 for the long series of diolenius. The internasals
are exceptionally variable, since in the series of four
lizards there are counts of 0, 1, 2, and 3. I include
these Haitian lizards with S. d. diolenius only pro-
visionally. The absence of ocelli in the males is cru-
cial, but on the other hand their absence may be
due to the length of time in preservative. The nearest
locality for diolenius (15 km E San Juan) lies some
1 20 km to the southeast of Hinche, and, although
it is not at all improbable that this subspecies con-
tinues from the Valle de San Juan across the inter-
national boundary onto the Plateau Central in Haiti,
more material from north-central Haiti may reveal
the presence there of another subspecies of S. dif-
ficilis.

No intergrades between diolenius and the adjacent
subspecies {difficilis, peratus) are known. The sub-
species difficilis and diolenius approach each other

at 7.5 mi NW Cruce de Pimentel and 1 km SE La
Mata, a distance of 20 km, and diolenius and peratus
occur at 1 km SE La Mata and 4 km N Azucey, a
distance of 30 km. S. d. diolenius also occurs ad-
jacent to S. clench i and sympatrically with S. sav-
agei\ details of these contacts are discussed in the
accounts of those species. S. d. diolenius is also sym-
patric with S. cochranae at the Bahia de San Lo-
renzo, where both species were taken in the Cuevas
de Cano Hondo.

Specimens of S. d. diolenius have been taken in
a variety of situations, including Cocos trash near
shore and under isolated palm fronds on mud ad-
jacent to a mangrove swamp, under the fallen thatch
of an abandoned native hut, under palm fronds and
trash on beach dunes and on sand flats behind dunes,
under rocks and cacao trash, under a log in a lime-
stone pasture, in rotting logs and under bark of fallen
dead trees, under a log in mesic woods, under a
thatch pile adjacent to a dry rice field, and in rock
rubble and human debris on a cave floor along the
edge of the cave wall. The lizards have been seen
actively running about in the morning in the leaves
on the floor of a mesic cacao grove. On three oc-
casions, S. d. diolenius were taken in human dwell-
ings. Altitudinal distribution is from sea level at
many localities to an elevation of 4 1 5 meters at San
Juan.

S. d. diolenius and S. altavelensis enriquilloensis
are broadly sympatric and under special circum-
stances syntopic in the southwestern portion of the
area occupied by S. d. diolenius. We have the
impression that in this region, diolenius is an in-
habitant of more shaded or mesic situations (al-
though these locales may not be wet) than does the
xerophilic enriquilloensis. East of Azua, both species
were taken together in the same piles of old Cocos
trash in a well shaded coconut grove— an artificial
oasis in an otherwise hot and dry cactus-desert. Of
the two species, S. a. enriquilloensis is much the
smaller (males to 26 mm, females to 28 mm) than
is diolenius (males 32 mm, females 33 mm). The
scale counts overlap broadly, but in general enri-
quilloensis has lower dorsal, ventral, and midbody
counts. In addition, enriquilloensis lacks a dark
scapular patch, a feature reduced but only at times
present in diolenius.

The short series from Isla Pascal in the Bahia de
Samana is referable to diolenius rather than to either
V. d. peratus or S. clenchi. The low scale counts,
and absent or very restricted scapular patch and
paired ocelli in females confirm that these specimens

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

22

are assignable to diolenius. We are unable to locate
Isla Pascal on any modem map but assume that it
is dose to the southern shore of the Bahia de Sa-
mana, adjacent to the known range of diolenius.

Etymology.— The name diolenius is from the Greek for “with
outstretched arms,” in allusion to the wide distribution of this
subspecies.

Sphaewdactylus difficilis anthracomus,

new subspecies

Holotype.—CM 52279, an adult female, from 1
km NE Paraiso, Rio Nizaito, Barahona Province,
Republica Dominicana, one of a series collected by
James A. Rodgers, Jr., Albert Schwartz, and James
B. Strong on 22 May 1969. Original number ASFS
V16941.

Paratypes.-A.STS VI 6937^0, ASFS VI 6942-45, same data
as holotype; ASFS V30438^1, same locality as holotype, 2 Sep-
tember 1971, J. R. Buffett, D. C. Fowler, A. Schwartz, B. R.
Sheplan; USNM 167294-97, 2 km SE Barahona, Barahona Prov-
ince, Republica Dominicana, 23 July 1963, D. C. Leber; LDO
7-5338—41, LDO 7-5364, 3 mi (4.8 km) S Barahona, Barahona
Province, Republica Dominicana, 11 July 1968, J. B. Ober, L.
D. Ober, R. A. Ober; ASFS V23422, 7 km SW Barahona, Ba-
rahona Province, Republica Dominicana, 4 January 1971, A.
Schwartz; ASFS V20958, 5 km S, 1 km W Barahona, ±200 ft
(6 1 m), Barahona Province, Republica Dominicana, 4 July 1969,
R. Thomas; ASFS V3 1014-1 8, Barahona, southern outskirts,
Barahona Province, Republica Dominicana, 1 2 September 1971,
D. C. Fowler, B. R. Sheplan; ASFS V30451-62, 3.3 mi (5.3 km)
NE La Cienaga, Barahona Province, Republica Dominicana, 2
September 1971, native collectors; LSUMZ 21933-36, MCZ
1 19374-80, 9 mi (14.4 km) SW La Cienaga, Barahona Province,
Republica Dominicana, 22 July 1963, D. C. Leber, A. Schwartz,
R. Thomas; AMNH 51466, Paraiso, Barahona Province, Re-
publica Dominicana, 1932, W. G. Hassler; CM 52280-81, 2 km
SW Paraiso, Barahona Province, Republica Dominicana, 1 Au-
gust 1963, P. A. Adams, R. F. KJinikowski; ASFS V1451-52, 2
km SW Paraiso, Barahona Province, Republica Dominicana,
hatched 25 September 1963 from eggs taken 1 August 1963;
ASFS V1615, 2 km SW Paraiso, Barahona Province, Republica
Dominicana, hatched 9 October 1963 from egg taken 1 August
1963; ASFS V30843-47, 1.9 mi (3.0 km) W Paraiso, 600 ft (183
m), Barahona Province, Republica Dominicana, 10 September
1971, D. C. Fowler, B. R. Sheplan; ASFS V35770, 3 km W
Paraiso, 500 ft (153 m), Barahona Province, Republica Do-
minicana, 25 December 1972, M. D. Lavrich; ASFS V31019,
1.9 mi (3.0 km) W Paraiso, 600 ft (183 m), Barahona Province,
Republica Dominicana, hatched 11 September 1971 from egg
taken 10 September 1971; ASFS V30948-49, 4.1 mi (6.6 km)
W Paraiso, 500 ft (153 m), Barahona Province, Republica Do-
minicana, 1 1 September 1971, D. C. Fowler; ASFS V30750-55,
0.5 mi (0.8 km) NE Caleton, 400 ft (122 m), Barahona Province,
Republica Dominicana, 8 September 1971, D. C. Fowler, A.
Schwartz, B. R. Sheplan; ASFS V30772-89, 1.5 mi (2.4 km) SW
Caleton, 200 ft (6 1 m), Barahona Province, Republica Domini-
cana, 8 September 1971, D. C. Fowler, A. Schwartz, B. R. She-

plan; LSUMZ 21932, 1.4 mi (2.2 km) NE Enriquillo, Barahona
Province, Republica Dominicana, 1 1 July 1968, L. D. Ober.

Associated specimens. — Repub\.ic\ Dominicana: Barahona
Province. 10 km SW Barahona (ASFS V40500); 3.3 mi (5.3 km)
NE La Cienaga (ASFS V39799-830, ASFS V399 16-24); 6 km
NE Paraiso (ASFS V39744^5, ASFS V39831-35); 6 km N En-
riquillo, 366 m (ASFS V42 166-67); 3 km N Enriquillo, 214 m
(ASFS V42233); 5 km SW Enriquillo (ASFS V42388-89); 7 km
SW, 0.5 km E Enriquillo (ASFS V425 19-22); 7 km SW, 1 km E
Enriquillo (ASFS V4249 1-500); 1 km SE La Lanza, 732 m (ASFS
V45027). Pedernales Province, 0.5 km E Juancho (ASFS V42485-
87).

Definition.— A subspecies of S. dijficilis charac-
terized by a combination of moderate number of
dorsal scales (24-33) between axilla and groin, high
number of midbody scales (45-56), female shoulder
pattern consisting of a pair of pale ocelli and usually
a very large dark scapular patch which is never bar-
like (Fig. 2H), and males with a pair of pale ocelli
and the scapular patch at times indicated.

Distribution.— The: eastern coast of the Peninsula
de Barahona, from the vicinity of Barahona on the
north, south to near Juancho; generally confined to
coastal situations or low elevations on the eastern
slopes of the Sierra de Baoruco, but also occurring
north of Enriquillo on the southern slopes of this
range and near La Lanza in the uplands.

Description of holotype.— An adult female with a
snout-vent length of 31 mm; tail length 22 mm,
almost completely regenerated; dorsal scales axilla
to groin 29, ventral scales axilla to groin 27, mid-
body scales 51, supralabials to mid-eye 3/3, 1 in-
temasal, fourth toe lamellae 1 1 , gular, chest, and
ventral scales smooth.

Dorsal ground color brown, randomly flecked with
black. Three cephalic lines dark brown on a buffy
ground, the lines very distinct and the median line
not touching the scapular figure posteriorly; scapular
patch very large and black, with a pair of large dis-
tinct white ocelli laterally, the entire complex fol-
lowed by a dark brown transverse line; ventrolateral
stripes prominent and extending behind shoulder;
venter fleshy gray, throat with very fine brownish
stippling along jaw margins but otherwise unpat-
temed.

Variation.— T\%hiy -one S. d. anthracomus have
the following measurements and scale counts: larg-
est male (ASFS V30843) 33 mm, largest females
(ASFS V30776, ASFS V30845) 32 mm; dorsal scales
between axilla and groin 24-33 (28.7); ventral scales
between axilla and groin 26-35 (30.1); midbody
scales 45-56 (49.8); supralabials to mid-eye 3/3 (47),

i983

SCHV< XRTZ- SPHAERODACTYLUS SYSTEMATICS 1

23

3/4 (2); intemasals 0(12), 1 (66), 2 (2), 3(1); fourth
toe lamellae 8-12(11.1; mode 11); gular scales usu-
ally smooth (73), very rarely keeled (1) or slightly
keeled (7); escutcheon 4-8 (5.7) by 15-27 (21.0).

The dorsal coloration of males varies from very
pale sandy to tan or dark brown or gray-tan, either
without a dorsal pattern (LSUMZ 21934), or with
a lightly flecked (LSUMZ 21933) or heavily flecked
(ASFS V 1 6937) dorsum. Paired white ocelli are reg-
ularly present and visible even in the palest males,
and the scapular patch is usually absent; two males
(LSUMZ 21934, USNM 167296) have a slightly
darker area, which Represents the patch in females
between the ocelli. One peculiar individual (ASFS
V30778) has the scapular patch as well developed
as females, and there are remnants of a female head
pattern; however, there is a moderately well devel-
oped escutcheon and the specimen appears, at least
on this structural feature, to be a male. The trilineate
head pattern persists in some large males (ASFS
VI 6940, snout-vent length 28 mm), but even in
such specimens there is a strong tendency for head
dotting to obscure the basic female pattern. Flead
dotting is a common feature in most males, although
it is absent in some fully adult specimens. The venter
is whitish, pinkish gray, or dull orange, and the throat
is either dotted with dark brown or not; both con-
ditions occur with about equal frequency.

Females were recorded dorsally as tan to brown,
with either little or no dorsal dark spotting or dotting
(LDO 7-5338) or with fairly prominent but random
dark flecks as in the holotype. The head pattern is
unusually well defined and distinctive, the dark lines
on a buffy ground color. The scapular patch is vari-
ably expressed, but in most females, especially in
those from the type-locality and its vicinity, the patch
is very large, dark brown, and regularly has two
white ocelli along its lateral margins; there is often
also a dark brown transverse line following the patch
itself, as in the holotype. In specimens from Bara-
hona and vicinity, the patch is somewhat more re-
stricted but still evident, and the ocelli are large and
conspicuous, even in pale individuals. The ventral
coloration is whitish to fleshy gray, and the throat
is concolor with the venter (not yellowish as in males)
and rarely has a few scattered dark flecks. The iris
color is yellow to golden yellow in both sexes.

Juveniles show the fullest expression of the female
pattern, with large block-like scapular patches and
associated white ocelli. Barahona juveniles have the
patch less well expressed than do juveniles from

farther south. One small specimen (ASFS VI 6944,
snout-vent length 20 mm) has a widely opened pale
chevron preceding the large scapular patch; this
chevron is characteristic of juveniles and females of
S. randi to the south. Juveniles have yellow throats
and gray venters, and the underside of the tail is
orange or coral red.

Comparisons. — In having a large and conspicuous
scapular patch and two ocelli, female S. d. anthra-
comus are easily distinguished from females of those
subspecies with one ocellus (difficilis) or with re-
duced scapular patches {lycauges, typhlopous, per-
atus, diolenius). From female euopter, which have
a large scapular patch and two ocelli, female an-
thracomus differ in not being lineate orange dorsally.
In having ocellate males, anthracomus differs from
difficilis, lycauges, typhlopous, and peratus. Only eu-
opter and diolenius males are ocellate, and males of
the former have a well developed scapular patch.
Male diolenius and anthracomus can be most easily
distinguished by the absence of a lineate body pat-
tern in the latter and its usual presence in the former.

In dorsal scales between axilla and groin, anthra-
comus has a higher mean than all discussed sub-
species except difficilis, typhlopous, and peratus. In
midbody scales, anthracomus has the highest mean
of all discussed subspecies except peratus-, the mid-
body mean of anthracomus (49.8 ± 0.78) differs
significantly from those of difficilis (48.1 ± 0.71),
lycauges (43.4 ± 0.62), euopter (45.3 ± 1.17),
typhlopous (46.8 ± 1.51), /?£’rarr«(51.8 ± 0.79), and
diolenius (46.6 ± 0.33), but is closest to that of
difficilis.

Remarks.— S. d. anthracomus is apparently a de-
rivative of V. d. diolenius. Only in the extreme east-
ern, mesic end of the Valle de Neiba has S. difficilis
been able to cross this xeric valley, with subsequent
penetration onto the eastern coast of the Peninsula
de Barahona near sea level, and thence onto the
southern portion of the Peninsula itself in appar-
ently especially favored situations. As has been re-
peatedly pointed out, the eastern littoral of the Pen-
insula de Barahona is narrow, and the Sierra descends
steeply to the ocean at many localities along this
coast. Most specimens of S. d. anthracomus were
taken in seaside situations — under palm trash in
coastal groves, in a banana grove, under very dry
palm thatch adjacent to an inhabited native house,
under dry Cocos trash piles on sand behind man-
groves, under palm trash in mesic woods, and (at
the type-locality) in the wet fallen thatch of a shelter

24

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 22

along the lower reaches of the Rio Nizalto and under
palm trash in a park-like grassy grove adjacent to
the river bank. One lizard was taken as it was active
in an abandoned schoolhouse in Acacia forest dur-
ing the morning. The highest elevation for anthra-
comus is 2,400 ft (732 m) at La Lanza in the uplands
of the Sierra de Baoruco.

Four eggs, taken 2 km SW Paraiso, had measure-
ments of 7.9 by 5.4, 7.4 by 5.4, 7.4 by 5.5, and 7.6
by 5.3; three hatchlings from these eggs had snout-
vent lengths of 14 and 15 mm.

It is possible that the specimens from the vicinity
of Barahona represent a subspecies different from
anthracomus\ as pointed out above, these lizards
are uniformly pale and females are unpatterned or
weakly patterned dorsally. Likewise, the scapular
patch is limited, although the ocelli are present and
prominent. Another possibility is that these few
geckos represent extreme intergrades between di-
olenius to the north and anthracomus. No certain
intergrades between these two subspecies are known:
anthracomus occurs at Barahona, whereas the near-
est diolenius locality is Barreras, distant some 20
km airline and on the northern side of the Valle de
Neiba. We have little doubt that S. difficilis occurs
between these two localities, but specimens are lack-
ing. The intervention of the large Bahia de Neiba
between the Barahona and Barreras localities ren-
ders the distance of 20 km between them equivocal,
since S. difficilis obviously must circumvent the bay
in this region.

The Sierra de Baoruco and the Peninsula de Ba-
rahona are complex as far as geographic interrela-
tionships between members of the difficilis complex
are concerned. Not only does S. difficilis itself occur
widely along the eastern margin of the Peninsula,
but in these same lowlands occur S. altavelensis and
S. armstrongi; S. difficilis has been taken syntopi-
cally with both these species— with altavelensis in
the north and armstrongi in the south. At the type-
locality of S. d. anthracomus and at other localities
in this more mesic portion of the range of the Ba-
rahona subspecies, anthracomus is syntopic with S'.
armstrongi. The latter species is distinctly more me-
sophilic than S. d. anthracomus and seems to be
associated with more upland areas in the Sierra de
Baoruco. The two species are syntopic in those sit-
uations where the mesic upland flora and conditions
descend to near the coast.

There are five species of sphaerodactyls which
occur either syntopically or allotopically along the
eastern margin of the Peninsula de Barahona— S.
difficilis, S. randi, S. armstrongi, S. altavelensis, and

S. streptophorus. Of these, we have taken difficilis
with armstrongi, difficilis with altavelensis, and dif-
ficilis with streptophorus, syntopically. Schwartz
(1977) discussed the geographic relationships of these
species and presented a map as well.

Of the four species, S. d. anthracomus is the larg-
est with males reaching a snout-vent length of 33
mm, females 32 mm. All other local species are
much smaller, with Y a. enriquilloensis the largest
(males to 26 mm, females to 28 mm). Sphaerodac-
tylus d. anthracomus and 5. a. armstrongi are sep-
arable on the basis of dorsal scales (24-33 versus
29-41), whereas this count in altavelensis and strep-
tophorus falls within the known parameters for S.
d. anthracomus. At the species level, midbody scales
in S. armstrongi are more numerous (49-64) than
in S. difficilis (45-56) and midbody scales in S. al-
tavelensis are less so (38-50), whereas midbody scales
in S. streptophorus are comparable in number (4 1 -
60) to those of S. difficilis. Of the four species, only
S. d. anthracomus has females with bold dark scap-
ular patches with pale ocelli on a lighter ground; this
condition stands most especially in contrast to that
of S. a. enriquilloensis wherein the scapular patch
is absent or, at its best expression, faint and weak.
Both S. armstrongi and Y. streptophorus lack scap-
ular patches but have ocelli. The head patterns of
all four species are so very distinctive that one has
little difficulty assigning individual specimens to the
proper taxon within this area of sympatry.

Etymology.— Tht name anthracomus is from the Greek for
“charcoal” and “shoulder” in allusion to the large scapular patch
in females.

Sphaerodactylus clenchi Shreve

Sphaerodactylus clenchi Shreve, 1968, Breviora, Mus. Comp.

Zool., 280:21.

Definition.— K species of Sphaerodactylus with
small, acute, strongly keeled, flattened, imbricate
dorsal scales (Fig. 3) axilla to groin 33 to 48; no area
of middorsal granules or granular scales; dorsal body
scales with two to five hair-bearing organs, each with
a single hair, around apex. Dorsal scales of tail keeled,
acute, imbricate, and flat-lying; ventral scales of tail
smooth, rounded, enlarged midventrally; gular scales
smooth, but occasionally weakly to strongly keeled;
ventral scales rounded, imbricate, axilla to groin 27
to 36, smooth; scales around midbody 53 to 71;
intemasals 0 to 3 (mode 1); upper labials to mid-
eye 3 (rarely 4); escutcheon with a relatively narrow
and compact central area and extensions onto thighs
to near underside of knee (3-7 by 16-30).

1983

SCHV^ ARTZ- SPIIAERODACTYLUS SYSTEMATICS I

25

Fig. 3. — Dorsal views of Sphaerodactylus clenchi and 5”. lazelli, as follow: A) S. c. clenchi (male, ASFS V21847, and female, ASFS
VI 996); B) S. c. apocoptus (holotype female, USNM 166965); C) S. lazelli (holotype male, MCZ 63281).

Color pattern very weakly to moderately sexually
diehromatie, and variable between the subspecies.
Males yellowish tan to brown, with scattered dark
brown flecks to dots and scattered creamy to orange
ocelli giving a coarsely salt-and-pepper eflect, no
head pattern but top of head often with scattered
dark brown flecks or dots, throats gray to yellow
(not orange) with a dark reticulum or scattered dots
or flecks; scapular patch and ocelli absent; venter
grayish yellow or flesh. Females with same dorsal
body pattern and color as males, but head either
almost completely without pattern except for some
vague darker spots, or with a quadrilineate pattern
of which the two inner lines correspond to the widely
separated edges of the median single line in 5. dif-
ficilis\ scapular patch and ocelli absent; ventral color
as in males. Iris color yellow to orange.

Distribution. — RtTpuhVicsi Dominicana, on the
Peninsula de Samana as far as west of Sanchez, at
Caba on the southwestern coast of the Bahia de
Samana, and on the eastern extremity of Hispaniola
at Playa El Coco and near La Vacama, La Altagracia
Province.

Sphaerodactylus clenchi clenchi Shreve

Sphaerodactylus clenchi Shreve, 1968, Breviora, Mus. Comp.

ZooL, 280:21.

Type- /ocaZ/fv.— Samana (= Santa Barbara de Sa-
mana), Samana Province, Republica Dominicana.

Holotype. -MCZ 43706.

Definition.— A subspecies of S. clenchi character-
ized by a combination of high number of dorsal
scales (34-48) between axilla and groin, high num-
ber (54-7 1 ) of midbody scales, and almost no sexual
dichromatism in head pattern with females having
only the barest indication of the more completely
expressed head pattern in the following subspecies.

Distribution. — The Peninsula de Samana, west as
far as 5.0 mi west of Sanchez, and at Caba at the
southwestern comer of the Bahia de Samana, Re-
publica Dominicana (Fig. 4).

Variation.— The series of 97 S. c. clenchi has the
following counts and measurements (means in pa-
rentheses): largest male (MCZ 43706) 33 mm snout-
vent length, largest females (ASFS V34317, ASFS
V34335, ASFS V34904, ASFS V34906, ASFS
V36153) 32 mm; dorsal scales between axilla and
groin 34-48 (40.7); ventral scales between axilla and
groin 27-36 (31.7); midbody scales 54-71 (61.4);
supralabials to mid-eye 3/3 (62 individuals), 3/4 ( 1 );
internasals 0 (1 individual), 1 (83), 2 (11), 3 (2);
fourth toe lamellae 7-14 (11.8; mode 13); throat
scales usually smooth (82 individuals) but at times
keeled (4) or partially keeled (11); escutcheon 3-7
(5.0) by 16-30 (22.4).

Males are brown, dark brown, tan, or yellowish
tan with scattered dark brown flecks, giving a finely
salt-and-pepper eflect; these dark flecks often alter-
nate (especially along the sides) with much larger
dull orange ocelli; these ocelli occasionally (ASFS
V2 1 859) fuse to form a ventrolateral line above the

26

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Fig. 4. — Map of extreme eastern Hispaniola, showing the distri-
butions of two subspecies of Sphaerodactylus clenchi.

forelimb insertion on each side (Fig. 3A). The top
of the head is either unicolor (at times with a yellow
wash) with the dorsum and unpattemed, or there
may be heavy' dark brown flecks, dots, or mottling
over the top of the head and including the snout.
One male (ASFS VI 995) shows the vaguest rem-
nants of the female head pattern. The venter is gray-
ish yellow; the throat ground color is yellowish gray
to yellow, and there is a heavy pattern of dark brown
(either dots or a reticulum). The iris is rarely yellow
to usually bright orange, and the underside of the
tail is orange. The upper side of the tail is marked
with transversely juxtaposed orange to bufly ocelli
giving, at least basally, a more or less crossbarred
appearance; a pair of orange ocelli occurs fairly reg-
ularly at the base of the tail.

Females are colored and patterned dorsally like
the males, even to the dotted head pattern. Young

females, however, as well as juveniles, often show
the vaguest indications of a quadrilineate head pat-
tern, much better expressed in females of the other
subspecies, which is represented by solid dark brown
flecks arranged along the positions of the four lines
(Fig. 3A). The venters of females were recorded as
yellow-orange, and the throats are patterned with
some scattered brownish to dark gray dots or flecks,
less bold than in males. The ocellate tails of females
are like those of males.

Juveniles are patterned like adults, and there is
little ontogenetic change in pattern with increased
size, since many small juveniles have the heads with
scattered dark spots.

Mertens (1939:43) described a single male from
Samana as “Schuppen sehr klein; . . . Hellgrau mit
vielen dunkelgrauen Recken, die kaum deutliche
Langsreihen bilden. Keine hellen Abzeichen auf den
Schultem Oder am Schwanzende.” Shreve (1968:21)
considered that N. clenchi was most closely related
to S. caicosensis Cochran from the Caicos Bank is-
lands, but it seems more likely that S. clenchi is a
derivative of S. dijficilis (see beyond).

Remarks.— S. c. clenchi is common on the Pen-
insula de Samana. Our specimens were secured in
palm trash in mesic Cocos groves, in a small, old,
and very decayed pile of palm trash about 3 m from
the inner edge of the mangrove border in a dense
growth of herbaceous halophytes, under palm trash
in an open and grassy Cocos grove, and under palm
fronds behind Coccoloba on a sandy beach. At the
other extreme are specimens from the Sierra de Sa-
mana at an elevation of 1,000 ft (305 m) and in a
mesic but somewhat lower cacaotal. The Sierra de
Samana locality is in high-canopy hardwood forest
with some cultivation.

Specific status for S. clenchi depends primarily
upon the higher scale counts of S. c. clenchi, the lack
of or different head pattern in females (as compared
with those of N. difficilis), and the apparent sympatry
of 5. clenchi and S. difficilis at Sanchez near the
isthmus of the Peninsula de Samana. The precise
situation at Sanchez requires confirmation. In a 1 2
day stay at Sanchez, Schwartz and Buffett secured
only S. clenchi, even as far distant as 5 mi (8 km)
northwest of that settlement (and thus toward the
isthmus). They were unable to secure any sphae-
rodactyls between this locality and Cano Abajo
(which lies on the mainland at the western extreme
of the isthmus) where all material was S. difficilis.
Thus the supposed sympatry remains unconfirmed;
because it is based exclusively upon old specimens

1983

AKTZ- SPHAERODACTYLUS SYSTEMATICS 1

27

Table \.—Meristic data on seven subspecies o/Sphaerodactylus difficilis and two ofS. clenchi; extremes and means are given of selected
counts except in S. lazelli where only one specimen is known. Diagnostic characters of the seven subspecies ofS. difficilis (both sexes) as

far as the shoulder pattern is concerned are likewise given briefly.

Dorsals

Venlrals

Taxon

N

axilla-groin

axilla-groin

Midbody scales

Male pattern

Female pattern

S. d. difficilis

65

25-34

24-37

42-55

patch and ocelli ab-

1 ocellus small patch

(29.2)

(30.2)

(48.1)

sent

S. d. lycauges

1 19

ocelli and patch usu-

2 ocelli small patch

22-33

24-35

37-50

ally absent, but

(26.4)

(28.7)

(43.4)

ocelli indicated

S. d. euopter

30

24-31

23-31

41-50

patch and ocelli

2 ocelli large patch;

(27.8)

(26.7 )

(45.3)

present

dorsum with orange
lines

S. d. typhlopous

85

25-35

24-36

41-53

ocelli and patch ab-

patch and ocelli absent

(30.3)

(30.0)

(46.8)

sent

to barely present

S. d. peratus

83

28^0

26-35

46-60

ocelli and patch ab-

patch absent to bar-like

(34.0)

(29.4)

(51.8)

sent

2 ocelli

S. d. diolenius

446

23-35

24-37

40-53

ocelli usually pres-

ocelli present, patch

(28.0)

(29.8)

(46.6)

ent; patch absent

absent to bar-like

S. d. anthracomus

81

24-33

26-35

45-56

ocelli usually pres-

ocelli large, bold; patch

(28.7)

(30.1)

(49.8)

ent; patch absent

variable usually
large, never bar-like

S. c. clenchi

97

34^8

27-36

54-71

(40.7)

(31.7)

(61.4)

S. c. apocoptus

60

33^3

27-36

53-63

(37.2)

(31.1)

(57.8)

S. lazelli

1

20

25

42

whose locality data may be imprecise, we are not
convinced that S. difficilis occurs on the Peninsula
de Samana. The higher scale counts in S. c. clenchi
versus those of S. difficilis are verified by our study,
but it should be recalled that S. d. peratus, that
subspecies of S. difficilis which occurs in the north-
eastern Republica Dominicana, has the highest scale
counts (midbody 46-60) of all the difficilis subspe-
cies (in contrast to midbody counts of 54-7 1 in S.

c. clenchi). It might be more proper to consider S.
clenchi as a subspecies of 5. difficilis (because un-
equivocal sympatry remains unknown), but the oc-
currence of the species to the east on the southern
side of the Bahia de Samana, where it is readily
distinguishable from and allopatric to adjacent S.

d. diolenius, suggests strongly that the species are
quite distinct.

Because of the general resemblance between S.
clenchi and S. difficilis, there seems to be no reason
to postulate a close relationship between S. clenchi
and S. caicosensis as Shreve has done. The latter is
a distinctly sexually dichromatic species with uni-
color and non-ocellate males, and transversely
crossbanded females — two pattern conditions which
are quite different from the ocellate and salt-and-

pepper patterns in S. clenchi (see Schwartz, 1968;
247 et seq., for details of variation in S. caicosensis).

The large Rio Yuna empties into the head of the
Bahia de Samana. The short series of S. c. clenchi
from Caba on the south side of the Bahia presents
a minor problem of distribution. Caba is a small
fishing village lying on a pair of tiny beaches at the
foot of the haitises which, on the southwestern por-
tion of the Bahia de Samana, come abruptly to the
coast. Between Caba and the nearest northern lo-
cality for S. c. clenchi (Sanchez), lie the extensive
swamps of the mouth of the Yuna. Although S.
clenchi is a mesophile, it seems hardly likely that
the species occupies these intermediate swampy re-
gions. On the other hand, the specimens from Caba
were all taken within and near the settlement; Caba
residents sell their products and make purchases
weekly in Sanchez, and it may well be that the Caba
S. clenchi population has been fortuitously estab-
lished through human agency. There are no other
Sphaewdactylus known from the interior haitises
(with the exception of S. darlingtoni near Gonzalo)
and none from farther east along this coast until the
Bahia de San Lorenzo, west of Sabana de la Mar,
where four species {difficilis, darlingtoni, cochranae.

28

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

samanensis) are sympatric. I once more reiterate
that the karst haitises region of northeastern Re-
publica Dominicana is one of the herpetologically
least known areas of the Repiiblica Dominicana,
and extensive work is demanded in that difficult area
before the geographic relationships of these four
geckos can be understood.

Specimens examined. — Repvblic a Dominicana: Samand
Province. 5.0 mi (8.0 km) NW Sanchez (ASFS V34332-37, ASFS
V34942); Sanchez (MCZ 43706); 2.2 km E Sanchez (ASFS
V141 18-19); 14 km E Sanchez (ASFS V2011-12); 4.5 mi (7.2
km) E Sanchez (ASFS V34174); 7.6 mi (12.2 km) NE Sanchez,
1,000 ft (305 m) (ASFS V34307-21, ASFS V34896-914); 5 km
E Las Terrenas (ASFS V21847); 7 km E Las Terrenas (ASFS
V2 1858-71); 3.8 mi (6.1 km) E Las Terrenas (ASFS V34125);
0.3 mi (0.5 km) NW El Limon (ASFS V36151-54); 5.1 mi (8.2
km) NW El Limon (ASFS V36 1 55-58, ASFS V36 1 63); ca. 2 km
W Los Cacaos (ASFS V21932); 8 km W Samana (ASFS V1994-
V2002); 6 km W Samana (ASFS VI 973-74, ASFS V 14 13 1-36);
4 km W Samana (ASFS V14124); Samana (MCZ 43706 — ho-
lotype); Puerto Escondido (ASFS V2978); Caba (ASFS V36075-
79).

Sphaerodactylus clenchi apocoptus, new subspecies

//o/o(vpc — USNM 166965, an adult female, from
Playa El Coco, 46 km N Higuey, La Altagracia Prov-
ince, Repiiblica Dominicana, one of a series col-
lected by James A. Rodgers, Jr., and James B. Strong
on 13 June 1969. Original number ASFS VI 7540.

Paratypes.-KS¥S V17521-25, ASFS V17536-39, ASFS
VI 7552-55, MCZ 1 1 935 1-58, CM 45895-900, USNM 167252-
57, LSUMZ 21896-902, same data as holotype; ASFS V29020-
36, 2.6 mi (4.2 km) NE La Vacama, La Altagracia Province,
Republica Dominicana, D. C. Fowler, A. Schwartz, native col-
lectors, 22 July 1971; ASFS V29039, 5.7 mi (9.1 km) SE La
Vacama, La Altagracia Province, Republica Dominicana, D. C.
Fowler, 22 July 1971.

Definition. — /S subspecies of N. clenchi character-
ized by a combination of low number of dorsal scales
(33-43) between axilla and groin, low number (53-
63) of midbody scales, and sexually dichromatic in
that females have a quadrilineate head pattern (Fig.
3B), whereas males have the head concolor with (or
more yellowish than) the dorsum, either unpat-
terned or with scattered dark brown spots or flecks.

Distribution. — }Lnown only from three coastal or
near-coastal localities on the eastern extremity of
Hispaniola in La Altagracia Province, Republica
Dominicana.

Description of holotype.— Kn adult female with a
snout-vent length of 31 mm, tail length 25 mm,
distal half regenerated; dorsal scales axilla to groin
36, ventral scales axilla to groin 33, midbody scales
59, supralabials to mid-eye 3/3, 1 internasal, fourth

toe lamellae 1 1 , gular scales keeled, chest and ven-
tral scales smooth.

Dorsal ground color brown with scattered darker
brown flecks alternating in a random fashion with
large creamy ocelli outlined in dark brown; head
pattern consisting of four longitudinal lines of which
the outer pair begins on the snout, passes through
the eye, across the temples and thence above the
forelimb insertion onto the anterior trunk, this en-
tire line outlined below by a creamy longitudinal
line; median head lines consisting of (apparently)
the outer and more widely separated dark edges of
the S. difficilis median head line, beginning as a pair
of narrow lines on the snout, becoming less con-
spicuous in the interocular region, having a lyre-
shaped configuration in the postocular region, and
progressing as a pair of well defined lines onto the
neck; limbs and tail brown, marked with buffy ocelli
or crossbars on the limbs and paired dark-edged
ocelli on the tail (unregenerated portion); venter and
throat flesh, throat with scattered brown flecks; iris
orange.

Variation. Sixty S. c. apocoptus have the follow-
ing measurements and counts: largest male (ASFS
VI 7522) 31 mm snout-vent length, largest females
(holotype, USNM 166965) 31 mm; dorsal scales
between axilla and groin 33-43 (37.2); ventral scales
between axilla and groin 27-36 (31.1); midbody
scales 53-63 (57.8); supralabials to mid-eye 3/3 (40),
3/4 (1), 4/4 (2); intemasals 0 (1 individual), 1 (58),
2 (1); fourth toe lamellae 9-14 (11.3; mode 12);
throat scales usually smooth (44 individuals), but
at times keeled (8) or weakly keeled (8); escutcheon
3-7 (4.5) by 16-27 (23.0).

Males are yellowish tan to brown dorsally, flecked
with dark brown, often with scattered creamy ocelli;
head unicolor with the dorsum (or washed with yel-
lowish) and at times with a few scattered dark brown
flecks or dots, the dark spots generally aligned along
the female quadrilineate head pattern. The bellies
are flesh to yellow, and the throats are yellowish
with scattered dark gray dots or spots, never with a
reticulum. Unregenerated tails have creamy ocelli
arranged in pairs to give a cross-barred appearance.
The iris color is orange (rarely) to yellow.

Females are like the males dorsally (except that
the heads are not yellowish), although there is a
tendency for the former sex to be more prominently
ocellate with creamy ocelli. The head pattern is as
described for the holotype, although some females
tend toward an obliteration of the two median lines.
The pale longitudinal lines paralleling the lateral

1983

SCHWARTZ- SPHAERODACTYLUS SYSTEMATICS I

29

pair of dark lines below is a common and easily
discernible feature. The ventral color is like that of
males, but the throats are not yellow and have scat-
tered gray flecks or dots.

Juveniles are unicolor brown with some dark
flecking. The juvenile head pattern is like that of
females, and there may be some emphasis of the
pattern by line-following dark brown dots.

Comparisons. — In all body counts, S. c. apocop-
tus averages less than nominate S. clench i. The mid-
body means of 61.4 ± 1.1 in clenchi and 57.8 ±
0.8 in apocoptus are statistically different. The two
subspecies are also easily distinguishable in pattern.
Male clenchi have heavily reticulate or spotted or
dotted throats, whereas those of male apocoptus are
not reticulate but are usually marked by discrete
gray spots. There is a tendency for male clenchi to
have the top of the head spotted with dark brown,
whereas the head is rarely so densely marked in male
apocoptus. Females of the two subspecies are very
easily distinguished; female clenchi have the quad-
rilineate head pattern very reduced or absent,
whereas the pattern is fully and boldly expressed in
female apocoptus. Although both subspecies have
yellow to orange i rides, those of clenchi are usually
orange and those of apocoptus usually yellow.

S. c. apocoptus is not known to be sympatric with
any other Sphaerodactylus\ in fact, the range of the
subspecies as presently known lies between that of
S. d. diolenius on the west (nearest locality, 20.2 mi
NW, 3.4 mi N La Vacama, El Seibo Province, 25
km to the west) and of S. savagei on the east (nearest
locality, 0. 1 mi SE El Macao, La Altagracia Prov-
ince, 1 5 km to the east); to the south also occurs S.
d. diolenius (7 mi W El Cuey, El Seibo Province; 30
km). Of the three eastern Dominican species of
Sphaerodactylus, sympatry is not surely known be-
tween difficilis and clenchi and is unknown between
clenchi and savagei, the precise geographical inter-
relationships of these species in this region are sub-
ject to further study. Sympatry between difficilis and
savagei is known only at the type-locality of the
latter La Romana).

Remarks.— The range of S. c. apocoptus lies about
55 km southeast of the main center of S. clenchi (on
the Peninsula de Samana). The original series was
collected by Rodgers and Strong in a coastal Cocos
grove under palm trash, and the series collected sub-
sequently by Fowler and Schwartz was in the same
precise situation. The single male from southeast of
La Vacama was taken during the early afternoon on
the trunk of a Musa in an open and rather mesic

Musa-Coffea grove. In general, the southern shore
of the Bahia de Samana between Miches and El
Macao is difficult of access, since no roads proceed
along it; this doubtless accounts for the fact that S.
c. apocoptus has not been previously taken in this
region. Interestingly, coastal Cocos groves between
Miches and El Macao yield only one of the three
species involved in this region — difficilis, clenchi, or
savagei— znd we were unable to secure any two
species syntopically. There are no major geographic
barriers to account for this peculiar situation. How-
ever, the Rio Nisibon lies between the ranges of
difficilis and clenchi, and the Rio Maimon and the
Rio Anamuya lie between the ranges of clenchi and
savagei. Whether these rivers are major and real
barriers for these three species remains to be deter-
mined.

If S. c. apocoptus is syntopic with either S. diffi-
cilis or S. savagei, the species can be easily distin-
guished: in apocoptus there are 53 to 63 midbody
scales, in local savagei 37 to 49 midbody scales, and
in local difficilis 40 to 53 midbody scales. The head
and body patterns of the three species are quite dif-
ferent, and specimens are not easily confounded.

Etymology.— Tht name apocoptus is from the Greek meaning
“cut off,” in allusion to the remote geographical position of the
subspecies in relation to Y c. clenchi.

Sphaerodactylus lazelli Shreve

Sphaerodactylus lazelli Shreve, 1968, Breviora, Mus. Comp. Zool.,

280:8.

Type-locality. — Cstp-WdiiXien (under bark of tree
in shady gully), Departement du Nord, Haiti.

Holotype. -MCZ 63218.

Distribution. — Known only from the type-locality
(Fig. 1).

Definition.— A. species of Sphaerodactylus with
large, acute, strongly keeled, flattened, imbricate
dorsal scales, axilla to groin 20 in only known spec-
imen; no area of middorsal granules or granular
scales; dorsal body scales with seven or eight hair-
bearing organs, each with one or two hairs, around
apex. Dorsal scales of tail keeled, acute, imbricate,
and flat-lying; ventral scales of tail smooth, rounded,
enlarged midventrally; gular scales keeled, chest and
ventral scales smooth; ventral scales rounded, im-
bricate, axilla to groin 25, smooth; scales around
midbody 42; intemasal 1; upper labials to mid-eye
3/3; escutcheon with a broad and compact central
area and extensions onto thighs to near underside
of knee (6 by 29).

30

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Male (only sex known; color from preserved spec-
imen) medium brown above with a very few scat-
tered darker brown scales arranged in a more or less
lineate pattern; a large prominent and sharp-edged
black scapular patch with two included tan ocelli,
followed by a transverse series of three tiny black
dots in the position of three of the vague dorsal lines
formed by the darker dorsal scales; a large black
sharp-edged elongate spot on the neck and another,
ovate, on the occiput, the former with a pair of fine
dark lines extending from its anterolateral comers
to pass around the occipital spot and end between
the eyes; a clear tan postocular line which joins its
mate posterior to the nuchal blotch to give a mod-
erately well defined pale U-shaped figure on the head
and neck (Fig. 3C); tail not ocellate; venter immac-
ulate pale tan with some seattered dark brown flecks
along the ventrolateral margins.

Remarks. — \n addition to the scale counts given
in the definition, the holotype of 5. lazelli has 12
fourth toe lamellae, a snout-vent length of 3 1 mm,
and a tail length of 30 mm.

S. lazelli is a very distinctive form, if for no other
reason than the head pattern which is strikingly dif-
ferent from the style of N. difficilis or of that of any
other member of the difficilis complex. Presumably
S. lazelli is sympatric with N. d. lycauges (of which

there is abundant material from Cap-Haitien), but
the two species differ in head pattern since lycauges
males lack the U-shaped figure, large black scapular
patch and ocelli, and occipital and nuchal spots of
lazelli. As far as scales are concerned, lazelli lies
below lycauges in dorsal scales between axilla and
groin (20 versus 22 to 33), but counts of ventral and
midbody scales of lazelli are included within the
observed variation of these counts in lycauges. The
width of the escutcheon in lazelli (29) is greater than
that of male lycauges ( 1 0-24). The distinctly keeled
throat scales in lazelli aid in separating the species
from local S. difficilis-, however, four of 105 S. d.
lycauges have the throat scales equally as keeled.

There is no reason to doubt that N. lazelli is a
satellite species derived from S. difficilis-, in having
the black scapular patch and ocelli, lazelli appar-
ently has fully retained a juvenile (and female) pat-
tern feature into adult males, a condition few S.
difficilis have done. Presumably, the species will be
found widely distributed along this mesic northern
Haitian coastal region, but many man-weeks of
search, with the aid of many Haitians, has failed to
turn up a second specimen.

Specimen examined. — Hmtv. Dept, du Nord, Cap-Haitien
(MCZ 63218-holotype).

Literature Cited

Barbour, T. 1914. A contribution to the zoogeography of the
West Indies, with especial reference to amphibians and rep-
tiles. Mem. Mus. Comp. Zool., 44:209-359.

. 1921. 5'p/zacro<fflCtv/Mi'. Mem. Mus. Comp. Zool., 44:

217-278.

Graham, E. D., Jr., and A. Schwartz. 1978. Status of the
name Sphaerodactylus cinereus Wagler and variation in
“Sphaewdactylus stejnegerr Cochran. Florida Sci., 41:243-
251.

Mertens, R. 1939. Herpetologische Ergebnisse einer Reise nach
Insel Hispaniola, Westindien. Abh. senckenberg. naturf. Ges.,
449:1-84.

Noble, G. K., and W. G. Hassler. 1933. Two new species of
frogs, five new species and a new race of lizards from the
Dominican Republic. Amer. Mus. Novitates, 652:1-17.

Ruibal, R. 1946. A new from the Dominican

Republic. Amer. Mus. Novitates, 1308:1-4.

Schwartz, A. 1966. Geographic variation in Sphaerodactylus
notatus Baird. Rev. Biol. Trop., 13:161-185.

. 1968. The geckos (Sphaerodactylus) of the southern

Bahama Islands. Ann. Carnegie Mus., 39:227-269.

. 1977. The geckoes (Sauria, Gekkonidae) of the genus

Sphaerodactylus of the Dominiean Peninsula de Barahona,
Hispaniola. Proe. Biol. Soc. Washington, 90:243-254.

Schwartz, A., and R. Thomas. 1977. Two new species of
Sphaerodactylus (Reptilia, Lacertilia, Gekkonidae) from
Hispaniola. J. Herpetology, 11:61-65.

Shreve, B. 1 968. The notatus group of Sphaerodactylus (Sauria,
Gekkonidae) in Hispaniola. Breviora, Mus. Comp. Zool.,
280:1-28.

Thomas, R., and A. Schwartz. 1966a. Sphaerodactylus (Gek-
konidae) in the Greater Puerto Rico region. Bull. Florida
State Mus., Biol. Sei., 19:193-260.

. 1966L The Sphaerodactylus decoratus compltx 'm ihc

West Indies. Brigham Young Univ. Sci. Bull., 7:1-26.

. 1977. Three new species of Sphaerodactylus (Sauria:

Gekkonidae) from Hispaniola. Ann. Carnegie Mus., 46:33-
43.

. MS. Variation in Hispaniolan Sphaerodactylus (Sau-
ria: Gekkonidae). Mus. Comp. Zool., in press.

PART 2. SPHAERODACTYLUS SAVAGEI, S. COCHRANAE,
S. DARLINGTONI, S. ARMSTRONGI, S. STREPTOPHORUS,

AND CONCLUSIONS

Richard Thomas and Albert Schwartz

ABSTRACT

Four species of Hispaniolan geckos of the difficitis complex are
discussed in detail, and a new subspecies of a fifth (S. strepto-
phorus) is also named. Sphaerodactylus savagei has two weakly
differentiated subspecies in extreme eastern Hispaniola, and 5.
cochranae is monotypic and is known from a very limited amount
of material and from a circumscribed area in eastern Hispaniola.

Sphaerodactylus armstrongi occurs on the Hispaniolan south is-
land in Haiti and the Republica Dominicana and has two sub-
species, whereas S. darlingtoni occurs on the Hispaniolan north
island and has three subspecies, all in the Republica Dominicana.
A phyletic schema of the relationships of all known members of
the difficilis complex is presented.

INTRODUCTION

The present paper continues the discussion of the phyletic trends within the entire group of these geck-

variation of Hispaniolan members of the difficilis os on Hispaniola,
complex (see Part 1), and further attempts to show

ACKNOWLEDGMENTS

In addition to all those persons who have given us their capable assistance of National Science Foundation grant SER 77-04629

assistance in both Haiti and the Republica Dominicana, whom to the senior author and grants G-7977 and B-023603 to the

we have thanked in the several previous papers in this series on junior author.
difficilis group geckos, we wish also to acknowledge the financial

SYSTEMATIC ACCOUNTS

Sphaerodactylus savagei Shreve

Sphaerodactylus notatus savagei Shreve, 1968, Breviora, Mus.

Comp. ZooL, 280:7.

Sphaerodactylus savagei: Schwartz and Thomas, 1975:161.

Definition.— A species of Sphaerodactylus with
large, acute, strongly keeled, flattened, imbricate
dorsal scales, axilla to groin 20 to 31; no area of
middorsal granules or granular scales; dorsal body
scales with seven to twelve hair-bearing organs, each
with one hair, around apex. Dorsal scales of tail
keeled, acute, imbricate, and flat-lying; ventral scales
of tail smooth, rounded, only slightly enlarged mid-
ventrally; throat, chest, and ventral scales always
entirely smooth; ventrals rounded, imbricate, axilla
to groin 25-38; scales around midbody 34-50; in-
temasals 0-2 (mode 1); upper labials to mid-eye
3/3 (occasionally 3/4 or 4/4); escutcheon small and
patchlike, without extensions far along the thighs to
knees (3-6 by 5-17).

Color pattern sexually dichromatic. Males some
shade of tan (yellowish to purplish) or gray, to brown
with scattered dark brown to black scales giving a

strongly and randomly salt-and-pepper effect which,
because of the generally pale ground color, often
gives a strikingly contrasting appearance (Fig. 2B);
head ground color yellow, usually overlaid by large
irregular dark brown to black spots with cream frost-
ing between them; a pair of white to bufiy scapular
ocelli interconnected by a dark brown bar or dash;
throats grayish, pale yellow, to bright orange, im-
maculate to heavily spotted with dark brown, the
variation correlated with the amount of dorsal head
spotting; venter pink to dark gray. Females with
dorsal color like males, but ground color greatly
obscured by the irregular dark brown to black mark-
ings, at times coarsely salt-and-pepper (Fig. 2A), at
times quinquelineate, at times with the dark pig-
ment organized into about five or six transverse
bands (Fig. 2C), with scattered buffy to cream scales
between the dark markings; head and scapular pat-
tern extremely variable, but basically consisting of
a pair of faint white to buffy ocelli in a clear area
surrounded on all sides by a large rectangular dark
brown to black spot, the anterior (preocellar) portion

31

32

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

of the spot joining anteriorly with a dark postocular
stripe; a vague postocular median dark stripe which
is variable in expression and intensity and often
restricted to a dilfuse dark occipital blotch; ventral
color like males except at times pale yellow; throat
concolor with venter, usually immaculate, at most
with some grayish lateral diagonal streaking. Iris
golden yellow to yellow.

Distribution. — Dominicana, coastal
areas of La Altagracia and La Romana provinces,
from El Macao on the north, around Cabo Engano
to La Romana on the south; inland as far as near
San Rafael del Yuma; Isla Saona; Isla Catalinita; an
apparently isolated population from near Sabana
Grande de Palenque, San Cristobal Province.

Sphaerodactylus savagei savagei Shreve

Sphaerodactylus notatus savagei Shreve, 1968, Breviora, Mus.

Comp. Zool., 280:7.

Sphaerodactylus savagei savagei: Schwartz and Thomas, 1975:

161.

Type- locality. — 'La. Romana, La Romana Prov-
ince, Republica Dominicana.

//o/o/.vpc. — California Academy of Sciences-
Stanford University 14695.

Definition.— \ subspecies of Y. savagei charac-
terized by a combination of slightly higher mean
number (34-50) of midbody scales, and female pat-
tern without dark crossbands and with either a dark
salt-and-pepper pattern or a pattern of five longi-
tudinal dark lines on a pale ground; pale ocelli var-
iably present in both males and females.

Distribution.— T\\q southeastern coast of the Re-
publica Dominicana, from La Romana in the west,
east to the east side of the Rio Chavon, in La Ro-
mana and La Altagracia provinces (Fig. 1 ).

Variation.— ThQ series of 103 S. s. savagei has the
following counts and measurements (means in pa-
rentheses); largest males (ASFS V35002, ASFS
V35006-07, ASFS V35009, ASFS V35019) 32 mm
snout-vent length, largest females (ASFS V35016,
ASFS V35028, ASFS V35032) 32 mm; dorsal scales
between axilla and groin 20-3 1 (26. 1); ventral scales
between axilla and groin 26-38 (30.6); midbody
scales 34-50 (44.3); supralabials to mid-eye 3/3 (69
specimens); internasals 0 (2 individuals), 1 (101);
fourth toe lamellae 7-13 (10.4; mode 11); gular,
chest, and ventral scales always smooth; escutcheon
3-6 (4.5) by 6-17 (12.3).

Shreve (1968:7-8), in naming both savagei and
juanilloensis (which, together, comprise a species
distinct from S. difficilis; see discussion beyond).

examined a total of 28 specimens, whereas we have
had access to 260 lizards. The two taxa, savagei and
juanilloensis, were partially diagnosed on the basis
of the dorsal coloration (= pattern). Although there
are some differences between these two taxa in dor-
sal pattern, both are so variable in their own right
that it is difficult to distinguish individuals; in series,
however, the differences in dorsal pattern are ap-
preciable. Accordingly, despite our original impres-
sions that savagei and juanilloensis were not dis-
tinct, we herein maintain both taxa but feel strongly
that the differences between them are not so obvious
as are those between, say, the subspecies of S. dif-
ficilis or S. clenchi.

Male S. s. savagei are brown to pale lavender
dorsally with a fairly dense salt-and-pepper pattern
of darker brown to black spots, these spots contin-
uous onto the head which may have an inter-spot
white frosting. Two white ocelli are variably present;
if present, they may be isolated or may have a small
dark bar between them. There is never a large black
scapular patch. Tails are reddish brown to pale cream
dorsally, the upper surfaces with a continuation of
the dorsal salt-and-pepper pattern. The venter is
pink to dark gray, and the throat varies between
immaculate to having a heavy black reticulum or
isolated circular dots; the throat ground color varies
between pale yellow and grayish. Some males (ASFS
V35003) have the dorsal dark spots aligned into
three or five longitudinal series of markings, typical
of some females of S. s. savagei.

The dorsal ground color of females is tan to brown
and there is a trilineate head pattern. The head and
neck pattern is complex and is best discussed by first
describing the most complete pattern available, that
of a female from Isla Saona (ASFS V3044; see Fig.
2C). This individual is not assigned to the nominate
subspecies, but it shows the head pattern most viv-
idly and may be used as a standard for comparisons
of the pattern of savagei versus that of juanilloensis.
In this female, there is a pair of small and incon-
spicuous pale (buffy to white) ocelli lying in a small
patch of dorsal ground color enclosed within a dark
brown scapular rectangle. The anterior comers of
the rectangle are more or less connected to a pair of
dark lines which begin on the lores, cross the eyes,
extend across the temples, and contact the dark rec-
tangle. There is a vague postocular median dark
stripe, often interrupted to form an interocular spot
and an occipital spot. The snout is variously mottled
or stippled with dark, and there is no indication of
a median snout line.

1983

THOMAS AND SCHWARTZ- SPHAERODACTYLUS SYSTEMATICS II

Fig. 1. — Map of Hispaniola showing the ranges of three species of Sphaerodactylus as follow; savagei, circles; S. cochranae, hexagon,

S. armstrongi. triangles. The known ranges of the two subspecies of S. savagei are indicated and the disjunct population of this species
in San Cristobal Province to the west is shown. The two subspecies of 5. armstrongi are indicated.

BULLETIN CARNEGIE MUSEUM OE NATURAL HISTORY

NO. 22

34

Fig. 2. — Dorsal views of Sphaerodactylus savagei. as follow: A)
S. s. savagei, female (ASFS X9296); B) S. s. juanilloensis, male
(ASFS VI 6255); C) S. s. juanilloensis, female (ASFS V3044).

The above description serves fairly well to denote
the head and neck pattern of female S. s. savagei,
but there are many modifications of this basic pat-
tern. The ocelli are variably present and may lie,
instead of within a clear area, in a dark transverse
bar comparable to that in males. The head lines may
be complete or virtually so, and well indicated, or
they may be fragmented or very obscure. In many
females, the top of the head and neck have a large,
pale, grossly U-shaped cephalic figure, without in-
terior markings. The variants seem almost endless,
and it is pointless to discuss all permutations of the
basic pattern described above. Perhaps the most
constant feature is the pale ocelli, but even these are
often absent.

The female dorsal body pattern is likewise vari-
able, but can be categorized into two sorts. Some
females are salt-and-pepper like males, although the
dark scales may be less contrasting than are those
of males. A second group of females is distinctly
quinquelineate dorsally. In addition to the dark dor-
sal markings, there are often scattered paler buffy
or white scales, especially on the upper eyelids. The
venter varies from pale yellow to pinkish gray or
gray. The tails are often very distinctly marked, with
a pair of buffy ocelli at the base of the tail and a
series of paired buffy ocelli, at times fused into com-
plete crossbands, on the tail. The throat is concolor
with the venter and usually unmarked; at best there

are some lateral diagonal grayish gular lines. The
iris is yellow to gold in both sexes.

Juveniles show the female pattern but with ad-
ditional intensity. The variation shown in adult fe-
males is duplicated in juveniles, however, and the
ocelli and scapular pattern are variably expressed.
Those juveniles which are salt-and-pepper are less
densely so than adults of both sexes. The tail ground
color is straw to orange and, as in females, there
may be series of paired caudal ocelli.

Grant (1952:187, pi. I) presented a photograph of
four adult females, three adult males, and seven
juveniles from La Romana (and thus topotypes of
5. s. savagei). Some of the features noted above for
this subspecies are clearly demonstrated in the pho-
tograph (the boldly crossbanded tails in females, for
instance), but his series on the other hand does not
show features such as the longitudinally lined fe-
males nor the strong sexual dichromatism in pat-
tern. Also, the juveniles all appear to be erossband-
ed, a condition which is never so clearly indicated
in our much longer series of 34 juveniles. This is all
the more remarkable since the thesis of Grant’s pa-
per was the ontogenetic change from crossbanded
juveniles to spotted adults. The situation is inexpli-
cable.

Remarks.— The long series (ASFS V35001-70) of
S. s. savagei was secured by Fowler and Sheplan in
a pile of lumber-trash in a hotel patio; two S. d.
diolenius were also taken in the same pile. Other S.
s. savagei were secured in piles of papers at the aban-
doned office of the sugar mill in central La Romana,
and one was secured behind a picture in the corridor
of another hotel in La Romana. Outside of the city,
the geckos have been collected in palm trash on the
beach and in semixeric woods (mouth of Rio Cha-
v6n), under old dry piles of Cocos trash in xeric
scrub on sandy soil (8 km E La Romana), and under
a rock in a dry pasture (2 km E La Romana). All
these natural situations are xeric. On the other hand,
the single speeimen from east of the Rio Chavon
was taken under a rock in mesic woods on a lime-
stone ridge. At this locality, a Tropidophis haetianus
regurgitated a S. savagei.

We already commented upon the syntopy of S.
s. savagei and S. d. diolenius at La Romana. Shreve,
in differentiating ^^Sphaerodactylus notatus savagei"
from other related difficilis-group sphaerodactyls,
correctly stressed the small escutcheon, without ex-
tensions onto the legs, of S. savagei (in contrast to
all other geckos associated nomenclatorially with

1983

THOMAS AND SCH^N \RTZ- SPHAERODACTYLUS SYSTEMATICS II

35

Sphaerodactylus not at us," with the exception of
juanilloensis). This feature is diagnostic of S. sav-
agei in contrast to N. diffidlis. In S. s. savagei, the
escutcheon is composed of 3-6 by 6-17 scales (means
4.5 by 12.3), whereas in N. d. diolenius it is com-
posed of 3-8 by 8-26 scales (means 5.5 by 17.8).
The escutcheon in S. s. savagei is more compact
than that of S. d. diolenius and extends less far onto
the legs. The two species dilfer markedly in dorsal
body pattern, the differences being more striking in
females than in males. Finally, the midbody scales
of N. d. diolenius vary between 40 and 53 (mean
46.6 ± 0.3), whereas in S. s. savagei the midbody
scales vary between 34 and 50 (mean 44.3 ± 0.8).
The means are significantly different, and in fact the
midbody mean of S. s. savagei is lower than that of
all subspecies of S. difficilis with the exception of
far northwestern S. d. lycauges.

Specimens examined. — Republica Dominicana; La Romana
Province, La Romana (MCZ 73900, ASFS V3074-76, ASFS
VI 4477-79, ASFS V14697, ASFS V35001-70); 2 km E La Ro-
mana (ASFS X9317); 8 km E La Romana (ASFS X9295-3 14);
mouth of Rio Chavon, west side (ASFS V1061, ASFS V28835-
37); 1 mi (1.6 km) NE Boca Chavon (ASFS V1075).

Sphaerodactylus savagei juanilloensis Shreve

Sphaerodactylus notatus juanilloensis Shreve, 1968, Breviora,

Mus. Comp. Zool., 280:8.

Sphaerodactylus savagei juanilloensis: Schwartz and Thomas,

1975:161.

Type-locality. — Juanillo, La Altagracia Province,
Republica Dominicana.

Holotype. -MCZ 73901.

Definition.— A. subspecies of S. savagei charac-
terized by a combination of mean low number (37-
49 on the mainland, 38-46 on Isla Saona) of mid-
body scales, and female pattern composed of dark
crossbands, this pattern often greatly obscured by
fragmentation, and never lineate; pale ocelli vari-
ably present in both males and females.

Distribution. — La Altagracia Province, Republica
Dominicana, from El Macao in the north, south to
the Boca de Yuma-San Rafael del Yuma region, and
Isla Saona (Fig. 1); also apparently Isla Catalinita.

Variation. — The series of S. s. juanilloensis may
be divided for convenience of discussion into two
subseries; 54 specimens from La Altagracia Prov-
ince and 103 specimens from Isla Saona. The main-
land specimens have the following measurements
and counts (means in parentheses): largest male
(ASFS V877) 33 mm snout-vent length, largest fe-

males (ASFS V28966, ASFS V35 1 1 0) 32 mm snout-
vent length; dorsal seales between axilla and groin
20-28 (22.8); ventral scales between axilla and groin
25-33 (29.5); midbody scales 37-49 (42.2); supra-
labials to mid-eye 3/3 (35 individuals), 3/4 (5), 4/4
(1); internasals 0 (1 individual), 1 (51), 2 (2); fourth
toe lamellae 8-13 (10.7; mode 1 1); gular, chest, and
ventral scales always smooth; escutcheon 3-6 (4.5)
by 5-12 (8.8). One hundred and three Saona spec-
imens have the following measurements and counts:
largest males (ASFS V 1 625 1 , ASFS V 1 6256) 29 mm,
largest females (ASFS V3044, ASFS V 1 6252-53) 30
mm snout-vent length; dorsal seales between axilla
and groin 20-27 (23.5); ventral scales between axilla
and groin 27-35 (30.4); midbody scales 38^6 (4 1.5);
supralabials to mid-eye 3/3 (35 individuals), 3/4 (4);
intemasals 0 (1 individual), 1 (101), 2 (1); fourth
toe lamellae 8-12 (9.9; mode 10); gular, chest, and
ventral seales always smooth; escutcheon 3-6 (4.4)
by 5-14 (8.3).

Males from the mainland vary in dorsal color-
ation from tan or purplish tan to gray or dark brown.
The heads and throats are yellow to bright orange;
the dorsum (and heads in many individuals) are
punctate with darker spots to give a salt-and-pepper
elfect. Rarely are the dark dorsal markings aligned;
an exception is ASFS V35 1 1 2, which shows vague
indications of transverse banding; this specimen also
shows some evidence of pale and widely spaced
transverse caudal bands, which are either very ob-
scure or entirely absent in other males. Scapular
ocelli are buffy to gray and are variably present; most
males laek them. The ventral coloration is gray to
pale grayish, and the throats vary from immaculate
to heavily spotted with dark brown.

Males from Isla Saona are comparable to main-
land males in both dorsal and ventral color and
pattern. However, of a series of 18 Saonan males,
only one has the throat spotted.

Mainland female juanilloensis are generally trans-
versely banded with dark brown to black on a pale
gray to grayish tan or yellow-brown ground, with
scattered white or bulfy scales accenting the cross-
banding. In some females (ASFS V35259), the tails
are vividly marked with four white spots, outlined
with black. The head pattern is variable, but there
usually are three eephalic lines, the central one ob-
scured or absent on the snout, connected to a scap-
ular rectangle or its remnants, and a pair of faint
pale ocelli in a more or less clear area. This pattern
is quite variable and in one extreme instance (ASFS

36

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

V28970) the head and shoulder patterns are com-
pletely obscured by overlying dark stippling and
marbling. Indeed, the body pattern itself is also vari-
able, and not all females are crossbanded; at the
other extreme are females which are male-patterned
salt-and-pepper dorsally and are distinguishable
from males only by the presence of a lineate head
pattern. As in the males, the ocelli are variably ex-
pressed; at best, they are faint and whitish, and at
the other extreme they are merely small clear areas
of dorsal ground color. The throats are immaculate
or may have some diagonal lateral gray markings.

Saona females were recorded in life as gray- or
black-spotted dorsally with yellowish to cream
crossbands and (if present) yellowish to cream ocelli,
these ocelli surrounded by black to brown pigment.
The tails were gray to dark yellow with dark-edged
cream crossbands and the venter was pinkish to
gray. The throats are unspotted, but a few females
have the throat indistinctly marbled with brown.
Head and scapular patterns are as described for
mainland material.

Juveniles from the mainland and those from Saona
are similar, except that there seems to be a stronger
tendency in Saona juveniles to be transversely band-
ed. The patterns in both samples are more often
widely spaced dark spots or dots on a tan to medium
brown ground. The iris color in both sexes is yellow
in adults.

Schwartz and Thomas (1975:161) noted the oc-
currence of S. savagei on Isla Catalinita. This record
is based on 12 specimens (in the USNM and the
Los Angeles County Museum) which we have ex-
amined only casually. Superficially they resemble
Isla Saona 5”. savagei in both color and pattern, and
scutellation. Since the two subspecies of Y. savagei
are weakly differentiated, we see no reason to suggest
that the Catalinita specimens are different from those
on Isla Saona, for which we use the name S. s.
juanilloensis.

Comparisons. — stated previously that savagei
and juanilloensis are rather weakly differentiated,
afthough we regard them as distinct. Part of the
problem is the material from Isla Saona which is
extremely variable in dorsal pattern (the only critical
character between the two taxa) and in many ways
seems to bridge even further the character differ-
ences between the two mainland subspecies. The
fact remains that, when living specimens of savagei
and Juanilloensis are observed together, the two sub-
species in series are seen to be quite distinctive, but

these differences are subtle and disappear to a large
extent in preserved animals. The fact that most
savagei and some juanilloensis females are salt-and-
pepper dorsally, whereas some savagei are longi-
tudinally lineate and some juanilloensis are trans-
versely banded has caused us to regard the two pop-
ulations as distinctive. As far as scutellation is
concerned, savagei averages higher than juanilloen-
sis in all dorsal scale counts, but the differences are
never striking. Nevertheless, midbody scales in sav-
agei (44.3 ± 0.8) are significantly different from
those of both the mainland (42.4 ± 0.8) and the
Saona (41.5 ± 0.3) populations of juanilloensis. It
seems likely that the Saona lizards are derived from
an as yet unsampled mainland population which
occupies the southeastern peninsula between Boca
de Chavon and Boca de Yuma. These two localities
mark, respectively, the known western and eastern
extremes of the ranges of savagei and juanilloensis,
and intergradient specimens presumably occur on
the peninsula proper. It seems likely that the Saona
population has been derived from this presumed
intergradient population, since Isla Saona lies across
the extreme southern end of the peninsula.

Remarks.— \n La Altagracia Province, 5". s. juan-
illoensis has been encountered under palm trash
along beaches, under rocks and logs in semi-mesic
woods, and in piles of fallen bark and in dead tree
stumps and among petiole bases of dead palmettos
in dense dry woods. Two juveniles were taken be-
neath boards adjacent to a shack in hot coastal scrub.
One juvenile fell from an overhanging living palm
frond onto the collector during the day, and an adult
was encountered at night 0.6 m above the ground
on a cycad (Zamia) after a heavy nocturnal rain.
Apparently S. savagei climbs upon occasion.

The previous discussion has dealt exclusively with
specimens from the eastern extreme of Hispaniola
and Isla Saona. There is another population of S.
savagei between 2 and 10 km NE Sabana Grande
de Palenque in San Cristobal Province, some 200
km west of the nearest record for S. s. savagei (La
Romana). The material from Sabana Grande (ASFS
V14425, ASFS V28514-18, ASFS V28462, ASFS
V42525-39) is composed of 1 3 juveniles, one adult
female (ASFS V285 14, snout-vent length 33 mm),
and five subadult males and two subadult females
(snout-vent lengths 25-29 mm). There is nothing
distinctive about the subadults and juveniles, and
scale counts are not unusual. The adult female is
heavily mottled (very grossly salt-and-pepper) with

1983

THOMAS AND SCH^ \RTZ- S PH AERO DACTYLUS SYSTEM ATICS II

37

black and with scattered white scales dorsally. There
is a pair of white ocelli but no solid dark scapular
area (although a hollow dark rectangle is present),
and the tail is contrastingly crossbanded with about
four white bands, broadly outlined with black. The
snout-vent length of the female is 1 mm greater than
that of any other female S. savagei.

Although the series resembles other S. savagei, it
seems unlikely that it is properly assigned to either
savagei or juanilloensis. But the lack of more adults
of both sexes deters us from taking any nomencla-
tural action. There is also of course the possibility
that this population, which appears to be rather lo-
cal, is an introduction. On the other hand, we cannot
cavalierly dismiss the possibility that there are other
(relictual ?) populations of S. savagei between La
Romana in the east and Sabana Grande in the west.
All of our collections in this intervening area have
yielded only S. difficilis, but this is negative evi-
dence. The Sabana Grande lizards were taken under
trash on a Coco^-planted beach and under trash in
an abandoned army post. At nearby (but interior)
localities (3.4 mi W Sabana Grande de Palenque; 2
km E Juan Baron) we secured only S. difficilis. The
only other nearby coastal locality (4.8 mi S Bani)
yielded only S. difficilis. We have thus been unable
to extend the known range or our knowledge of the
distribution and interrelationships of this western
population of S. savagei with S. difficilis. Its taxo-
nomic status likewise remains very uncertain and
depends upon securing more adults and upon the
status of the two more eastern subspecies.

The specimens from Isla Saona were taken in Co-
cos trash along the beach (Mano Juan) and in rock
piles and Cocos trash adajcent to the beach and
among the houses of the local cuartel at the north-
west comer of the island. The species is exception-
ally common on Isla Saona. As yet there are no
Sphaerodactyl us known from Isla Catalina, which
lies closely offshore of La Romana. Either S. savagei
or S. difficilis (or both) may well occur on that islet.

S. s. juanilloensis lies between the ranges of S. c.
apocoptus and S. d. diolenius. It should be recalled
that S. s. savagei and S. d. diolenius are syntopic at
La Romana, but there are no records of syntopy
between diolenius and juanilloensis. These two taxa
approach each other most closely at 1 2 km E Otra
Banda {diolenius) and 19 km SE El Macao (juanil-
loensis), a distance of about 17 km. Most records of
juanilloensis are coastal, the exceptions being the
two localities near San Rafael del Yuma, about 10

kilometers inland. The closest approach of S. s.
juanilloensis and 5. c. apocoptus is 0. 1 mi SE El
Macao and 5.7 mi SE La Vacama, a distance of
about 15 km.

Although Shreve regarded savagei as a subspecies
of S. notatus, along with difficilis, we are convinced
that savagei is a species distinct from S. difficilis.
We had this conviction prior to the collection of the
two species syntopically at La Romana. The two
species are comparable in size and in male color-
ation and pattern (although male S. savagei tend
toward more pale colors than do male S. difficilis),
but the small and almost patch-like escutcheon of
S. savagei is quite different from that of S. difficilis
with its poplitead extensions. Likewise, in having a
compact escutcheon, S. savagei differs from S. ocoae,
S. clench i, and S. lazelli. Female Y. savagei are quite
different from female S. difficilis, S. ocoae and S.
clenchv, the complex head-neck-scapular pattern of
S. savagei, although derivable from the trilineate
difficilis head pattern and scapular patch with in-
cluded ocelli, differ in many details, not the least of
which are the absence or reduction of the median
snout-head line and the occurrence of ocelli within
a pale area inside a darker scapular rectangle. S.
savagei differs in scutellation in having only slightly
enlarged median subcaudal scales (in contrast to
usually widened scales in this position in difficilis,
clenchi, and lazelli). In scale counts, S. savagei shows
no overlap in dorsal or midbody counts with S.
clenchi, and very little overlap in dorsal scale counts
with S. ocoae (see Schwartz and Thomas, 1977:62).
The overlap in midbody scale counts between S.
savagei and S. difficilis is fairly great (total of 13
scales overlap), but the means in both subspecies of
S. savagei are lower than those of all subspecies of
difficilis except lycauges. The dorsal and midbody
counts of the single N. lazelli fall within the known
range of these counts in S. savagei. The lack of throat
keeling in all specimens of S. savagei differs from
the strongly keeled throat scales of S. lazelli, and
the occasionally keeled throat scales of S. difficilis,
S. ocoae, and S. clenchi.

Specimens examined. — Rep\jbi.ica Dominicana: La Altagra-
cia Province. 0.2 km SE El Macao (ASFS V28959-61); 1.1 km
SE El Macao (ASFS X7882-89); 1.4 km SE El Macao (ASFS
V886); 8 km SE El Macao (ASFS V131 15-17); 5.3 km SE El
Macao (ASFS V28964-83); 19 km SE El Macao (ASFS V909);
1.2 km SSW Puma Cana (ASFS V35092, ASFS V35107-12); 5.5
km SSW Punta Cana (ASFS V35257-61); 14.1 km NE Juanillo
(ASFS V29180-81); Juanillo (MCZ 73901 -holotype); 0.8 km
SE San Rafael del Yuma (ASFS V13810); 6.4 km SE San Rafael

38

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

del Yuma (ASFS V877); 0.8 km NW Boca de Yuma (ASFS
V960). Isla Saona, environs of Mano Juan (ASFS V3041-52,
ASFS V16187-90, ASFS V16251-302); 0.8 km W Mano Juan
(ASFS V16203-29, LDO 7-5538-39); northwest comer (ASFS
V34976-81).

Sphaerodactylus cochranae Ruibal

Sphaerodactylus cochranae~RM\ha\, 1946, Amer. Mus. Novitates,

1308:1.

Type-locality. — Bdi]\i2L de San Lorenzo, El Seibo
Province, Republica Dominicana.

Holotype. — AMNH 50233.

Definition.— A. species of Sphaerodactylus with
large, acute, strongly keeled, flattened, imbricate
dorsal scales, axilla to groin 20 to 23; no area of
middorsal granules or granular scales; dorsal body
scales with hair-bearing scale organs, three to eight
organs around free edge of scales and at times three
organs set back from the edge, each organ with one
hair. Dorsal scales of tail keeled, acute, imbricate,
and flat-lying; ventral scales of tail smooth, rounded,
slightly enlarged midventrally; gular scales smooth
to keeled; ventral scales smooth rounded, imbricate,
axilla to groin 20 to 26; scales around midbody 38
to 41; intemasals 0 or 1; upper labials to mid-eye
3; escutcheon 5 by -.

Color pattern not sexually dichromatic, consisting
basically of three prominent dark lines on a pale
ground; scapular patch and ocelli absent. Iris color
unknown.

Distribution. — K.no'wn only from the vicinity of
the type-locality (Fig. 1).

Variation. —Sphaerodactylus cocranae is known
from the holotype and paratype and another sub-
adult specimen taken by Danny C. Fowler in 1971.
The holotype is a female with a snout-vent length
of 28 mm; the paratype is a very badly damaged
male with a snout-vent length of 30 mm; and the
freshly taken specimen is a young female with a
snout-vent length of 25 mm. Combining data from
all three specimens (and recognizing that the data
from the male paratype are subject to inaccuracies)
yields the following counts: dorsal scales between
axilla and groin 20-23; ventral scales between axilla
and groin 20-26; midbody scales 38-41; suprala-
bials to mid-eye 3/3; intemasals 0 (2 specimens) or
1 ; fourth toe lamellae 1 1 and 1 3 (no data from para-
type); gular scales keeled in paratype, partially keeled
in new specimen, smooth in holotype; chest scales
partially keeled in paratype, smooth in two other
specimens; ventrals smooth.

Ruibal’s (1946) data on coloration and figure of

the holotype pattern describe the pattern of S. coch-
ranae excellently: “A middorsal stripe extending
from the frontal region to the base of the tail varies
in width but at mid-body embraces three scale rows.
It is flanked on either side by a light stripe approx-
imately two scales wide at mid-body. Dorsolateral
blackish stripes extend from the nostril, through the
eye, broadening above the ear opening, and con-
tinuing onto the sides of the tail. At mid-body the
dorsolateral stripes fall on the ninth and tenth row
of keeled scales. Ventrally, the lizard is immaculate,
although scattered melanophores are present on the
lighter areas of the dorsum. The legs are mottled
brownish, the hind legs having a light-colored spot
on the femur and knee. . . . Mr. Hassler in his field
notes described the fresh specimen as having ‘lon-
gitudinal black stripes on a pale tan or yellowish
brown ground color; hind legs with slightly reddish
spots; belly white.’”

Concerning the color and pattern of the freshly
taken specimen, Fowler noted that the pale dorsal
stripes were tan and the dark stripes dark brown.
All stripes begin on the snout and continue to the
tail base, where the upper surface of the tail is con-
color with the pale dorsal stripes but has some brown
stippling. The venter was tan with pink on the chest
and at the vent. The forelimbs were patternless tan
above, whereas the hindlimbs were tan with orange
spots. The ventral surfaces of the limbs were tan.

Remarks.— Wxhovigh Shreve (1968) did not in-
clude S. cochranae in his discussion of the dijficilis
complex, there seems little reason to exclude it. Rui-
bal ( 1 946:3-4) suggested that S. cochranae was most
closely related to N. macrolepis Gunther from Puer-
to Rico, but that the species was also similar in some
scale features to S. samanensis (which has the same
type-locality as S. cochranae). Ruibal also noted
some similarities with S. copei, an exclusively Hai-
tian species which has a middorsal region of small
granules, a feature which S. cochranae lacks. We
consider S. copei to be most closely related to the
predominantly Cuban scaber group of sphaerodac-
tyls (Schwartz, 1961; Thomas, 1966), with which
we cannot associate S. cochranae.

By virtue of its very strongly lineate dorsal pat-
tern, S. cochranae differs from all other Hispaniolan
members of the dijficilis complex. Some specimens
(females) of S. dijficilis are lineate but never so
strongly and contrastingly as S. cochranae. Like-
wise, the dorsal scales of S. cochranae are much
larger (and thus dorsal counts are lower) than other
members of the complex (or lie at the lower extremes

1983

THOMAS AND SCHWARTZ- SYSTEMATICS II

39

of dorsal scale variation in other species). Further
comparisons are hardly necessary, since S. coch-
ranae is a very distinctive species.

S. cochmnae is now known to be sympatric with
S. difficilis and S. samanensis, and it may be sym-
patric with S. clenchi. Fowler and Sheplan collected
in four caves on the Bahia de San Lorenzo; in two
they secured no geckos, in another S. cochranae and
S. difficilis, and in another S. samanensis. Interest-
ingly, although some female Bahia de San Lorenzo
S', d. diolenius are somewhat lineate dorsally, they
in no way approach the condition in S. cochranae
in intensity of their markings, and we do not inter-
pret them as intermediate between S. cochranae and
S. difficilis.

The details of distributions of the species coch-
ranae, darlingtoni, difficilis, clenchi, and samanen-
sis along the southern shore of the Bahia de Samana
and farther east (where also savagei occurs) remain
to be clarified. Although S. clenchi has not been
taken sympatrically with cochranae, samanensis, or
difficilis in this region, clenchi is known from Caba,
only 23 km to the west at the head of the Bahia de
Samana. Since the San Lorenzo locality lies at the
foot of the haitises region of the eastern Republica
Dominicana, it is possible that the occurrence here
of such “rare'” species of cochranae and samanensis
is due to the fact that collections at San Lorenzo are
peripherally sampling a much more extensive dis-
tribution of cochranae and samanensis south of the
Bahia de Samana. These two geckos may be pri-
marily inhabitants here of mesic forested limestone
hills; however, Thomas and Leber secured only S.
darlingtoni in the haitises proper (near Gonzalo),
and at Caba on the coast we secured only clenchi.
This entire region remains so poorly known her-
petologically that further speculation is impossible.

Ruibal ( 1 946), quoting field notes by Hassler, stat-
ed that the holotype was taken from the breakwater,
and the paratype from a limestone wall near the
caves. Fowler’s specimen was taken from trash just
within the mouth of a cave immediately south of
the mouth of the Bahia de San Lorenzo. This and
other local caves are now used by Sabana de la Mar
fishermen as temporary retreats, and thus there has
accumulated in them a quantity of human refuse,
in which the cochranae was secured. Wetmore and
Swales (1931) have a photograph (PI. 13) of one of
the Lorenzo caves, which have long been well known
as fossil sites.

Our examination of the two original specimens
of S. cochranae showed that there were only knob-

like scale organs present on the dorsal scales. The
freshly taken individual definitely shows hair-bear-
ing organs. Although it is barely conceivable that
the differences are ontogenetic, we suspect that the
older specimens (which have long been preserved
and repeatedly handled) no longer possess hairs in
the scale organs merely because of their history.
Only additional material will verify this suggestion.

Specimens examined. — Kepubuca Dominicana; El Seibo
Province. Bahia de San Lorenzo (AMNH 50233 — holotype); 1.2
km inland from Bahia de San Lorenzo (AMNH 50133); Cuevas
de Cano Hondo (ASPS V35284).

Sphaerodactylus darlingtoni Shreve

Sphaerodactylus darlingtoni Shreve, 1968, Breviora, Mus. Comp.

Zool., 280:15.

Sphaerodactylus noblei Shreve, 1 968, Breviora, Mus. Comp. Zool.,

280:17.

Definition. — \ species of Sphaerodactylus with
large, acute, flattened, imbricate dorsal scales, axilla
to groin 24-36; no area of middorsal granules or
granular scales; dorsal body scales with four to six
hair-bearing organs around apex, each with a single
hair. Dorsal scales of tail keeled, acute, imbricate,
and flat-lying; ventral scales of tail smooth, rounded,
enlarged midventrally; ventral scales smooth,
rounded, imbricate, axilla to groin 24 to 3 1 ; scales
around midbody 48 to 59. Snout moderately blunt
in dorsal aspect, slightly depressed, not decurved;
snout scales small, elongate, keeled, subimbricate to
imbricate; postnasals 1 or 2; intemasals 1 to 3; su-
pralabials to mid-eye 3; dorsal head scales small,
elongate, keeled, imbricate to juxtaposed; temporal
scales small, rounded, keeled, raised on posterior
edge. First infralabial slightly broader anteriorly than
posteriorly, subrectangular; gular scales between in-
fralabial rami very small to small, keeled, subim-
bricate; central gulars small to very small, almost
granular, subimbricate, keeled; fourth toe lamellae
6 to 1 1 ; escutcheon with a narrow central area (mode
2-3 scales in length) with extensions onto thighs to
popliteal region (2-5 by 14-28).

Coloration not sexually dichromatic in pattern.
Dorsal pattern of a central, bilobed, light mark on
head, followed by paired or bilobed light marks
on occiput; paired scapular ocelli and scapular patch;
light border of scapular figure divided into separate
anterior and posterior portions, the anterior portion
roughly U-shaped with median anteriad indenta-
tion, the posterior portion roughly U-shaped, set off
from scapular patch. Sacral figure of dorsolateral
light lines; tail pattern of light chevronate markings;

40

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

venters pale to heavily pigmented; throats un-
marked to heavily marked with dark pigment.

Basic Pattern

The pattern of S', darlingtoni is conveniently described with
reference to the pattern elements of S. altavelensis.

Anterior cephalic figure. — Clearly demarcated, light centrally
with dark edge, cleft or bilobed with each lobe posteriorly di-
rected, ending slightly posterior to auricular opening.

Posterior cephalic figure.— Two transversely positioned occip-
ital lobes (separate or fused) having bold dark edges laterally and
posteriorly but fading anteriorly and in some specimens confluent
with the light central areas of the anterior cephalic figure.

Scapular figure. — Scapular patch (not present in all specimens)
small, irregular, often bilobed, enclosing two small buff ocelli and
(in its most complete development) with an anteriorly directed
extension to the midregion of the scapular figure border.

Scapular figure border.— The border is a light vitta that en-
closes the scapular figure. In 5'. darlingtoni it is divided into two
unconnected anterior and posterior parts. The anterior border is
concave anteriorly (more or less U-shaped), at times with a mid-
line anteriad extension or indentation; the ends extend forward
on each side and curve ventrad, fading out in the vicinity of the
auricular opening, where they may be closely juxtaposed to the
posterior end of a light postauricular stripe. The posterior border
is likewise concave anteriorly; its ends curve anteriorly and ven-
trally, fading out just anterior and dorsal to the forearm insertion.
This posterior scapular figure border may be continuous or bro-
ken into dashes or ocelliform marks. The name “scapular figure
border” is retained for the two markings described here because
of their evident homology with the scapular figure border of S.
altavelensis (Thomas and Schwartz, MS), which is a closed light
figure surrounding the scapular figure. Hereinafter we refer to the
anterior and posterior parts of the scapular figure border as “an-
terior border” or “posterior border.”

Dorsal body pattern. — Waries from uniform brown through
various degrees of flecking and spotting with dark brown or black
to coalescence of darker pigment into indistinct crossbands or
imperfect scattered ocelli.

Sacral figure. — Pairs of short, dorsolateral dark lines enclose a
light zone on each side to give a lineate sacral figure. The sacral
lines fade a short distance anteriorly but posteriorly merge into
the tail pattern, in which (in its best expression) the dark dor-
solateral lines have mesiad diagonal extensions with associated
light areas producing a posteriorly directed chevronate-ocellate
pattern. The dark-edged light areas may be especially prominent
on the base of the tail and even on the posterior sacral region,
producing an adjunct to the sacral pattern.

Ventral coloration.— \ enters have various degrees of uniform-
ly dispersed dark pigment ranging from nearly uniformly pale
(off-white) to dark gray in intensity; some specimens show in-
dications of ventral lines reminiscent of S. altavelensis. Throats
vary from nearly unpigmented to heavily marbled with dark
pigment, producing a pattern of light spots on a dark ground that
merges laterally with the light and dark labial and postocular
markings.

Distribution. -North-central and eastern Hispan-
iola, from Valverde Province, east and south onto
the Peninsula de Samana, south to the vicinity of
Esperalvillo in San Cristobal Province and east to

La Romana Province. To the west and southwest,
populations distantly removed from the rest of the
known range of the species occur in the north range
of the Sierra de Neiba south of Elias Pina and in
the Sierra Martin Garcia (Fig. 3).

Sphaerodactylus darlingtoni darlingtoni Shreve

Sphaerodactylus darlingtoni Shreye, 1968, Breviora, Mus. Comp.

Zool., 280:15.

Sphaerodactylus darlingtoni darlingtoni: Schwartz and Thomas,

1975:149.

Type-locality.— ¥ico Diego de Ocampo, summit
dome, ca. 4,000 ft, between Puerto Plata and San-
tiago, Santiago Province, Republica Dominicana.

Holotype. — MCZ 44380.

Distribution. — The western part of the Cordillera
Septentrional of the Republica Dominicana, where
known from Pico Diego de Ocampo west to Val-
verde Province.

Definition.— A subspecies of S. darlingtoni char-
acterized by the presence of a single postnasal, by a
posterior cephalic figure not strongly connected with
the anterior eephalic figure, and by posterior border
set off from a small scapular patch.

Variation. — 'LdsgQsi male (MCZ 44380) 25 mm
snouth-vent length, largest female (ASFS V34223)
29 mm snout-vent length; dorsal scales axilla to
groin 24-36 (30.0); ventral scales axilla to groin 24-
29 (26.3); midbody scales 49-59 (53.5); supralabials
to mid-eye 3/3; intemasals 1 (30), 2 (4); fourth toe
lamellae 9-11; throat scales strongly to weakly
keeled; escutcheon 2-5 by 11-21.

Coloration is basically as described for the species.
Males show a tendency towards uniform (unspotted)
dorsal body coloration. The posterior cephalic figure
may be represented by a single transverse or straight
shallow V-shaped mark, or two, three, or more iso-
lated light marks (Fig. 4). Scapular patches vary from
well developed, completely enclosing ocelli and with
anteriad extension to scapular figure border, to re-
duced and not completely surrounding ocelli or re-
duced to isolated black margins to each ocellus. In
several specimens the anterior scapular figure bor-
der is most prominent in a short middorsal segment,
but the ends, although faint or interrupted, can be
seen following approximately the normal course. One
individual (ASFS V32252) shows extreme reduction
of head pattern. The median anteriad indentation
of the anterior scapular figure border is lacking or
indicated by only a slight widening; in four it is
reversed (that is, the border is V-shaped). Throats
are uniform pale to heavily marbled with dark gray

1983

THOMAS AND SCH^N \KTZ- SPHAERODACTYLUS SYSTEMATICS II

41

Fig. 3. — Map of eastern Hispaniola showing the distributions of the subspecies of Sphaerodactylus darlingtoni.

and white; there is a greater tendency for males to
have marked and females unmarked throats. The
holotype is almost completely patternless except for
a smudge indicating the scapular patch and some
ambiguous markings, possibly not true pattern rem-
nants, on the head. The paratype (MCZ 44379),
although faded, has sufficient pattern evident to as-
sociate it with the small dark geckos of northeastern
Hispaniola. An anterior cephalic figure is plainly
evident, and, although not bilobed, this character-
istic is variably expressed in the species. The spec-
imen possesses an anterior scapular border, the ex-
tremities of which proceed auriculad, and a posterior
border, the extremities of which fade out near the
forearm insertion. An irregular but bold scapular

patch is in contact with the anterior scapular figure
border; the ocelli, not completely enclosed by the
patch, are not evident, probably because of fading
of the ground color. The paratype additionally shows
faint, irregular, dark-edged, light crossbars on the
dorsum, a pattern variant found in occasional spec-
imens (ASPS V32664, ASPS V33505); a lineate sa-
cral pattern; and a marbled throat pattern.

Color in life: the dorsal ground color dark wood
brown (almost black) to medium brown; the scap-
ular patch is black and the ocelli white. Light head
and scapular markings are buffy or whitish, and ven-
ters dark brown overlying deep yellow.

Remarks. — Shreye (1968) regarded the “absence
of distinctly keeled scales on the throat” as a diag-

42

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Fig. 4. — Diagrammatic dorsal views of cephalic and scapular patterns of Sphaerodactylus darlingtoni: A) S. d. darlingtonr, B) S. d.
noblei; C) S. d. bobilinr, D) 5. d. mekistus.

nostic feature of darlingtoni versus noblei, but faint
keeling is evident on the throats of the holotype and
paratype of darlingtoni (part of the dilficulty in dis-
cerning the keeling is likely a result of the age and
poor condition of the specimens). Keeling is vari-
able, however, in noblei (see below) and, although
modally present, it may be very weak or absent.
There is no difference in keeling between the two
taxa. Except for the presence of a single postnasal,
the other meristic differences noted by Shreve be-
tween darlingtoni and noblei have been swamped
in the longer series of these geckos presently avail-
able. The incidence of single postnasals is geograph-
ically circumscribed, and the overwhelming simi-
larity of these geckos in other aspects of scutellation
and in coloration to the more eastern populations
leaves us with little doubt that the names darlingtoni
and noblei apply to a single species.

Specimens examined. — Kevv^vic/k Dominicana: Santiago
Province, Pico Diego de Ocampo, summit dome, ca 1,220 m,
between Puerto Plata and Santiago (MCZ 44380— holotype; MCZ
44379). Valverde Province, 9.4 km N Cruce de Guayacanes, 427
m (ASFS V22251-68); 9.9 km N Cruce de Guayacanes, 427 m
(ASFS V27617-18, ASFS V27633-37).

Sphaerodactylus darlingtoni noblei Shreve

Sphaerodactylus noblei Shreve, 1 968, Breviora, Mus. Comp. Zook,

280:17.

Sphaerodactylus darlingtoni noblei: Schwartz and Thomas, 1975:

149.

Type-locality. — \.os Bracitos, Duarte Province,
Republica Dominicana.

Holotype.- 45216.

— Northeastern Republica Domini-
cana from Salcedo and northern La Vega provinces
east to the Peninsula de Samana, south to central

San Cristobal Province (Esperalvillo) and southeast
to El Seibo and La Romana provinces.

Definition.— A subspecies of Sphaerodactylus
darlingtoni in which the nasal scale is contacted pos-
teriorly by 2 postnasals, the posterior cephalic figure
is separate from the anterior cephalic figure, and the
posterior border is set off from the small scapular
patch (Fig. 4B).

Variation. — TstrgQsi male (ASFS V33545) 25 mm
snout-vent length, largest female (ASFS V32259)
27 mm snout-vent length; dorsal scales between
axilla and groin 25-33 (28.7); ventral scales axilla
to groin 24-3 1 (26.9); midbody scales 47-56 (5 1 .6);
supralabials to mid-eye 3/3; intemasals 1 (48); 2
(39), 3 (10); fourth toe lamellae 8-11; throat scales
strongly to weakly keeled; escutcheon 2-5 by 11-
28.

Coloration generally as described for the species.
The anterior cephalic figure varies from strongly
bipartite with acuminate lobes to weakly cleft with
rounded or truncated lobes. The posterior cephalic
figure is modally separate from the anterior figure,
each of the two parts being roughly oval with a
median anastomosis, but it may be fused into a
transverse oval with no indication of the paired con-
dition, or the two parts may be completely separate,
resulting in a pair of occipital ocelli. In the modal
bilobate configuration the figure may be open an-
teriorly (no dark margin and interior light area fad-
ing to ground color or anastomosing with the an-
terior cephalic figure). The anterior border of the
scapular figure varies from complete as described
for the species, to abbreviated without extensions
onto the side of the head; the median anteriad ex-
tension is prominent to absent and when present
may (rarely) join with the posterior cephalic figure.

1983

THOMAS AND SC\m XKTZ- SPH AERO DACTYLUS SYSTEMATICS II

43

The scapular patch is typically as described for the
species but may be reduced to a small smudge with-
out ocelli or may be absent. Ocelli alone may be
present without indication of a scapular patch or
with only a broad dark margin around each ocellus
as a remnant of the scapular patch. The posterior
border is incomplete in most specimens, frequently
broken into dashes or weak ocelli which follow the
outline of a regular border with various degrees of
fidelity.

Color in life: The dorsal ground color is very dark
brown, in some animals almost black. Anterior and
posterior cephalic figures and scapular figure borders
are buff to yellowish brown. Venters are dark brown
with a dull orange underlayer or gray to cream with
darker flecking, or yellowish brown or yellowish.
Tail ground color is the same as that of the venters.
Scapular ocelli are white or gray, and the iris color
is golden.

Remarks. — (1968) commented on speci-
mens of S. darlingtoni from the Bahia de Samana
region (USNM 74943 from Rio San Juan on the
Peninsula de Samana and USNM 74974-77 from
Boca del Infiemo on the south side of the bay),
noting that they might belong to a distinct race be-
cause of differences in scale counts from the holo-
type and five paratypes of S. nobiei (=S. darlingtoni
noblei). However, the specimens in question do not
differ in scale counts from our much larger sampling
of the species from surrounding areas. The color-
ation of the Samana region specimens is very faded,
but what is visible does not differ radically from
darlingtoni. In three of the specimens (USNM
74974-75, USNM 74977) the posterior cephalic fig-
ure is broad and arcuate, following the curvature of
the posterior edge of the anterior cephalic figure.
This condition is not typical of darlingtoni but is
approached in some specimens (ASFS V33546,
ASPS V34152, ASFS V33486 and others). A pe-
culiarity of two of these three specimens (USNM
74974-75) is the presence of a light spot just pos-
terior to the lateral extremities of the posterior ce-
phalic figure on either side, that is, just dorsal and
slightly posterior to the auricular opening. In the
other Boca de! Infiemo specimen (USNM 74976)
the posterior cephalic figure is irregular and, al-
though broad, is not arcuate; it also lacks light pos-
terolateral head spots. In the Rio San Juan specimen
the posterior cephalic figure is not clearly discernible
but does not seem extensive or arcuate. Otherwise
the five specimens show indications of, although not
in every specimen, anterior borders, and sacral pat-
terns, all typical of S. darlingtoni. Our specimens

from 10 km NE Gonzalo (ASFS V3 134-36), the
locality for new specimens closest to Boca del In-
fiemo, show no peculiarities that would ally them
with the specimens in question. The pattern as a
whole is rather typical of S. darlingtoni. The pos-
terior cephalic figures are small, centrally posi-
tioned, and Y-shaped; temporal light spots are ab-
sent.

The Peninsula de Samana is an established region
of endemism; and the haitises region on the south
side of the Bahia, although poorly known herpeto-
logically, is distinctive physiographically. We can-
not presently exclude the possibility of there being
one or more distinct subspecies of S. darlingtoni on
the Peninsula de Samana or the area surrounding
the bay; but on the basis of the few specimens avail-
able, we cannot affirm it either.

It is interesting that USNM 74943 is the only
Peninsula de Samana specimen of S. darlingtoni.
Our extensive collections on the Peninsula in recent
years have yielded no additional material. The very
mesic habitat seems ideal, although many of our
collections have come from the coastal areas of the
Peninsula, especially Cocos plantations along the
south coast. However, S. darlingtoni has been taken
in a variety of situations, including cultivation of
various sorts, and we cannot account for its scarcity
on the Samana.

Specimens examined. — Kevijulicx Dominicana: Salcedo
Province, 3 km NW Salcedo (ASFS V2942-43). La Vega Prov-
ince, Guaigiii, 4 km S La Vega (ASFS VI 8390); Guaigui, 4.8 km
S La Vega, 92 m (ASFS V33533-38). Duarle Province, 10.2 km
SETenares, 214 m (ASFS V33499-508); 8 km NE Ponton (ASFS
V3 1 94, ASFS V3633); 7.2 km E Cruce de Pimentel, 122 m (ASFS
V33486-88); 6.1 km NW San Francisco de Macoris (ASFS
V28115-21, ASFS V28276-363); 29.9 km E San Francisco de
Macoris (ASFS V28128). Marta Trinidad Sdnchez Province, 2
km S El Factor (ASFS VI 846-52); El Barro, 1.4 km N El Factor
(ASFS V28097); 1.6 km S Cano Abajo (ASFS V34 149-60, ASFS
V342 1 8-38); 4 km N Azucey (ASFS V 1 6070-76). Samand Prov-
ince, Rio San Juan (USNM 74943). San Cristdbal Province, 1 1
km W Esperalvillo (ASFS VI 440 1-03); 10 km W Esperalvillo,
143-144 m (ASFS V28624-29, ASFS V40947-53); 2 km W Es-
peralvillo (ASFS VI 4396-98); 10 km NE Gonzalo (ASFS V3134-
36). El Seibo Province, Boca del Infiemo (USNM 74974—77);
10.5 km N Hato Mayor (ASFS V35296-98, ASFS V35305). La
Romana Province, 1 3.4 km NE La Romana, 3 1 m (ASFS V28907,
ASFS V35071-80).

Sphaewdactylus darlingtoni bobilini,

new subspecies

Holotype. — USNM 194045, adult female, from
the west slope of Mt. Busu, between 610 and 854
m, Barahona Province, Republica Dominicana, one
of a series collected by Danny C. Fowler and Bruce

44

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Fig. 5. — Dorsal views of subspecies of Sphaerodactylus darlingtoni, as follow; A) 5. d. darlingtoni (ASFS V3225 1); B) 5. d. noblei (ASFS
V33499); C) Y d. bobilini (USNM 1 94045 -holotype); D) S. d. mekistus (USNM 1 94046 -holotype).

R. Sheplan on 15-17 September 1971. Original
number ASFS V31236.

Paratypes.—ASiFS V21173, Sierra Martin Garcia, about 915
m, above Barreras, Azua Province, Republica Dominicana, Rob-
ert K. Bobilin, 20 July 1969; ASFS V3 1233-35, ASFS V31237-
40, same data as holotype.

Definition. — A geographically disjunct subspecies
of S. darlingtoni characterized by a fusion of each
part of the paired posterior cephalic figure with the
corresponding paired lobes of the anterior cephalic
figure to form a slender posterior extension of each
lobe of the anterior figure, posterior scapular figure
border set olf from scapular patch (Fig. 5C), and 2
postnasals.

Distribution. — Highlspcids of the Sierra Martin
Garcia in the southwestern Republica Dominicana.

Description of holotype.— An adult female with a
snout-vent length of 26 mm, tail 21 mm (regener-
ated); dorsal scales axilla to groin 26, ventral scales
axilla to groin 27, scales around midbody 53, fourth
toe lamellae 9. Snout moderately blunt in dorsal
aspect, slightly depressed; snout scales moderately-
sized, flattened, keeled, subimbricate; 1 intemasal;
2 postnasals; supralabials to mid-eye 3/3; dorsal head
scales small, elongate, keeled, juxtaposed to sub-
imbricate; temporal scales small, oval, swollen,
keeled and only slightly raised on posterior edge.
First infralabial slightly broader anteriorly than pos-
teriorly, roughly rectangular in shape; gular scales
between infralabial rami moderately small, juxta-
posed to subimbricate, smooth to keeled; central

gulars very small, smooth, imbricate; chest and ven-
tral scales smooth.

Dorsal ground color brown, dorsum finely spotted
with black one-scale-sized spots; buff anterior ce-
phalic figure with weak basal bilobations and paired,
narrow, elongate posterior extensions onto occiput.
Anterior border curves anteriorly across base of neck
and ventrally towards auricular opening; black scap-
ular patch has narrow anterior contact with anterior
border and bilobed posterolateral expansions to in-
clude a pair of buff scapular ocelli; posterior border
asymmetrical, nearly complete but broken (that is,
not continuous but composed of more or less dis-
crete light dots). Lineate sacral figure composed of
dark-edged light dorsolateral lines; dorsum of tail
with series of paired elongate ocelli, each pair jux-
taposed or fused to form a chevronate pattern. Ven-
ter pale (pinkish gray in life), scale edges stippled
with black pigment; throat heavily stippled with
black (Fig. 5C).

Variation.— Of the holotype and eight paratypes,
the largest male (ASFS V31233) is 25 mm snout-
vent length, largest female (ASFS V21 173) 27 mm;
dorsal scales axilla to groin 26-33 (29.5); ventral
scales axilla to groin 27-31 (28.8); midbody scales
50-55 (52.5); supralabials to mid-eye 3/3; intema-
sals 1 (7), 2 (2); fourth toe lamellae 9-10; throat
scales weakly keeled in five specimens, smooth in
four; escutcheon 3-5 by 24-26.

Three males show fainter (paler, not contrasting)
cephalic figures and scapular figure borders and were
noted as having a yellow-tan ground color in life in

1983

THOMAS AND SC\m KRTZ-SPHAERODACTYLUS SYSTEMATICS II

45

contrast to the darker brown color of the females;
one male with a poorly developed escutcheon (and
therefore possibly immature) and included with the
females in the color notes has a contrasting “female”
pattern. As noted, the female ground color is darker
and contrasts more sharply with the pale head mark-
ings and scapular figure borders; the scapular patches
are more extensive. Venters were pinkish gray,
throats heavily stippled black, and scapular ocelli
white.

Remarks. — Sphaerodactylus d. bohilini is essen-
tially an insular subspecies on a mesic island, the
highlands of the Sierra Martin Garcia, surrounded
by xeric lowlands. As with any insular subspecies,
its taxonomic position must be somewhat arbitrarily
assigned. Sphaerodactylus d. bobilini could be de-
scribed as a separate species with little danger of this
decision being upset by the discovery of intergrades.
On the grounds of proximity and the sharing of
certain pattern elements, it might otherwise be pre-
sumed that bobilini is a high elevation derivative of
S. altavelensis, abundantly present in the surround-
ing lowlands (Thomas and Schwartz, MS). How-
ever, the similarity in several respects (particularly
the breaking of the scapular figure border into an-
terior and posterior parts and the configuration of
these parts) of the bobilini pattern to that of north-
eastern Hispaniolan S. darlingtoni convinces us of
the strong relationships of these two forms.

Etymology.— The: subspecies is named in honor of Robert K.
Bobilin who collected the original specimen.

Sphaerodactylus darlingtoni mekistus,

new subspecies

Holotype. — U SNM 194046, adult female, from
1 1 km S Elias Pina, 854 m, La Estrelleta Province,
Republica Dominicana, taken by Richard Thomas
on 13 August 1963. Original number ASFS V418.

Paratype. — ASFS VI 449, same locality as holotype, hatched
22 September 1963 from egg taken on same date as holotype by
Richard Thomas.

Definition.— \ subspecies of S. darlingtoni char-
acterized by modification of the posterior cephalic
figure to form a pair of divergent, winglike lobes
juxtaposed to the anterior cephalic figure, a compact
scapular figure in which the anterior and posterior
borders are in broad contact with the scapular patch
(Fig. 4D and 5D), and the presence of 2 postnasals.

Distribution. — Known only from the type-locality
in the highlands of the Sierra de Neiba.

Description of holotype (data on paratype in pa-

rentheses).—An adult female, 25 mm snout-vent
length ( 1 3), tail 20 mm ( 1 2), regenerated at tip. Dor-
sal scales axilla to groin 27 (28), no axilla to groin
ventral count due to injury (25 for paratype), scales
around midbody 47 (51). Snout moderately blunt
in dorsal aspect, slightly depressed; snout scales
moderately-sized, keeled, subimbricate to imbri-
cate; dorsal head scales small, elongate, keeled,
weakly subimbricate; temporal scales small, broad,
rounded, raised posteriorly and keeled; 2 (2) post-
nasals; 2 (2) intemasals; upper labials to mid-eye 3/
3 (3/3); dorsal head scales small, elongate, keeled,
subimbricate to weakly imbricate; temporal scales
small, rounded, keeled, raised posteriorly; first in-
fralabial slightly broader anteriorly, subrectangular;
gular scales between infralabial rami medium to
small, keeled to smooth, subimbricate to imbricate;
central gulars very small, imbricate; chest and ven-
tral scales smooth; fourth toe lamellae 10(11). An-
terior cephalic figure bilobed with broadly rounded
lobes and interlobal cleft marked by massing of dark
pigment, not by an indentation; posterior cephalic
figure bipartite, in direct contact with posterior edge
of anterior cephalic figure and with elongate pos-
terolateral extensions. Anterior border broad, dark-
edged, straight but with slight anteriad indentations;
ends with faint anteroventral extensions towards au-
ricular opening. Scapular patch oval with anterior
pedicelate extension to anterior scapular figure bor-
der. Ocelli round, completely contained within patch.
Posterior border broad and dark-edged, in broad
contact with posterior edge of scapular patch. Left
end of posterior border can be seen to proceed an-
teroventrally to just beyond forearm insertion. Dor-
sum brown in life with scattered sparse spotting
(spots about one scale in size). Sacral figure of brief,
dark-edged dorsolateral light lines; tail pattern of
few, short, diagonal elements on base and diffuse
distal spotting and banding (Fig. 5D). Venter stip-
pled with dark pigment, particularly around scale
edges (dull grayish orange in life, underside of tail
brighter orange).

The paratype, a hatchling, has faint but complete
cephalic and scapular patterns that agree in consid-
erable detail with that of the holotype. Anterior and
posterior cephalic figures are identical; the anterior
scapular figure border shows the anteroventrad lat-
eral extension on one side; the scapular patch seems
to cover a greater area but does not completely in-
clude both ocelli. The posterior scapular figure bor-
der is in contact with the scapular patch, but the
lateral extension toward the forearm insertion is very

46

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

faint and evident on one side only. The dorsum is
uniform brown in coloration with a suffusion of dark
pigment around the scale edges.

Color in life: Dorsal ground color dark brown with
darker brown to black flecking; head markings buff
to light brown; scapular spots white to buffy (para-
type). Venter of holotype grayish, throat light gray;
iris brown.

Remarks.— The holotype and the egg containing
the paratype were taken beneath rocks along a road-
side through montane rain forest. Another egg was
taken that did not hatch; the eggs measured 7.8 by
5.3 mm and 8.1 by 5.5 mm.

Etymology.— The name mekistus is Greek and means “re-
mote,” in allusion to the geographic distance of this subspecies
from other populations of the species.

Sphaerodactylus armstrongi Noble and Hassler

Sphaerodactylus armstrongi Noble and Hassler, 1933, Amer. Mus.

Novit., 652:5.

Definition. — Dorsal scales small, acute, strongly
keeled, flattened, imbricate, axilla to groin 29 to 46,
no area of middorsal granules or granular scales;
dorsal body scales with four to eight hair-bearing
scale organs, each with one hair, around apex on
distal margin; dorsal scales of tail keeled, acute, im-
bricate, flat-lying to slightly raised; ventral scales of
tail smooth, rounded, enlarged midventrally; chest
and ventral scales smooth, rounded, imbricate, ax-
illa to groin 25 to 35; scales around midbody 49 to
75. Snout short and blunt, not depressed or de-
curved; rostral high and rounded with very small
dorsal depression or flat area; snout scales small,
narrow, swollen, strongly keeled, subimbricate to
imbricate; intemasals 0 to 3; postnasals 1, rarely 2;
upper labials to mid-eye 3; dorsal head scales small,
narrow, keeled, imbricate, raised to almost conical;
temporal scales small, broad, obtuse, keeled, raised,
subconical; gular scales between infralabial rami
small, acute, keeled, imbricate; central gulars very
small, keeled to smooth, imbricate; first infralabial
broad anteriorly, subtriangular; fourth toe lamellae
7 to 14; escutcheon with broad central area and
extensions onto thighs to popliteal region (3-9 by
1 0-30); ear diameter large (Fig. 6).

Sexual dichromatism incomplete; some males with
a unicolor pattern. Pattern variable within and be-
tween subspecies but having, basically, paired light
postocular stripes, occipitonuchal chevron or Y, and
scapular ocelli without scapular patch; scapular ocelli
sometimes transversely elongate and sometimes
fused with stem of occipitonuchal Y ; sacral pattern

lineate; tail pattern of intermittently constricted
middorsal zone derived from sacral middorsal zone;
venter pale to stippled with dark pigment; throat
pattern of dark lines.

Basic pattern.— The dorsal head pattern consists,
of a pair of light postocular stripes that may have
various irregularities or emarginations of their up-
per (median) margin and a mesiad branch which
may or may not meet with its mate of the opposite
side to form a transverse light bar across the parietal
region; the main branch of the postocular stripe may
proceed posteriorly to join with another pattern ele-
ment, a transverse crescent or shallow V or Y in the
occipitonuchal region. In the scapular region or just
immediately anterior to it occur two ocelli which
show various stages of transverse elongation and
fusion culminating in a scapular or posterior neck
chevron or shallow V, which in some cases is joined
to the stem of the more anterior Y figure producing
a connected double chevron. The lateral head pat-
tern includes one or two additional light postocular
stripes (the dark edges of these stripes may appear
as a series of dark postocular lines). The dorsum is
brown and uniformly covered with an irregular spot-
ted pattern of darker brown. In some specimens
additional series of paired ocelli follow posterior to
the scapular pair. The sacral pattern is lineate (a pair
of dorsolateral dark-edged, light lines); the dark-
edged, brown middorsal zone of the sacral pattern
continues onto the tail after which its margins be-
come irregularly scalloped with indentations which
at times are marked by ocelli. Venters show some
degree of dark stippling and vary from pale gray
(off-white) to dark gray; undersides of throats have
a pattern of dark lines (bold to very faint) or mar-
bling, which are continuations of labial suture marks
and postocular stripe borders.

Distribution.— Eastenx half of the south island
montane chain from the lowlands in the vicinity of
Barahona and Paraiso, Republica Dominicana, west
as far as the Massif de la Selle, Haiti (Fig. 7).

Sphaerodactylus armstrongi armstrongi Noble and
Hassler

Sphaerodactylus armstrongi Noble and Hassler, 1 933, Amer. Mus.

Novit., 652:5.

Type-locality. — Mountain top of property of G.
Herrmann, near Paradis (=Paraiso), 2,400 ft (732
m), Barahona Province, Republica Dominicana.

Holotype.- KMND 5 1 470.

Definition. — As. subspecies of Sphaerodactylus

1983

THOMAS AND ^RTZ-SPHAERODACTYLUS SYSTEMATICS II

47

1 1 1 T-

5.0 7.0

MM HL

Fig. 6. — Ear diameter (ED) versus head length (HL) for Sphaerodactylus armstrongi (circles) and Sphaerodactylus streptophorus (triangles);
both subspecies of both species are included; the heavy dashed line encloses values for 5. armstrongi armstrongi, solid hexagons indicate
specimens from Mome I’Hopital and Soliette.

armstrongi having a V-shaped, light occipitonuchal
figure frequently with a median posterior extension,
ends of V usually not clearly joined to light post-
ocular stripes; frontal region with a light transverse
bar or crescent connecting the postocular stripes;
scapular pattern of paired ocelli or ocelli fused to
form a transverse bar or shallow V-shaped mark;
one pair of ocelli on dorsum of trunk; sacral figure
of dorsolateral light lines that indent on base of tail
to (often) form a pair of ocelliform spots; midbody
scale counts low to moderate (49-64); size moder-
ate, to 27 mm SVL.

Distribution. — from the eastern extremity
of the south island from the vicinity of Barahona
south along the coast and foothills to the vicinity of
Enriquillo, from sea level to at least 3,600 ft (1,098
m) (Figs. 1 and 7).

Variation (ranges and means given first for Ba-
rahona area, then for Paraiso area samples). — Larg-
est male (ASFS V30950) 25 mm snout-vent length,
largest females (ASFS V3095 1 , V44 1 3) 27 mm; dor-
sal scales axilla to groin 29-41 (35.6), 32-40 (35.3);

ventral scales axilla to groin 25-34 (29.5), 25-34
(29.1); midbody scales 49-64 (56.9), 58-65 (61.1);
supralabials to mid-eye 3/3; internasals 0-3; fourth
toe lamellae 7-12; escutcheons 4-7 by 10-30.

The majority (68%) of the specimens have the
occipitonuchal V with a median stem extending to
the scapular region where it is flanked on each side
by a prominent ocellus. These scapular ocelli may
themselves fuse to form a posteriorly directed chev-
ron attached centrally to the stem of the V (1 1%);
intermediates between complete separation of scap-
ular ocelli and complete fusion occur. The occipi-
tonuchal chevron occurs without a stem in about
20% of the specimens; scapular ocelli or the fused
(chevronate) ocelli occur in 100% of the specimens,
about 10% having ocelli alone with none of the other
pattern elements. (The pattern may not be detect-
able in some specimens because of fading.) The oc-
eipitonuchal figures and other parts of the pattern
are generally more contrasting in the specimens from
the lowlands between Paraiso and Enriquillo. The
transverse head (frontal) bar is present in the ma-

48

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Fig. 7. — Distributions of four species of Sphaerodactylus in southeastern Haiti and southwestern Republica Dominicana, as follow:
triangles = S. armstrongi (solid = S. a. armstrongi, hollow = S. a. hypsinephes); circles = 5'. streptophorus (solid = S. s. sphenophanes,
hollow = 5. 5. streptophorus)-, solid hexagons = S. sp. 1 ; hollow diamonds = S. cryphius.

jority of specimens, although it is frequently faint,
and in some cases the mid-region of the bar is in-
distinct; at times its posterior dark edge is all that
is evident, the anterior edge being absent or frag-
mented. One male from Paraiso (ASFS V30443)
shows the unicolor male pattern with only a faint
trace of the basic pattern. Body ocelli are small and
indistinct (without prominent dark edges) when
present; 33% have just one pair on the dorsum
slightly anterior to midbody; one heavily pigmented
specimen (ASFS V20579) has about eight irregularly
arranged small dorsal ocelli. The first postsacral in-
dentation of the tail pattern often produces the effect
of a pair of light ocelli on the tail base.

The series of 42 specimens from 1 1 km N Enri-
quillo at 1,098 m (ASFS V42121-62) is of especial
interest. The size is larger than the other samples of
this subspecies. The fused scapular ocelli occur in
only one specimen; the occipitonuchal mark is sin-
uous (with a postauricular extension) rather than V-
shaped in eight specimens; and all of the light head

and neck markings are narrow rather than broad,
as in the more coastwards populations. Thirteen
specimens have a series of dorsal body ocelli. Mid-
body scale counts (55-69, mean 61.8, standard de-
viation 2.82) are much higher than the Barahona
area specimens but not significantly higher than the
Paraiso lowland specimens. These specimens show
the influence of the more western, high-elevation
subspecies described below.

Dorsal ground color in life is dark brown, very
dark brown (nearly black) to reddish tan. The nuchal
chevron is gray to buffy-orange or yellowish. Scap-
ular ocelli are orange, dull orange to cream or white;
and body ocelli, when they occur, have been re-
corded as yellowish. Light markings on the head are
pale gray. Venters are grayish, and throats of males
are yellow (ground color). Iris color is brown.

Remarks. — \i is likely that the type-locality is in
a somewhat intergradient area, similar to the above
mentioned specimens from 1 1 km N Enriquillo, but
since on balance of characters the holotype and

1983

THOMAS AND SCU^N \KTZ- SPHAE ROD ACTYLUS SYSTEMATICS II

49

paratype agree more with the eastern coastal pop-
ulations, we attribute the nominate trinomen to the
easternmost subspecies.

The Barahona area specimens differ in certain re-
spects from the lowland specimens of the Paraiso
to Enriquillo region, the latter group having some-
what bolder head, neck, and scapular patterns. The
two groups also differ in midbody scale counts, al-
though there is considerable overlap (Table 1); by
Student’s /-test the difference is statistically signif-
icant {P < .001).

Specimens examined. — KEPil%l^c^ Dominicana: Barahona
Province, 5 km S Cabral (MCZ 51820-21); 16.2 km S Cabral, ca
1,098 m (ASFS V4413); 4 km NW, 2 km SW Barahona, 153 m
(ASFS V213-14); 3 km N, thence 3 km W Barahona (ASFS
V20471-72); 2 km SE Barahona (ASFS X9521-25); 5 km S
Barahona (ASFS V20555-83); 4.8 km S Barahona (ASFS V 14060,
EDO 7-5351, EDO 7-5355, EDO 7-5358-60); 6.7 km S Barahona
(ASFS ¥4398^00); 5.3 km NE Ea Cienaga (ASFS V39796-98);
mountain top on property of G. Herrmann, near Paraiso (AMNH
5 1470— holotype); 1 km NE Paraiso, Rio Nizaito (ASFS VI 6946,
ASFS V30442-53); 3.0 km W Paraiso, 183 m (ASFS V30848-
54, ASFS V35758-69); 6.6 km W Paraiso, 153 m (ASFS V30950-
51); 11 km N Enriquillo, 788 m (ASFS V42121-62); 3 km N
Enriquillo, 212 m (ASFS V42 179-80).

Sphaerodactylus armstrongi hypsinephes,

new subspecies

//o/o/vpc. — USNM 194047, an adult male, from
32 km N Pedemales, ca 4,000 ft (1,220 m), Ped-
emales Province, Repiiblica Dominicana, taken by
Richard Thomas on 28 July 1969. Original number
ASFS V21482.

Paratypes (all from Pedemales Province, Republica Domini-
cana, except as indicated).— ASFS V2609, 5 km NE Eos Arroyos,
5,800 ft ( 1 ,769 m), David C. Eeber, 27 June 1 964; ASFS V2 1479-
81, ASFS V21483, same data as holotype; ASFS V21506-08, 25
km N Cabo Rojo, Richard Thomas, 29 June 1969; ASFS V29906,
4.4 mi (7.0 km) N Eos Arroyos, 5,600 ft (1,708 m), Bruce R.
Sheplan, 1 8 August 1971; ASFS V9749, 4 mi (6.4 km) SW Seguin,
4,400 ft (1,342 m). Dept, de I’Ouest, Haiti, Richard Thomas, 27
March 1966; ASFS V37663-37723, 6.1 km SW Seguin, 4,000 ft
(1,212 m). Dept, de I’Ouest, Haiti, J. W. Norton, A. Schwartz,
W. S. Sommer, 2 June 1974; RT 5046-5279, 3.3 mi (5.3 km)
SW Seguin, 3,700 ft (1,121 m). Dept, de POuest, Haiti, 30 June
1978, natives, Richard Thomas.

Definition. — Ps. subspecies of Sphaerodactylus
armstrongi having postocular stripes that join or
nearly join an occipitonuchal chevron with ventrad
postauricular marks; frontal region irregularly pig-
mented or with a light X- or V-shaped mark; a U-
shaped (or other non-ocelliform) scapular marking;
a series of dorsal trunk ocelli or ocelliform spots;
sharply constricted (or bridged), non-ocellar post-

Table \ . — Comparison ofSphaerodactylus armstrongi armstron-
gi and S. a. hypsinephes.

Characters

S. a. armstrongi
Barahona : Paraiso

S. a. hypsinephes
Seguin : Los Arroyos

Occipitonuchal figure

V-shaped

Sinuous (with
postauricular
extension)

Frontal pattern

Eight interocular

Mottled or light X

bar or crescent

or V

Median neck line

49%

10%

Scapular pattern

Paired ocelli 59%

U-shaped 61%

V-figure 11%

Irregular 38%

Dorsal body ocelli

1 pair

Dorsolateral series
or irregular
scattering

Caudal pattern

Ocelliform light

Complete or near-

indentations of

ly complete

dorsal zone

bridge; non
ocelliform

Snout-vent length

Maximum (3/V)

23/24 : 25/27

30/31

Mode

22 ; 24

27

Midbody scales

56.98 ± 3.14 :

68.89 ± 2.81 :

(Mean ± 1 SD)

61.14 ± 3.95

63.36 ± 3.47

N = 47

N = 87

N = 36

N = 1 1

Range

49-64 58-65

61-75 58-69

sacral middorsal zone; large size (males to 30, fe-
males to 3 1 mm snout-vent length); and high num-
ber of scales around midbody (61-75).

Distribution. — Yiigher elevations {ca. 2,000 to
6,000 ft— 606 to 1,830 m) of the western Sierra de
Baoruco and the Massif de la Selle. The exact east-
ward extend of the subspecies is unknown but may
extend to the Valle de Polo (see comparisons below).
Its western range is apparently limited by the Vallee
de Trouin in Haiti (Figs. 1 and 7).

Description of holotype. — Kn adult male, 27 mm
snout-vent length, tail 25 mm; dorsal scales axilla
to groin 38, ventral scales axilla to groin 30, mid-
body scales 62. Snout blunt, not depressed; snout
scales small, acute, keeled, swollen, slightly imbri-
cate; supranasals broader than long, roughly trian-
gular; 1 intemasal; 2 postnasals; first infralabial broad
anteriorly, subtriangular; dorsal head scales small,
narrow, keeled, raised posteriorly, almost conical;
temporal scales small, broad, keeled, raised; gular
scales between infralabial rami small, acute, keeled,
imbricate; central gulars very small, acute, keeled,
imbricate; chest, ventral scales smooth; supralabials
to mid-eye 3/3; fourth toe lamellae 9.

50

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Dorsal ground color brown with irregular darker
brown mottlings or vermiculations; light upper
postocular stripe proceeds posteriorly and stops just
before joining the lateral extremities of a broad, light
occipitonuchal Y, the lateral ends of which turn ven-
trally and posteriorly for a short distance (postau-
ricular mark). The stem of the Y extends posteriorly
in the midline of the neck to a point between two
faint, transversely oriented light marks that form an
indistinct posterior nuchal U; the central region of
the head has a faint patterning formed by lightening
of the ground color in the parietal and frontal areas.
The sacral figure consists of dark-edged, light dor-
solateral lines that extend for a short way onto the
base of the tail, where they form a series of inden-
tations. The venter is pale yellowish in preservative,
the throat pale with dark streaking and marbling as
described for the species.

Variation. — Largest male 30, largest female 31
mm snout-vent length; dorsal scales axilla to groin
33-46; ventral scales axilla to groin 26-35; midbody
scales 58-65; supralabials to mid-eye 3/3 (65), 4/4
(1); intemasals 1 (12), 2 (4), 3 (3); fourth toe lamellae
8-14; throat scale keeling present in all; escutcheons
have large central areas with extensions onto thighs
to vicinity of knees (4-9 by 21-27).

A head pattern of postocular stripes that join or
abut an occipitonuchal V having short, ventrad
postauricular extensions is a consistent feature of
this subspecies. A stem to the V occurs in a minority
of specimens (about 1 0%). A more or less symmet-
rical U-shaped neck or scapular figure can be seen
as the “ideal” condition but in reality occurs only
modally (about 60%), being highly asymmetrical,
reduced, fragmented, or divided in the rest. Depar-
tures from the symmetrical scapular U include: ab-
sence (sometimes with a dorsolateral series of body
ocelli continuing onto the neck); a diagonal line; a
straight transverse “collar;” a pair of reversed pa-
rentheses or comma-like marks; an apically inter-
rupted V; and various other fragmentary conditions
(Fig. 8). All of these pattern elements are finely lin-
eate, except that the postocular stripes may be
broadened (but are always smooth-edged). The fron-
tal pattern consists of irregular mottling or a light
X- figure or (if the posterior half of the X is obscure)
a pale V; a straight light connection between the
postocular stripes is never found, and rarely a com-
plete connection of any sort. The presence of a series
of ocelli (or less regular ocelliform spots) on the body
is a consistent feature, although the number and
arrangement is not. The sacral light lines frequently

extend well forward of the groin region; the first
postsacral indentations may meet to form a com-
plete bridge (43%), nearly meet (41%), or be sepa-
rated by a distinct gap (6%); they are not ocelliform.
Throat marbling is distinctly heavier in males, al-
though there is overlap in intensity between the sexes.
Males also tend to have a contrasting, vermiculate
dorsal head pattern that does not obliterate the post-
ocular and occipitonuchal pattern.

Comparisons. — Table 1 compares the features that
most differentiate the two subspecies. Qualitative
differences are also seen in the head, neck, and scap-
ular markings, those of hypsinephes being formed
of sharply delineated, narrow light markings, where-
as those of armstrongi are broader and less well-
defined. The two subspecies are markedly distinct,
but the sample from 1 1 km N Enriquillo, noted
under the account of S. a. armstrongi, indicates that
intergradation does occur, very likely along an el-
evational zone in the eastern part of the Sierra de
Baoruco. However, it is possible that “pure” hyp-
sinephes is not encountered until farther west in the
Sierra de Baoruco, perhaps west of the Valle de Polo,
the main physiographic feature separating the east-
ern from the western Baoruco (and west of which
elevations are generally greater).

Remarks.— h certain amount of intrasubspecific
variation is present, particularly in scale counts. The
western Sierra de Baoruco sample of hypsinephes
(Los Arroyos vicinity) differs significantly in mid-
body counts from the Haitian Seguin area sample
{P < .00 1 , Student’s /-test).

Etymology. —Hypsinephes is from the Greek meaning “dwell-
ing high in the clouds.”

When Thomas and Schwartz (1977) described
Sphaerodactylus streptophorus, they included under
that species, but not as paratypes, a few specimens
from low to moderate elevations of southern Haiti,
from the Jacmel region through the Trouin Valley.
These specimens differed in some scale characters
and in coloration from the samples from the Re-
publica Dominicana. We have since obtained sig-
nificantly more specimens from that region of Haiti
and find that they represent an undescribed sub-
species of S. streptophorus, for which we propose
the name

Sphaerodactylus streptophorus sphenophanes,
new subspecies

Holotype.— CM 83100, a gravid female from
Ravine Normande, 1 1.4 km (6.8 mi) E, thence 1.3

1983

THOMAS AND \KTZ- SPHAERODACTYLUS SYSTEMATICS II

51

Fig. 8. — Head and anterior body patterns of Sphaerodactylus armstrongi hypsinephes.

km (0.8 mi) N Jacmel, 77 m (254 ft). Dept, de I’Ouest,
Haiti, one of a series collected on 26 June 1978 by
natives and Richard Thomas.

Paratypes (all from the Dept, de I’Ouest, Haiti). — ASFS V43590,
4.7 mi (7.5 km) E Petit-Goave, 400 ft (121 m), 15 July 1976,
native collector; ASFS V45347-53, same locality as preceding,
9 July 1977, native collectors; ASFS V43593-604, 2.5 mi (4 km)
E Petit-Goave, 15 July 1976, native collectors; RT 7559-66, 0.8
mi (1.3 km) NE Bainet, ca 200 ft (60 m), 21 July, 1979, Luis
Rivera Cruz, R. Thomas; ASFS V9788-90, ca 2.4 km S Trouin,
244 m (805 ft), 28 March 1966, R. Thomas; RT 5479, 7.9 mi
(12.6 km) SE La Vallee de Jacmel (=3.6 mi (5.8 km) NW ford
of Riviere Gauche on La Vallee road), 1,000 ft (303 m), 7 July
1978, R. Thomas; RT 5473, 9.5 mi (15.2 km) SE La Vallee de
Jacmel, 750 ft (227 m), 6 July 1978, R. Thomas; RT 5535-39,
same locality as preceding, 9 July 1978, R. Thomas; RT 5293-
94, Nan Sumac, 1.8 mi (2.8 km) S Decouze, 1,400 ft (424 m), 2
July 1978, R. Thomas; RT 5445-53, same locality as preceding,
5 July 1978, natives, R. Thomas; RT 5532, 5 km (3 mi) airline
W Jacmel, ravine of Petite Riviere de Jacmel, 1,000 ft (303 m).

8 July 1978, R. Thomas; RT 4974-76, same data as holotype;
ASFS V37433-35, same locality as holotype, 30 May 1974, J.
W. Norton, W. W. Sommer, A. Schwartz.

Associated specimens (n\\ from Dept, de I’Ouest, Haiti).— ASFS
V25295, RT 4905-08, ca 5 km (airline) W Petionville, north
versant, Mome I’Hopital, ca 2,000 ft (606 m); ASFS V37297,
Soliette, 3.8 mi (4.0 km) NW Fond Verrettes, 2,000 ft (606 m);
MCZ field numbers 0691, 0692, Savane Mouton, east of Foret
des Pins, ca 4,884 ft (1,480 m).

Definition. — A. subspecies of Sphaerodactylus
streptophorus characterized by the presence of paired
ocelli or paired dashes on the neck rather than a
continuous collar; a reduced incidence of paired
scapular ocelli; lack of a contrasting, vermiculate
head pattern in males; high number of midbody
scales (52-62); smaller, more heavily keeled and
raised (less cobblelike) snout scales; large size (males
to 24, females to 26 mm snout-vent length).

52

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Distribution. — Southern Haiti from the vicinity of
Petit-Goave south to Bainet and east to Ravine Nor-
mande, 1 1 km E Jacmel; elevational distribution
from near sea level to 1,400 ft (424 m) at Nan Su-
mac. This subspecies possibly extends into the up-
lands near the Dominican border at Savane Mouton
(4,884 ft; 1,480 m) (Fig. 7).

Description of holotype. — An adult female, 23 mm
snout-vent length, tail 18 mm, partly regenerated;
dorsal scales axilla to groin 36; ventral scales axilla
to groin 32; scales around midbody 56. Snout blunt,
not depressed; rostral with distinct flattened dorsal
portion; snout scales small, acute, keeled, swollen,
slightly imbricate ( 1 5 across snout between posterior
ends of 2nd upper labials); supranasals broader than
long, roughly triangular; 2 intemasals; 2/2 postna-
sals; 3/3 upper labials; first infralabial broader an-
teriorly, subtrapezoidal; dorsal head scales and tem-
poral scales small, broad, keeled, raised on posterior
edges, subconical; gular scales between infralabial
rami small, acute, keeled, imbricate; central gulars
very small, acute, keeled, imbricate; chest and ven-
tral scales smooth; fourth toe lamellae 10.

Dorsal ground color brown with darker brown
mottling of irregularly interconnected scales that
form the edges of irregular dorsolateral stripes. Dark-
edged light dorsolateral stripes begin on posterior
trunk and continue along sacrum onto tail, where a
series of indentations constrict the dark interspace.
Postocular stripe broad, light-centered, upper dark
edge continuing posteriorly to weakly demarcate a
central cephalic figure with indented sides. Mesially
oriented temporal wedge formed by diagonal pos-
terolateral margin of cephalic figure and a light pos-
teroventral postauricular stripe. Indistinct narrow
light transverse occipitonuchal line present. Ground
color of entire head light, darkening immediately
posterior to occipitonuchal line. A pair of small oval
ocelli at mid-neck; no scapular ocelli; several small
ocelli between ear opening and forearm insertion.
Throat with indistinct light longitudinal stripes on
a brown ground color that are continuous with light
subocular stripes across labial region. Venter with
frosting of dark pigmentation concentrated around
scale edges.

Variation.— The temporal wedge, visible either as
a complete wedge or as only the light posterior (post-
auricular) limb is present in 84% of the specimens.
The bilobed cephalic figure is found in 26%, and an
occipitonuchal mark, somewhat in the shape of a
stretched out “M” is seen in 7%. (Contrary to the
statement of Thomas and Schwartz, 1977:37, the

transverse neck line or collar is not a homologue of
the armstrongi occipitonuchal mark; rather, the oc-
cipitonuchal mark is obscured in most streptopho-
rus.) The mid-nuchal marks are paired ocelli, sim-
ilar to the condition in the holotype, in about 42%
of the specimens; more elongate marks (paired dash-
es) occur in 26%; and an unpaired transverse collar
occurs in 1 8%. Paired scapular ocelli occur in about
33%; 13% have irregular scapular markings (one
ocellus or more than two), whereas 53% have no
scapular markings. Body ocelli are generally not
present (98%), but a few have irregular ocellations,
not as well defined as, for example, the condition
in S. armstrongi hypsinephes. A postsacral constric-
tion of the middorsal zone is present in about 3%,
although it is not extreme and does not result in
paired ocelli on the base of the tail nor in a complete
bridge of light pigment. Ventral lines are present
(but not well-developed) in about 1 6% of the spec-
imens. The throats of males are more boldly and
contrastingly striped or marbled than those of fe-
males.

The largest male is 24 mm snout-vent length, the
largest female 26 mm; dorsal scales axilla to groin
are 28-36 (mean 30.28); ventral scales axilla to groin
25-33 (28.5); scales around midbody 52-62 (56.84);
intemasals 2 (44) or 3 (6); postnasals 2/2 (43), 2/1
(1), or 1/1 (1); scales across snout between posterior
ends of second upper labial 14-17 (14.53); upper
labials to mid-eye 3/3 in all; fourth toe lamellae 8
(1), 9 (16), 10 (29), or 1 1 (4); escutcheons 3-7 scales
long by 1 0-26 scales wide with extensions onto thighs
as far as popliteal region in some specimens.

Comparisons. — The principal comparison is with
the nominate subspecies. S. s. sphenophanes is larg-
er (modal snout-vent length 23 mm versus 20 for
nominate streptophorus', maximum size 26 versus
23 mm). Midbody scale counts are significantly
higher (52-62, mean 56.84 ±2.19 standard devia-
tion versus 41-54, mean 47.45 ± 10.66; means sig-
nificantly different by Mest at R < .001), as are the
number of scales across the snout (14-17, mean
14.53 ± 2.65 standard deviation versus 1 1-14, mean
1 1.35 ± 1.10; means significantly different by /-test
at P < .001).

In coloration sphenophanes differs in possessing
ocelli or paired dashes on the neck (79%) instead of
the continuous collar (21% versus 93% in strepto-
phorus). Over half (53%) lack any scapular mark,
although 33% have a pair of ocelli, and the remain-
ing 14% irregular or unilateral ocelli; 93% of nom-
inate streptophorus have paired scapular ocelli, the

1983

THOMAS AND \KTZ- SPHAERODACTYLUS SYSTEMATICS II

53

Fig. 9. — Length of rostral flat area (ordinate) versus total length of rostral (abscissa) for Sphaerodactylus armstrongi (circles) and 5.
streptophorus (squares); triangles indicate values for Mome I’Hopital and Soliette specimens.

rest have none. The head pattern of nominate strep-
tophorus is boldly vermiculate in males, but this is
not true of sphenophanes, where vermiculation is
mueh fainter and does not obliterate the head pat-
tern.

We were not initially sure whether sphenophanes
was closest to streptophorus or to armstrongi. The
presence of armstrongi-Mko. occipitonuchal marks in
some specimens suggested a possible closest rela-
tionship with that species. But we (Thomas and
Schwartz, 1977) have remarked on the strong affin-
ities of all these geckos {armstrongi, streptophorus,
cryphius) and have noted that they all share some
homologous pattern features. We still cannot com-
pletely eliminate the possibility that these forms (plus
another undescribed species from Haiti) constitute
a single Rassenkreis but lacking evidence of inter-
gradation we feel that there are distinct gaps between
our species-level groupings.

The wedge marks on the head of sphenophanes
and streptophorus are vestiges of the more complete
occipitonuchal marks and abutting postocular stripes
of S. armstrongi hypsinephes, the wedge being
formed by the meeting of the postocular and post-
auricular stripes. As noted above, a few spheno-
phanes possess an M-shaped occipitonuchal mark.

The possession of markings at mid-neck (ocelli,
dashes, or a collar) further ally sphenophanes and
nominate streptophorus, as does the possession of
scapular ocelli by both forms (only some spheno-
phanes) along with lack of body ocelli. The lack of
pronounced basal caudal constriction of the dark
median zone of pigment is also a feature contrasting
5. streptophorus and S. armstrongi. The diameter
of the opening of the external auditory meatus is
proportionately smaller in both streptophorus and
sphenophanes than it is in S. armstrongi (Fig. 6).
The rostral of streptophorus has a large, semicir-
cular, dorsal flat area bordered by a distinct rim; the
flat area slopes towards the tip, giving the scale a
sloping profile. The rostral of S. armstrongi does
not slope towards the tip so pronouncedly (or at all)
and has a smaller flat or depressed dorsal area, the
sides of which are usually rounded rather than ridged.
A comparison of the distance from the rostral-supra-
nasal sutures to the anterior edge of the flat area
with the total length of the rostral (Fig. 9) shows a
quantitative aspect of this difference. These char-
acteristics of rostral shape and ear size should be
added to the definition of S. streptophorus.

Remarks. — Yht habitat of 5. streptophorus
sphenophanes contrasts with that of the xerophilous

54

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

nominate subspecies. All specimens of spheno-
phanes came from mesic shaded situations, where
they were taken in leaf-litter or under rocks. The
type-locality is a shaded ravine, and all of the spec-
imens were taken in Musa and Artocarpus altilis
litter along the ravine sides (those collected by or
with Thomas). At other localities sphenophanes was
taken in coffee plantings (Nan Sumac); in an Arto-
carpus grove (9.5 mi SE La Vallee de Jacmel); in
the mesic shady ravine of the Petite Riviere de Jac-
mel) among leaf-litter and limestone jumble; in ba-
nana and coffee plantings (7.9 mi SE La Vallee de
Jacmel). At 0.8 km NE Bainet the region is some-
what more xerophytic than the other localities, al-
though basically wooded rather than scrubby; the
specimens were taken in leaf-litter in a grove of
Mangifera indica. Along the south coast of Haiti,
between the type-locality of sphenophanes and the
range of the nominate subspecies, we took no strep-
tophorus of any kind, even though there were hab-
itats appropriate for either subspecies. The habitat
at the localities reckoned from Petit-Goave is ba-
sically mesic and in an area of mixed agriculture
(including coffee, bananas, and com) and mixed open
to shaded habitats; the specimens were native col-

Fig. 10 shows our tentative inference of phylo-
genetic relationships in the Sphaerodactylus difficUis
complex. Our ideas on a possible sequence of evo-
lution of dorsal patterns in the difficilis complex are
presented in Fig. 1 1 , in which we show only how
certain patterns may have been derived from others.
We have relied largely on pattern in constructing
our dendrogram (Fig. 10), particularly for apomor-
phous features. Scale and other external structural
features, while useful in defining taxa, have been of
little use in assessing evolutionary sequences. It is
obvious that, for example, scale size has evolved
repeatedly in different directions (that is, reversals
have occurred); likewise such characters as post-
nasals or intemasal number could easily have under-
gone reversals. In the complex and variable color
patterns it is easy to see trends, although not always
easy to identify derived features. With some pattern
elements homology is patent, in others it is far from
certain. For evolutionary analysis we have used pat-
tern features even though they may not characterize
a majority of a population; they serve as an indicator
that the population made a certain evolutionary step.

lected, and we did not see the precise situations from
which they came.

The specimens from the Mome I’Hopital (ASFS
V25925, RT 4905-08) and Soliette (ASFS V37297)
have faint markings, although they show the post-
auricular stripes, indications of an occipitonuchal
mark and indications of mid-nuchal marks, al-
though the pattern of the Soliette specimen is frag-
mented. Two specimens (RT 4905-06) show faint
or partial indications of an armstrongi-\i\.c prescap-
ular U, and the others have scapular or prescapular
ocelli. So they show some resemblances to arm-
strongi hypsinephes and are not completely typical
of S. streptophorus sphenophanes. Midbody scale
counts (51-61, mean 55.2) coincide with the range
in sphenophanes-, the rostrals are streptophorus-\i\ae,
although their measurements (Fig. 6) are somewhat
intermediate; ear size (Fig. 6) agrees more with
streptophorus. Very likely these north-slope speci-
mens represent another somewhat differentiated
population, but more specimens will be needed to
determine the variation. The fragmented pattern of
the Soliette specimen is reminiscent of that of the
Savane Mouton specimens (Thomas and Schwartz,
1977).

Our reason for postulating that something like the
difficilis lineate pattern with scapular patch and ocel-
li is primitive is that patterns of this sort are wide-
spread in the complex (and beyond), occurring in
members of the Puerto Rico bank and Cuban ra-
diations. Our belief that the Cuban (notatus group)
and Puerto Rican (macrolepis group) sphaerodactyls
are derived from the Hispaniolan difficilis radiation
is based upon the observation that there is not a
great amount of diversification in either area (Puerto
Rico bank and greater Cuban-Bahamian area), which
implies recency of entry into both areas. Of course
an area-limitation on diversity must affect Puerto
Rico. Rough equivalence of evolutionary rates is an
unsubstantiated premise. We will not be apodictic
that our dendrogram necessarily be rooted in Puerto
Rico but offer it as a reasonable suggestion subject
to further evidence. It is principally the left side of
the diagram in which we have the greatest confi-
dence. The right side of the diagram is based mostly
on minor changes in plesiomorphous characters of
pattern; for this reason we have less confidence in
it.

1983

THOMAS AND ^CH^N \KTZ- SPH AERO DACTYLUS SYSTEMATICS II

55

Fig. 10. — Dendrogram of evolutionary relationships of difficilis complex Sphaerodactylus—sa., savagev, da, darlingtonr, al, altavelensis\
ny, nycteropus', st, streptophorus\ cr, cryphius\ sp, undescribed species (sp. 1 ); ar, armstrongi, zy, zygaena; co, cochranae, ma, macrolepis
group; no, notatus group; la, lazellv, ra, randr, cl, clenchi\ oc, ocoae\ di, difficilis.

The taxa armstrongi, streptophorus, nycteropus,
cryphius, and omoglaux seem to be the result of
relatively recent allopatric differentiation and so re-
tain allopatric or parapatric distributions with only
limited or as yet questionable sympatry. This group-
ing of taxa, including the subspecies of armstrongi
and streptophorus is polythetic with respect to a pool
of characters. We have not been successful in ana-
lyzing their relationships beyond the polychotomy
illustrated; forcing the characters into a dichoto-
mous representation may be unrealistic.

The following is a list of the apomorphous char-
acters used in constructing the dendrogram along
with the species possessing each.

A. Anterior cephalic figure

Al. Simple, bilobed: savagei, altavelensis, dar-
lingtoni, cryphius, streptophorus, nyctero-
pus, sp. 1

A2. Bilobed, laterally indented: streptophorus,
nycteropus

B. Occipitonuchal mark

Bl. V-shaped: armstrongi
B2. M -shaped (with postauricular stripe): arm-
strongi, streptophorus, cryphius, sp. 1

C. Posterior cephalic figure (probable homologue
of occipitonuchal mark): savagei, altavelensis,
darlingtoni, sp. 1.

D. Mid-nuchal marks

D 1 . Ocelli or paired dashes: streptophorus, cry-
phius, sp. 1

D2. Collar: streptophorus

E. Border of scapular figure

El. Large (not in contact with patch): altavel-
ensis, darlingtoni

E2. Bipartite: altavelensis, darlingtoni, arm-
strongi (?)

F. Small ear size: streptophorus, cryphius, nycter-
opus, sp. 1

Ecological Notes on the Difficilis Complex

Sphaerodactylus altavelensis enriquilloensis is an
inhabitant primarily of xeric habitats, ranging from
wooded situations to thorn scrub and palm savan-
nah, where it frequents terrestrial microhabitats —
leaf-litter, trash, and rocks. The more mesic situa-
tions in which it has been collected occur in the
eastern Valle de Neiba, where it was taken in banana
groves, but in this case the mesicity was artificial
(irrigation). The immediate situations in which we
have often taken this form are relatively mesic oases
within otherwise inhospitable xeric scrub. In the
Barahona region enriquilloensis occurs in somewhat
intermediate situations that tend towards the xeric.

56

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

Fig. 1 1. — Hypothetical sequence of dorsal pattern evolution in the Sphaerodactylus difficilis complex.

1983

THOMAS AND SCU^N \KTZ- SPH AERO DACTYLUS SYSTEMATICS II

57

The specimen tentatively referred to this subspecies
from 7 km NW Vallejuelo was taken in dense semi-
xeric woods. It is sympatric with S. difficilis in the
eastern portion of its range and has been taken syn-
topically with difficilis at localities east of Azua and
in the vicinity of Barahona. Also in the Barahona
region it has been taken with S. armstrongi at one
locality (3 mi S Barahona, a garbage dump). Al-
though both species have been taken extensively in
the Barahona region, they are seldom found at the
same locality. In some instances long series of either
species were taken without the other being found.
To the collector, their microhabitats are similar, both
being inhabitants of ground litter and trash or rocky
situations. In the western Valle de Neiba S. a. en-
riquilloensis was collected at the same locality with
S. cryphius but only at localities 6. 7-7.0 km NW
Duverge. At the other localities for cryphius in the
foothills of the Sierra de Baoruco, no specimens of
altavelensis were taken, although the habitat appears
ideal. In this region S. rhabdotus is known from a
single locality and might eventually be found to be
syntopic with altavelensis.

S. a. brevirostratus inhabits much the same kinds
of situations as does enriquilloensis over most of its
known range but appears to occur in some naturally
more mesic situations. At 2.2 km SW Lascahobas
two specimens were taken in banana and coffee
growing areas, and the region is generally mesic.
There is also a record from Furcy, a high elevation
locality. It is at least broadly sympatric and no doubt
syntopic with S. elegans (see Graham and Schwartz,
1978) over much of its range, although elegans prob-
ably has different microhabitat preferences. It is also
sympatric with S. cinereus in the Cul de Sac Plain,
with S. copei on the south side of the Cul de Sac
(the two were collected together in rocks and trash
in a cave at Diquini) and around Jacmel and at
Jeremie, and possibly with S. armstrongi hypsi-
nephes on the northern slopes of the La Selle. It also
occurs macrosympatrically with S. streptophorus in
the area between Trouin and Cayes Jacmel, but the
two have not been taken at the same locality. If S.
altavelensis brevirostratus is wide-ranging on the Ti-
buron Peninsula, it may be sympatric with several
other species of sphaerodactyls, including S. zy-
gaena, S. elasmorhynchus, and S. nycteropus.

The major part of the distribution of 5. darling-
toni is in the wettest part of the island, the north-
eastern comer. There it is most often encountered
in the gloomy pseudoforest of moist cacaotales and

cafetales, where it inhabits leaf litter and piles of
cacao husks, rotten logs, rocks, and piles of palm
trash. The only recently collected specimens of S.
d. darlingtoni were taken under rocks, palm trash,
and logs in a mixed coffee-and-cacao planting in
well-developed deciduous montane forest (1 ,400 ft).
The holotype and paratype were taken at 4,000 ft
on Pico Diego de Ocampo. North of Cruce de Guay-
acanes, it seemed to be primarily a rock-rubble in-
habitant. Specimens of S. d. noblei were found ac-
tive in leaf litter at 1600 hrs in a cafetal-cacaotal
6.4 mi SE Tenares at 700 ft. Two places where S.
d. noblei was encountered in natural situations were
in the forested region at Guaigiii, where the geckos
were found under rocks and in rock rubble on a
densely shaded forested hillslope of a ravine and
under a rock on the floor of high-canopied forest,
and within the forested edge of the haitises at 10
km NE Gonzalo, where specimens were taken under
limestone rocks.

The holotype of S. d. mekistus was taken under
a rock along a roadcut through montane rainforest
at 2,800 ft. The holotype and paratypes of S. d.
bobilini were taken in deciduous montane forest at
elevations of 2,000 to 2,800 ft, under rocks and in
leaf litter. The range of S', d. bobilini is essentially
insular, as the surrounding lowlands are xeric scrub,
certainly inhospitable to this form.

S. armstrongi is primarily an inhabitant of mesic
situations, because much of its range is in the up-
lands, although much of the range of the nominate
subspecies at low elevations tends towards the xeric.
The high elevations records include, in addition to
the type-locality, which we have not knowingly seen,
a coffee grove at 3,600 ft (10.1 km S Cabral). Our
supposition that S. a. hypsinephes would be a forest
dweller has not entirely been borne out. Certainly
it does inhabit forested situations, whence came most
of the Dominican specimens (a forested ravine at
about 1,400 ft; palm and deciduous forest at about
4,000 ft in a pile of thatching material), broadleaf
rainforest at 5,800 ft, and in pinewoods at 5,600 ft.
Our first Haitian specimen came from a coffee grove
at 4,400 ft, but much of the subsequent long series
from the same area (Seguin) came from open fields,
where Haitians took them from little cairns of lime-
stone that had been stacked to provide access to the
soil for planting corn and other crops. It is still pos-
sible that it is a sciophile that can exist in open
situations only when something other than vegeta-
tion (rocks) provides sufficient shade. One specimen
from Boutillier road came from beneath rocks (but

58

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

many rocks) in an open situation not far different
from the Seguin situation; the others came from
small patches of shade (Clusia rosea, Mammea
americana). Along Boutillier Road (=5 km W Pe-
tionville), it occurs syntopically with S. copei and
S. altavelensis.

Sphaerodactylus cryphius is a xerophile, known
from the southern edge of the Valle de Neiba and
the northern foothills of the Sierra de Baoruco. At
localities NW of Duverge it was taken in a shaded
oasis surrounded by very xeric habitat and syntop-
ically with S. altavelensis. At the type-locality (13
km NW Puerto Escondido) it was taken in piles of
thatch-palm fronds in xeric scrub with the type-
series of S. rhabdotus; it was also taken not far away
in xeric cactus-acacia scrub. It would not be unrea-
sonable to expect S. copei in this region, but it has
not been taken there yet.

Sphaerodactylus omoglaux (Thomas, 1982) is
presumably a xerophile, because it is known only
from the eastern Cul de Sac Plain of Haiti, although
most specimens have been taken in a shaded oasis
near Fond Parisien, Haiti.

Sphaerodactyl us streptophorus streptophorus is
known primarily from xeric situations of the Ba-
rahona lowlands, although it occurs at mesic inter-
mediate elevations of the Baoruco and La Selle
mountains. It is syntopic with S. difficilis and S.

(Schwartz, 1977), probably with S. thompsoni,
and possibly with S. copei, although the latter species
has not yet been taken within the range of S. s.
streptophorus.

S. streptophorus sphenophanes has been found
mostly in mesic situations (coffee groves, banana-
breadfruit forests, etc.) but also in somewhat xeric
areas, although not extreme ones.

Sphaerodactylus nycteropus seems confined to the
xeric woods of the Mome Dubois Peninsula; it was
not found in the extremely cut-over xeric coastal
region southeast of Aquin, where S. copei was abun-
dant. It occurs syntopically with V. copei at the type-
locality.

Sphaerodactylus ocoae was collected in mesic but
non-natural situations {Cocos groves) at 153 to 214
m (504-706 ft) in the Sierra de Ocoa.

Sphaerodactylus zygaena has been taken near the
ocean (Dame Marie) and in a mesic Musa grove
(Les Anglais).

The sphaerodactyls of this complex seem to fit
into four broad climatic habitat types: (1) Steno-
xerophiles, species that are pretty much confined to

extreme (or distinctly) xeric situations (even though
they may at times be found within artificial or nat-
ural “oases” within such situations). (2) Eury-xe-
rophiles, species that inhabit either (a) semi-xeric
(but more towards the xeric) situations or (b) both
xeric and intermediate or mesic situations (but with
the greater part of their range in the xeric). (3) Eury-
mesophiles, species that inhabit (a) mesic but open
situations (for example, savannahlike), (b) both me-
sic and open situations (but more of the range is
mesic), or (c) mesic and xeric (but more of the range
is mesic). (4) Steno-mesophiles, those species that
are confined largely to moist areas. Ranges of steno-
xerophiles will generally not extend upgradient from
the 750 mm (30 inch) isohyet, and steno-mesophiles
generally fall within (upgradient of) the 1,750 mm
(69 inch) isohyet (See Santiago de la Fuente, 1976:
165 for a rainfall map of the Dominican Republic).
Between these limits fall most of the habitats of
Hispaniola, the eury-xerophiles inhabiting the parts
more towards the xeric, and the eury-mesophiles
inhabiting those parts more towards the mesic (or
combinations of more extreme habitats as noted
above). Very generally, eury-xerophytic conditions
prevail in western Hispaniola and eury-mesophytic
conditions in eastern Hispaniola. It should be added
that some taxa are difficult to place with exactitude
in the two middle categories (eury-xerophiles and
eury-mesophiles), since the ranges may be incom-
pletely known or the amount of the range in different
habitats may be nearly equal.

Geographic Relationships of the Species

The species armstrongi, cryphius, streptophorus,
nycteropus, omoglaux, and zygaena are south island
species in the sense of Schwartz (1980), whereas
altavelensis and difficilis were categorized (Schwartz,
1980) as south island invaders from the north island;
darlingtoni, clenchi, and savagei are wholly north
island species, as presumably is lazelli. Although
altavelensis appears now to be restricted on the south
island and thus to have the appearance of a recent
invader, its distribution there also appears relictual
(its occurrence on Alto Velo but not the Barahona
Entrapment and its apparently isolated occurrence
near Jeremie at the end of the Tiburon Peninsula).
The species altavelensis and darlingtoni are a pair
that would conform to the north island/south island
dichotomy, if it be assumed that altavelensis has
secondarily invaded the north island along xeric cor-
ridors into northwestern Haiti and into the Valle de

1983

THOMAS AND \RTZ- S PH AERO DACTYLUS SYSTEMATICS II

59

San Juan of the Republica Dominicana. Although
altavelensis has been found in mesic situations, it is
principally a species of xeric or semi-xeric habitats.

The following summarizes the geographic and
habitat categorizations for the difficiiis complex
sphaerodactyls (NI = north island; SI = south is-
land). Omitted from consideration are S. lazelli and
an undescribed species from ^a Soleil in Haiti
(Thomas and Schwartz, MS).

Steno-xerophiles:

S. streptophorus streptophorus (SI)

S. cryphius (SI)

S. nycteropus (SI)

S. omoglaux (SI)

S. altavelensis altavelensis (SI)

S. altavelensis enriquilloensis (NI?)
Eury-xerophiles:

5. altavelensis brevirostratus (NI?)

S. altavelensis lucioi (NI)

N. ocoae (NI)

Eury-mesophiles:

5. difficiiis (NI)

S. savagei (NI)

S. zygaena (SI)

S. streptophorus sphenophanes (SI)
Steno-mesophiles:

N. clenchi (NI)

S. darlingtoni (NI)

S. armstrongi (SI)

Relationships with other Sphaerodactylus

Shreve (1968) used the term ^"notatus group” to
include the then-known Hispaniolan members of
the difficiiis complex (as well as S. samanensis, which
we exclude from the group). By such an action (and
by uses of trinomina — under S. notatus— for several
of the forms included therein) he correctly indicated
that there exists a large group of geckos in the An-
tilles and on the extreme southeastern portion of the
North American continent that are allied to each
other. These are, with their respective ranges and
including species described since Shreves paper: S.
notatus (Cuba, Bahamas, southern Florida, Swan
Islands); S. bromeliarum (Cuba); S. inaguae (Great
and Little Inagua islands); S. underwoodi (Turks Is-
lands); S. macrolepis (Puerto Rico, Virgin Islands,
northern Lesser Antilles south to St. Barthelemy);
S. roosevelti (Puerto Rico); S. nicholsi (Puerto Rico);
S. parthenopion (Virgin Islands); S. beattyi (St.

Croix); S. monensis (Isla Mona); S. micropit hecus
(Isla Monito); and S. levinsi (Isla Desecheo). In-
spection of this list shows that Puerto Rico (includ-
ing the Virgin Islands, Mona, and Desecheo) and
Hispaniola account for 25 of the 29 species that we
associate in this group. It seems reasonable to as-
sume that Hispaniola has been the center of evo-
lution of the notatus group ( 1 5 species) with “Puerto
Rico” a secondary center (10 species). From this
center, there has been very limited “spillage” into
adjacent areas and islands; S. inaguae and S. un-
derwoodi occur in very circumscribed insular areas
in the southern Bahamas. Of them, S. underwoodi
most closely resembles S. difficiiis, whereas S. in-
aguae seems most closely related to S. notatus. Of
the extra-Hispaniolan-Puerto Rican species, surely
S. notatus has been the most successful, because it
is not only island-wide in Cuba and the Isla de la
Juventud and the Archipielago de los Canarreos, but
it has also successfully invaded both the Great and
Little Bahama banks, where it is also widely dis-
tributed, the southern Florida mainland and the
Florida Keys, and the Swan Islands.

To combine difficiiis, notatus, and macrolepis all
under one name {notatus, which has priority) ob-
scures the quite distinctive pattern features and the
variation in each of these species. Such a nomen-
clatural union also results in a single species which
ranges from the Little Bahama Bank and continental
North America in the north, south throughout Cuba,
Isla de la Juventud, the Swan Islands, Hispaniola,
Puerto Rico, and the northern Lesser Antilles. Such
an agglomeration is to us unsatisfactory. On each
of the geographic areas concerned (the Bahamas and
Cuba, Hispaniola, and the Puerto Rico Bank) there
has been a radiation of geckos in this assemblage;
these radiations constitute natural groupings. The
criterion of intergradation versus reproductive iso-
lation is not applicable. It is our judgement that the
interests of a more manageable taxonomy are served
by according specific rank to S. notatus, S. difficiiis,
and S. macrolepis.

We would not be completely averse, on the other
hand, to consider S. inaguae conspecific with S.
notatus and S. underwoodi with S. difficiiis. Differ-
ences in each case are not particularly trenchant;
nevertheless, there are differences and these do re-
flect (in the cases of these two southern Bahamian
species) divergences from basic stocks which may
well be long-standing (see Schwartz, 1 968, for a dis-
cussion of the southern Bahamian herpetofauna).

60

BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY

NO. 22

perhaps equal in time to several of the Hispaniolan
species which seem to have diverged to a greater
extent than have either inaguae or underwoodi from
their original stocks. Anyone who examines speci-
mens of inagiiae and underwoodi cannot fail to note
their similarities with (and also differences from) S.
notatus and S. difficilis. If these similarities demand
nomenclatural recognition, then “S', n. inaguae” and

“S. d. underwoodi” may be used with propriety. We,
on the other hand, prefer to emphasize the differ-
ences between the satellite taxa and their basal stocks
and to take into consideration the apparent history
of the notatus group in the southern Bahamas; there-
fore our course is to consider S. inaguae and S.
underwoodi as species distinct from their relatives
elsewhere in the Antilles.

LITERATURE CITED

Graham, E. D., Jr. and A. Schwartz. 1978. Status of the name
Sphaerodactylus cinereus Wagler and variation in “Sphae-
rodactylus stejnegerf' Cochran. Florida Sci., 41:243-251.

Grant, C. 1952. Sexual and growth changes in the pattern of
the gecko Sphaerodactylus difficilis Barbour. Copeia, 1952:
187-188.

Noble, G. K., and W. G. Hassler. 1933. Two new species of
frogs, five new species and a new race of lizards from the
Dominican Republic. Amer. Mus. Novitates, 652:1-17.

Ruibal, R. 1946. A new Sphaerodactylus from the Dominican
Republic. Amer. Mus. Novitates, 1308:1-4.

Santiago de la Fuente, G. 1976. Geografia Dominicana. Pub-
lished by the author, Santo Domingo, Republica Domini-
cana. X + 272 -1- 50 -1-25 pp.

Schwartz, A. 1961. A review of the geckoes of the Sphaero-
dactylus scaber group of Cuba. Herpetologica, 17:19-26.

. 1968. The geckoes {Sphaerodactylus) of the southern

Bahama Islands. Ann. Carnegie Mus., 39:227-269.

. 1977. The geckoes (Sauria, Gekkonidae) of the genus

Sphaerodactylus of the Dominican Peninsula de Barahona,
Hispaniola. Proc. Biol. Soc. Washington, 90:243-254.

. 1980. The herpetogeography of Hispaniola, West In-
dies. Stud. Fauna Curasao Carib. Is., 189:86-127.

Schwartz, A., and R. Thomas. 1975. A check-list of West
Indian amphibians and reptiles. Spec. Publ., Carnegie Mus.
Nat. Hist., 1:1-216.

. 1977. Two new species of S'p/zacroifac/y/iti (Reptilia,

Lacertilia, Gekkonidae) from Hispaniola. J. Herpetology, 1 1 :
61-65.

Shreve, B. 1968. The notatus gyovrpoi Sphaerodactylus
Gekkonidae) in Hispaniola. Breviora, 280:1-28.

Thomas, R. 1966. A reassessment of the herpetofauna of Na-
vassa Island. J. Ohio Herp. Soc., 5:73-89.

. 1982. A new dwarf from Haiti (Lac-
ertilia, Gekkonidae). Proc. Biol. Soc. Washington, 95:81-88.

Thomas, R., and A. Schwartz. 1977. Three new species of
Sphaerodactyl us (Sauria: Gekkonidae) from Hispaniola. Ann.
Carnegie Mus., 46:33^3.

. MS. Variation in Hispaniolan Sphaerodactylus (Sau-
ria: Gekkonidae). Mus. Comp. Zool.

Wetmore, A., and B. H. Swales. 1931. The birds of Haiti and
the Dominican Republic. Bull. U.S. Nat. Mus., 155:iv -I- 1-
483.

Copies of the following Bulletins of Carnegie Museum of Natural History may be obtained at the prices
listed from the Publications Secretary, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pittsburgh,
Pennsylvania 15213.

1. Krishtalka, L. 1976. Early Tertiary Adapisoricidae and Erinaceidae (Mammalia, Insectivora) of North

America. 40 pp., 13 figs $2.50

2. Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the
late Pleistocene paleoecology of the central Appalachian Mountains of Virginia. 88 pp., 21 figs. . . .

$12.00

3. Wetzel, R. M. 1977. The Chacoan peccary, Catagonus wagneri (Rusconi). 36 pp., 10 figs

$6.00

4. Coombs, M. C. 1978. Reevaluation of early Miocene North American Moropus (Perissodactyla,

Chalicotheriidae, Schizotheriinae). 62 pp., 28 figs $5.00

5. Clench, M. H., and R. C. Leberman. 1978. Weights of 151 species of Pennsylvania birds analyzed

by month, age, and sex. 87 pp $5.00

6. Schlitter, D. A. (ed.). 1978. Ecology and taxonomy of African small mammals. 214 pp., 48 figs.

$15.00

7. Raikow, R. J. 1978. Appendicular myology and relationships of the New World nine-primaried

oscines (Aves: Passeriformes). 43 pp., 10 figs $3.50

8. Berman, D. S, and J. S. McIntosh. 1978. Skull and relationships of the Upper Jurassic sauropod

Apatosaurus (Reptilia, Saurischia). 35 pp., 1 1 figs $3.00

9. Setoguchi, T. 1978. Paleontology and geology of the Badwater Creek area, central Wyoming. Part

16. The Cedar Ridge local fauna (Late Oligocene). 61 pp., 30 figs $4.50

10. Williams, D. E. 1978. Systematics and ecogeographic variation of the Apache pocket mouse (Ro-

dentia: Heteromyidae). 57 pp., 23 figs $4.00

11. Guilday, J. E., H. W. Hamilton, E. Anderson, and P. W. Parmalee. 1978. The Baker Bluff Cave

deposit, Tennessee, and the Late Pleistocene faunal gradient. 67 pp., 16 figs $5.00

12. Swanepoel, P., and H. H. Genoways. 1978. Revision of the Antillean bats of the genus Brachyphylla

(Mammalia: Phyllostomatidae). 53 pp., 17 figs $4.00

13. Schwartz, J. H., and H. B. Rollins (eds.). 1979. Models and methodologies in evolutionary theory.

105 pp., 36 figs $6.00

14. Engesser, B. 1979. Relationships of some insectivores and rodents from the Miocene of North

America and Europe. 68 pp., 12 figs., 20 plates $5.00

15. Bentz, G. D. 1979. The appendicular myology and phylogenetic relationships of the Ploceidae

and Estrildidae (Aves: Passeriformes). 25 pp., 5 figs $2.00

16. Mares, M. A. 1980. Convergent evolution among desert rodents: a global perspective. 51 pp., 25

figs $3.50

17. Lacher, T. E., Jr. 1981. The comparative social behavior of Kerodon rupestris and Galea spixii and

the evolution of behavior in the Caviidae. 71 pp., 40 figs $6.00

18. McIntosh, J. S. 1981. Annotated catalogue of the dinosaurs (Reptilia, Archosauria) in the collections

of the Carnegie Museum of Natural History. 67 pp., 22 figs $6.00

19. Bristow, C. R., and J. J. Parodiz. 1982. The stratigraphical paleontology of the Tertiary non-marine

sediments of Ecuador. 52 pp., 24 figs $5.00

20. Williams, S. L. 1982. Phalli of Recent genera and species of the family Geomyidae (Mammalia:

Rodentia). 62 pp., 30 figs $5.00

21. Li Chuan-kuei and Ting Su-yin. 1983. The Paleogene mammals of China. 98 pp., 16 figs

$15.00

i

L

1. .

]

■!

i

\

, .T

r'A

I

I ' fl l'v5' i?

i '’ 'i' ,. '

I . ■ * 1 ; ■'

t *, •■<' 4Y'

-

(A-. '

>

^ _

m w rn ^ m ^ ■■ ' m

1; c/5 _ c/5' — (/) Z ^

UliSNJI NVINOSHillAJS SBiyvySIl libraries SMITHSONIAN INSTITUTION NOlifliliSNI NVlNOSHillMS SBiaVB

z. 00 z: ^ z » oo 2 CO

S xriiSife- < 'V S < \v 2 <

I %r | i'^

c/5 ^ w/ ^ 2 '-,)C?.'' z CO ^

AR I Es'^SMITHSONIAN^INSTITUTION “^f^Olini!lSNI_ NVlNOSHillMS S3 I aVH 8 11 LI BRAR I ES^^SMITHSONIAN "^INSTITUl

z \ :z (/) —

o

2

o

-.2

liliSNI NVlNOSHillMS S3iaVH8n LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVlNOSHillMS SBiaVd
r- , z ^ ’Z ^ f~ . z

m ^ rn x^\ ^ m

C/5 *-. _ CO — CO “ [/) . _

ARIES SMITHSONIAN INSTITUTION NOliOiliSNI NVlNOSHillMS SBiaVaail LIBRARIES SMITHSONIAN INSTITU

CO

>

* Z — CO

UliSNI NVlNOSHillMS S3iava8
• rx fh ^

sA# I S

nH<tZ o '

i .< s „< ^ s ^

CO

O

? Vc

> 2 ^ 5 5

1 LIBRARIES SMITHSONIAN INSTITUTION NOliOiliSNI NVINOSHillMs'^S 3 I a V ^

CO — CO — , , CO

— /o':

O _ xiOiij^jx o

ARIES SMITHSONIAN~'lNSTITUTION NOliOiliSNI “"NVlNOSHillMS S3iava8ll L I B R A R I E S^ SMITHSONIAN‘S INSTITU

Z <~ Z_f“ \. z r- 2 r-

m ^ rn ^ rn ^ * ' rn

c/5 _ co "' — CO _ CO

UliSNI NVlNOSHillMS SBIdVaail LIBRARIES SMITHSONIAN INSTITUTION NOlifliliSNI NVlNOSHillMS S3iava

^ ^ CO z <t '"'

2 x4yos>i^

CO Z CO ■ Z CO ' ■* * Z CO

ARIES SMITHSONIAN INSTITUTION NOIiniliSNI_ NVlNOSHillMS S3iava8ll LIBRARIES SMITHSONIAN INSTITU'

c/) rr c/) \ ~ c/)

c

UliSNI NVlNOSHillMS S3iava8ll LIBRARIES SMITHSONIAN INSTITUTION NOliOiliSNI NVlNOSHillMS S3iav:

- z r- _ z r- z

3RARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHIIINS S3IHVyail LIBRARIES SMITHSONIAN INSTIT
^ ^ ^ z \ (f) z » ^ ^ ^ 2:

2 < \v S -ife, < 2 < V S

3 m^A i /#

C/J

o
2: ,

X

J, g 2 Xivos^ >

z £/) ■-•■<■' 2 to z (/)'■' ^ CO

liflillSNI_NVINOSHillMS S3iavyail LIBRARIES SMITHSONIAN INSTITUTION NOIinillSNI_ NVINOSHillNS S3 IB'

^ X (O ~ CO ^ z \ ^

^ ^ W ^

■A H X^x ..Si. “ AWf -I -c — A'iSYFK A

<
q:

^ - - - i

3RARIES SMITHSONIAN INSTITUTION NOlifUliSNI NVINOSHillNS S3iaVMai1 LIBRARIES SMITHSONIAN INSTIl
2 ^ z r*.z: [3 z |3_

2 CO 9 m 0 m 9 . rr.

73
>

73

— \K>,^

m XgvDt^ ^ rn c/j m xi;;<DC^ ^ m

to ± (/> \. £ C/5 X CO

iiniiiSNi NviNOSHiiiMS S3iavaan libraries Smithsonian institution NoiiniiiSNi nvinoshii/js S3ia'

Z CO Z ... to Z V to Z to

< s ,.< ^ 2 < x^s ? •-

i «i\ i I'f 8 8

5 %% i 'clc^l ? i

2 ’CTUV^ J. ^ i N^ooisriX 2 ^ ^ >

3RARIES SMITHSONIAN INSTITUTION NOlifUliSNI NVINOSHillMS^S 3 I d V H a 11 ^ L I B R A R 1 ES SMITHSONIAN INSTI'

m '■ >^v " xjvAsu^ m m %^>vas>^ m

to ' — CO _ to X to

3RARIES SMITHSONIAN INSTITUTION NOliOiliSNI NVlNOSHilWS S3iavaail LIBRARIES SMITHSONIAN INSTI

"' Z . CO Z CO Z ^ ^ Z r

° m 3 E 3 § %% g ==

S 8 S m §} 8 8

o

CO

> 5 xovos^ > 2 \T“- > 5

CO ■**“ Z (O 2 CO 2 CO

iiniiiSNi_NViN0SHims S3iavaan libraries Smithsonian institution NouniiiSNi_NviNOSHiiws S3ia

to — tn — CO ^\tO

1

o 5 ^ ^ 5 ^ XiVA^ O

^ * Z _j 2 _j ^

3 RAR I ES_ SMITHSONIAN_ INSTITUTION NOliniliSNI NVINOSHilWS SBiaVaan LIBRARIES SMITHSONIAN_ INSTI'

- / c,.'” P i P ,./ ™

m 's-Acot^ wc to m ^ m

to £ — CO ? to _ CO

uniiiSNi NviNosHiiws S3iavaaii libraries Smithsonian institution NouniiiSNi NViNOSHiiws S3ia'

Z to Z .. to Z V ^ Z to

< S .< .vv^,. S 2 .< E

I «

> \OOlX52X 5 X > 2

— - — — to ‘ z CO ^ Z CO z

3RAR1ES SMfrUsONIAN INSTITUTION NOliOiliSNI NVINOSHillNS SBiavaail LIBRARIES SMITHSONIAN INSTI'

to — CO — CO X CO

ui .xr:?:7r>^ z .xv ... ^--rrrr*^ ^

CO

CO

UJ

CO

CO