Full text of "Bulletin of the Museum of Comparative Zoology at Harvard College"

See other formats

HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

BULLETIN

OF THE

MUSEUM OF COMPARATIVE ZOOLOGY

HARVARD COLLEGE, TN CAMBRIDGE

VOL. 114

CAMBRIDGE, MASS., U. S. A.
1955 - 1956

The Cosmos Press, Inc.
Cambridge, Mass., U. S. A.

/'^

CONTENTS

PAGE

No. 1. — The Size of Parietal Foramen and Organ in
Reptiles. A Rectification. By Tilly Edinger. Oc-
tober, 1955 . 1

No. 2. — A Classification of Therapsid Reptiles. By D.

M. S. Watson and A. S. Romer. February, 1956 . 35

No. 3. — The Placenta of the Wolverine {Gulo Gulo
Luscus (Linnaeus)). By George B. Wislocki and
Emmanuel C. Amoroso. (8 plates.) February, 1956 89

No. 4. — The Functional Anatomy of the Delphinid
Nose. By Barbara Lawrence and AYilliam E. Sche-
vill. (30 figures.) February, 1956 . . . .101

No. 5. — Silicified Middle Ordovician Trilobites : The
Odontopleuridae. By H. B. Whittington. (24
plates.) May, 1956 153

No. 6. — A Cat Skeleton with an Anomalous Third Hind
Leg and Abnormal Vertebrae. By Thomas S.
Parsons and John M. Stein. (4 plates.) May, 1956 291

No. 7. — Post-Triassic Nautiloid Genera. By Bernhard

Kummel. (28 plates.) July, 1956 .... 319

Bulletin of the Museum of Comparative Zoology

AT HARVARD COLLEGE
Vol-. 114, No. 1

THE SIZE OF PARIETAL FORAMEN AND ORGAN IN
REPTILES. A RECTIFICATION

By Tilly Edinger

CAMBRIDGE, MASS., U.S.A.
PRINTED FOR THE MUSEUM

October, 19.").")

Publications Issued by or in Connection

WITH THE

MUSEUM or COMPARATIVE ZOOLOGY
AT HARVARD COLLEGE

Bulletin (octavo) 1863 — The current volume is Vol. 114.

Breviora (octavo) 1952 — No. 48 is current.

Memoirs (quarto) 1864-1938 — Publication was terminated with Vol. 5.")

JoHNSONiA (quarto) 1941 — A publication of the Department of Mollusks.
Vol. 3, no. 34 is current.

Occasional Papers of the Department of Mollusks (octavo) 1945 —
Vol. 1, no. 18 is current.

Proceedings of the New England Zoological Club (octavo) 1899-
1948 — Published in connection with the Museum. Publication terminated
with Vol. 24.

The continuing publications are issued at irregular intervals in numbers
which may be purchased separately. Prices and lists may be obtained on
application to the Director of the Museum of Comparative Zoology,
Cambridge 38, Massachusetts.

Of the Peters "Check List of Birds of the World," volumes 1-3 are out
of print; volumes 4 and 6 may be obtained from the Harvard University
Press; volumes 5 and 7 are sold by the Museum, and future volumes will be
published under Museum auspices.

Bulletin oi the Museum of Comparative Zoology

AT H A R V A E D COLLEGE
Vol. 114, No. 1

THE SIZE OF PARIETAL FORAMEN AND ORGAN IN
REPTILES. A RECTIFICATION

By Tilly E dinger

CAMBRIDGE, MASS., U.S.A.
PRINTED FOR THE MUSEUM

October, 195r>

No. 1 — The Size of Parietal Foramen and Organ in Reptiles.

A Rectification.

By Tilly Edinger^

CONTENTS

Page
Introduction 3

Opinions 6

Foramen-eye relations in lizards 11

Literary data 11

Figured data 16

Foramen and skull .......... 19

The size question 19

An answer ........... 20

Application to fossils 24

Summary 29

Literature cited 30

INTRODUCTION

The pineal apparatus consists of one or two organs evaginated
from the roof of the diencephalon and their nerve fiber connec-
tions with and inside the brain ( Fig, 1 ) . Study of such an organ
complex would seem a subject altogether inaccessible to students
of extinct vertebrates, of which only the hard parts are preserved.
This is in great part true of the pineal organ proper (glandula
pinealis, epiphysis cerebri), the pineal body which develops in
almost all vertebrates. In the extant forms, at least, the epiphysis
has no relation to osseous elements of the head, and it is only in
some very ancient fishes that the skull bones clearly indicate the
presence, above the brain, of two organs in the pineal complex
(Edinger, in press). The parapineal organ fails to develop in
the vast majority of living vertebrates. In most Lacertilia, how-
ever, it is a stalked, dorsomedian, photoreceptive sense organ
whose presence is reflected in the structure of the skull roof.
Whenever, in a lizard, this parietal organ is well developed as a

1 This study Is one of a series aided by a grant from the Penrose Fund of the
American Philosophical Society.

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

parietal eye, its influence causes and maintains an opening in the
bony shield over the brain. This opening is the parietal foramen
— always median but variable in position even within species (El-
Toubi, 1945, p. 5), lying either between the two parietal bones,
between the two frontal bones, or in the fronto-parietal suture
(Fig. 4). A corresponding foramen is of frequent occurrence in
Paleozoic Agnatha and fishes, and in Paleozoic and Mesozoic
amphibians and reptiles.

BICTAL OdCAN y PINEAL OWCAN

|<Vllllll]IIIIIIIIIIIIIMIIIIIII|IIJ

"EPIOCftMIS-

"•CORIUM

- SKULL" --

PABAPHYSIS ,^p„„^^

[A CyCLOSTOME|H*BENULA«' •

MIOBAAIN
INCAL NERVE

ITERfOft
COMMISSURE

PARAPHYSIS
DORSAL SAC

PARIETAL EYE

JINIIIIIII

PINEAL OACAN

B. ANURAF31
SK{JU.\

PINEAL TRACT

CORIUM ^EP'OERwrS

SKULL

EPIPMTSIS

PARIETAL NERVE

PINEAL TRACT

MIDBRAIN
POSTERIOR COMMISSURE

DORSAL SAC-^^

SUPERIOR MABENULAR/

COMMISSURE

POSTER IC
COMMISSURE

DSRAIN

ID mammaD

IC- REPTILEI

Fig. 1. Diagrams of pineal (epiphysial) and parapineal (parietal) organs
in the midsagittal plane. (From Neal and Band.)

In order to understand the significance of the parietal foramen
and its variations, without direct knowledge of the related organ,
paleontologists must necessarily have recourse to study of the
conditions in the extant Lacertilia. Lizards are the only group
in which the variable relations between the parietal organ and
the parietal foramen can be observed.

There are two other extant types whose pineal apparatus in-
cludes parietal eyes with access to them of light through the
skull roof. In the lowest of living vertel)rates, the lampreys,
both the pineal and the parapineal organs have elements of an

EDINGER : PARIETAL FORAMEN 5

eye as well as of a gland. The petromyzontid condition is im-
portant to paleontologists in evolutionary speculations, but not
in the study of paleo-anatomical detail ; the lamprey skull is
not ossified but consists of cartilage and connective tissue, that
is, of tissues almost never found in a fossilized state. The
other extant type with a parietal eye is another reptile, Spheno-
don. But this last of the Rhynchocephalia survives into the
Present with only one species. In contrast, the Lacertilia still
flourish with several thousand species, of which numerous speci-
mens from almost a hundred species have been used for studies
of the pineal complex.

In the course of preparing a survey of the parietal foramen
through the ages, I have become acquainted with much of the
literature on the pineal apparatus. A recent review lists 1762
papers (Kitay and Altschule, 1954, pp. 107-271). Extensive
surveys from Agnatha to man have been made, respectively, by
a zoologist-librarian (Studnicka, 1905), a neurologist and a
physician (Tilney and Warren, 1919), an embryologist and a
surgeon (Gladstone and Wakeley, 1940). ''The solution of the
pineal riddle"^ (Bargmann, 1943, p. 311) has been attempted,
further, by endocrinologists and pathologists, physiologists and
psychiatrists, anatomists and comparative anatomists. While
paleontology has contributed not more than occasional observa-
tions and some speculations, neontologists have freely drawn
conclusions as to the state of parietal organs in various stages
of vertebrate history. Their concepts were, of course, based on
the pineal complex in the scale of living vertebrates, its embryo-
genesis, its manifest involution in individual life, and sometimes
also on fossils. "A relatively or actually large foramen in the
skulls of the more ancient types of fossils and the tendency for
it to lessen in size or disappear in the more recent types and
living species indicate that the pineal organ itself was more
highly evolved in the more ancient and primitive types, and has
become vestigial in the living representatives of these types"
(Gladstone and Wakeley, 1940, pp. 236-237). Paleontologists,
so far as I know, have never even protested that the types so
connected to demonstrate this supposed trend are not ancestors
and descendants. They have actually been discouraged from
participating in the discussion : they have been warned that, in

1 Foreign language quotations are translated.

6 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

the case of parietal eyes, fossil specimens cannot reveal the con-
dition, or even the presence of the vanished organ.

The first purpose of the present paper is to reject this warning.
Those pineal organs in which the gland potentiality was dominant
have vanished without trace in fossils ; but the condition in living
lizards shows that reptilian fossils do provide a reasonably reliable
record of those organs in the pineal complex in which the sense
organ character was dominant.

OPINIONS

In the literature on the pineal complex, various statements
of a general nature are repeated over and over again. One
reason for this appears to lie in the fact that pineal problems
have attracted students in so many fields. For example, a zoolo-
gist (Stadtmiiller, 1936, p. 597) perpetuates the opinion of a
neo- and paleo-zoologist (Doderlein, 1890, p. 595) who, in a
general description of the tetrapod skull, wrote as follows: "A
parietal foramen is present only in the more primitive forms. It
is highly improbable that this should have corresponded to an
unpaired parietal eye." It is true that the term "parietal fora-
men" has been applied to mid-dorsal openings in the skull of
certain Eocene and Recent teleosts, fishes which indeed lack
dorsal visual organs; but these median gaps actually are fon-
tanelles. The view that pineal organs and foramina in archaic
fishes had no visual function probably goes back to Dean's
homologizing the mid-dorsal foramen of placoderms with one
of the fontanelles of siluroids (1891; 1895, pp. 55-56). The tiny
"open or overgrown hole," which was first discovered but not
named by Wolterstorff (1886, p. 21) in some but not all fronto-
parietal bone specimens of a Miocene frog, has now been found
in a total of three Recent anuran genera (Griffiths, 1954). This
opening is invariably called a parietal foramen; it is, however,
a microscopic, oblique canal containing the vessel and nerve of
the mysterious pineal, or pineal+parapineal, extracranial frontal
organ of the Anura (Fig. IB). Irregular venous foramina in
mammalian parietal bones are likewise called parietal foramina.
But all these structures — in teleosts, anurans, and mammals —
are easily distinguished from the foramen associated with a

EDINGEB: PARIETAL FORAMEN

parietal eye. This quite distinctive feature last occurred, in
Pisces, in the Carboniferous period ; iu Amphibia, in the Triassic.
Among the extant vertebrates with an ossified cranial roof, the
genuine parietal foramen is found only in the tuatara and in
lizards, that is, only in animals that have a photoreceptive para-
pineal organ. If one doubts that this association existed also
in extinct vertebrates, one may as well doubt that the orbits
of fossil skulls contained eyes.

Fig. 2. A, Pariecal eye filling parietal foramen (Anolis sp., after Spencer;
enlargement not known). B, Parietal eye in exceptionally wide foramen
{Agama sanguinolenta, after Schmidt; x 30).

Functions other than visual have been attributed to the parie-
tal organ of lizards for almost a century. It has also been
regarded as an inactive rudiment of a larger organ which, in
Reptilia of the Past, had an unknown, presumably sensory func-
tion, related to conditions not occurring in the Present. The
majority of students, however, have realized that its structure
clearly points to one function : photosensitivity. Thermosensitiv-
ity may be associated with this, and granules apparently secre-
tory in nature have been found in the organ; but, mainly, the

b BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

parietal organ is a photoreceptor. The names of lateral eye
elements are used in its description. Wrapped in loose, vascular-
ized connective and some mucous tissue ("sclera"), the parietal
vesicle lies below or in the parietal foramen, above which a spot of
skin may be conspicuously light in color ("cornea"). When the
organ is fully developed and has not undergone involution, its
distal wall is in great part translucent ("lens"). Its equator
and proximal wall ("retina") contain supporting cells with
light-absorbent pigment, sensory cells, and ganglion cells whose
centripetal processes form the parietal nerve.

Opinions long proven to be erroneous keep reappearing because
the relevant papers are published in journals serving fields of
science too varied to be noticed by workers in any one field.
Recent authors are victims of this situation when they write :
"There is no experimental proof that parietal eyes of lizards or
Sphenodon can perceive light rays" (Stigler, 1950, p. 316), or:
' ' Clearly it cannot be an organ of sight ; and equally clearly it
cannot be an eye in the true sense" (Burton, 1954). Images, of
course, cannot be perceived by the parietal vesicle ; and it is true
that experiments with only negative results were reported in a
1911 paper quoted by Stigler and in a 1907 paper quoted by
Burton. But in later work reactions to light and darkness of
the lizard parietal organ have been demonstrated by several
methods (shift of pigment, e.g., Trost, 1953, p. 202 ; change in
oxygen consumption, Clausen and Mofshin, 1939). Sufficient
proof now exists that in some lizards the parietal eye helps to
distinguish light and dark.

Actually it is to be expected that in some individuals the
parietal organ is found to function as a photoreceptor while in
other individuals light rays concentrated on the parietal skin
spot fail to produce any reaction of the animal. To explain the
variable results of experiments one has only to note the variabil-
ity of structure reported by the histologists. The parietal nerve,
for example, is clearly seen in the embryo and very young, but
has rarely been found in adult lizards (discussed by Bargmann,
1943, p. 455). Absence of the cranial opening is commonly re-
garded as a specific or generic character; but I have come to
realize that the parietal foramen can close during individual
life, a rough macula of bone being built up in its place (Fig.
4Z>).

EDINQER: PARIETAL FORAMEN 9

Even Spencer, the first to compare parietal eye preparations
from many (28) lacertilian species, observed that in every speci-
men there is at least some one feature in which the eye or the
parietal skin spot is less perfect than in other specimens. He
concluded "that the pineal eye in Lacertilia is a rudimentary
structure — that at the present time it is not so highly developed
as it must have been at some previous period when fully func-
tional" (1886, p. 216). This idea found much support, authors
citing very large parietal foramina in some reptiles of the Past.

One must realize that, in contrast to the foramen, the structure
of the parietal organ is revealed only in microscopic sections.
Students of such preparations have not in general been concerned
with the size of the organ. I found only the following data in
the literature. Spencer was amazed to find the axis of the eye
0.4 mm. long both in a 2-foot long Sphenodo7i and in a 6-foot long
Varanus (1886, p. 183). Longitudinal and transverse diameters
varied from 0.18 to 0.20 mm. and from 0.06 to 0.07 mm. in the
Anniella specimens of Coe and Kunkel (1906, p. 393) ; medio-
dorsal skull length is about 8 mm., and length of the animals
105 to 152 mm. from snout to cloaca, plus a post-cloaeal portion
varying from 16 to 75 mm. {ibid., p. 351). Slightly larger parietal
eye diameters in Phrynosoma have been expressed in microns :
258pi and ni[K (Ritter, 1891, p. 212). In short, the diameters of
measured lizard parietal eyes were fractions of millimeters.

In the vast majority of lizards the foramen — correspondingly,
it appears — ranges from pinpoint size to diameters of about
1 mm. Diameters "of considerable size", namely 2.2 and 2.6
mm., were noted by Schmidt (1909, p. 373) in two agamid in-
dividuals. Also among the dozens of lizard skulls I studied,
diameters of more than 1 mm. are rare. I found 4 Agamidae and
3 Iguanidae with diameters of 1.1 to 2.0 mm. ; and the largest, by
far, in the MCZ Department of Herpetology is the parietal fora-
men of the lizard giant, Varanus komodoensis, which is 3 mm. long
and 2 mm. broad (the skull, 190 mm. long). On the other hand,
among extinct Reptilia one can encounter a parietal foramen
as much as 50 mm. in length and 20 mm. in breadth — in a
Jurassic plesiosaur, Pliosaurus, whose skull, lacking tip of muz-
zle, is 1100 mm. long (Linder, 1913, p. 394).

Certainly one must conclude that an extinct reptile such as

BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

this plesiosaur had a larger parietal eye than have the Laeertilia.
On the other hand, there are no data of any kind to support the
idea that in the Past parietal eyes were different, were more
elaborate in structure, than they are in the Present. If there had
existed, for example, auxiliary structures such as muscles and
extrinsic nerves to move the parietal eye, their traces should be
found in the walls of parietal foramina of at least the type shown

Fig. 3. Tapinocephaloid. Meaiau section through back part of skull.
F.m., foramen magnum, P.p., parietal foramen. (After Haughton 1918;
xl/6).

in Figure 3. Further, the ontogenesis of the lacertilian pineal
complex has been thoroughly investigated, over the past 70 years,
by a great number of students, in innumerable embryos of many
species. Although the process is by no means always the same
in every detail, never has anything been observed which could
be interpreted as indicating an originally more complicated
structure of the parietal eye.

In the concept that in fossil reptiles with a large parietal
foramen the organ had great photoreeeptive power, it need not
be assumed that the structure was different from that in lizards.
Size obviously has no necessary relation to function in the lateral
eyes of vertebrates ; the big eyes of a cow see less well, the tiny
eyes of an eagle see far better than human eyes. The parietal
photoreceptor, however, in contrast to the highly differentiated
lateral eye, has no provisions to control the access of light rays.
The amount of light received by its retina must depend only on
the size of the lens, and thus of the vesicle. The tacit assumption

EDINGER : PARIETAL FORAMEN 11

is reasonable that the larger a parietal organ is, the more light
stimuli it will absorb and transmit.

Although the volume of the parietal vesicle must in some
lizards be several times that in others, one cannot well imagine
that among the Recent Lacertilia photosensitive capacity is very-
much greater in larger than in smaller parietal eyes. It can
hardly matter much whether in these minor aids to the major
photoreceptors (later eyes and skin) the lens diameter is a greater
or lesser fraction or, possibly, all of 1 mm. Far greater lenses,
however, must be assumed to kave existed in several extinct
branches of the Reptilia. It is with regard to these that the
questions of parietal eye size become interesting.

In consequence, great importance must be attributed to the
widespread opinion according to which a large parietal foramen
does not necessarily reflect the presence of a large parietal organ.
It was formulated as follows by Camp (1923, p. 395). "The size
of the foramen has no relation to the size of the eye in the latter
group [Lacertidae]. Those who would relate fiLnctional signifi-
cance of the eye with size of the foramen in fossil forms should
note this."

This challenge to paleontology has never been met. It has not
even been discussed. Yet it can be successfully combatted —
with, of course, the material of neontology.

Camp's studies of actual specimens, like my own, were not of
the organ but of skulls. His conclusion as to the absence of size
relations between organ and foramen was drawn from zoological
papers that he quotes. During my more extensive perusal of the
literature I came to recognize this opinion as one of the many
generalizations about the pineal apparatus that have acquired
seeming overall validity merely by being repeated for decades.

FORAMEN-EYE RELATIONS IN LIZARDS

Literary data. The papers dealing specially with parietal eyes
contain in their texts very little information about size relations
between eye and foramen. With the exception mainly of Stud-
nicka (1905) and Schmidt (1909), those who described the organ
did not mention the foramen — although bony foramen margins
occasionally appear in one or both upper corners of their figured
microscopic preparations.

12 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

The consequences are evident whenever a general statement has
seemed necessary, such as in the big monographs mentioned
in the introduction. In one of them the authors agree with Camp
and others : ' ' The size of the foramen differs and bears no direct
relation to the size of the parietal eye" (Tilney and Warren,
1919, p. 131). Another survey, in a chapter entitled "The
Parietal Foramen in Reptiles", says of the parietal eye that
"as a rule ... its diameter is about one-third the diameter of
the foramen" (Gladstone and Wakeley, 1940, p. 270). No
references are given. That 1 :3 proportion has indeed been re-
ported— namely, in one Phrynosoma individual. "While not
stressed by the describer of the specimen (Ritter, 1891), the
condition was mentioned by Studnicka in his monograph — but
mentioned as a case of quite exceptional discrepancy between
organ and foramen (1905, p. 160).

In Phrynosoma, Studnicka 's own specimen also showed "an
extraordinarily broad parietal foramen. It is a big gap in the
bony skull roof, about four times as broad as the parietal eye"
(1905, p. 181). A foramen with an abnormally large transverse
diameter seems quite likely to occur when the opening lies in the
crossing of the frontoparietal and mediosagittal sutures in a thin
skull, particularly in a young lizard. It can then even have the
shape of a T. In such a case, the transverse bar of the T is an
unclosed middle section of the frontoparietal suture, the longi-
tudinal bar a gap either between the two frontal bones (Engel-
brecht, 1951, fig. 10) or between the two parietal bones: a
parietal foramen not fully framed. In none of the fossil reptiles
I know is there a particularly broad parietal foramen ; it is either
a longitudinal oval or a circle, very rarely a little broader than
long. Should a T-shaped opening be found in a fossil, it would
certainly be recognized as a fontanelle rather than a foramen.
There could be no misunderstanding the fact (Engelbrecht, 1951,
p. 21) that only the middle section is related to the parietal organ.

The concept of independent size variation of the parietal organ
and its foramen is, admittedly, a generalization, based on data
reported in special studies of the organ. The basis is, I have
found, psychological rather than factual.

When Spencer (1886, p. 21.")) wrote that in his speeiiiiens of
Laccrta ocellata and Cyclodus "the eye is placed within the

EDINGER: PARIETAL FORAMEN 13

parietal foramen, fitting it closel}'", this was an incidental ob-
servation in a discussion of positions of the parietal eye. Schmidt
was the only author who, with a large amount of material, paid
attention to and discussed size proportions of eye and foramen.
Understandably, he noted and reported the conspicuous and
the exceptional conditions — as did Studnicka in his survey. Only
the following data are mentioned in Schmidt's text (1909).

Much as in a previously figured Anolis, in which "the parietal
eye occupies the entire foramen space in thick bone" (p. 374;
our Fig. 2A), the foramen was only slightly wider than the eye
in one of Schmidt's Anolis cristellatiis specimens, but "approxi-
mately double the size of the eye" in another specimen (p. 381).
In one of Schmidt's other Iguanidae, a specimen of Hoplurus,
the diameter of the foramen was approximately 6 times that of
the eye (ibid.). In some but not all specimens of Agama the
diameter of the foramen was "a multiple" of the eye diameter.
It was because the foramen was almost filled by the eye in the
Lacerta ocellata specimens of other authors that Schmidt men-
tioned one case of a foramen "more than five times as large as
the parietal eye" (p. 392). And it was in connection with this
case that Schmidt quoted, and gave enduring character to the
opinion laid down, without any example, by Carriere (1889, p.
146) : "There is no relation at all between the size of the parietal
foramen and the size and development [Ausbildung] of the
parietal organ, nor between the size of the foramen and the
position of the organ. . . . Bone structure in this case allow^s
no conclusions as to the soft organ we regard as related. The
common conclusion, based primarily upon the larger size particu-
larly of a fossil foramen, that the organ was on a higher level
of development, is inadmissible."

While this statement, being carried on and on in the literature,
must have discouraged paleontologists, certain observations of
Schmidt himself teach the student dealing solely with bones how
he can recognize cases in which the foramen may be considerably
larger than was the organ.

In the Hoplurus specimen, Schmidt reports, the parietal bone
tapers to a very thin lamina as it approaches the foramen. In
all the Agama specimens with a wide foramen, the parietal bones
were thin plates, thinning further towards the foramen (Fig.

14 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

2B). The Laeerta ocellata specimens which Schmidt found
described in the literature had parietal bones far thicker than
the specimen in which he found the wider gap. The numerous
specimens of 40 lizard species which he studied, taken together,
convinced Schmidt (p. 374) that the parietal foramen is small,
more or less fitting the organ, in thick bone. A foramen much
wider than the organ occurs in bone which is thin and which,
further, tapers toward the foramen. "It seems plausible that in
a thick bone the tendency and capacity to lessen or close an
opening are greater than in a thin bone."

One wonders whether the latter condition (thin bone : wide
foramen) is a phase preceding the former (thicker bone: close-
fitting foramen) in ontogeny. Apparently the question has never
been raised as to whether in the Lacertilia a wide parietal fora-
men is a character of very young individuals. In descriptions of
post-embryonic specimens, body length is hardly ever given,
and individual age never. Probably, as little was known of the
age of these Kecent specimens as is the case of all fossil indi-
viduals. Hanitsch (1889) did compare two age stages; but his
specimens were, besides one adult, two newborn Anguis whose
parietal bones were still unossified. Significantly, however, in
a nearly mature embryo of Laeerta the parietal foramen was
described as "a very large pentagonal fontanelle" (Parker,
1880, p. 630).

Among the specimens cited above as examples of absolute si^e
of the parietal organ is a Phrynosoma in which the organ diam-
eters were 0.26 and 0.17 mm. Two other Phrynosoma specimens
were cited among those with foramina much wider — 3 x and
4 X wider — than the organ. However, in one of the two MCZ
Phrynosoma skulls, both 15 mm. long dorsomedially, I could not
find the foramen with a magnifying glass, and in the other the
foramen is a minute circle, just visible to the naked eye : it is not
larger than the organ whose diameters were measured. The
conclusion forcibly presents itself that in the two individuals
with great size discrepancies between organ and foramen the
parietal bone was not fully ossified. Presumably they were
younger than either the individual in which the organ was meas-
ured or those represented by the MCZ skulls. (Perhaps very
young heads were purposely chosen for sectioning in preference

EDINGER: PARIETAL FORAMEN

to heads containing such thickened, knobby, and strongly spined
skulls.)

Studying the skulls alone can not, of course, answer the
question as to whether the relation between the soft organ and
the skull foramen is different in the very young from that in
the adult. Study of the MCZ collection of lizard skulls, all of
them well ossified, showed only that there is much individual
variation in foramen size, and that fortification of the foramen

C - -I

Fig. 4. Iguana iguana {A, C) and Iguana fhinolopha {B, D). D2, Ventral
view of cranial roof, showing remnant of dorsally closed parietal foramen.
X about 2/3.

by a ring of dense bone (Fig. 4) develops at different ages in
different individuals; and it suggested that closure of the fora-
men (Fig. 4Z>) occurs late in individual life.

It may well be because specimens of fossil reptiles are not, in
general, those of very young individuals that I know of no
fossil parietal foramen with tapering margins. Many types of

16 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

extinct reptiles have, on the contrary, the foramen margins raised
and buttressed, as have some lizards. In several kinds of mam-
mal-like reptiles, such fortification developed into "a sort of
Cyclopean periscope" (Gregory, 1926, p. 186) rising above the
general level of the parietal region — a large perforated boss
(e.g., Moschops; ibid., pi. XV) or a chimney-like structure
{Hippo saur us ; Haughton, 1929, fig. 20). A strikingly large
parietal opening is often found in very thick bone — indeed, a
canal rather than a foramen (Fig. 3). Thick bone, be it noted,
is a condition in which, according to Schmidt's considerable ma-
terial of Lacertilia, major size discrepancies between parietal
foramen and organ do not occur.

While occasionally highlighting cases in Avhich the parietal
foramen was found strikingly larger than the organ, the descrip-
tions of the parietal eye in Lacertilia have nowhere answered
the question of what, if any, are its normal size relations to the
foramen.

Figured data. Not the texts of the parietal eye studies but their
illustrations constitute a record of relations between the organ
and the parietal foramen.

In all suitable figures that I could find, I have measured the
maximum diameter of the eye (without the surrounding
"sclera") and the minimum diameter of the foramen. The ratios
of eye diameter to foramen diameter are listed in the table
(p. 00).

It is unfortunate that these ratios must be interpreted without
regard to age factors, as the latter are unknown. Moreover, the
list is rather short. The big surveys by their rich illustration
suggest that the literature they summarize must contain many
figures well-suited to the purpose, showing in section not only
the eye but also both sides of the foramen. Actually, a 1940 figure
for example is likely to be a copy, from Studnicka 1905, of a
19th century figure. The data I brought together were found
in figures published by the following authors: Boveri (B), de
Graaf (G), Hanitsch (H), Leydig (L), Owsiannikow (0),
Schmidt (S), Spencer (Sp), and Studnicka (St).

EDINGER: PARIETAL FORAMEN

TABLE OF RATIOS
Between PARIETAL EYE (=1) and parietal foramen

Seps (L)

0.4

Lacerta (L)

1.4

Chamaeleo (S)

0.5

Lacerta (St)

1.4

Chamaeleo (S)

0.7

Lacerta (Sp)

1.5

Anguis (G)

0.8

Varanus (Sp)

1.5

Seps (Sp)

0.8

Chamaeleo (Sp)

1.5

Lacerta (L)

0.8

Lacerta (L)

1.6

Lacerta (L)

0.8

Chamaeleo (Sp)

1.7

Far anus (Sp)

1.0

Varanus (Sp)

1.7

Maiuia (S)

1.0

Pseudopus (St)

1.8

Lacerta (L)

1.1

Varanus (B)

1.8

Lacerta (L)

1.1

Lacerta (L)

1.1

Anolis (Sp)

1.1

Lacerta (Sp)

2.1

Eumeces (S)

1.1

Caiofes (Sp)

2.1

Lacerta (0)

1.2

Leiodera (Sp)

2.2

Chalcides (S)

1.2

Moloch (Sp)

2.3

Pseudopus (St)

1.3

Lacerta (L)

1.3

Agama (B)

1.3

Chamaeleo (B)

1.3

Agama (S)

3.2

Anguis (H)

1.4

PZica (Sp)

5.8

0-1,2 I.3-L5 1.6-1.8 19-2.1 2.2-24 2.5-2.7 2B-30 3.l-i3

5.8

Fig. 5. Distribution of diameter ratios between parietal eye and foramen;
data from table.

The illustrations in which I measured diameters of parietal
eye and foramen represent 37 individuals. The table, and the
graph illustrating the table (Fig. 5), show that the ratios fall

18 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

into three groups. The small gap between the first and second
group is, however, probably not a genuine separation but merely
due to the paucity of specimens.

By far the largest group is the first. Thirty-one ratios form
a continuous series in which the foramen diameter varies from
about one-half^ to one and four-fifths times the diameter of the
eye. One or both the Anolis specimens mentioned by Schmidt
also belong in this group. The majority of the specimens (19)
are in the 1 :1 to 1 :1.5 range. This is a span of size relations to
be expected between an organ and an orifice formed by the more
or less remote influence of the organ to keep it open and by the
tendency of bone growth to close it. One must remember that
for capacity function of the parietal eye the foramen need not
be much larger than the "lens"; that there is great variation
in the thickness of the "scleral" tissue outside the eye proper,
as also in the amount of its blood capillaries ; and that the eye
is never directly in touch with the foramen margins. One also
has to consider a feature of the material on which the table is
based. "When the direction of a figured section is mentioned at
all in the caption, it is identified as medio-sagittal, i.e., longi-
tudinal. As the foramen is either circular or a longitudinal oval
or slit, I must have, in general, measured the longest diameter
of the foramen, while transversely the eye : foramen ratio was
presumably smaller.

In the second group, in 4 specimens, the foramen diameter is
slightly more than twice the diameter of the eye. This brings to
35, out of 37, the cases which, to me at least, demonstrate that,
although the two related structures are not in contact, there does
exist a definite if variable size relationship between foramen
and eye.

Only 2 specimens are in the third group, that with foramina
3 to 6 times as long as the eyes. This, however, is the group to
which must be added cases described in the literature, quoted
above : 2 Phrynosoma (x3, x4), 1 Hoplurus (x6), 1 Lacerta (more
than 5 x) and possibly some of Schmidt's Agama specimens ("a
multiple"). As such great differences in size were observed in
particularly thin skulls, they probably occur in particularly
young individuals.

1 The Chamaeleo cases are abnormal : the degenerated eye lies above the
foramen.

EDINGEB: PARIETAL FORAMEN 19

Taking together the evidence of published figures and the data
reported in texts, we see that there are 35 cases of what may be
taken as normal size relations, against the 6 definite cases of
considerable discrepancy which have been so highlighted in the
literature as to put in doubt all paleontological research on the
pineal complex.

From the structure of the lacertilian parietal organ we have
above concluded that it must be a type of photoreceptor whose
capacity is proportionate to its size. In my opinion, the size
relations between foramen and eye in the extant Lacertilia
entitle the paleontologist to conclude that the larger a parietal
foramen is in a fossil skull, the larger was the organ it served.

FORAMEN AND SKULL

The size question. What do the authors mean when they write
about "small" and about "large" parietal foramina ? It is
odd that this question should have to be asked, but I have not
found it answered in either the neontological or the paleonto-
logical literature.

Wettstein, zoologist, stated that Sphenodon, "representative
of the lowest evolutionary stage among the living reptiles . . .
has a relatively large parietal foramen" (1931, pp. 2, 33). This,
presumably, is meant to say that a tuatara skull has a larger
parietal foramen than a similar-sized skull of other living rep-
tiles. This is indeed the ease ; the foramen is 3 x 1 mm. both in
an MCZ tuatara skull (mediosagittal length 56 mm.) and in
Wettstein 's specimen (ibid., fig. 37; skull 51 mm.). Among the
comparable lizard skulls I measured, the largest foramen is
1x1 mm.

Huene, paleontologist, made a survey as to "which stego-
cephalians and reptiles have a relatively large-sized parietal
foramen" (1933, pp. 372-373). He concluded from the distribu-
tion of ' ' large ' ' foramina that the parietal eye was a more potent
organ in certain underwater conditions than it was, and is, in
terrestrial habitats. From this it would seem that by size Huene
meant absolute size, and by ' ' relatively large ' ' he meant that the
foramen is absolutely larger in one group than in another.

20 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

I realized, however, that this interpretation was wrong when
I checked on one of Huene's examples — the suborders of the
Sauropterygia, an extinct order on which Huene is an expert,
and of which I have measured the foramen parietale in numerous
skulls (Edinger, MS, in preparation). "In the placodonts that
foramen is large, in the nothosaurs smaller, and similar in the
plesiosaurs" (Huene, loc cit.). But, in e.g. the Middle Triassic,
Placodns skulls (median length 158 to 200 mm.) have the parietal
foramen 10 to 19 mm. long, with breadths from 10 to 14 mm. ;
and in Nothosaurus skulls (125 to about 800 mm. long), length
of the foramen ranges from 4 to as much as 23 mm., breadth
from 2.5 to 23 mm. Parietal foramina still larger developed later
in at least one infraorder of the Plesiosauria. A 50 mm.-long
foramen was mentioned above ; there have further long been on
record another Jurassic Pliosaurus (maximum skull length 1160
mm.) with a 40 x 15 mm. foramen (Linder, 1913, p. 394), and
a Cretaceous Brachauchenius (skull about 800 mm.) vrith a fora-
men approximately 40 mm. in length (Williston, 1907, pp. 479,
484) . It thus appears that Huene's statements regarding the size
of the opening did not imply absolute size but, rather, size rela-
tive to skull size.

The fact is that the parietal foramen is a far more conspicuous
feature in the center of a short, broad Placodus head than in the
back part of the slender, elongated head of Nothosaurus or a
plesiosaur. I believe that this is what most authors have in mind
when they call a parietal foramen large or small ; they mean the
size of the foramen relative to the size of the whole skull. Others
apparently intend to say that the foramen they describe is
larger or smaller than in some other form whose head, or
perhaps whose body is of comparable size. Those few who paid
attention to the foramen: organ ratio called "small" a parietal
foramen approximately the size of or smaller than the organ,
while "large" in such a study means that a foramen is larger
than the organ. There has, further, been occasional comment on
small size of the parietal foramen as compared with the lateral
eye orbits. Still further, some authors must have been thinking
merely of absolute size of the parietal foramen.

A7i answer. Great differences in absolute size of the parietal

EDINGER: PARIETAL FORAMEN 21

organ and hence of the parietal foramen presumably are, we
have seen, significant physiologically. But in the comparative
anatomy of a group as vast as the Class Reptilia the meaning
of absolute size of the foramen differs enormously. "We have
to compare tiny animals with animals whose body is a hundred
or a thousand times larger. Obviously, the size of the parietal
organ relative to the size of the body would be not only difficult
to calculate but also of dubious value. The same is true of the
size of the foramen relative to the skull as a whole. We find
the foramen in skulls compactly constructed, and in skulls drawn
out into short or long muzzles; but the parietal foramen is a
part of the cranium, the ''Hirnschadel" of German terminology
which, like the brain, is never involved in the process of skull
elongation as much as the facial portion ("Kieferschadel").

A meaningful expression of parietal eye size can be found
through the fact that this organ, in contrast to the lateral eyes, is
entirely a part of the brain. Skulls do not build a separate orbit
to lodge the parietal eye. This lies either within the brain chamber
or, if in the parietal foramen, in an extension of the brain
chamber.

Seeing the parietal eye as one of the brain organs, one is not
surprised by the discovery of Spencer quoted above — namely,
that this organ can be of the same size in a two-foot long and in
a six-foot long reptile. The entire brain, including the extensive
olfactory tracts whose length necessarily varies with skull length,
was found to be 70 mm. long in a Fython of about 6.5 m. length,
and 40 mm. in a PytJion of about 2.5 m. length (Edinger, 1929,
p. 100). The difference in brain-to-body length ratios (1:93 and
1:63, respectively) would be far greater if only the contents of
the braincase proper had been measured.

It follows that calling the parietal foramen small in large
animals or in large skulls may be as unfortunate as the common
usage of calling the brains of elongated heads smaller than the
brains in short heads. It is even assumed that the former are less
efficient than brains which constitute a large portion of the
head. Efficiency of the various brain functions depends, of
course, much more on the composition than on the size of the
brain, relative or absolute. Greater or lesser relative size, within
the brain, of one or more of its components, is in general the

22 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

anatomical correlate of greater or lesser importance of the func-
tions they serve.

In consequence, a parietal organ should be regarded as large
when it is a relatively large appendage of the brain. Its osteologi-
cal reflection, the parietal foramen, is large when it is a consider-
able interruption of the braincase roof, otherwise closed.

In many reptiles an approximate size relation of the parietal
organ to the rest of the brain can be shown three-dimensionally
by endocasts of the braincase (Fig. 6). This method, however, is
not practicable in the Lacertilia and the many other groups whose
braincase is not a complete bony capsule.

But even in species whose braincase is incompletely ossified,
an index of brain size is available. The occipital area, which
lodges the hindbrain, is usually well ossified. Consequently,
macerated or fossil skulls in which the foramen parietale is also
present include the foramen occipitale magnum. This surrounds
the posterior end of the brain. The lumen of the foramen mag-
num is somewhat larger than but corresponds approximately to
a cross section of the medulla oblongata. The relation between
bone and medulla is here similar to that between parietal foramen
and parietal eye. In both these cranial openings, size differences
roughly represent size differences of the related organs. Com-
paring the diameters of foramen magnum and foramen parietale
is a method not as exact as paleoneurology can use in other in-
vestigations but preferable, I believe, to the vague methods of
comparison hitherto used. I regard as a fortunate circumstance
the fact that the brain area whose osseous counterpart we can
compare with that of the parietal eye is in the most conservative
portion of the brain. The medulla, in contrast to the brain as a
whole, is of similar shape in all reptiles ; and its thickness varies
mainly, or exclusively, with body size.

For the appraisal of parietal organ size by measuring and
comparing the diameters of foramen parietale and foramen
magnum, an immense amount of material is available in extinct
and extant reptilian skulls. However, for large-scale comparisons
of the two foramina, paleontology is under more than its usual
handicaps. Crushing of course can practically eliminate even the
foramen magnum (for example, in all plesiosaurs preserved in
slates). A fragment may contain one but not both foramina. The
good fossil skulls are held in collections all over the world; de-

EDINGER: PARIETAL FORAMEN 23

scriptions of them rarely list foramen diameters, and figures
show but indistinctly that small feature, the parietal foramen.

A few examples will be given here of this means of comparison.
We will first reconsider from this viewpoint animals mentioned
above in one or another context. We can then examine with this
method two of the general evolutionary concepts mentioned
which, incidentally, are contradictory (as is typical of the pineal
complex literature). Is the size of the parietal organ correlated
with the habitat of a reptile? Do large parietal foramina occur
only in very ancient and/or very primitive reptiles ?

The parietal foramen of Sphenodon is larger than in lizards
relative to the brain, as expressed by the approximate breadth
and height of the medulla oblongata as well as by other standards
of comparison. In the two skulls mentioned, in which the parietal
foramen is 3 x 1 mm., the foramen magnum is 5 x 5 (MCZ) and
5x4 mm. (Wettstein, 1931, fig. 32).

The relatively smaller size of the parietal foramen in Lacertilia
may be illustrated by two series of skulls in the MCZ Department
of Herpetology. One series consists of 8 Iguana skulls with
dorsomedian skull lengths from 38 to 100 mm. Nos. I and IV- VII
are I. igua^ia; II, III and VIII are I. rhinolopha (see Fig. 4).
The foramen maguum is 5 x 4 mm. in the smallest, 5x5 mm. in
the next four, 6x6 mm. in a 64 mm.-long skull, but 7x7 mm. in a
60 mm.-long and in the largest skull. The parietal foramen meas-
ures fractions of 1 mm. in I, IV and VI, and reaches diameters
of 1 mm. in II (length even 1.2 mm.). Ill, V, VII. The foramen
is dorsally closed in the 100 mm. skull ; its remnant is a ventral
circular pit with 1 mm. diameters.^

The other series is one of six skulls of Ctenosaura pectinata,
55 to 84 mm. long. Breadth of the foramen magnum rises from
5 to 6 mm. and so does height except that, while it is 6 mm. in
specimens IV and VI, height is only 5 mm. in V. The parietal
foramen is 1 x 1 mm. large only in IV and in VI, the largest skull.
Length is 1 mm. also in the smallest specimen, but breadth is
less than 1 mm., as are the diameters in the other three speci-
mens.

1 As the closure phenomenon is apparently not previously noted in lizards, I
mav add here that it is also shown in a series of eight Phifsignathiis cocinciiius
skulls ."G to 52 mm. long. The parietal foramen, invisible to the naked eye in the
smallest, tiny in all others except in one of the five 45 to 47 mm.-long skulls
(1 xO.5 mm.), is closed in the largest siieeimen.

24 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

The parietal foramen of the giant, Varanus komodoensis, so
conspicuously large with diameters of 2 and 3 mm., is similar
in relative size to that of some of the smaller lizards ; the foramen
magnum is 14 mm. broad and 12 mm. high.

If an attempt is made to compare the areas rather than
diameters of the two foramina — by the method of multiply-
ing with each other the diameters of the openings — the difference
between the rhynchocephalian and the lizards becomes quite
striking. The parietal foramen then compares to the foramen
magnum as 1 -.7 and 1 .-8 in Sphenodon, but as 1 :28 in Varanus
komodoensis, and the largest parietal foramina in the Iguana and
Ctenosaura series have ratios of 1 :21 and 1 :36, respectively.

Briefly, in the lizards I measured, the relatively largest diam-
eters of the parietal foramen are only about one-fifth of the
diameters of the foramen magnum ; in most specimens they
are far less. This is the osteological expression of a parietal eye
so small that its existence remained unknown for many decades
of research on lizard brains.

Application to fossils. Much larger than in any of the Kecent
Reptilia was the parietal foramen in specimens of the two
sauropterj^gian genera discussed above. There are late Middle
Triassic (Upper Musehelkalk) skulls both of Placodus and of
Nothosaurus in which the foramen parietale is almost as large as
the foramen magnum. Before this phase, however, parietal
foramen diameters in Lower Musehelkalk Nothosaurus averaged
only about half those of the foramen magnum ; and they seem
to have shrunk to about one-third in an Upper Triassic Notho-
saurus. In the placodonts, the foramen ultimately closed {Hen-
odus, Upper Triassic; Edinger, MS. in preparation). Similarly,
the foramen had disappeared in the last of the plesiosaurs (at
least Hydrotherosaurus, late Upper Cretaceous; Welles, 1943, p.
133).

The Placodontia and Nothosaurus were semi-aquatic seashore
dwellers at the times when their parietal eyes were small and
when they were large. It is really not possible to correlate evolu-
tion of a large parietal eye with the environment of the animals
in which it developed. The history of this structure in placodonts
and Nothosaurus constitutes but two examples of a phenomenon

EDINGER: PARIETAL FORAMEN 25

which occurred repeatedly in the history of the reptiles. The
parietal foramen was enlarged for some time in several branches
of the Reptilia, in different habitats, in different periods of
earth history.

The story of Nothosaurus may. further, serve as one of the
actual records which do not fit into the theory that the parietal
foramen is larger in primitive than in progressive forms. In
this long-lived genus, neither the evolutionary enlargement nor
the reduction of the parietal eye was paralleled by any notice-
able change in other structures, except that some of the Upper
Muschelkalk individuals were very large animals.

When it is said that the genuine foramen parietale occurs
only in primitive vertebrates, this is true in so far as neither
foramen nor eye are left in birds and mammals. On the other
hand, the statements are unacceptable which assert that among
those vertebrate groups in which the foramen occurs it is present,
or it is large, only in primitive forms. Such assertions do not
agree with the record. The vertebrates with parietal foramina
are extinct Agnatha, Pisces, and Amphibia, and both extinct and
extant Reptilia. I have not yet found a large parietal foramen
in a fish; among amphibians, the foramen appears to be large
in those labyrinthodont skulls which represent young individuals
— but my studies are far from concluded. At present it seems
to me that parietal foramina of strikingly large relative size
occurred, in gnathostomes, only among Reptilia. Apparently
the related ancient photoreceptor came to play a larger role in
this class than in the other classes. This view finds support, I
believe, in differences between the dorsal eyes of living verte-
brates, that is, of lampreys and of reptiles (Fig. 1 A and C).
Only in the reptiles is the ocular structure perfected ; only the
reptilian parietal eye has a distinct, clear lens.

The various extinct Reptilia in which the parietal eye became
large cannot all be regarded as primitive. Permian Therapsida,
mammal-like reptiles of the suborders Theriodontia and Dino-
cephalia were mentioned above as having the parietal opening
as a wide canal through a thick skull roof (Fig. 3), or having
it strongly buttressed by a superstructure of the parietal bone.
In the parietal cone of Delphinorinathus conoccphalus, Seeley
found a circular parietal foramen "nearly 2 cm. in diameter";
the foramen magnum "appears to have been not more than 2.7

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

cm. high by 1.3 cm. wide" (1892, pp. 469, 471). A certain im-
portance of the parietal eye in some kinds of therapsids is also
demonstrated by the endocranial cast of a later, early Triassic
representative of still another suborder, Dicynodontia. In a
141 mm. -long skull of Lystrosaurus (Fig. 6^), the parietal
foramen does not appear to be particularly large. In the brain
(Fig. 6B), the parietal organ must have been an appendage of
considerable size. The channel it produced in the skull is 18
mm. high caudally where it joins the midbrain chamber. On the

Fig. 6. Lystrosaurus sp. A, Skull in dorsal view, from Huene 1931; x 1/3.
B, Its endocranial cast in left side vieWj M.C.Z. 2124 (duplicate of specimen
in Geol. -paliiont. Inst. d. Univ. TUbingen; x 1).

EDINGER: PARIETAL FORAMEN 27

outer skull surface the diameters of the opening (9x7 mm.) are
about half those of the foramen magnum (18 x 13 mm.) ; but
below that, the east has revealed a 10 x 10 mm. swelling which
presumably indicates the position of the parietal eye. As ap-
parently everywhere among the Reptilia when the parietal fora-
men is known in more than one individual, its variability can
be seen in Lystrosaurus. In a 171 mm. -long skull, the parietal
canal is 17 mm. deep, and its diameters on the surface are 10
and 7 mm., much as in our figured specimen; but the diameters
of the foramen magnum, 17 and 23 mm., are more than twice
those of the foramen parietale in this larger skull (van Hoepen,
1913, pp. 25, 34). To judge from Brink's figures (1951, 3D
and E), the parietal foramen was still smaller, relative to the
foramen magnum, in the largest of complete Lystrosaurus skulls.^

The Therapsida are a reptilian order greatly specialized in
the direction of mammals. It contains many forms in which
the parietal foramen was very small, and the foramen was lost
before the therapsids gave rise to the new class. But in this order
there also developed at least some individuals in which the
parietal eye was half to fully as large in section as the medulla
oblongata.

The most primitive reptiles are included in the order Cotylo-
sauria. According to the general concept of parietal eye devolu-
tion, the early cotylosaurs should have the largest parietal fora-
mina among the Reptilia.

A paragon, in the literature, of parietal foramina ' ' exception-
ally large" (Gregory, 1951, p. 338), is that of the early Permian
cotylosaur Diadectes. It is indeed huge as seen in the short and
massive skulls, a perpendicular tunnel through an extremely
thick skull table. Less impressive is its endocast, seen as a dorsal
extension of an endocast of a cranial cavity (Gregory, 1951, figs.
12. 13. Al). As I have reason to suspect that this endocranial
cast is not that of one specimen but a composite which Cope
took from several cranial fragments, the size of the parietal
organ is better appraised by the method used above. Among
the Diadectes specimens in the American Museum of Natural
History, the often-figured, undistorted skuU No. 4839 contains

1 Postscript. Six Lystrosaurus skulls were measured in the paleontological col-
lection of Bavaria and in the British Museum (Nat. Hist.). The foramen magnum
diameters are in one case more than three times, in three cases more than twice,
and in two cases less than twice those of the foramen parietale.

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

the foramen magnum as well as the foramen parietale. The
former is 17 mm. high and 19 mm. broad. The latter is prac-
tically the same size : 16 mm. long and 20 mm. broad (the respec-
tive "areas" being 323 and 320 mm.^). A parietal foramen far
larger (at least in an absolute sense) is seen in the Diadectes
fragment AMNH 4352, the partial roof of a skull which ap-
parently was but little larger than No. 4839, and also in the
excellent natural (stone-) mould of its ventral side (Gregory,
1951, figs. 12.13. A 2 and A3). While in other specimens the
parietal foramen is more or less circular, in this specimen it is

Fig. 7. Right halves of sectioned cotylosaur skulls.
Price; x 3-1/3. B, Diadectes, after Gregory; x 1/3.

A, Captorhinus, after

15 mm. broad and as much as 28 mm. long. Thus some Diadectes
may have had a parietal organ larger, relatively, than e.g.
Placodus. To judge from figures in the literature, some later
diadectomorphs indeed had a parietal foramen larger than the
foramen magnum (e.g. Pareiasaiirus, Boonstra, 1930, figs. 1-2).
These later Diadectomorpha, nota bene, were among the types
more advanced than Diadectes; and Diadectes itself was not
a truly primitive but a rather large and in several respects a
specialized cotylosaur.

EDINGER: PARIETAL FORAMEN 29

It is the other, generally more primitive branch of the Cotylo-
sauria, the Captorhinomorpha, which "may lie close to the stem
of the reptiles, if, indeed ,the stem reptile itself was not a member
of it" (Romer, 1945, p. 173).

Figure 7 serves to compare L. I. Price's (1935) reconstruction
of a medio-sagitally sectioned Captorhinus braincase, a careful
combination of his findings in six specimens, and W. K. Gregory's
(1951) similar reconstruction of Diadectes, based on the AMNH
material. The figure shows that the length of the foramen parie-
tale was much less in the primitive than in the specialized coty-
losaur, relative to the height of the foramen magnum. It is true
that the foramen magnum is so high in Captorliinus that surely
not its entire height was occupied by the medulla; its breadth,
however, is also greater than the diameters of the parietal fora-
men, which is circular. In Protorothyris, Romeria and other
earlier and more primitive captorhinoids, the relative sizes of
foramen parietale and foramen magnum appear to be essentially
the same as in Captorhinus, according to Mr. John B. Clark, who
is studying this group. He further informs me that the occiput
is generally distorted so that his conclusion is in part based on
observations of the braincase as a whole and its relationship to
the whole skull, where there is a very strong resemblance to
Captoi'hinus.

In Seyynouria, primitive in combining amphibian and reptilian
characters, AYhite (1939, p. 333) described the parietal foramen
as shaped like an inverted tunnel, its diameter on the ventral
surface of the bone about one and one-half times that on the
dorsal surface ; reconstructing the braincase on the basis of sev-
eral specimens, White shows (his fig. 10) the external parietal
orifice about one-third as long as the foramen magnum is high.

Thus in at least some of the truly primitive Reptilia the
parietal foramen was relatively smaller than in various progres-
sive and in later forms. It was, however, far larger than it is in
any surviving reptile.

SUMMARY

It has been claimed that paleontology is unable to contribute
reliable data to the evolutionary history of the pineal complex,

30 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

although it was in the Paleozoic and Mesozoic eras that numerous
types of vertebrates had the parietal foramen. In modern times,
this foramen occurs only in Lacertilia and the tuatara, and it is
invariably associated with a parapineal photoreceptive organ,
the parietal eye.

The idea that the size of the foramen can not reveal the size
of the organ is here rejected. The size proportion between organ
and foramen does not vary as much in the Lacertilia as was
assumed because a few (possibly juvenile) cases of exceptional
discrepancies were repeatedly highlighted in the literature. The
conditions found in lizards, rather, give confidence in paleonto-
logical study — not, of course, of the entire pineal apparatus, but
of the parietal eye. The large parietal foramina in some fossil
reptilian skulls are valid evidence as to the one-time presence
of parietal eyes far larger than have survived into the Cenozoic
era.

The concepts that large parietal foramina and organs de-
veloped in a particular habitat, or occurred only in very ancient
or primitive Reptilia, are put to a test. They appear to be mis-
conceptions due in great part to the lack of a definition as to how
the size of the foramen should be judged. As the organ is a part
of the brain, it is suggested that the parietal foramen be com-
pared with another cranial opening, the foramen magnum, cor-
responding approximately to the circumference of another brain
part, the medulla. Whereas the diameters of the foramen mag-
num are, in lizards, at least five times those of the parietal fora-
men, parietal foramina relatively far larger were evolved in
various other branches of the Reptilia. Examples here given of
extinct reptiles with parietal foramina close to the foramen
magnum in size include terrestrial Permian, and semi-aquatic
Triassic forms.

LITERATURE CITED

Bargmann, W.

1943. Die Epiphysis cerebri. Mollendorff 's Haudb. mikrosk. Anat. d.
Menschen, 6 (IV) : 309-502.

BOONSTBA, L. D.

1930. A contribution to the cranial osteology of Fareiasawrus serridens
(Owen). Ann. Univ. Stellenbosch, 8 (A): 1-19.

EDINGEB: PARIETAL FORAMEN 31

BOVERI, V.

1925. Untersuchungen iiber das Parietalauge der Eeptilien. Acta Zool.,
6: 1-57.
Brink, A. S.

1951. On the genus " Lystrosauru^ " Cope. Trans. Eoy. Soc. S. Africa,
33: 107-120.
Burton, M.

1954. Misleading ' ' Third Eye. ' ' 111. London News, 1954 : 270.
Camp, C. L.

1923. Classification of the lizards. Bull. Amer. Mua. Nat. Hist., 48:
289-481.

Carriere, J.

1889. Neuere Untersuchungen iiber das Parietalorgan. Biol. Cbl., 9
(1890) : 136-149.

Clausen, H. J. and B. Mofshin

1939. The pineal eye of the lizard (Anolis carolinensis) , a photorecep-
tor revealed by oxygen consumption studies. Jour. Cell. Comp.
Physiol., 14: 29-41.

CoE, W. R. and B. W. Kunkel

1906. Studies of the California limbless lizard, Anniella. Trans. Conn.
Acad., 12: 349-403.

Dean, B.

1891. Pineal fontanelle of placoderm and catfish. Rep. Conun. Fisheries

New York, 19: 306-363.
1895. Fishes, living and fossil. Columbia Univ. Biol. Sex., 3. New York

and London.

Doderlein, L.

1890. Vertebrata. In Steinmann, G. und L. Doderlein, Elemente der
Palaontologie, pp. 516-827. Leipzig.

Edinger, T.

1929. Die fossilen Gehirne. Erg. Anat. Entwicklungsgesch., 28: 1-249.

(In press). Paired pineal organs. Folia Psychiatr. Neurol. Neurochir. Neerl.,
Suppl.

El-Touibi, M. R.

1945. Notes on the cranial osteology of TJromastyx aegyptiam (For-
skal). Bull. Fac. Sci. Fouad I Univ., 25: 1-12.

Engelbrecht, D. v. Z.

1951. Contributions to the cranial morphology of the chamaeleon
Miorosaura pumila Daudin. Ann. Uuiv. Stellenbosch, 27 (A) :
3-31.

32 BULLETIX: MUSEUM OF COMPARATIVE ZOOLOGY

Gladstone, R. J. and C. P. G. Wakelet

1940. The pineal organ. The comparative anatomy of median and
lateral eyes, with special reference to the origin of the pineal
body; and a description of the human pineal organ considered
from clinical and surgical standpoints. XVI + 528 pp. London
and Baltimore.
Graaf, W. H. de

1886. Bijdrage tot de kennis van den bouw en de ontwikkeling der
epiphyse bij amphibian en reptilien. Leiden.
Gregory, W. K.

1926. The skeleton of Moschops capensis Broom, a dtnocephalian rep-
tile from the Permian of South Africa. Bull. Amer. Mus. Nat.
Hist., 56: 179-251.
19.")1. Evolution emerging. 2 vols. New York.
Griffiths, I.

1954. On the nature of the fronto-parietal in Amphibia, Salientia.
Proe. Zool. Soc. London, 123: 781-792.
Hanitsch, E.

1889. On the pineal eye of the young and adult Anguis fragilis. Proc.
Liveipool Biol. Soc, 3: 87-95.

Haughton, S. H.

1918. Some new carnivorous Therapsida, with notes upon the brain-
case of certain species. Ann. S. Afr. Mus., 12: 175-216.
1929. On some new therapsid genera. Ann. S. Afr. Mus., 28: 55-78.
HoEPEisr, E. C. N. v.

1913. Bijdragen tot de Kennis der Eeptielen van de Karrooformatie.
Ann. Transvaal Mus., 4: 1-46.
HUENE, F. V.

1931. Beitrag zur Kenntnis der Fauna der siidafrikanischen Karroofor-

mation. Geol. palaont. Abh., 22 (N.S. 18) : 157-228.
1933. Zur Lebensweise und Yerwandtschaft von Placodus. Abh.
senckenb. naturf. Ges., 38: 365-382.
KiTAY, J. I. and M. D. Altschule

1954. The pineal gland. A review of the physiologic literature. Cam-
bridge, Mass.
Leydig, F.

1890. Das Parietalorgan der Amphibien und Eeptilien. Anatomisch-
histologische Untersuchung. Abh. senckenb. naturf. Ges., 16:
441-551.

LiNDER, H.

1913. Beitrage zur Kenntnis der Plesiosaurier-Gattungen Peloneustes
und Pliosaurus. Geol. palaont. Abh., 15 (N.S. 11) : 337-409.
Neal, H. Y. and H. W. Rand

1936. Comparative Anatomy. Philadelphia.

EDINGER: PARIETAL FORAMEN 33

OWSIANNIKOW, P.

1888. Ueber das dritte Auge bei Petromyzon fiuviatilis nebst eiaigen

Bemerkungen iiber dasselbe Organ bei anderen Thieren, M6m.

Acad. Sci. St. Petersbourg, (7) 36 (9) : 1-26.
Parker, W. K.

1880. On the structure and development of the skull in the Laeertilia.

I. On the skull of the common lizards (Lacerta agilis, L. viridis

and Zootoca vivipara). Phil. Trans. Roy. Soc. London, 170:

595-640.
Price, L. I.

1935. Notes on the brain cai5e of Captor'hinus. Proc. Boston Soc. Nat.
Hist., 40: 377-386.

RiTTER, W. E.

1891. The parietal eye in some lizards from the western United States.
Bull. Mus. Comp. Zool., 20: 209-228.

ROMER, A. S.

1945. Vertebrate Paleontology. Chicago.

Schmidt, W. J.

1909. Beitrage zur Kenntnis der Parietalorgane der Saurier. Zeitschr.
wiss. Zool., 92: 359-426.

Seeley, H. G.

1892. On Delphinognathns conoccphalus (Seeley) from the Middle
Karoo Beds, Cape Colony, preserved in the South African Mu-
seum, Capetown. Quart. Jour. Geol. Soc. London, 48: 469-475.

Spencer, W. B.

1886. On the presence and structure of the pineal eye in Laeertilia.
Quart. Jour. Microsc. Sci., (N.S.) 27: 165-238.
StadtmItller, F.

1936. Kranium und Visceralskelett der Stegocephalen und Amphibien.
Bolk's Handb. vergl. Anat. Wirbelt., vol. 4, pp. 501-698.

Stigler, R.

1950. Versuche iiber das Parietalauge von Agama colonorum. Zool.
Anz., 145: 316-318.

Studnicka, F. K,

1905. Die Parietalorgane. Oppel's Lehrb. vergl. mikrosk. Anat. Wir-
belt., vol. 5, VIII + 254 pp.

TiLNEY, F. and L. F. Waiiren

1919. The morphologj' and evolutional significance of the pineal body.
Amer. Anat. Mem., 9: 1-257.

Trost, E.

1933. Die Histogenese und Ilistologie des Parietalauges von Anguis
fragilis und Chulcides oceUatus. Zeitschr. Zellforsch., 38: 185-217.

34 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Welles, S. P,

1943. Elasmosaurid plesiosaurs, with a description of new material
from California and Colorado. Mem. Univ. Cal., 13: 125-254.
Wettstein, 0. V.

1931. Ehynchocephalia. Kiikenthal's Handb. Zool., 7 (I) : 1-235.
White, T. E.

1939. Osteologj' of Seymouria haylorensis Broili. Bull. Mus. Comp.
Zool., 85: 325-410.
Welliston, S. W.

1907. The skull of Brachauchenius, with observations on the relation-
ships of the plesiosaurs. Proc. U.S. Nat. Mus., 32: 477-489.

WOLTERSTORFF, M.

1886. Ueber fossile Frosche, insbesondere das Genus Palaeotatrachus.
I. Jahresber. Abh. naturw. Ver. Magdeburg, 1 (1885) : 1-93.

Bulletin of the Museum of Comparative Zoology

AT HARVARD COLLEGE
Vol. 114, No. 2

A CLASSIFICATION OF THERAPSID REPTILES

by D. M. S. Watson and A. S. Romer

CAMBRIDGE, MASS., U. S. A.
PRINTED FOR THE MUSEUM

February, 1956

Publications Issued by or in Connection

WITH THE

MUSEUM OF COMPARATIVE ZOOLOGY
AT HARVARD COLLEGE

Bulletin (octavo) 1863 — The current volume is Vol. 114.

Breviora (octavo) 1952 — No. 51 is current.

Memoirs (quarto) 1864-1938 — Publication was terminated with Vol. 55.

JoHNSONiA (cjuarto) 1941 — A publication of the Department of ]\Iollusks.
V^ol. 3, no. 34 is current.

Occasional Papers of the Department of Mollusks (octavo) 1945 —
Vol. 1, no. 18 is current.

Proceedings of the New England Zoological Club (octavo) 1899-
1948 — Published in connection with the Museum. Publication terminated
with Vol. 24.

Bulletin of the Museum of Comparative Zoology

AT HARVARD COLLEGE
Vol. 114, No. 2

A CLASSIFICATION OF THERAPSID REPTILES

by D. M. S. Watson and A. S. Romer

CA^iIBRIDGE, MASS., U. S. A.
PRINTED FOR THE MUSEUM

February, 1956

No. 2 — A Classification of Therapsid Reptiles
By D. M. S. Watson and A. S. Romer

CONTENTS

Page

Introduction 38

Order Therapsida 42

Suborder Theriodontia 46

Infraorder Titanosuchia 47

Family Brithopodidae 48

Anteosauridae 49

Jonkeriidae 49

Infraorder Gorgouopsia 51

Family Aelurosauridae 52

Aelurosauropsidae 52

Arctognathidae 53

Arctognathoididae 53

Broomisauridae 54

Burnetiidae 54

Cynariopsidae 54

Galerhinidae 55

Galesuchidae 55

Gorgonognathidae 55

Gorgonopsidae 56

Hipposauridae 56

Inostranceviidae 56

Pachyrhinidae 57

Phthinosuchidae , 57

Eubidgeidae 57

Scylacocephalidae 58

Scylaeopsidae 59

Scymnognathidae 59

Sycosauridae 60

Gorgonopsians lacking family characteristics ........ 60

Gorgonopsians represented by inadequate material ... . . 61

Infraorder Cynodontia 61

Family Procynosuchidae 62

Thrinaxodontidae 63

Cynognathidae 63

Diademodontidae 64

Gomphodontosucliidae 65

Cynodontia incertae sedis 65

38 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

Infraorder Ictidosauria 66

Family (unnamed) 66

Tritheledontidae 66

Tritylodontidae 67

Microcleptidae 67

Possible mammals from the ' ' Ehaetic. " 67

Infraorder Therocephalia 68

Family Pristerognathidae 68

Trochosuchidae 69

Whaitsiidae 69

Euchambersiidae 71

Incertae sedis 72

Infraorder Bauriamorpha 72

Family Lycideopsidae 73

Ictidosuchidae 73

Nanictidopsidae 74

Silpholestidae 74

Scaloposauridae 74

Ericiolacertidae 75

Bauriidae 75

Incertae sedis 76

Suborder Anomodontia 76

Infraorder Dinocephalia 77

Family Deuterosauridae 77

Mosehopidae 78

Tapinocephalidae 78

Infraorder Venyukovioidea 79

Family Venyukoviidae 79

Infraorder Dromasauria 80

Infraorder Dicynodontia 85

Therapsida int;ertae sedis 86

INTRODUCTION

In 1838 a Scottish engineer, Andrew Geddes Bain, who con-
structed military roads in South Africa, and was the first to
determine the general geology of that region, discovered fossil
reptiles in rocks north of Fort Beaufort in the eastern part of the
colony of the Cape of Good Hope. These were the first therapsids
to be found (unless some of those from the Ural mountains were
collected earlier) and were later described by Richard Owen.
Bain and his son Thomas, Dr. Atherstone and some of his friends.

WATSON AND ROMER : THERAPSID CLASSIFICATION 39

and some others made large private collections, most of which
ultimately reached the British Museum and were also described
by Owen, In 1889 H. G. Seeley, aided by a grant from the Royal
Society, made a special expedition in order to collect fossil rep-
tiles in the Karroo. This was the first expedition of such a kind
ever to leave England, and the first to go to the Karroo. Since
then local collectors and visitors from many universities have
sought reptiles there. The vision and enthusiasm of Robert
Broom, then in medical practice, led to the collection by him,
and by men inspired by him, of immense numbers of fossil skulls,
few of which are adequately prepared.

Broom described these materials, founding on them several
hundred species, each with a short .statement of matters wherein
it differed from others, and a figure, or more recently several
figures. Other men added — and are still adding — to the collec-
tions and to the list of conventionally diagnosed species. Thus
there are now some 688 species described from the Karroo. Of
these the great bulk are therapsids. The rocks which have yielded
this immense number of terrestrial vertebrates cover a period
from some point in the Middle Permian to the end of Triassic
time, about 45 million years.

The therapsids fall essentially into two groups, plant-eating and
animal-eating, the first in effect serving as the food supply of the
second series. The first group comprises in fact the only plant-
eating terrestrial vertebrates of their time, and with a few ex-
ceptions (essentially their ancestors) all fall into the group
"Anomodontia," typified by Dicynodon. The range in size of
anomodonts is very great ; adult skulls vary from less than 3
cm. to more than 100 cm. in length. There exist very large num-
bers, certainly several thousand, of complete but usually unpre-
pared skulls, but skeletons are of the greatest rarity; not many
more than a dozen are to be found in the museums of the world.
The attempts made up to the present to divide this great mass
into genera have not reached far, and evidently their understand-
ing will depend not only on much further preparation, but on a
greatly improved precision of stratigraphical knowledge. Nothing
is so important at the present time as a detailed study of the
stratigraphy.

The remaining therapsids, Gorgonopsia, Theriodontia, etc., are
more susceptible to classification, and we felt that for our own

40 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

purposes some attempt should be made to draw up a scheme, with
short diagnoses, isolating groups of species which could thus be
handled in the literatui'e.

The great bulk of the species were described by Robert Broom,
a man of genius with a unique visual memory and a critical mind.
He recognised some hundreds of therapsids which were not
anomodonts. Such a Avealth of species is in no way improbable ;
it should be compared with the number of mammalian species
found in the Tertiary of the United States from Upper Eocene
to Pleistocene times. What this number may be we do not know,
but it must vasth^ exceed the number of Karroo reptiles. Thus
we are inclined to believe that the majority of Broom's species
may be well founded.

Both the present authors have handled many Karroo fossils,
including a considerable part of the type materials. Thus it
seemed that direct comparison of figures would allow us to draw
up a classification, which could make no pretence to complete-
ness or even probability, but would at least pro\nde some sort of
framework for further studies.

The material to which the following classification applies is
comparable in variety and complexity to the Class Mammalia.
The classification is therefore designed on the broadest lines, the
families, apart from the Gorgonopsia, being wide groups includ-
ing a range of often very varied forms, for no useful purpose
could be served by establishing small, tightly defined ones.

Our experience of considering systematically the whole litera-
ture of the therapsids has shown us that the continued descrip-
tion of new species of these reptiles on the level which is now
customary is most undesirable. What is wanted is a complete
preparation, by the acetic and formic acid techniques, of all the
existing type skulls. These should then be fully described, and
illustrated by figures representing dorsal and palatal views of
the skull, lateral views of skull and lower jaw, an occipital view,
and probably a detailed figure of the otic region. The figures
should distinguish fractures from genuine margin; restorations
of broken outlines (which are often useful) should be made in
dotted lines outside the damaged edges. The locality and horizon
of the specimen must be stated.

The classification of such very varied groups can only progress
when the rocks from which they are derived have been adequately

WATSON AND ROMER : THERAPSTD rLASSIFTOATION 41

subdivided into stages, and these mapped. The good exposures
usual in the Karroo make such mapping easy (in a technical
sense) though it involves time and much walking. Such work is
by far the most urgent requirement of students of these reptiles,
and should have complete priority over the description of any
more new species.

Miss Townend made for us tracings of all published figures,
with the lateral views of skulls drawn facing the same way.
Spreading these out side by side, sorting them into groups and
making use of the published descriptions, and of our personal
knowledge of the types, we grouped them into "clumps" and
tried to draw up a workable definition of each.

Recently we have had the very great advantage of a publication
by S. H. Haughton and A. S. Brink, "A bibliographical list of
Reptilia from the Karroo Beds of Africa," Palaeontologia Af-
ricana, II, 1954. This is an admirable work, listing each species
with a full set of bibliographic references, and giving the horizon
and place of discovery of the type specimen. It provides also a
series of families, and gives a diagnosis for each genus. The
families we have set up were made independently, and chiefly
before the publication of Haughton and Brinlv's list.

We were able to collaborate in this way because the senior
author held an Alexander Agassiz Professorship in the Museum
of Comparative Zoology in Harvard University in 1952, and was
thus able to discuss the drawings Miss Townend (with help from
the Royal Society) made for us. Thanks are due to these two
bodies for their generosity.

In general we have carried the classification down to the genus
only. In the case of the gorgonopsians, however, generic distinc-
tions are none too certain in many cases and hence we have listed
all species. We have not attempted subdivision of the dicynodonts,
since much greater knowledge of skull structure than we possess
today is needed before the systematics and evolution of this
group can be unravelled. The vast majority of the therapsids
listed are from South Africa ; we have noted localities only for
forms found in other regions. Stratigraphic horizons for types
from South Africa and Russia are given in terms of the zones
currently accepted for the Permian and Triassic of those regions.

42 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

ORDER THERAPSIDA

Advanced theropsid reptiles derived from the Pelycosauria.
Characterised by the relatively great size and downtiirning of
the paroccipital process so that it abuts directly against the
inner surface of the squamosal opposite the inner side of the
quadrate, not far above the articulation for the lower jaw. The
temporal fossa is widened by an outpushing of the zygomatic
arch, which is associated with a lateral extension of the squamo-
sal. The posterior surface of this bone faces backward lateral
to the ridge which bounds the area of attachment of neck
muscles and lies at the outer end of the paroccipital process.

The quadrate is always attached to the quadratojugal in a
characteristic theropsid manner, the latter bone resting upon
the upper surface of the projecting outer condyle of the quadrate,
being then separated from it by a quadratojugal foramen, and
finally rising as a thin sheet closely applied to the posterior
surface of the quadrate. It thus, to a greater or lesser extent,
separates the quadrate from that depression on the front face
of the lateral extension of the squamosal in which it lies.
The suture between the epicondyle of the quadrate and the
quadratojugal closes early, and has very rarely been seen. The
quadrate-quadratojugal complex becomes much reduced, finally
forming little more than a flattened nodule providing the articu-
lar surface for the lower jaw. In early forms the distal end of
the stapes lies in a depression, the stapedial recess, in the inner
edge of the quadrate, and in some later forms actually articu-
lates directly with it, the stapedial recess vanishing.

The braincase is a massive structure, its component bones
usually fusing early in life. The condyle always includes contri-
butions from the exoccipitals, and in very advanced forms these
alone remain functional, the basioccipital withdrawing from con-
tact with the atlantal intereentrum. The prootic (which tends
to be more expanded than in pelycosaurs), opisthotic and supra-
occipital usually fuse early in life. The paroccipital process is
massive and distally has a large abutment with the squamosal.
Proximally the opisthotic joins with the prootic to form a fenes-
tra ovalis completed by the posterolateral corner of the basi-
sphenoid. The fenestra is placed very ventrally so that it is
usually completely below the base of the brain. The floor of the

WATSON AND ROMER : THERAPSID CLASSIFICATION 43

braincasp, orioinally thick, becomes much thinner with time in
the history of some groups, and may eventually be extremely
reduced.

The basipterygoid process in the most primitive Russian form
still articulates with the pterygoid by a smooth cartilage-covered
surface. In all later forms the articulation is immobile and
usually sutural. In general the process is converted into a
flattened sheet of bone projecting horizontally and laterally well
in front of the i)ituitary fossa. To it the pterygoid is attached
below, and the epipterygoid rests on it. The anterior part of
the brain usually lies in a sphenethmoid, a mesial bone resting
on the upper border of a deep sheet of bone which is part of the
basisphenoid extending forward below the pituitary fossa. Ex-
(•ej)tionally the sphenethmoid resolves itself into a pair of orbito-
sphenoids.

Superficial membrane bones may, as in Dinocephalia, be rough-
ened by an ornament, but in the later forms they are usually
smooth (with such exceptions as the snout of Cynognathus) .
The membrane bones of the temporal region in early forms lie
superficially, but later come to be closely attached to the brain-
case and largely buried by muscles. There is no supratemporal,
and the dermosupraoccipital and tabulars, which long survive,
lose all superficial exposure and become thin films "plastered"
onto the posterior surface of the widened supraoccipital and
parietals and the squamosals which border them. They never
transgress the flange which marks the limits of attachment of
the neck muscles on the squamosal. The parietals in early forms
are widely expanded on the roof of the head, the brain cavity
being limited by ridges on their ventral surfaces which may lie
much mesial of the free border of the bone. They rest on the
supraoc('ii)ital and are tied to it by the dermosupraocciptals and
tabulars. The lateral border of the parietal is primitively, and
always to some extent, attached to a process from the postorbital
which extends backward mesial to the temporal fossa, towards
and in more primitive forms to the squamosal. In most forms
the parietals separate to surround a pineal foramen, and in
some groups are separated by a median preparietal bone. The
parietals may form a deep and very narrow sagittal crest. The
frontals are relatively small bones usually forming some part of
the orbital margin. They are bordered by a postfrontal and, on

44 BULLETIN: MUSEUM OP COMPARATIVE ZOOLOGY

the final loss, b}^ reduction in width, of that bone, by the post-
orbital which may itself be lost. Anteriorly they join the nasals
and are bordered by the prefrontals. The lacrimal, short but
always present, retains its superficial position.

The face may be very long, or very short, in specialised forms.
It is seldom, if ever, pointed and usually is parallel sided, or
even bulbous, so that the premaxillae are wide. In consequence
the nostrils, in many groups, face as much forward as outward,
and are often overhung by a projection of the nasal. There is a
septomaxilla, with a superficial exposure in most early forms,
withdrawn within the nasal cavity in some advanced animals.
Sclerotic plates are occasionally present. There is some evidence
which suggests that the latest therapsids possessed mammal-like
turbinal bones in their very large muzzles, implying that they
may have been warm blooded. The presence of large foramina
on the face of late forms suggests that the skin was muscular
and sensory and possibly had hairs, wdtli the glands associated
with them.

Primitively the quadrate ramus of the pterygoid, arising from
the attachment to the basipterygoid process, pa.sses directly to
the quadrate, but in some advanced forms it may be cut short,
scarcely extending l^ehind the basipterygoid process, its place
being taken by an epipterygoid process which may itself fail to
reach the quadrate. The rest of the pterygoid stretches outward
to make a transverse flange against the lower jaw, and forward
to unite with the vomers and palatines to roof the mouth. The
interpterygoid vacuity is reduced, often to a very small hole
between the transverse flanges, or a narrow slit in advance of the
basipterygoid processes. The palate so formed is greatly varied :
it may be essentially flat, it may be largely vaulted, or it may
bear a deep but narrow gulley passing backward from the
posterior nostrils. In some forms, in differing ways, a secondary
palate is formed so that the point at which inspired air enters
the mouth becomes placed far back, hi primitive forms the
palate is continuous l)elow the eye but in some a suborbital
fenestra, between the pterygoid, ectopterygoid and palatine,
makes its appearance. And this suborbital fenestra may in turn
vanish with a re-expansion of the bones which surround it. A
hyoid bone is not uncommonly seen ; it stretches forward and
inward from the neighborhood of the jaw articulation and pre-

WATSON AND ROMER : THERAPSID CLASSIFICATION 45

snraably implies the presence of a muscular tongue.

The lower jaw is always compouud, but the dentary may be-
come very large and a place of muscle attachments. There is
always a reflected lamina of the angular. There is usually a
marginal dentition, and often palatal teeth, but all these may
be lost and functionally replaced by horny sheets. The jaw
articulation is opposite or anterior to the level of the occipital
condyle. In some cases the lower jaw is capable of very free
movements so as to grind up food in the mouth.

The vertel)ral column usually lacks intercentra, except an-
teriorly; the centra are usually not notochordal. To an in-
creasing extent it is diversified, cervicals tend to he sharph-
distinguished from the dorsals which follow them, and a lumbar
series with short fused ribs becomes established in later forms.
There are usually at least three sacrals, and there may be as many
as seven. The tail varies immensely in length, being sometimes
as long as the rest of the body, sometimes a mere conical hinder
end.

The limb girdles steadily become modernised, so that although
a precoracoid and coracoid are always present the screw-shaped
glenoid cavity of pelycosaurs is lost, and the precoracoid ceases
to contribute to its surface at all. The clavicles are little ex-
panded ; short, broad interclavicles are universal. The cleithrum
finally disappears. An ossified .sternum is often present shewing
the points of articulation of sternal ribs. The pelvis, at first pely-
cosaur-like, gains a thyroid fenestra; the iliac blade lengthens
and approaches the mammalian position. The short pubis is
still deflected anteriorly, and the ischium shortened.

Both fore and hind limbs become more advanced by bringing
the elbow and knee in toward the body ; thus the humerus becomes
narrower and less twisted than in pelycosaurs and the femur
gains an inturned head. The humerus always has an entep-
icondylaj-. and sometimes an ectepicondylai- foramen. The del-
topectoral crest is elongate. The number of digits is five in both
hand and foot of all known thera'psids. The numbers of phal-
anges, at first 2.3.4.5.4(3), is reduced to 2.3.3.3.3 in both fore and
hind foot, an intermediate stage being known in which the primi-
tive number of phalanges is retained but the third phalanx of
the third toe, and the third and fourth of the fourth toe are
reduced to verv short discs. There is one centrale in the tarsus.

46 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

Distal carpal and tarsal five are absent. The effect of these
changes is to raise the body of the animal off the ground, so that
with the development of a digitigrade foot and the inturning
of the elbow and knee, the animal's gait becomes very mammal-
like, and its speed presumably largely increased.

There are apparently never any epiphyses in therapsids, so
that residual cartilages of some substance seem to have remained
(throughout life ?) in most members of the group.

The therapsids are customarily divided directly into a number
of groups such as dinocephalians, dromasaurs, dicynodonts, theri-
odonts, etc. It appears, however, highly probable that at a very
early stage in tlierapsid evolution there occurred a dichotomy
resulting in two main lines, one including mainly carnivorous
types and leading toward and to the mammals, the other con-
sisting of herbivores, with the dicynodonts as its most conspicuous
component. We therefore consider the order Therapsida as
consisting of two suborders, here termed Theriodontia and
Anomodontia. The former group is an expansion of the Therio-
dontia, as usually defined, to include, on the one hand, the car-
nivorous dinocephalians and, on the other, the ictidosaurs. In
the Anomodontia are here included the herbivorous dinocephal-
ians and uroiuasaurs as well as the dicynodonts. The present
classification necessitates the division of the Dinocephalia as
currently conceived. This last term is here restricted to the
herbivorous forms (tapinocephaloids). The carnivorous forms
(titanosuchids) fall into two groups, one of which may be
called the Titanosuchia whilst the other includes certain very
primitive members of the Gorgonopsia, the Phthinosuchidae.

SITBORDER TI1P]RI0D()NT1A

Therapsids generally wdth a carnivorous dentition in which
there is a series of incisor teeth in the premaxilla, while the
maxillary series contains towards its anterior end one or more
enlarged teeth, the canines. The latter are usually oval in sec-
tion and deep rooted. In some late forms this dentition may
be converted (as in many mammals) into an herbivorous den-
tition by loss or reduction of the canine and modifications of the
cheek teeth. Palatal teeth, usually restricted to thickened re-
gions of the palatine and pterygoid, are commonly present in

WATSON AND ROMER : THERAPSID CLASSIFICATION 4:7

members of the group. A tranverse flange, composed maiiily of
pterygoid but with a modest contribution from the ectopterygoid
bone, always occurs ; the quadrate ramus of the pterj^goid arises
abruptly from the flange, usually considerably dorsal to the
admesial part of its lower border, and extends outwards toward
and generally to the quadrate. This ramus, for a greater or less
part of its length, touches the basisphenoid (in which the para-
sphenoid is included). At some point between the pterygoid
flange and the posterior end of the basisphenoid (at the tubera
which support the front part of the border — usually cartilagi-
nous— of the fenestra ovalis) the pterygoid passes ventral to
and is attached to the lower surface of a horizontal sheet of
bone projecting outwards from the side of the basisphenoid.
This projection is the basipterj-goid process, whose upper sur-
face is in contact with the base of the epipterygoid. The oc-
cipital aspect of the skull has a central area composed of the
supraoccipital, the exoccipitals, and the usually powerful opis-
thotics, largely coated by the fused dermosupraoccipitals and
tabulars. It is bounded by a baclvwardly projecting ridge, made
jointly by the tal)ulars and squamosals with contributions from
the parietals and interparietal in advanced forms. The supra-
temporal has completely vanished, and the posterior surface of
the squamosal (lateral to the ridge) is usually of considerable
width. The face is never greatly shortened; the lacrimal never
reaches the septomaxilla.

In contrast to anomodonts, the face is not notably bent down-
ward on occiput and suspensorium. The epipterygoid is usually
broad. The suspensorium is never of great height, the jaw
articulation being essentially in line with the tooth row and
about at the level of the occipital condyle. The under surface
of the skull tends to be flattened. The fenestra ovalis lies close
to the ear cavity. A coronoid bone is present and a "coronoid
process" frequently developed in the jaw.

Infraorder Titanosuchia

The parietal region broad, the temporal fossa lying largely on
the outer side of the skull in early forms, its margin carried
backwards on to a backwardly turned border of the squamosal.
The face is long. There is no preparietal bone. Typically four

48 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

stout incisors, a prominent canine, usually circular in section,
and a variable number — typically 8 to 16 — of small cheek
teeth in both upper and lower jaws. The quadrate is a bone
inserted in front of the lower border of the squamosal, but
largely visible from behind. The external nostrils are narrow
slits, floored largely by the septomaxilla, just touching the nasal,
but largely surrounded by the premaxillae, whose internarial
processes slope steeply backwards, separating the anterior ends
of the nasals for a considerable distance. The lower jaw is shal-
low, with the articular surface directed largely backward and
below the line of insertion of cheek teeth into the dentary. The
reflected lamina is quite deep and its region of insertion lies far
forward on the angular. Postfrontal present. No secondary
palate. Paired vomers. No suborbital vacuities. Quadrate ramus
of pterygoid well developed and of some height. Quadratojugal
relatively large with laterally exposed area. Narrow ascending
ramus of epipterygoid. Occipital condyle single. No projecting
coronoid process. Cleithrum present. Little iliac expansion ;
no thyroid fenestra.

The titanosuchians have generally been included in the Dino-
cephalia, of which the typical members are large herbivores.
Both are unquestionably primitive in many regards and in their
most generalised members close to the base of the therapsid stock.
But apart from primitive features the two groups have little in
common except a clumsy build and a tendency toward thicken-
ing of the skull roof, both of which characters appear to be accom-
paniments of the usual large size.

Family Brithopodidae. Primitive titanosuchians, characterised
by a face which is rather deep in proportion to its width and re-
tains something of the triangular section found in Dimetrodon.
The zygomatic arch is shallow and the temporal fossa large.
The parietal foramen is very far back and the plane of the
occiput nearly vertical. The quadrate is relatively short and not
appreciably carried forward ventrally ; the condyle is separated
into two by an oblique spiral depression. Thickening and rough-
ening of the bones of the skull is incipiently deA'eloped. Teeth
are developed on the palate. The precoracoid enters the glenoid
cavity, which retains a trace of its original screw shape. Phalan-
geal formula 2.3.3.3.3 in both limbs. The tail very long.

WATSON AND ROMER : THERAPSID CLASSIFICATION 49

Very primitive therapsids from the Russian Permian. They
were long known only from fragmentary remains, but Efremov
has recently described a good skull of Titanophoneus and Orlov
has given a general description of the skulls of Admetophonens
and Syodon.

Admetophoneus Efremov 1954, Russian zones 1 and 11 ; Briiho-
pus Kutorga 1838 {Dinomurus Fischer 1847, Eurosaurus Fischer
1842, Orthopus Kutorga 1838, fUhopalodon Fischer 1841), Rus-
sian zone I; Syodon Kutorga 1838 (Cliorhizodon Twelvetrees
1880), Russian zones I and 11 ; Titanophoneus Efremov 1938,
Russian zone II.

Limb bones found with brithopids are brigaded together as
Phreatosuchidae by Efremov, and are remarkable because their
structure includes many features which characterise pelycosaurs.
They are described as Phreatophasma Efremov 1954, Phreato-
saurus Efremov 1954, and Phreatosuchus Efremov 1954.

Family Anteosauridae. Titanosuchians of large size with skulls
usually greatly thickened in aged individuals. The pineal fora-
men lies relatively far forward, but the intertemporal surface
is wide, the postorbital strips being largely covered by muscle.
The face is short, deep, and triangular in section (with a rounded
median ridge). The frontal area is swollen and usually forms a
l)row over the snout. The facial processes of the premaxillae
have a nearly straight dorsal border and extend far backward
as narroAV points betAveen the nasals. Their dentigerous border
slopes upward in front. The incisors are simple conical teeth
which, with the canines, point somewhat forward ; the numerous
small maxillary teeth lie on a festooned border. The quadrate
and attached quadratojugal are largely exposed below the
squamosal, but continue the general direction of the hinder sur-
face of that bone. The basis crauii is thin ; the paroccipital proc-
ess, though wide from back to front, is shallow.

Anfeosauriis Watson 1921 (Titanognathus Broili and Schroe-
der 1935, Dinosuchus Broom 1936, Broomosuchus Camp, Taylor
and Welles 1942), Micranteosanrus Boonstra 1954, Paranteosau-
rus Boonstra 1954, Pseudanteosaurus Boonstra 1954 ; all from the
Tapinocephalus zone.

Family Jonkeriidae. Large titanosuchians with a very elon-
gated face, wider than high and of nearly the same depth at the

50 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

nostril and the front of the orbit. The face is rounded in section
and in no way triangular. The temporal region has a wide bar
between the fenestrae, composed very largely of parietals, with
the pineal foramen far forward. The temporal fossa is much
smaller than in the Anteosauridae, and the quadrate very much
larger, the quadratojugal attached to it being largely visible
below the zygoma in a side view of the skull. The quadrate con-
dyle faces largely forward and lies so far in front of the occiput
that the temporal fossa appears floored by bone in direct dorsal
view.

The palate lacks the tooth-covered thickenings found on the
palate of the anteosaurs, and has enormously long and wide pala-
tal nares, but is generally of similar type. The teeth carried by
the widened premaxilla differ from those of Anteosaurus in that
they possess basin-shaped crushing surfaces (like those of the
tapinocephaloids) behind the high labial spike. The upper sur-
face of the parietal and hinder part of the frontal are raised
as a boss, a forehead standing out above the nearly straight
mid-dorsal line of the face.

The jonkeriids as above described have a resemblance, which
may be of importance, to the tapinocephaloids. Each group has
an expanded cranial region from which a relatively small face
projects forwards, and in each group certain members, at any
rate, have the highly characteristic crushing premaxillary teeth
best explained by Efremov in Deuterosaurus. These unique
crushing incisors, taken with the development in each of a
similar association of a swollen posterior part of the skull with a
relatively slender face, suggest a community of origin, the two
groups having parted from one another not very long before
their first recognition at the base of the fossiliferous Karroo.

lArchaeosuclius Broom 1905 (not identifiable) ; Dinartamus
Broom 1923; Dinocifnodon Broom 1929; Dinophoriens Broom
1923; Enohiiis Broom 1923; Glaridodon Seeley 1888 (not identi-
fiable) ; Jonkeria van Hoepen 1916 {Dinospliageiis Broom 1929,
Dinopolus Broom 1936); ILamiasaurus Watson 1914; Phoneo-
suchus Broom 1929; Scapanodon Broom 1904 (not identifiable) ;
SdiHiki. Broom 1929; TitanosHchus Owen 1879, all from the
Tapinocephalus Zone.

watson and romer : therapsid classification 51

Infraorder Gorgonopsia

Muzzle typically long and heavj'. Temporal region relatively
short, low and moderately expanded. Greatest skull width some-
what anterior to the quadrates. Dorsal processes of premaxillae
relatively short here and in remaining theriodonts (in contrast
to the titanosuchians).

The roof of the skull in the temporal region is wide, the
postorbital meeting the squamosal and excluding the parietal
from the border of the temporal fossa. There is in all later forms
a preparietal bone, not usually in contact with the parietal fora-
men. The postfrontal is always present. The squamosal is at-
tached medialty to the parietal, tabular, and the end of the
paroccipital process and laterally forms a thick transversely
placed plate which turns forward to form the upper border of
the zygomatic arch. Immediately lateral to the end of the paroc-
cipital process the anterior face of the lower part of the squamosal
is excavated into a recess, with a sharply defined upper border
which may actually overhang the recess. Into this the quadrate
and the adherent quadratojugal are received in such a way that
much of the height of both bones is hidden by the squamosal in
occipital view; in more advanced forms only a small proportion
of the height of the quadrate can be so seen. There is no sec-
ondary palate, the choanae being very extensive openings which
extend far back and in all later forms are enlarged on their outer
sides anteriorly to form a hole through which the lower canine
passes when the mouth is closed. In this and the following
theriodont groups the vomers are fused. No suborbital vacuities.
In Gorgonopsia, in contrast to the Therocephalia, the lateral
border of the quadrate ramus of the pterygoid instead of follow-
ing a straight line from its origin at the transverse flange, passes
back parallel to the principal plane of the skull for a long way
before turning out to join the quadrate, often with a curved
border. This condition, though less well developed, anticipates
the "girder" existing in higher cynodonts. The strap-shaped
epipterygoid often has a quadrate ramus which may very nearly
reach the quadrate. In the lower jaw the hinder end of the
dentary projects freely above the surangular. Incisor teeth,
usually 5 in number, are of large size ; a precanine tooth in the
maxilla very rarely occurs; there is a single canine and the

52 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

number of cheek teeth rarely exceeds 5. Palatal teeth occur. The
digital formula is 2.3.4.5.4. The preeoracoid is excluded from
the glenoid cavity in this and the remaining theriodont groups.
Little acromial development. Ilium little expanded; thj'roid
fenestra present.

The gorgonopsians are relatively primitive theriodonts, sug-
gestive of ancestry to the cynodonts; however some features,
such as the reduced cheek dentition indicate that typical gor-
gonopsians are aberrant to at least some extent. Most gorgonop-
sian skulls exhibit but minor variations on a fairly uniform
pattern. For systematic purposes most might either be "lumped"
in a single large family or divided into a considerable number
of small families separated by relatively slight distinctions. The
latter course is followed here, although in pursuing it there is
found to be a number of genera in which proof is lacking of
the distinctive qualities of any of the families erected.

Family Aelurosaurldae. Gorgonopsians in which the inter-
temporal region is wide and the zygomatic arches probably
do not spread widely. The snout is long but rather narrow in
proportion to its depth, and the transverse section across it has
a rounded upper margin. The dentition is 5.1.4-5, the dentiger-
ous surface of the premaxilla sloping downward from the first
incisor tooth, and the molar series inserted in a gently convex
border of maxilla. The palate possesses teeth in irregular rows
along the transverse flanges of the pterygoids, and there are
sharply limited patches of teeth on the pterygoids and palatines
posterior to the hinder ends of the internal nares.

This is a good recognizable group, but difficult to define.

""Aelinoynathus" serratidens Haughton 1915, L. Cistecephalus
zone; " Aelurognatlius" cf. serratidens Broili and Schroeder
1934, Cistecephalus zone; '* Aelurogtiathus" sollasi Broili and
Schroeder 1935, U. Endothiodon zone; Aelurosauroides watsoni
Boonstra 1934, Endothiodon zone; Aelurosaurus felirius Owen
1881 (Aelurosaurva hreviceps Broom 1931, Aelurosaurus striati-
dens Broom 1912, Aelurosaurus tenuirosiris Broom 1911, Aeluro-
saurus whaitsi Broom 1911), Endothiodon zone; "Gorgonopsier"
Broili and Schroeder 1936, E. Africa, "Lower Bonebed."

Family Aelurosauropsidae. Small gorgonopsians character-
ised by the flattened face, which is nearly twice as wide as it is

WATSON AND ROMER : THERAPSID CLASSIFICATION 53

high. The premaxilla below the nostril is deep, the face very
little less than half the total length. The intertemporal width
exceeds the interorbital. The pineal foramen touches the pre-
parietal and is well behind the postorbital bar. The occipital
condyle is the most posterior point of the skull.

Aelurosauropsis {Aelurusauroides) ivilmanae Broom 1940, M.-
U. Cistecephalus zone.

Family Arctognathidae. Advanced gorgonopsians with a short
muzzle, rounded in section and with its tooth rows widely
separated. The tooth-bearing lower border of the maxilla forms
a large segment of a circle, the point of origin of the Z3'goma
lying high up. Upper teeth 4.1.6. The intertemporal surface is
narrow ; the occipital surface meets it far forward ; the pre-
parietal lost or very small. The palate has a deep median trough
extending back from the posterior nares, overhung by small
tooth-bearing projections on the palatines. Behind the ptery-
goid flanges the pterygoid and parasphenoid form a narrow
girder, the quadrate rami of the pterygoids taking off from it
very far back and not reaching the quadrates. The lower sur-
face of the basisphenoid forms a gently concave triangular area
recalling that of cynodonts. The lower jaw is deep anteriorly,
with a pronounced chin, and the dentary extends upward, free
from the surangular, for a considerable distance.

\'ery recognizable forms.

ArctognafJnis curvimola Owen 1876, Cistecephalus zone; Arc-
toytwthus ivhaitsi Ilaughton 1924, M. Cistecephalus zone; Lycae-
nodontoides haihijvhinus Ilaughton 1929, Cistecephalus zone.

Family Arctognaflioididae. Large gorgonopsians, character-
ised by a broad face which is about half the total skull length.
The orbit is placed high up with a deep jugal below it. The
temporal fossae are of moderate size and the occiput stretches
forward between the fossae to their anterior ends. The pineal
foramen, when present, lies just behind the parieto-frontal su-
ture. There is usually no preparietal. Upper dentition 4-5.1.4-6.
There are patches of teeth on the pterygoid and palatine.

"Aelurognathus" haughtoni Huene 1950, Lower Cistecephalus
zone, E. Af r. ; Arctognatiwides hreviceps Boonstra 1934, Cistece-
phalus zone; Leontocephalus cadlei Broom 1940, V. Cistecephalus

54 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

zone; Leontosaurus vanderhorsti Broom and George 1950, Cis-
tecephalus zone.

Family Broomisauridae. Gorgonopsians with a relatively
short, -wide, and high face with a rounded upper surface. The
continuous tooth-bearing border of the upper jaw is bowed
dowuAvard, there being scarcely any trace of a step. The frontal
enters the orbital margin by a narrow, parallel-sided process.
The temporal fossa is short and wide, the temporal roof forming
a third, or little more, of the extreme width. The occipital con-
dyle lies far back.

Broomisaurus planiceps Broom 1913, Tapinocephalus or Ciste-
cephalus zone ; Broomisaurus ruhidgei Broom 1940, Cistecephalus
zone; " Lycaenops" pricei Broom and Robinson 1948, L. Ciste-
cephalus zone.

IFamily Burnetiidae {Burnetiamorpha). Small theriodonts
in which the roof of the skull is thickened, shows in general no
sutures, and is composed of fine cancellar bone, with essentially
no superficial table of hard bone. There is a wide temporal roof,
with laterally placed temporal fossae, presumably of gorgonop-
sian pattern. The palate lacks suborbital vacuities, has teeth
on the pterj^goids and palatines, and apparently a median
groove ; there may be 11 post-canine maxillary teeth.

The two animals reported to this family differ greatly in age.
They may have been produced by parallel evolution from differ-
ent parent stocks.

Burnetia miraMlis Broom 1923, U. Cistecephalus zone ; Styra-
cocephalus platyrhynchus Haughton 1929, U. Tapinocephalus
zone.

Family Cynariopsidae. Gorgonopsians with a face which is
never greatly elongated, rounded on its upper surface and with
rather short temporal fossae. The most characteristic feature is
the smooth curve made by the tooth-bearing margins of the pre-
maxilla and maxilla which bring the anterior incisor so high up
that the nose aljove it is extraordinarily shallow. Dentition 5.1.5
or thereabouts. There is a marked groove in the palate behind the
posterior nares.

Cynarioides gracilis Broom, 1930, U. Endothiodon zone;
Cynarioides laticeps Broom 1935, ?L. Cistecephalus zone ; Cyiia-

WATSON AND EOMER : THERAPSID CLASSIFICATION 55

rioides teyuns Broom 1925, V. Endotliiodon zone; Cynariops ro-
hustus Broom 1925, U. Endothiodon zone.

Family Galerhinidae. Small gorgonopsians with a rather flat
skull in which the temporal fossa is markedly short, the occiput
extends far forward, and the occipital condyle is the most pos-
terior part of the whole skull. The family may be distinguished
by the lack of a special, narrow, parallel-sided, orbital process
of the frontal from some Tapinocephalus zone forms with other-
wise comparable features. The zygomatic arch is noticeably
shallow. The dentition is ?.1.5 in the upper jaw.

Galerhinus ruMdgei Broom 1936, L.-M. Cistecephalus zone ;
Galerhinus ruhidgei of Huene 1950, East Africa, ?Cisteeephalus
zone; Galerhinns polyodon Broom 1935, ?Cistecephalus zone.

Family Galesuchidae. Gorgonopsians in which the frontal
enters the orbital margin as a more or less parallel-sided process
between the prefrontal and very wide postfrontal. The temporal
fossa is large and is not carried back as far as the level of the
occipital condyle.

In Galesiichus the basicranial region forms a broad expansion
ventral to the condyle, unknown in other gorgonopsian groups.
In Eoarctops, which may belong here, there are 4 or 5 upper
incisors, a canine and 3 molars arising from a continuous jaw
margin which seems to show no sign of a step ; in the lower jaw
the coronoid process of the dentary projects above the surangu-
lar, the hinder part of the jaw being unusually deep.

Cerdodon tenuidens Broom 1915, Tapinocephalus zone;
Eoarctops vanderlyli Haughton 1929, Tapinocephalus zone;
Galesuchus gracilis Haughton 1915, Tapinocephalus zone; Scyla-
cognathus major Broom 1935, L. Cistecephalus zone ; Scylaco-
gnathus parvus Broom 1913, Tapinocephalus zone.

Family Gorgonognafhidac. Large gorgonopsids, with a very
long face exceptionally wide at the premaxilla-maxilla suture.
The orbit is small (because of the animal's size), the temporal
fossa rather small, Avith a very deep zygoma below it, the ventral
border of the skull having a deep rounded notch below the orbit.
The frontal reaches the orbit by a narrow process, as in Dimetro-
don. Distinguished from Rubidgeidae by little except the en-
trance of the frontal into the orbit.

Gorgonognaihus longifrons Haughton 1915, Cistecephalus

56 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

zone; Gorgonognathus maximus Huene 1950, E. Africa, L. Ciste-
cephlaus zone; Gorgonorhinus luckhoffi Broom 1937, ?L. Ciste-
cephalus zone; Gorgonorhinus minor Broom 1948, ?L. Cistece-
phalus zone ; Tigricephalus kingwilli Broom 1948, Cistecephalus
zone.

Family Gorgonopsidae. Gorgonopsids in which the intertem-
poral region is comparatively wide and the zygomatic arches
spread posteriorly. The snout is long and heavy, being both wide
and high and the section across it a little in front of the eyes is
noticeably square-cut. The squamosals are considerably laterally
extended at the sides of the occiput and are shallow. The denti-
tion is 5.1.1-5. A very marked step in the maxilla, raising the
incisors above the canines and cheek teeth. A small sharply de-
fined area of the pterygoid, near the midline, and on the level of
the transverse flanges, bears small, sharp-pointed teeth. The
basis eranii is very shallow, so that the condyle and paroccipital
processes have very little height, and the posttemporal fossae lie
very near the lower border of the occiput.

Gorgonops forvia^ Owen 1876, Endothiodon zone; Leptofra-
chelus eupachygnafhus Watson 1912, Endothiodon zone; "Lycae-
nops" kingoriensis Huene 1950, E. Africa, L. Cistecephalus zone.

Family Hipposauridae. Gorgonopsians of small to middle size
which are characterized by possessing a markedly curved dorsal
contour; a very wide parietal region w^ith much excavated occi-
put. The interorbital width is great and the orbits, always large,
may become relatively enormous. The quadrate region is deep,
produced forward as well as downward, and the quadrate itself
projects very far below the ventral border of the squamosal
and of the ridge made by the tabular. The articular border of
the quadrate lies far below the basis eranii, the stapes passing
as much downward as outward.

Hipposauroides ruhidgei Broom 1 940, Cistecephalus zone ;
Hipposaurus hoonstrad Haughton 1929, U. Tapinocephalus zone ;
Hipposaurns major Boonstra 1952, Tapinocephalus zone; Ictido-
rhiv})s warfiiisi Broom 1913, U. (Usteeephalus zone; LeDiiirosaii-
rus pricei Broom 1949, L. Cistecephalus zone ; Pseud ohipposaur us
kifcliingi Broom 1948, ?Endothiodon zone.

Family Inostrancfviidae. Gorgonopsians which reach a very
large size. They have a very wide parietal region, with relatively

WATSON AND ROMER : THERAPSID CLASSIFICATION 57

small temporal fossae. The occiput is very wide, stands nearly
vertical, and the pineal foramen is immediately in front of its
dorsal border. The paroccipital processes are deep. The face
is long, its tooth-bearing border without a noticeable step. The
dentary has a well marked "chin." The reflected lamina may
bear a deep pocket on its lateral surface.

Distinguished from Rubidgeidae by lacking the extreme
localised widening of the temporal region in later members of
that family and the nearly circular widely separated temporal
fossae.

Amalitzkia and Inostrancevia may be distinct genera.

Amalitzkia annae Pravoslavlev 1927, U, Permian, Russia;
Amalitzkia wladimiri Pravoslavlev 1927, U. Permian, Russia ;
Inostrancevia alexandri Pravoslavlev 1927, U. Permian, Russia ;
Inostrancevia latifrons Pravoslavlev 1927, U. Permian, Russia ;
Inostrancevia parva Pravoslavlev 1927, U. Permian, Russia;
Inostrancevia procJivis Pravoslavlev 1927, U. Permian, Russia.

Family Pachyrhinidae. Large gorgonopsids, in which the face
is about as high as widp ; the orbits small (no doubt owing to the
overall size of the animal) ; the temporal fossae are relatively
large, the bony roof which separates them being relatively nar-
row. The occipital condyle lies far in front of the posterior
borders of the squamosals. The frontal reaches the orbit by a
definite process between the pre- and large postfrontals. The
dentition is 5.1.4 and there is no sign of a step. There is a small
median groove behind the posterior nares.

Differs from Galesuchidae in the anterior position of occipital
condyle. This may be largely a size effect. The frontal shape
seems to be common in the Tapinocephalus zone.

Pachyrhinos kaiseri Broili and Schroeder 1934, U. Tapinoce-
phalus zone.

Family Phthinosuchidae. The number of teeth is higher than
in typical Gorgonopsia. The lower jaw .shallow. The step in the
dentition is little developed. The exposure of the squamosal
lateral to the occipital surface is relatively very small. There is
no preparietal? The pineal foramen is as posterior as possible.

Phthinosaiirus Efremov 1938, Russian zone II (or III?);
Phthinosnchus Efremov 1954, Russian zone I.

Family Rubidgeidae. A family of gorgonopsians, the mem-

58 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

bers of which attain very large size. The skull has a relatively
large and narrow snout; the upper dentition being composed of
a series of very powerful incisors, usually 5 in number, a canine,
and a variable number of cheek teeth, ranging from 5 to 0. The
maxilla has a rounded tooth-bearing border and is very deep.
With increase in size, the whole temporal region increases in
width ultimately to a very great extent, and this increase in
general width goes with a corresponding and disproportionately
enlarged broadening of the intertemporal region. The occipital
region inclines forward and in large individuals reaches the
pineal foramen, which lies immediately in front of the temporal
openings. The occiput is necessarily very wide and the quadrates
and squamosals do not seem to extend ventrally to the level of
the occipital condyle. The palate lacks a suborbital vacuity and
is of standard gorgonopsian pattern. A feature which may be
rather characteristic is that the immensely deep anterior end of
the mandible sweeps into the lower border of the jaw. The re-
flected lamina of the angular often bears a deep ridge, behind
which are one, or two, deep pockets.

Dinogorgon {Broomicephalus) laficeps Brink and Kitching
1953, Cistecephalus zone; Dinogorgon ondehergensis Brink and
Kitching 1953, Cistecephalus zone ; Dinogorgon quinquemolaris
Huene 1950, Cistecephalus zone, E. Africa; Dinogorgon ruhidgei
Broom 1936, M. Cistecephalus zone ; Dinogorgon (Dracocephalus)
scheepersi Brink and Kitching 1953, Cistecephalus zone; Pro-
ruhidgea maccahei (pugnax) Broom 1940, ? Cistecephalus zone;
Prornhidgea rohusta Brink and Kitching 1953, Cistecephalus
zone ; Buhidgea atrox Broom 1938, Cistecephalus zone ; Ruiidgea
hitchingi Broom 1938, Cistecephalus zone; Buhidgea laticeps
Broom 1940, II. Cistecephalus zone ; Buhidgea majora Brink and
Kitching 1953, Cistecephalus zone ; Buhidgea platyrhina Brink
and Kitching 1953, Cistecephalus zone; fSmilesaurus ferox
Broom 1948, Endothiodou zone; ?S!}ni]esanrus maccahei Broom
1948, Endothiodon zone; fTangagorgon tenuirostris Boonstra
1953, E. Africa, Lower Bone Bed ; Tigrisaurus pricei Broom and
George 1950, fCistecephalus zone.

Family Scylacocephalidae. Small gorgonopsians characterised
by a long temporal fossa, the temporal roof being about twice as
wide as the interorbital width. The snout is less than half the

WATSON AND ROMER : THERAPSID CLASSIFICATION 59

total skull length, and is high and narrow. The frontal enters
widely into the orbit. The lower jaw has a raised symphysial
region, the canine socket being above the line of insertion of the
cheek teeth. The hinder part of the jaw is unusually long and
the reflected lamina deep and far forward.

Scylacocephalus wafermeyeri Broom 1940, Cistecephalus zone.

Family Scylacopsidae. Small gorgonopsians in which the face
forms about half the skull length. The face is rounded in cross
section and widens gently, passing smoothly into the orbital re-
gion and so to the widest point of the skull about half way along
the zygomatic arch.

The interorbital and intertemporal regions are of nearly equal
width. The postorbital bones are of some width and pass down
to rest on the zygoma by a widened arc. The temporal fossa is
longer than is usual in gorgonopsids.

The occiput slopes forward; the squamosals are turned back-
ward at their temporal border. There is a basioccipital with
large tubera which join with those from the basisphenoid to
support the cartilage surrounding the fenestra ovalis. The
basisphenoidal tubera are separated by a pit but anteriorly are
confluent and continued by a deep narrow ridge nearly to a very
small interpterygoid vacuity.

Dental formula 4-5. 1. 4 (-7?).

Cyniscopoides hroomi Brink and Kitching 1953, Cistecephalus
zone; fCyniscops hroomtanus Huene 1950, Cistecephalus zone,
E. Afr. ; fCyniscops cookei Broom 1948, Cistecephalus zone;
fCyniscops kitching i Broom 1948, Cistecephalus zone ; fCyniscops
longiceps Broom 1941, Cistecephalus zone; fCyniscops ruhldgei
Broom 1937, M. Cistec'e])halus zone; Galerhynchus ruhldgei
Broom 1937, Cistecephalus zone; " Gorgonopsid " Watson 1913,
Cistecephalus zone; Scylacops capensis Broom 1913, Cistece-
phalus zone ; Sycocephalus })igendens Brink and Kitching 1953,
L. Cistecephalus zone.

Family Scymnognaihidae. Gorgonopsids with a comparatively
wide intertemporal region, and spreading zygomata. The snout
is relatively long, square-cut in transverse section, and there is
a marked step from the premaxilla to the maxillary tooth-bearing
border. The paroccipital process is robust, deep below the post-
temporal fossa, and the condyle and floor of the braincase thick.

60 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Arcfops willistoni Watson 1914, ?Endothiodon zone ; Chiweta-
saurus dixeyi Haughton 1926, Cistecephalus or V. Endothiodon
zone, E. Africa; Dixeya nasuta Huene 1950, Cistecephalus zone,
E. Africa; Dixeya quadrata Haughton 1926, Cistecephalus zone,
E. Africa; Lycaenoides anyusticeps (minor) Broom lf)13, Ciste-
cephalus zone ; Scymnognathus hohnesi Broom 1948, Endothiodon
zone ; Scymnognathus major Olson and Broom 1937, Endothio-
don zone; Scymnognathus parringtoni Huene 1950, Endothiodon
zone, E. Africa; Scymnognathus whaitsi Broom 1912, Endothio-
don zone.

Family Sycosauridae. Gorgonopsians of medium size, with a
face as long as the rest of the skull and rather heavily built,
with a semicircular anterior end. Nostril very far above lower
border of premaxilla. The postorbital bar is broad (cf. Rubid-
geidae), the zygoma rather deep, the temporal fossa very short.
The occiput is steeply inclined and makes a large bay with dorsal
surface. The frontal is excluded from the orbital margin by a
meeting of the pre- and postfrontals. There is no preparietal.
Dentition 5. 1. 5.

Sycosaurus hrodiei Broom 1941, Cistecephalus zone; Syco-
saurus laticeps Haughton 1924, ?L. Cistecephalus zone.

Gorgonopsians represented hy adequate skulls hut lacking
definite family characteristics: — Aelurognathus microdon Boon-
stra 1934, Cistecephalus zone; Aelvrognathus nyasaensis Haugh-
ton 1926, Cistecephalus zone ; Aelurognathus tigriceps Broom and
Haughton 1913, Cistecephalus zone; "Aelurognathus" minor
Brink and Kitching 1953, Cistecephalus zone ; Alopecorhynchus
rubidgei Brink and Kitching 1953, Cistecephalus zone ; " Arctops ' '
watsoni Brink and Kitching 1953, Cistecei)halus zone ; Cerdor-
hinus parvidens Broom 1936, U. Tapinocephalus zone ; Clelandina
major Broom 1948, Endothiodon zone; Clelandina rubidgei
Broom 1948, U. Endothiodon or L. Cistecephalus zone ; Cyono-
saurus longiceps Olson 1937, L. Cistecephalus zone; Lycaenops
ornatus Broom 1925, U. Endothiodon zone; '' Lycaenops" alt i-
ceps Brink and Kitching 1953, Cistecephalus zone ; Nanogor-
gon gracilis Broom and Robinson 1948, ?Cistecephalus zone ;
Sauroctonus progressus Hartmann-Weinberg 1938, Cistecephalus
zone, Russia.

WATSON ANT) ROMER : THERAPSID CLASSIFICATION 61

Gorgonopsians represented hy material inadequate for proper
assignment : — " Aelurosaurus" brevirostris Broom 1948, Endothi-
odon zone; Aloposauroides tenuis Brink and Kitching 1953, Cis-
tecephalus zone ; Aloposaurus gracilis Broom 1910, U. Endothio-
don zone; Arctusuchus tigrinus Owen 1876, ?Endothiodon zone;
Ccrdognathus greyi Broom 1915, fCistecephalus zone; Cgna-
rioides grimbeeki Broom 1935, U. Endothiodon zone; Cy7iiscodo7i
lydekkeri Broom 1915, Tapinoeephalus zone; Cynodraco major
Owen 1876, h. Endothiodon zone; Cynodraco serridens Owen
1876, fEndothiodon zone; Delpliaciognathus {AstJienognathus)
paucidens Broom 1915, Cisteeephalus zone; Eriphostoma micro-
do)) Broom 1911, ■Tapinoeeplialus zone; Genovum hroilii Huene
1950, Cisteeephalus zone, E. Africa; Lycaenodon longiceps
Broom 1925, L. Cisteeephalus zone ; Lycosaurus pardalis Owen
1876, ? Cisteeephalus zone; Pardocephalus waUacei Broom 1948,
Cisteeephalus zone; " Scy)}})w</nafJiHs" parvus Broom 1915, U.
Endothiodon zone; Tetraodon nowaki Broili and Schroeder 1936,
Endothiodon zone, E. Africa; Tigrisuchus simus Owen 1876.
Cisteeephalus zone; "unnamed gorgonopsian" Haughton 1924,
M. Cisteeephalus zone.

Infraorder Cynodontia

Advanced theriodonts, which possess a secondary palate but
have no suborbital vacuities. There is a relatively narrow (or
very narrow) sagittal crest formed almost entirely by the pari-
etals, the postorbitals making only an intinitesimal addition to it.
A parietal foramen is present. Postorbital bar always present.
The temporal region is much expanded, with the greatest width
posteriorly. The temporal opening faces more dorsally than
laterally. The nasals are expanded posteriorly. The postorbitals
do not meet the squamosals above the temporal fenestrae. The
epipterygoid, widening dorsally, rises from the basipterygoid
process to a long contact with the parietal. The posterior part of
the basicranium has suffered so great a reduction of the basis-
phenoid tubera that they are no more than sharp-edged ridges
bounding a triangular surface which extends more and more

62 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

posteriorly as time goes on. The occipital condyle is double or
incipiently double. The posterior maxillary teeth have usually
advanced beyond the stage of simple cones. The dentary is very
large, posterior elements are much reduced, the reflected lamina
of the angular usually reduced to a narrow process. There is no
cleithrum ; the anterior edge of the scapula is outturned, ending
below in an acromial process. Coracoid elements are small. The
ilium is greatly expanded. Generally a large obturator fenestra.
Except in one primitive family the lumbar ribs are fused to the
centra, and are shortened and interlocking. The phalangeal
formula is always 2.3.4.5.3(4), but one intermediate phalanx
in digit III and two such phalanges in digit IV are reduced to
thin discs.

Cynodonts are present and not uncommon in the Cistece-
phalus, Lystrosaurus and Cynognathus zones of the Karroo
series ; these forms are here arranged in five families, those from
the two lower zones being the more primitive. In addition,
cynodonts are present in the late Permian of East Africa and
northern Russia, and the Middle Triassic of South America and
East Africa. A number of poorly known late Triassic therapsids
may be late survivors of the group.

Family ProcynosucJiidae. Cynodonts in which the secondary
plates of the maxillae and palatines do not always meet in the
midline of the palate. The vagal foramen is relatively high up
on the posterior surface of the skull. The occipital condyle is
incipiently double in most forms. The qviadrate margin lies
very ventrally and the whole height of the quadratojugal is
freely exposed in side and posterior views below and lateral to
the posterior plate of the squamosal. In the lower jaw the hinder
end of the dentary does not rise as a free process overhanging
the surangular from their point of contact. Precanines present.
Maxillary teeth normally all crenulated or tricuspidate and may
be further elaborated.

Members of this primitive family are confined to the Cistece-
phalus zone of South Africa and presumably equivalent horizons
in East Africa and Russia. " Nanicfosaurus" robitsfus is prob-
ably a species of "NanictosucJius."

Dvinia Amalitzky 1922, U. Permian, N. Russia ; Galeophrys
Broom 1948 (Galecraniiim Broom 1948), Cisteeephalus zone ; Lea-

WATSON AND ROMER : THERAPSID CLASSIFICATION 63

vachia Broom 1948 (Aleurodraco Broom and Robinson 1948),
Cistecephalus zone ; "Nanictosuchus" Broom 1940, Cistecephaliis
zone; Paracynosuchus Broom 1940, Cistecephalus zone; Para-
thrinaxodon Parrington 1936, tCistecephalus zone, E. Africa;
Permocynodon Sushkin 1929, IT. Permian, N. Russia ; Pro-
cynosuchus Broom 1937, Cistecephalus zone, E. Africa.

Family Thrinaxodontidae (Galesaurldae) . In this family and
those which follow the secondary palate is complete and no
precanine teeth are present. The vagal foramen is almost on the
plane of the ventral surface of the basisphenoid. The occipital
condyle is essentially double. The quadratojugal is not widely
exposed from the side, its outer surface being covered by a
ventral extension of the squamosal so that the bone is received
in a special notch and is visible from behind. The dentary has a
free coronoid process rising above the dorsal border of the sur-
angular and directed backward. The posterior maxillary teeth
are essentially conical, with minor cusps arising from the an-
terior and posterior borders of the central cone, these cusps not
exceeding four in number. The teeth are in no way widened.

More advanced than the procynosuchids, this family is char-
acteristic of the upper part of the Cistecephalus zone and the
Lystrosaurus zone; Sysphinctostoma is a surviving Cynognathus
zone form. Some material previously assigned to Nythosaurus
larvatus and to Galesaurus pertains to Thrinaxodon.

fBaurocynodon Brink 1951, Cistecephalus zone; Cynosaurus
Schmidt 1927, {Cyno.suchoides Broom 1931, Cynosuchtis Owen
1876), Cistecephalus zone; Galesaurus Owen 1859 {Glochinodon
Hoepen 1916, Glochinodontoides Haughton 1924), Lystrosaurus
zone ; Micrictodon Broom 1937, Lystrosaurus zone ; Nanicto-
saurus Broom 1936, IT. Cistecephalus zone; Notictosauriis Broom
1936, base of Lystrosaurus zone; Nythosaurus Owen 1876, Lystro-
saurus zone ; Plaiycraniellns Iloepen 1917 {Platycranion Hoepen
1916), Lystrosaurus zone; Sysphinctostoma Broili and Schroeder
1936, Cynognathus zone; Thrinaxodon Seeley 1894 {Ictidopsis
Broom 1912), Lystrosaurus zone.

Family Cynognathidae. Dentition carnivorous in type, the
post-canine maxillary teeth apparently divided into "premolars"
with simple crenulate crowns and "molars" with a series of
cusps arranged nearly symmetrically about an enlarged central

64 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

member in a single anteroposterior row. The zygomatic arch is
deep and broadly continuous posteriorly with the side of the
braincase ; its upper border not extending up to the level of the
sagittal crest, and with a groove for the external auditory meatus
some distance below its upper Ijorder. This family and the
theriotlont groups which follow (including the Ictidosauria) are
distinguished from all other reptiles by the fact that the angle
of the dentary is carried downwards as a free standing structure
below the level of the ventral border of the hinder part of the
jaw.

Of the common cynodonts of the Cynognathus zone with un-
expanded "molars'" most appear to belong to the genus Cyno-
gnathus, and supposed generic differences are mainly based on
minor differences in the degree of development or stage of re-
placement of post-canine maxillary teeth; Cynidiognathus is
distinguished by its short epipterygoid, which does not reach the
quadrate. Triholodon is, however, a very distinct form. Karoomys
is perhaps the jaw of. a young and indeterminate cynognathid.
Chiniquodon and probably Belesodon of the Brazilian Middle
Triassic appear to be late survivors of this family.

fBelesodon Huene 1936, M. Trias., S. Brazil; Chiniquodon
Huene 1936, M. Trias., S. Brazil; Cistecyuodon Brink and Kitch-
ing 1953, Cynognathus zone; Cynidiognathus Haughton 1922,
CViiognathus zone; Cynognathus Seeley 1895 {Cynogomphius
Broom 1932, Lycaenognathus Broom 1925, Lycochampsa Broom
1915, Lycognathus Broom 1913), Cynognathus zone; Karoomys
Broom 1903, Cynognathus zone; " Nythosaurus" hrowni Broom
1912, Cynognathus zone ; Triholodon Seeley 1894, Cynognathus
zone.

Family Diademodontidae. The post-canine maxillary dentition
consists of a series of simple peglike "premolars," followed by
a series of "molars" with transversely widened crowns, with a
central ridge which in effect connects an outer and an inner
cusp. The deep zygomatic arch is connected with the braincase
through a shallow root. Its upper border lies on the level of
the sagittal crest and descends abruptly behind, the whole being
separated from the braincase, in posterior view, by a V-shaped
notch. The external auditory meatus is deeply impressed; it is
overhung by the outwardly rolled superior border of the zygoma.

WATSON AND ROMER : THERAPSID CLASSIFICATION 65

The common ^enus of cynodont with expanded molars iu the
Cynognathus beds is that known alternatively as Diademodon or
Govipiwgnathus; since the former has page priority, it is prefer-
ably used. C ydogomphodon and Octagomphus are not generically
separable from Diademodon; supposed generic and specific dif-
ferences are largely l)ased on dental differences which have to
do with the state of wear or replacement of the 4 to 5 small
"premolars," and variations, apparently in part ontogenetic,
in the three posterior members of the molar series (which may
run to a high count of ten). Frotacmon is close to Diademodon
but it differs in that the epipterygoid does not reach the quadrate.
Trirachodon is a quite distinct tjqje ; Tr'irachodontoides is close
to the last but distinct. "Trirachodon" browni is also distinct
from that genus and merits a generic name. Theropsodon and
Traversodon appear to be late surviving diademodontids.

Diademodon Seeley 1895 {Cyclogompliodon Broom 1919,
fCynochampsa Owen 1860, Diasteniodon Seeley 19U7, Gonipho-
gnathns Seeley 1895, Microhelodon Broom 1931, Octagomphus
Broom 1919), Cynognathus zone; Gomphodontoides Brink and
Kitching 1951, Cynognathus zone; Inusitatodon Brink and
Kitching 1953, Cynognathus zone; Protacmon Watson 1920,
Cynognathus zone; Theropsodon Huene 1950, M. Trias., E. Af-
rica; Traversodon Huene 1936, M. Trias., S. Brazil, N. Argen-
tina; Trirachodon Seeley 1895, Cynognathus zone; "Triracho-
don" browni Broom 1915, Cynognathus zone; Trirachodontoides
Broom 1932, Cynognathus zone.

Family Gomphodont osuchidae . This family is founded for the
anterior part of a skull and lower jaw, which show a very short
and extremeh' massive face, with relatively small incisors and
canines and an entirely irregular series of post-canine teeth,
about 6 in number, markedly angular in transverse section and
with their grinding surfaces worn into a concavity.

Gomphodoniosuchus Huene 1928. M. Trias,, S. Brazil.

Cynodontia ( ?) incerfae sedis.

The following are known from teeth, jaw fragments, jaws and
other fragmentary remains from the Middle and Upper Triassic.
^Yhile most, at least, appear to be late 'surviving cynodonts, it
is not feasible to assign them systematically in the light of pres-
ent knowledge.

66 BULLETIN: MUSEUM OP COMPARATIVE ZOOLOGY

Dromatherium Emmons 1857, U. Trias. NA. ; Exaeretodon Ca-
brera 1943, M. Trias., N. Argentina ; Kunminia Young 1947, IT.
Trias., E. Asia; Lycorhinus Haughton 1924, Redbeds, S. Africa;
Microconodon Osborn 1886 {Tytthoconua Pabner 1903), U. Trias.,
NA.; Pachygenelus Watson lf(13, Redbeds, S. Africa, ( ?Cyno-
gnatliidae) ; Theropsis Cabrera 1943, M. Trias., N. Argentina, S.
Brazil; Tricuspes E. Hiiene 1933, Rhaetic, Germany.

Infraobder Ictidosauria

Advanced therapsid reptiles with a secondary palate, no
suborbital vacuities. The root of the zygomatic arch arises from
the side of the maxilla well dorsally to and in advance
of the last maxillary teeth. Tooth rows parallel or sub-paral-
lel. No postorbital bar ; no pre- or postfrontal or postorbital. The
ascending ramus of the dentary arises far forward, lateral to
and in front of the posterior cheek teeth. In contrast to cyno-
donts, the jaw articulation is far above the line of the teeth. The
posterior elements of the lower jaw more reduced than in cyno-
donts.

This group was erected by Broom for the reception, primarily,
of two small skulls, still unnamed, one with a partial skeleton,
from the late Triassic Cave Sandstone of South Africa, which
were obviously of a type advanced beyond the stage of any
typical theriodont. With these specimens he later associated
several other fragmentary remains, most of which, however,
appear to be on a lower, cynodont level. Recent work indicates
that the Tritylodontidae, long thought to be mammals, are ad-
vanced reptiles with a jaw construction of ictidosaurian type.

Family {unnamed) . Short-snouted ictidosaurians, apparently
without diastema. Maxillary teeth but little transversely wid-
ened, and incipiently bicuspidate transversely.

Including the "type" ictidosaurian material described but not
named by Broom (1932) ; this material is at present being
described by Mr. A. W. Crompton.

Family Trithelcdoiitidae. Ictidosaurians with a long diastema
and maxillary teeth placed very far posteriorly; cheek teeth
numerous, transversely widened and bearing a single row of
sharp-pointed cusps.

WATSON AND ROMER : THERAPSID CLASSIFICATION 67

Tritheledon Broom 1912, Redbeds.

Family Tritylodontidae. Ictidosaurians with a narrow muzzle,
on the lower surface of which is a very long diastema. One
incisor both above and below, the lower procumbent. Cheek teeth
quadrilateral in plan ; the upper ones with three rows of wedge-
shaped cusps, the lowers with two similar rows which bite into
the grooves between the upper rows. Propalinal movement of the
jaw results in longitudinal wear striae on the sides of the cusps
and eventually in their complete flattening. The parasphenoid
and basisphenoid form processes extending far below the level
of the base of the braincase.

Our knowledge of these forms was long confined essentially
to the fragmentary skull of Trifylodo7i from South Africa ; re-
cent discovery of a Chinese form described by Young and of
numerous remains of Oligokyphus by Kiihne indicates that we
are dealing, not with multituberculate mammals as once thought,
but with a somewhat parallel development from the ictidosaurian
level of therapsid evolution. Stereognathus is the latest known
theriodont.

Bienotherium Young 1940, U. Trias. Redbeds, E. Asia; Chale-
potherium Simpson 1928, Rhaeto-Lias, Europe; Oligokyphus
Hennig 1922 {Mucrotherium E. Huene 1933, Vniserium E.
Huene 1933), Rhaeto-Lias, Wiirttemburg and England; Stereo-
gnathus Charlesworth 1855, Stonesfield Slate, England ; TrityJo-
do7i Owen 1884 ( fTriglyphvs Fraas 1866) , Redbeds, South Africa,
and ?Rhaeto-Lias, Wiirttemburg.

fFamily Microcleptidae (Microlestidae) . Known only from
molar teeth, which are oval, with a rounded basin bordered by
cuspidate ridges.

These poorly known forms may be either advanced therapsids
or archaic mammals, and are noted here merely for the sake of
completeness (cf. Simpson).

Microcleptes Simpson 1928 ( fHypsiprymnopsis Dawkins
1864), Rhaeto-Lias, England; Thomasia Poche 1908 {Micro-
lestes Plieninger 1847, PJieningeria Krausso 1919), Rhaeto-Lias,
Wiirttemburg and England.

Possible Mammals from the "Rhaetic." Archaeodon Huene
1925 {Archaeotherium Huene 1925), S. W. Africa; Eozostrodon
Parrington 1941, England ; Morganucodon Kiihne 1949, England.

68 bulletin : museum of comparative zoology

Infbaorder Therocephalia

Theriodonts with a massively built skull in vt^hich the parietal
forms a large part of the border of the temporal fossa, separating
postorbital and squamosal by a large distance. Zygomatic and
postorbital arches usually broad. There is no preparietal. The
postfrontal usually present. The quadrate and quadratojugal
are reduced and supported by the squamosal, as in gorgonopsids.
The temporal region little expanded. Palatal plate long, basi-
cranium short. Vomers paired ; lower canines fit into pits in
choanae. Usually a slit-like interpterygoid vacuity. The mid-
ventral line of the basisphenoid-parasphenoid complex behind
the interpterygoid vacuity usually bears a very deep narrow
midventral plate. There is no secondary palate but in some more
advanced forms the central palatal strip is depressed so as to
provide conditions which ultimately lead (cf. whaitsiids) to an
arrangement which secures a comparable effect. The palate does
not usually bear teeth. Suborbital vacuities usually present.
The quadrate ramus of the pterygoid extends directly poste-
rolaterally toward the quadrate, sending in a triangular flange
to articulate with the basicranium. Occiput generally low and
broad ; condyle single. The jaw massive, the dentary deep. A
coronoid process of the dentary present, rising well above the
level of the surangular. Dentition strongly heterodont, with
one or two large canines sharply set off from the adjacent teeth.
Up to 7 incisors ; usually no precanine maxillary teeth ; up to 9
simple cheek teeth. A cleithrum primitively present. Little acro-
mial development ; scapular blade flat. Obturator fenestra little
developed. No heel on calcaneum. The phalangeal formula is
2.3.3.3.3.

We include here a series of relatively primitive Permian
theriodonts, generally of large size and massive build, the Pris-
terognathidae and Trochosuchidae, and their specialised de-
scendants such as the Whaitsiidae and Euchambersiidae. More
progressive forms, generally classed as therocephalians, are
here brigaded with Baurio and its relatives in the Bauriamorpha.

Family Pristerogvaihidae. Large therocephalians with a nar-
row parietal crest, usually elevated ; a parietal foramen situated
well forward. A postfrontal, generally f«airly large. The face is
heavy and rounded in transverse section. The temporal fossae

WATSON AND ROMER : THERAPSID CLASSIFICATION 69

are large, the quadrates unexpectedly small. The palate is
essentially flat. There are large suborbital vacuities. The occiput
is commonly deep, triangular in general plan and with a dermo-
supraoccipital which does not reach the summit of the back of
the skull. The dentition usually includes a large number of
incisors, up to 7 ; there is a single canine ; the number of molars
may reach 9 but may be as few as 3. There may be a small
interpterygoid vacuity between the transverse flanges and some-
times another at the point where the deepened median para-
sphenoidal ridge separates the borders of the pterygoid. The
lower border of the dentary is usually straight, rounding off to
its anterior extremity. The two dentaries are usually not fused.
These are the characteristic therocephalians of the Tapinoce-
phalus zone. Pristerognathiis, Scymnosaurus and Scylacosaurvs
are known from good skulls, several others from skulls less com-
plete ; a number of genera are based on snout fragments, recog-
nisable as belonging to this family but for the most part other-
»vise indeterminate.

Alopecideops Broom 1932; Alopecodon Broom 1908; Alope-
i'ognathm Broom 1915; Alopecorhinus Broom 1912; Gynario-
(jvathus Broom 1931: GJanosuchns Broom 1904; fHyorhynrluix
Seeley 1889; Irficloparia Broom 1925; IcfirJosaiinis Broom 1903;
Lycedops Broom 1935; Maraisaurus Boonstra 1953 ; PardosHchus
Broom 1908; Pristerognathoides Boonstra 1954; Prisferognathus
Seeley 1895 ; Pristerosaurus Boonstra 1954 ; Ptonialestes Boon-
stra 1954; Scylacoides Broom 1915; Scylacorhinus Broom 1915;
Scylacosaurus Broom 1903 ; Scymnosaurus Broom 1903 ; The-
rioides Boonstra 1953 ; Walteria Brink and Kitching 1951 ; all
from the Tapinoeephalus zone.

Family Trochosuchidar. Large therocephalians which re-
semble the Pristerognathidae in fundamental features of their
structure but differ in having a much lower skull, Avith a broad
and rather flattened snout, a sagittal crest never elevated, and the
occiput transversely widened. They may havp 6 incisors and
normally 2 canines, each separately replaced.

Hyaenasuchiis Broom 1008; Lycosiich us Broom 11)03; Tro-
chorhinus Broom 1936 ; Trochosuchus Broom 1908 {Trochosaurus
Haughton 1915) ; all from the Tapinoeephalus zone.

Family Whaitsiidae. Therocephalians usually with a broad

70 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

shallow skull and a bulbous muzzle. The palate has a raised
central area, roofed by the vomers and pterygoids. The sides of
the deep groove are bordered by the palatines, which extend far
forward, and the maxillae. In some forms special processes of
the maxillae grow inwards from the admesial surface of these
bones, in advance of the palatines, to meet the lateral margins
of the vomers and thus divide the originally very long internal
nostril into an anterior division for the reception of the lower
canine and a functional nostril posteriorly ; the anterior division
may be reduced to a pit. The upper dentition contains 5 incisors
and a large canine : there are normally no post-canine maxillary
teeth, although in some forms (e.g., Moschorhinus) 2-5 small
molars are to be seen, which are apparently early lost. As a
result the lower border of the maxilla is much narrowed, the
lateral surface of the face turning inward behind the canine
to meet the palatal surface in such a manner as to suggest that
the jaw was sheathed by a horny plate. In most forms a subor-
bital vacuity is absent ; in some it is represented by a small
opening, in others merely by a pit. The epipterygoid is widened.
There is no interpterygoid vacuity. The lower border of the
dentary forms a continuous, very well pronounced curve from
the incisor border to its contact with the angular. The posterior
part of the jaw is both large and deep ; the reflected lamina of
the angular is exceptionally well developed and sometimes fuses
with the surangular so as to separate two foramina which pass
down into the pocket which it occludes. The limbs are heavily
built.

A Cistecephalus zone group, with molar reduction the most
characteristic diagnostic feature. Hofmeyria of the Upper
Endothiodon zone is apparently a forerunner.

Alopecopsis Broom 1920. Cistecephalus zone; Hofmeiiria
Broom 193o, V. Endothiodon zone; Hyenosaurus Broom 1935,
M. or U. Cistecephalus zone ; Moschorhinus Broom 1920, U. Cis-
tecephalus and Lystrosaurus zones; Moschorhynchus Broom
1936, U. Cistecephalus zone ; Notaelurops Broom 1936, L. or M.
Cistecephalus zone, S. Africa, U. Perm., E. Africa; NotosoUosia
Broom 1925, U. Cistecephalus zone, S. Africa, U. Permian, E.
Africa; Promoschorhynchus Brink 1954. Cistecephalus zone;
Theriognathus Owen 1876, Cistecephalus zone; Whaitsia Haugh-

WATSON AND ROMER : THERAPSID CLASSIFICATION 71

ton 1918 {lAnengompJiius Broom and Robinson 1948), U. Cis-
tecephalus zone.

Family Euchamhersiidae. A monotj-pic gronp, to include only
Euckamhersia. A therocephaliau of medium size, -with a face
very broad for its length. The snout as far back as the canine
has a rounded anterior border and a normal constitution, the
septomaxilla being large and the septomaxillary foramen be-
tween it and the maxilla small. Behind the canine region the
maxilla is very short and tliere is no indication of any articulation
with a lacrimal or jugal ; instead, its outer surface, over almost
the whole of its height, is excavated into a great hemispherical
cavity which communicates over a narrow notch with the palatal
surface immediately behind the canine tooth. As this tooth,
which otherwise is circular in section, bears a narrow ridge
which arises abruptly from its outer surface, it is reasonable
to regard this great chamber, which opens within the position
of the lips, as housing a poison gland.

On the palate, the premaxillae, with indication of the presence
of some 5 incisors on each side, have a narrow exposure continu-
ing that of the anterior part of the maxillae. Nearly the whole
of the posterior surface is in contact with the indented anterior
ends of the vomers, which narrow as they pass backward to
form the inner borders of the very elongate internal nares. These
openings are widened and laterally produced anteriorly, pre-
sumably for the reception of the lower canines, but narrow
posteriorly, where they are bordered on their outer sides by the
palatines. The hinder border of the choana lies a considerable
distance behind not only the canine but the point of entry of the
poison gland duct. The vomers end posteriorly by expanding
into a relatively large surface bordered by the palatine and
pterygoid. The suborbital vacuities are reduced to foramina.
There is the customary transverse flange, not carried far down-
ward, but composed of a posterior pterygoidal and a somewhat
larger transverse component. The flange passes forward to
articulate powerfully with the posterior end of the maxilla im-
mediately Ijehind the notch for the poison gland. The condition
is such as to make it most improbable that there was any zygo-
matic arch. The pterygoid extends backwards to a contact with
the under surface of the basisphenoid. There are well-marked
basisphenoidal tubera and a single occipital condyle. A widened

72 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

epipterygoid passes upward to the parietal.

The orbit is small and lacks any lower or posterior border.
The only bone associated with it is the large prefrontal, which
extends back to end by a contact with a reduced postorbital ; the
latter passes backward in contact with the frontal and parietal
to form something of an inner wall of the temporal fossa. There
is no parietal foramen. The frontal is exceptionally small ; there
is no lacrimal, postfrontal, jugal, squamosal or quadratojugal.

Euchambersia Broom 1931, Cistecephalus zone.

Therocephalia incertae sedis

The genera listed below are therocephalians which are not
assignable to any of the families listed above, and do not show
sufficiently distinctive features to merit the erection of further
family groups for them. Most are from the Cistecephalus zone.

Anna Amalitzky 1922, U. Permian, N. Russia (cf. primitive
whaitsiids) ; Cerdops Broom 1948, Cistecephalus zone; fCerdo-
suchoides Broom 1936, ?Lystrosaurus zone; fCerdosuchus Broom
1936, Cistecephalus zone; Chthonusaurus Vjuschkov 1955, U.
JVrmian, liussia (cf. primitive whaitsiids) ; Notaelurodon Broom
li)36, Cistecephalus zone ; Proalopecopsis Brink and Kitching
1951, Cistecephalus zone; fTheriodesmus Seeley 1887, ?Ciste-
cephalus zone.

Inpraobder Bauriamorpha

Theriodonts descended from the Therocephalia and inheriting
many of tht' featui'cs of that group, but characterized by the
tendency to develop many advanced characters. The skull is
usually lightly built, the face generally long and cylindroid ; the
temporal fossa generally relatively small. The zygomatic and
postorbital arches are slender and the latter may be lost. No
postfrontal. The parietal foramen is small or absent ; the parie-
tals frequently form a flat plate. Vascular foramina are usually
present on the maxilla and may include a large forwardly di-
rected opening leading to a depression on the premaxilla. A sec-
ondary palate is present or incipient in most cases. Suborbital
vacuities large. An interpterygoid vacuity of large size whose
anterior end lies between the transverse flanges is very commonly
present. In many forms it extends backward onto the basi-
sphenoid, separating the pterygoids quite widely and sometimes

WATSON AND ROMER : THERAPSID CLASSIFICATION 73

ending abruptly at the level of a transverse suture visible on
the ventral surface of the basisphenoid. To this point the per-
fectly straight, deep, quadrate ramus is attached by a slender
flange of bone, usually with a concave ventral surface. The
posterior border of this flange forms the end of a considerable
space through Avhieh the projecting anterior part of the prootic
may be visible. The basicranial tubera tend to spread widely
and the prootic has a depression on its lateral surface whose
admedian wall has an opening into the pituitary fossa, the roof
lying horizontally and passing abruptly into the outer surface
of the bone. The dentary is long but slender, with a continuously
curved lower border which rises into a long ascending process,
the prearticular forming a large internal flange below the artic-
ular. Dentition usually less heterodont than in therocephalians;
the canines frequently reduced in prominence and. less sharply
marked off from the adjacent teeth ; there may be several pre-
canine maxillary teeth. The cheek teeth are usually small but
cuspidate and may be numerous. The maxillary tooth row is
markedly convex medially. Usually an ossified sternum and a
very long, narrow, scapula blade. A small obturator fenestra.
Calcaneum with heel; phalangeal formula 2.3.3.3.3.

We include here not only Bauria and other progressive Triassic
forms with a secondary palate, but in addition a series of re-
lated families (scaloposaurids, etc.) which are usually included
in the Therocephalia but appear to bear much the same relation-
ship to the bauriids that the early cynodonts do to the typical
Triassic members of that group.

Family Lycideoysidae. A monotypic family for Lycideops.
• A remarkable, large baurianiorph, with an elongated face and
relatively small supratemporal fossa. The premaxillary teeth
lie at a level considerably above that of the molars, there being a
marked step in the dentition. Two upper canines are present and
10 small, uniform cheek teeth. In the lower jaw there are no
incisors, but a canine and 5 minute, sparsely spaced cheek
teeth.

Lycideops Broom 1931, U. Cistecephalus zone.

f Family Ictidosuchidae. A bauriamorph with a narrow face
of considerable depth and a large and transversely wide temporal
fossa. The parietal crest is high with a very sharp upper border.

74 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

There is an upper canine, and eight molars above and below, the
anterior teeth being unknown.

Ictidosuchus Broom 1900, U. Endothiodon zone.

Family Nanictidopsidae. Bauriamorphs of medium size which
lack a complete secondary palate. The head is not as broad or as
depressed as in the Scaloposauridae. The interpterygoid vacuity
is variable. The parietals form a crest; a parietal foramen is
usually present. The postorbital arch is sometimes interrupted
and is always slender. The hasipterj'goid is widened posteriorly.
The dentition varies in count, but there are normally more than
4 and commonly 6 incisors ; there are usually 2 precanine maxil-
lary teeth, a distinct upper canine, and 6-10 cheek teeth (usually
9). All the teeth appear to be simple cones of circular section.

These Cistecephalus zone forms may represent the ancestors of
Bauria and related types and, on the other hand, appear to be
closely related to the Scaloposauridae.

Ictidosuchoides Broom 1931, Cistecephalus zone; Ictidosuchops
Broom 1938, Cistecephalus zone ; Nanictidops Broom 1940, Up-
per Cistecephalus zone; Pelictosuchus Broom 1940, ?Cistece-
phalus zone.

Family Silpholestidae. Small bauriamorphs with short tem-
poral fossa and narrow but continuous postorbital bars. The
parietal region, with a small parietal foramen, is broad and
flattened. The zygomatic arch is slender and the articular sur-
face of the quadrate not greatly depressed below it. The skull
tapers to a narrow snout. The dentition retains many teeth —
6 incisors, 2-0 precanine maxillary teeth, a canine, and 10 or
more cheek teeth. The angular is very deep and has an ornament
of radiating ridges, a point of distinction from the somewhat
similar scaloposaurids.

Ictidodraco Broom and Robinson 1948, ?L. Cistecephalus zone ;
Scaloporhinus Boonstra 1953, Cistecephalus zone; Silphoictidoi-
des Huene 1950, L. Endothiodon zone, E. Africa; i^ilpholestes
Broom 1948, U. Endothiodon zone; Tetracynodon Broom and
Hobiusoii 1948, Cistecephalus zone.

Family Scaloposauridae. Small bauriamorphs usually with an
elongated face and a very short temporal region. Parietals nor-
mally form a broad flat area not perforated by a parietal fora-
men. Postorbital bar nearly always incomplete and postfrontal
absent. The head as a whole low and very broad. No secondary

WATSON AND ROMER : THERAPSID CLASSIFICATION 75

palate. The basisphenoid widens abruptly posteriorly and the
prootic has a horizontal face directed downward, above the lateral
opening to the pituitary fossa and mesial of the facial nerve
foramen. The dentition always contains a large number of teeth
which are little varied in that upper canines, though usually
recognisable, are relatively small, and the incisors are not en-
larged. There are typically 6 incisors; the canine is preceded by
smaller maxillary teeth, numbering 2 or 3 ; there may be one or
two enlarged canines ; the postcanines are about 10 in number.
The teeth are normally simple cones, but the cheek teeth may be
incipiently tricuspidate. The lower jaw is exceedingly slender;
the posterior end of the dentary rises little if any above the
surangular.

Silphedociinodon and Silphedestts are not improbably scalopo-
saurids, in which both zygomatic and postorbital arches, always
slender in this group, have been lost — either naturally, in the
process of fossilization, or in preparation.

Choerosaurus Haughton 1929, M. Endothiodon zone ; fHomo-
dontosaurus Broom 1949, L. Cistecephalus zone ; Icticephalus
Broom 1915, Tapinocephalus zone ; Ididodon Broom 1925, Ciste-
cephalus zone; fNanictocepkalus Broom 1940, Cistecephalus
zone; Scalopocephalus Huene 1937, U. Endothiodon zone; Scalo-
posaurus Owen 1876, U. Cistecephalus or Lystrosaurus zone;
Scaloposuchus Broom 1940, Cistecephalus zone; Silphedestes
Broom 1949, M. Cistecephalus zone, ?Cynognathus zone; Sil-
phedocynodon Brink 1951. M. Cistecephalus zone.

Family Ericiolacerfidae. Very small bauriamorphs closely re-
lated to the Scaloposauridae but possessing a secondary palate.
Suborbital vacuities and a relatively large interpterygoid vacu-
ity present. The basisphenoid is expanded posteriorly. The skull
is low. The postorbital bar is incomplete ; there is no postf rontal.
The parietals are wide. There is no parietal foramen. The den-
tition varies, but there is no well emphasized canine ; the maxil-
lary teeth may be elaborated.

fCyrhasiodon Broom 1931, ?Lystrosaurus zone; Ericiolacerta
Watson 1931, Lystrosaurus zone; Scaloposaurid indet. Broom
1932, ^Lystrosaurus zone.

Family Bauriidae. Advanced bauriamorphs with a secondary
palate. There is a parietal crest, sometimes narrow; no pineal

76 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

foramen in most forms. The postorbital bar is slender and may
be interrupted. There are well marked and separated basisphe-
noidal tiibera, and the occipital condyle is notched as an incipient
division. The lower incisors tend to be procumbent. The tooth
rows diverge posteriorly, but are closer together than the width
of the face above them, the surface of the maxilla flaring outward
over a concavity to the lower border of the orbit and the face
in front of it. The hinder part of the jaw is relatively large.
The jaw articulation generally lies above the level of the cheek
teeth. There are 4 large pointed incisors in the premaxillae,
1 canine, not noticeably larger than the incisors, and no pre-
eanine maxillary tooth. The cheek teeth (4 to 12) have crowns
whose low irregular cusps were soon completely destroyed by
wear. The crowns may be circular or transversely widened.

Aelurosiichus Broom 1906, Cynognathus zone ; Bauria Broom
1909 {Baurioides Broom 1925), Cynognathus zone; Melinodon
Broom 1905, Cynognathus zone; Microgomphodon Seeley 1895,
Cynognathus zone; Sesamodon Broom 1905 (Sesamodonfoides
Broom 1950), Cynognathus zone; WafsonielJo Broili and Schroe-
der 1935, Cynognathus zone.

Bauriaynorpha incertae sedis

Akidnognnflius Haughton 1918, Cisteeephalus zone; Blaftoide-
alestes Boonstra 1954, Tapinocephalus zone; Haitghtoniscus
Kuhn 1937 (Macroscelesaurus Haughton 1918), zone uncertain;
Ictidochampsa Broom 1948, Cisteeephalus zone; Icfidognathus
Broom 1911, Cisteeephalus zone; Ictidostoma Broom 1931, En-
dothiodon zone; Nanictosuchus Broom 1936, Cisteeephalus zone;
Polycynodon Broom and Robinson 1948 {Ocfocynodon Broom
1940), Cisteeephalus zone; Protocynodon Broom 1949, Cisteee-
phalus zone.

SUBORDER ANOMODONTIA

Therapsid reptiles usually of herbivorous habit. They are
distinguished by the exi.stence in them of a downwardly directed
anterior part of the basioccipital whose morphologically ventral
surface is directed backward below the condyle. The fenestra
ovalis lies very ventrally, lateral to the downturned part of the
basioccipital, and is connected to the rest of the labyrinth by a
long downwardly directed tube filled with perilymph. Teeth
are found only on the premaxilla, maxilla and dentary. Post-

WATSON AND ROMER : THERAPSID CLASSIFICATION 77

orbital and squamosal meet above the temporal fenestra. Ptery-
goid flanges generally reduced and not sharply set off from
the quadrate ramus. Deep suspensorium, extending far below
tooth line and at a considerable angle to it. Quadrate and
quadrato.iugal less reduced than in theriodonts. Ventral surface
of skull not flattened as in many theriodonts. Jaw deep pos-
teriorly, with a strongly curved dorsal margin descending pos-
teriorly to the articulation. No coronoid process of dentary; no
coronoid bone. Dentition usually highly modified or reduced.
Phalangeal formula never exceeds 2.3.3.3.3.

Infeaorder Dixocephalia

Primitive anomodonts in which the premaxillary teeth are
often modified for cru.shing; the canine, at first large and of
circular outline, is reduced to a point where it is no longer
recognizable as a canine; the other maxillary teeth, at first
simple cones, acquire, at least anteriorly, something of the
crushing structure of the incisors. The postorbital part of the
skull is short, the temporal fossa often wider than long, and
the quadrate condyle carried far down below the base of the
brain. The face is short, and often depressed and of feeble ap-
pearance, whilst the skull roof may be immensely thickened.
Broad parietal table. Cheek deep, not greatly excavated in
front of suspensorium. Squamosal not expanded laterally as in
dicynodonts. Postfrontal always present. Premaxillae with long
dorsal processes between nasals. No preparietal. Never any
secondary palate, the ehoanae anteriorly placed. Body and
limbs of massive build. Scapular blade flat ; no acromion. Iliac
blade little expanded, no obturator fenestra. Digits short, unguals
broad.

As noted above we are restricting the term Dinocephalia to
the tapinocephaloids.

Family Deuterosanridae. Primitive dinocephalians in which
the skull is not pachyostotic to any great degree. The skull is
very deep in proportion to its length, the relatively small orbit
lying in the middle of its height, the face in front of it being
of rather triangular section and largely made by a huge lacri-
mal. The temporal fossa is of great size, its ventral border lying
on the same level as that of the orbit. It extends inwards so as

78 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

to be separated from its fellow only by a very narrow pair of
parietals ; these form a crest in front of the pineal foramen,
which lies at the summit of the occiput. It is uncertain whether
the squamosal and postorbital meet. The palate is triangular
in plan with narrow premaxillae, which bear about 4 incisor
teeth, each with a characteristic piercing cusp in front of a
crushing heel. There is a large canine of nearly circular section
and a few relatively simple maxillary teeth.

Deuterosaurus Bichwald 1846 (Mnemeiosaurus Nopcsa 1923,
Uraniscosaurus Nopcsa 1928), Russia, zone I.

Family Moschopidae. Dinocephalia in which the skull is rela-
tively high and narrow, the pachyostosis extending to many of
the bones of the face and head, but being not very intense.
There is no sharp distinction between the posterior part of the
head and the face in the outline of the skull seen from the side.
There is no enlarged canine, and the premaxillary and some
maxillary teeth are of the type with a crushing heel, but some
simple teeth are found in the maxilla behind the region where the
canine should be, up to to the number of perhaps a dozen.

DelpJmiognathus Seeley 1892, probably Tapinocephalus zone;
Moschognathus Broom 1914, U. Tapinocephalus zone; Moschops
Broom 1911, Tapinocephalus zone ; perhaps Pnigalion Watson
1914, Tapinocephalus zone.

Family Tapinocephalidae. Skull roof greatly thickened and
rugose, often forming a brow above the apparently slender face
which is always markedly separated from the postorbital region.
The canine is never distinct, but the incisors vary from large
teeth with a single cusp and a crushing heel to quite small
structures (cf. Mormosaurus). There is a great variety of form
expressed in a large number of generic names which may repre-
sent to some considerable extent different developmental stages.
This family consists of most South African forms from the
Tapinocephalus zone, and Ulemosaurus from zone II of Russia.

Agnosaurus Boonstra 1952, Avenantia Boonstra 1952, Crioce-
phalvs Broom 1928, Eccasaurus Broom 1909, Ke/ra.tocephalus
Huene 1931, Mormosaurus Watson 1914, Moschoides Byrne 1937.
Moschosauriis Haughton 1915, Pelosuchns Broom 1905, Phoco-
saurus Seeley 1888, Riebeeckosaiirus Boonstra 1952, Strufhio-
rephaloides Boonstra 1952, StruthiocephaleMiis Boonstra 1955,

WATSON AND ROMER : THERAPSID CL.ASSIFICATION 79

Struthiocephahis Haughton 1915, Struthionops Boonstra 1952,
Tapinocephalus Owen 1876, Taurocephalus Broom 1928, Taurops
Broom 1912, Vlemosaurus Riabinin 1932.

Infraorder Vknyukovioidea

Little or no paehyostosis. Lacrimal large and long. Suspen-
sorium approaching dicynodont condition. Choanae placed in a
deep pit, with the maxillae excluded from their borders. Teeth
differentiated, with large anterior incisors, retention of a stout
upper canine, and an irregularly placed series of small cheek
teeth. Transitional in structure to the dicynodouts (cf. Watson
1948, Proc. Zool. Soc. London, vol. 118, pp^. 823-77).

Family Venyiikoviidae. This contains reptiles which are clearly
intermediate between Dinocephalia and Dicynodontia. They are
known from lower jaws and the anterior part of a skull. The
dentition in the upper jaw is composed of three or four large
grinding teeth in the premaxillae, the maxilla having a row of
blunt teeth with rounded conical points. Posteriorly there is an
incipient introduction of a second row of teeth. The posterior
nares open into a deep depression on the palate surrounded by
the palatines and premaxillae and divided into two by the
vomers. The ectoptyerygoid bone is quite large and meets the
pterygoid so as to form a low but characteristic flange against
the inner surface of the lower jaw. Behind this the pterygoid
soon ends ; it is separated from its fellow by a small vacuity
behind which it is attached to the basiscranium by a joint
sutural surface. The rod-like quadrate ramus diverges from the
basiscranium a short distance back of the transverse process.
On the face, the maxilla is deep and appears to touch the pre-
frontal, excluding the lacrimal from the orbital border. The
nostril lies some distance l)eliind the end of the snout, and is
underlain by a septomaxilla well exposed on the side of the face ;
the premaxillary processes separate the nasals almost to the
level of the anterior part of the orbit. In the lower jaw the
structure is essentially that of a dicynodont, but the dentary
bears a row of teeth beginning with two enlarged procumbent
teeth whose upper surfaces are worn by contact with the pre-
maxillary teeth in the upper jaw! Farther back there is a pit
on the widened upper surface of the dentary into which the

80 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

upper canine bites and behind this a series of pits for more
posterior maxillary teeth. This part of the jaw must have been
covered by horn, the dentarj' tooth row mesial to which is
doubled. This genus provides a perfect intermediarj- stage in
the development of dicynodonts from deuterosaurids.

fRhopalodon Fischer 1841, Russia, zone 1; Venyukovia Ama-
litzky 1922 {Myctosuchus Efremov 1937, Venjukovia Efremov
1940), Russia, zones I and II.

Infeaorder Dromasauria

Small anomodonts with or without teeth, but with a lower
jaw of dicynodont pattern. The squamosal wraps round the
quadrate in a normal way and the zygoma arises on the most
lateral part of the bone. The temporal fossa is short. The tail is
long. There is no such development of the acromion as in dicyno-
donts. Coracoid elements large. Ilium little expanded ; a small
obturator fenestra. Feet slender, unguals claw-shaped.

The group Dromasauria was founded by Broom as a division
of the Therapsida of equal value to the Dinocephalia, Anomo-
dontia, etc. The entire group is represented by four individual
animals belonging to three genera and species. The known speci-
mens range in date from the middle of the Tapinocephalus zone
to a horizon which appears to be in the middle of the Cistece-
phalus zone. The skull and lower jaw are known by unsatis-
factorily preserved remains which do not allow of a complete
description but do establish certain fundamental facts.

The only complete, unerushed skull, the type of Galepus
jouherti, has a very short muzzle with laterally directed nostrils;
the orbit is very large, the frontals being bowed above it and
passing backwards to the broad temporal region, which itself is
directly continuous with a vertical occiput. The squamosals pass
downward far below the level of the paroccipitals and occipital
condyle so that the quadrate articular surface is carried down
to a point well below the level of the teeth. The skull is pre-
served only as an internal mould in not very fine sandstone, and
the condition of this mould makes it impossible to determine
the position of the sutures between bones of the skull, though its
general nature is obvious.

The lower jaw of this individual has a dentary whose upper
border is slightly concave and bears a short row of some 6-7 teeth,

WATSON AND ROMER: THERAPSID CLASSIFICATION 81

none larger than any other, not extending to the anterior end
of the dentary and largely represented by casts of the alveoli in
which they were placed. These alveoli are, in effect, impressed
into the admedian surface of the dentary. The hind part of the
jaw is of the same nature as that better seen in other specimens
belonging to the group. Taken as a whole, the skull is character-
ized by the great size of the orbit, the short snout, and the very
short temporal fossae.

Galechirus is represented by two specimens, one of which con-
tains the impression of the outer surface of part of a skull and
the lower jaw. The fragmentary skull resembles that of Galeops
but show^s evidence of the existence of teeth in both premaxilla
and maxilla; there is no enlarged canine. In this specimen the
lower jaw has no coronoid process on the dentary ; the hinder
part of the jaw resembles that of anomodonts in its possession of
a fenestra between surangular, angular and dentary. The angu-
lar has a rounded lappet from its lower border which is no doubt
a reflected lamina.

The single specimen of Galeops is contained in a split slab of
very hard sandstone on which no preparation, apart from certain
fractures, has been carried out. The skull is very high, appar-
ently^ narrow, and short. The face is short, the external nostril
very large, the orbit enormous and the temporal fossa deep, very
narrow from back to front and separated from the orbit by a
very narrow postorbital bar. The squamosal is continued far
below the level of the base of the orbit, having a long free lateral
border which extends downwards to the region of the condyle.
Towards its dorsal end, where the squamosal comes up to the
little broadened intertemporal roof, its posterior surface is
rounded from side to side and produced a little backwards, as it
is in many early therapsids. The suborbital bar, prt^suiuably
largely composed of jugal, is exceedingly shallow. The quadrate
is clearly visible but is not in a state to be described in any de-
tail, though there is some reason for believing that it had a
condyle divided by a median groove. It extends upwards, in con-
tact with the front face of the descending process of the squamo-
sal, to about the level of the lower border of the orbit.

The best preserved and most comprehensible part of the whole
skeleton is the lower jaw. The dentary is a massive bone showing
no signs whatever of teeth, unless a very curious notch, sur-

82 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

rounded by a peak in front and behind, which lies at the extreme
front of the bone on its upper surface, is to be associated with
a tooth insertion. The dentary extends backward to end in a
small, backwardly projecting coronoid process, extending about
a millimeter beyond the attachment to the surangular. The
hinder part of the jaw includes an articular bone whose condylar
face is directed backward and upward at an angle of about 45°
with the length of the jaw and faces towards the condylar sur-
face of the quadrate. As the bone is seen only in broken longi-
tudinal section, the nature of the articular surface cannot be
further described other than that it is convex as seen in side
view, as in Dicynodon, is very extensive, and apparently ends
immediately above a short, narrow, downturned process which is
presumably an attachment for a posterior pterygoidal muscle,
and not a retroarticular process. The surangular, which is con-
tinuous posteriorly with the articular, has a gently concave
margin for the greater part of its length, and is overlapped on
its outer surface by the upper border of the angular, the two
bones separating to surround a foramen. The angular has a
large reflected lamina passing downwards to a rounded lappet
which extends several millimeters ventrally to the lower border
of the body of the bone, when it extends backwards to overlap
the articular at the root of the pterygoidal muscle process. In
other words, the lower jaw is almost completely anomodont in
structure.

The general nature of the postcranial skeleton is similar in all
four dromasaur specimens. The essential features are that there
is a distinct neck, with 5 to 7 cervical vertebrae ; there is a back
of some 20 vertebrae ; the number of sacral vertebrae is not cer-
tainly known (12) and the tail is a relatively immense structure,
with some ;^0-40 vertei)rae, tapering very slowly, and probably
as long as the presacral column and head together. Ribs in the
cervical region are scarcely shown; they were presumably short.
The dorsal ribs were long, fairly well curved and indicating the
presence of a relatively slender body lacking, except perhaps in
Galeops, the large size of the anomodont trunk. There is no
definite evidence of specialized lumbar ribs. The caudal verte-
brae seem to lack transverse processes, though they do carry
chevron bones, at any rate in the proximal half of the tail. In
three of the four specimens the tail is wrapped around the rest

WATSON AND ROMER : THERAPSID CLASSIFICATION 83

of the body in sueh a way as to suggest a flexibility analogous
to that of a squirrel or lemur, and is entirely unlike the lizard
tail. The shoulder girdle has a large, rather wide, interclavicle,
with an expanded and rounded hinder end; the scapula has a
narrow, curved, blade expanding widely at the lower articular
end, where it is attached to a rather small, perforate precoracoid
and a larger coracoid. There is no sign of an ossified sternum or
eleithrum. The humerus is a long, relatively narrow bone, with
a slender shaft and not very expanded extremities, which lie
only at a very small angle with one another in the later forms ;
there is an entepicondylar foramen. Radius and ulna are a
little shorter than the humerus, subequal in size. There are 4
proximal and 5 distal carpals, with two centralia, one of which
is exposed on the radial side of the carpus. The phalangeal
formula is 2.3.3.3.3. The ilium is short, little expanded at the
dorsal margin, and articulates directly with the pubis and
ischium — rather flat, plate-like bones not separated by an
obturator fenestra but with an unusually large pubic foramen.
There is evidence that the ventral surface of the pelvis rose to a
medial ridge. The femur is a slender bone, with no recognizable
peculiarities; the fibula has a somewhat expanded upper end,
extends slightly farther distally than the tibia and leaves a
rather wide interosseous space between the two bones. The as-
tragalus is a bone of considerable size which articulates with a
face on the admedian border of the ealcaneum, there being some
evidence of a notch between the two. The ealcaneum is a large
bone, extending laterally to its contact with the fibula and the
fifth metatarsal. There are four distal tarsals, the fourth being
much the largest ; a single centrale separates the astragalus from
the first to third distal tarsals. The phalangeal formula is
2.3.3.3.3. There is evidence of the existence of some al)dominal
ribs a little in front of the pelvis.

Broom, in his early discussions, held that the Dromasauria
was a group of equivalent rank to the Dinocephalia or the dicyno-
donts, but in the "Mammal-like Reptiles of South Africa"
(1932) which appears to be his last statement on the matter, he
withdrew a little from this view, stating that it is impossible to
say whether the dromasaurians are a primitive group or dwarf
aberrant descendants of early therocephalian ancestors.

To anyone who is familiar with dic}Tiodont structure, or even

84 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

with the structure of Venynkovia, there is evident a remarkable,
rather detailed dicynodont appearance in the lower jaw of
Galeops. This is, in effect, so striking that we do not think any-
one presented with the lower jaw of this animal would come
to any conclusion other than that it was a dicynodont of slightly
unusual character. This resemblance depends in part on the lack
of teeth in the dentary, largely on the remarkably massive den-
tary as a contrast to the rather less massive hinder part of the
jaw; even the texture of the bone of the dentary, as seen in
fracture, is extraordinarily like that of a dicynodont. The sur-
angular and angular are precisely those of a small dicynodont,
even to the rather characteristic shape of the reflected lamina.
The general nature of the articular, so far as it can be seen in
the longitudinally fractured specimen is, again, dicynodont-like,
and the dorso-ventrally convex hinder part of its articular sur-
face is a character known in no therapsid group except the
anomodonts.

The way in which the alveoli of the cheek teeth of the Galepus
JHAv are as it were, notched into the inner surface of the dentary
can be matched only in endothiodonts, where the condition has
arisen as a result of the development of a dicynodont horny beak
lateral to the teeth. The long ventral process of the squamosal,
which supports the quadrate, is another dicynodont quality, but
the external view of the battered hinder part of the squamosal
which is seen on the back of the counterpart of the slab of
Galeops seems to show no trace or, at any rate, no more than a
beginning of that lateral production of the descending part of
the squamosal which is characteristic of dicynodonts. The
extraordinarily short temporal fossa of all dromasaurians is a
point of extreme contrast to the dicynodonts, in which the tem-
poral fossa is (except secondarily, in such forms as Lystrosaurus)
of quite exceptional length.

The dicynodonts as a whole are animals with a head of normal
size which is connected by a neck which would appear short to an
extraordinarily broad and deep trunk. Even in small forms
comparable in size to the Dromasauria the tail is short and
conical, the limbs short and heavily built ; further, all the bones
from humerus to phalanges are thick in proportion to their
length and the claws in which both fingers and toes end, appear
never to be laterally compressed and slender but are usually

WATSON AND ROMER : THERAPSTD CLASSIFICATION 85

markedly broad and flat. There is indeed no resemblance be-
tween the postcranial skeletons of individuals of the two groups
other than that implied by the statement that they are therapsids.
Galechirus Broom 1907, zone uncertain; GaJeops Broom 1912,
U. Tapinocephalus zone; Galepus Broom 1910, zone uncertain.

Infraorder Dicynodontia

Anomodonts in which the dentition, no doubt complete and
marginal in their ancestors, tends to disappear as it is supple-
mented and ultimately replaced by a horny beak, analogous to
that of a tortoise, in both upper and lower jaws. The animals
were certainly herbivorous, and the skull becomes a unique struc-
ture designed to house a very powerful musculature which en-
ables the heavy lower jaw to crush food against the palate, and,
in all probability, to bite a large mass from such a thing as a
cycadean trunk. In the great majority of known forms very
powerful canine tusks are found in the males, such structures
being completely lacking in females, and therefore forming no
part of the feeding mechanism. The dentaries are always fused
at a short and very heavy symphysis whose lower surface stands
largely vertical. The squamosal has a unique structure in that
its body forms a flat sheet continuing the occipital surface.
From the front face of the squamosal the zygoma projects
abruptly so that its upper border meets that of the body of the
squamosal at a point, whilst its lower border passes into the
middle of the body. There thus arises a special large space for
a masseteric muscle cut off from the larger temporal mass by
the flat sheet of the posterior part of the zygoma. The temporal
fossa is long.

In the more primitive dicynodonts a wide roof of parietals bor-
dered by postorbitals and sf(uamosals separates the very large
temporal fossae, the pineal foramen which perforates it often im-
pinging on a preparietal bone. The face is short (in the more
advanced forms very short) in comparison with the huge tem-
poral region. The face is liounded anteriorly by the premaxillae
which descend to form a I'idge round the anterior end of the
palate. They are always fused with one another and their inter-
narial process, which rises at first directly upwards and then
passes round a segment of a circle towards the upper surface of
the skull, separates the deeply impresserl nostrils by a consider-
able space, and reaches the nasals. The maxillae, from their

86 BULLETIN: MUSEUM OP^ COMPARATIVE ZOOLOGY

suture with the premaxillae, rise to meet the nasals and form
part of the border of the nostril, though the septomaxilla, which
is usually well developed, has often a considerable facial expos-
ure, and may exclude them from that opening. The maxilla never
has any marginal teeth except for the very large canines of the
males, whose long root extends upwards wdthin the borders of the
bone so as to gain support for the projecting tusk. The tooth
is circular in section. The point on the maxilla where the tooth
arises is often projected down as a very long process which
also occurs in females in which the tusk is missing. In the more
primitive forms, " endothiodonts, " teeth occur on the palate
mesial of the canine (if present). They are actively replaced and
usually have ridged pointed crowns, often with a saw edge on
one side. Thev occur usually in the maxilla, and occasionally
also in the premaxilla. It is evident that this tooth row is not
palatal, but is the original therapsid series. A similar develop-
ment of teeth in the lower jaw^ is usually inserted in the lingual
side of the dentary. The body is ordinarily very bulky, but with
a definite neck. The fore and hind feet have a digital formula of
2 3 3 3 3 The limbs are short and strong and the claws flattened
and obviouslv powerful. The stride is short compared with the
width of track. There is on the scapula a definite acromial proc-
ess within which lay a prespinous muscle. The precoracoid is
excluded from the glenoid cavity and there is a sternum. The
pelvis has an ilium widened dorsally and a sacrum of from 3 to
7 vertebrae. There is an obturator foramen and the ventral bor-
ders of the pubis and ischium do not always meet their fellows

in symphysis. ^ ^ ^. t

We have made no attempt at an arrangement of the genera ot
dicynodonts, of which more than fourscore have been described.

THERAPSIDA INCERTAE SEDIS
Arnognathus Broom 1907, ?Cistecephalus zone; Dimacrodon
Olson and Beerbower 1953, L. or M. Permian of North America;
Herpefochirus Seeley 1895, ?Lystrosaurus zone; Mygalemurns
Broom 1942, Cistecephalus zone ; Mygalfsnchus Broom 1942, Cis-
tecephalus zone; Rubklcjina Broom 1942, Cistecephalus zone ;
Simorhinella Broom 1915, Tapinocephalus zone; Tamhoerta
Seeley 1904, Tapinocephalus zone ; Theromiis Seeley 1895, ILy-
strosaurus zone.

A&metop}ioneus, 49
Aelurodraco, 63
AelurognathuK, 52, 53
Aelurosauroides, 52, 53
Arlurosauropsis, 53
Aelurosaurn.s, 52, 61
AelurosucJius, 76
Agnosaurus, 78
Akidnognathus, 76
Alopeoideops, 69
Alopeoodon, 69
Alopecognathus, 69
Alopecopsis, 70
Alopecorhinus, 69
Alopecorhynchus, 60
Aloposauroldes, 61
Aloposauriis, 61
AvmlitzMa, 57
Aneugomphins, 71
^UTia, 72
Anteosaurus, 49
Arcliaeodon, 67
Arcliaeosuchus, 50
Archaeotherium, 67
Arctognathoides, 53
Arctognathus, 53
Arctops, 60
Arctosuchus, fil
Arnognathun, 86
Asthenognathus, 61
Avenantia, 78
Bauria, 76
Baarioides, 76
Baurocynodon, 63
Belesodon, 63
Bienolherrum, 67
Blattaidealentes; 76
Brithopus, 49
Broomicephalus, 58
BrooviLsauriis, 54
Broomosuchus, 49
Burnetia, 54
Cerdodon, 55
Cerdognathus, 61
Cerdops, 72
Cerdorhinns, 60
Cerdosuchoides, 72
Cerdosuchus, 72
Chalepotherium, 67
ndniquodon, 63
Chiiuetasauriis, 60
Choerosaurus, 75
Chthonoaaurus, 72

INDEX TO GENERA

Cutecynodon, 64
Clelandina, 60
Cliorhizodon, 49
Crioeephalus, 78
Cyclogomphodon, 65
Cynariognathus, 69
Cynarioides, 54, 61
CynariopK, 55
Cynidiognailiu.9, 64
Cyniscodon, 61
Cyniscopoides, 59
Cyniscops, 59
Cynochampsa, 65
Cynodraco, 61
Cynognathus, 64
Cynogomphius, 64
Cynosaurus, 63
Cynosuchoides, 63
Cynosuchus, 63
Cyonosaurtts, 60
Cyrhasiodon, 75
Delphaciognathus, 61
Delpliinognathu.^, 78
Deuterosaurns, 78
Diademodon, 65
Diastemodon, 65
Dimacrodon, 86
Dinaitnmus, 50
Dinocynodov, 50
Dinogorgon, 58
Dinophoneus, 50
Dinopolus, 50
Binosaurus, 49
Dinosphageus, 50
Dinosuchits. 49
Dixpyn, 60
Brarucephalus, 58
Dromatherium, 66
Dvinia, 62
Eccasaurus, 78
Enohius, 50
Eoarctops, 55
Eozostrodon, 67
Ericiolacerta, 75
Eriphostoma, 61
EiicJiambersia, 72
Eurosaurus, 49
Exaeretodon, 66
Galechirus, 85
Galecranium, 62
Galeophrys, 62
Galeops, 85
Galepus, 85

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Galerhinus, 55
Galerhynclius, 59
Galesaurus, 63
GalesucJiUS, 55
Genovum, 61
Glanosuchttti, 69
Glaridodon, 50
Gloohinodon, 63
Gloehinodontoides, 63
Gomphodontoides, 65
Gomphodontosuchus, 65
Gcmipliognathus, 65
G orgonognathus, 55
Gorgonops, 56
Gor gonorhinus, 56
Ilaiightoniscus, 76
Ilerpetochirus, 86
Hipposauroides, 56
Kipposaurvs, 56
Hofmeyria, 70
Eomodontosaurus, 75
Eyaenasudhu^, 69
Hyenosaurns, 70
HyorhyncJius, 69
Hypsiprymnopsis, 67
lotieephalus, 75
Ictidochampsa, 76
Ictidodon, 75
Ictidodraco, 74
lotidogruitJius, 76
Ictidoparia, 69
Ictidopsis, (53
Ictidorhinus, 56
Ictidosaur-u^i, 69
Ictidostoma, 76
Irtidosuchoides, 74
Ictidosuchops, 74
Ictidosuclius, 74
Inostrancevia, 57
Inusitatodon, 65
Jonkeria, 50
Karoomys, 64
Keratocephalus, 78
Kunminia, 66
Lmniasaurus, 50
Leavachia, 62-63
Lemurosaurus, 56
Leontocephalus, 53
Leontosaurus, 54
Leptotrachelus, 56
Lycaenodon, 61
Lycaenodontoides, 53
LyoaenognatJius, 64

Lycaenoides, 60
Lycaenops, 54, 56, 60
Lycedops, 69
Lycideops, 73
Lycochampsa, 64
Lycognathus, 64
Lycurhinus, 6G
Lycosaurus, 61
Lycosuchus, 69
Macroscelesaurus, 76
Maia'Utaurus, 69
MeMnodon, 76
Micranteosaunts, 49
Mirrirtodon, 63
Microcleptes, 67
Microconodon, 66
MicrogompJiodon, 76
Microhelodon, 65
Microlestes, 67
Mnemeiosaurus, 78
Morganucodon, 67
Mormosaurus, 78
Moschognafhus, 78
MoscJwides, 78
Moschops, 78
Mosdwrhinus, 70
MoscJwrhyndhus, 70
Moschosauruii, 78
Mucrotherium, 67
Myctoauiiius, 80
Mygalesauni.s. 86
M ygalesuchu.s, 87
Nanictidops, 74
N anictocephalus, 75
Na7iictosaurm, 63
NanictosuchU'S, 63, 76
Nanogorgon. 60
Notaelurodun, 72
Notaelurops, 70
Notictosaurus, 63
Notosollasia, 70
Nythosaurus, 63, 64
Octagomphios, 65
Octocynodon, 76
OligoTcyphus, 67
OrtJwpus, 49
Pachygenelus, 66
P achy rhinos, 57
Paracynosuchus, 63
Paranteosaurus, 49
Parathrinaxodon, 63
PardocepJuxlus, 61
Pardosuchus, 69

WATSON AND ROMER : THERAPSID CLASSIFICATION

Pelictosncliu.'i, 74
Pelosuclius, 78
Perinoc;iiu)don, 63
Phocosaurus, 78
Phoneosuvhus, 50
PhreatopJidsma, 49
Phreatnsatiriis, 49
Phrrato.siuhus, 49
Plithinosattrus, 57
J'lithino-'otchits, 57
Platycraniellus, 63
Platycranion, 63
Plieningerki, 67
Pnigalion, 78
Polycynodon, 76
Pristerognathoides, 69
Pristerognathus, 69
Pristerosaurus, 69
Proalopccopsis, 72
Procynosuclms, 63
Promoschorhynchus, 70
Prorubidgea, 58
Protacmon, 65
Protocynodon, 76
Psendanteosaiirns, 49
Pseudohipposaurus, 56
Ptomalestes, 69
Fiwpalodon, 49, 80
BiebeecTcosaurus, 78
Eubidgea, 58
Subidgina, 87
Sauroctonus, 61
Scalopocephalus, 75
Scaloporliinus, 74
Scaloposaurus, 75
Scaloposurliiis, 75
Soapanodon, 50
Soullya, 50
Scylacocephalus, 59
Scylacognnfhiis, 55
Scylacoidrs, 69
Scylacops, 59
Scylncorhivvs, 69
Scylaccsanriis, 69
Scymnognathus, 60, 61
Soymnosaurvf!, 69
Sesamodon, 76
Sesamodovtoides, 76
Silphedesles, 75
Silpliedocynodon, 75
Silphoictidoides, 74
Silpholestcs, 74
SvmorhineUa, 87

Smilesaurus, 58
StereognathU'S, 67
Struthiocephalellus, 78
Struthiocephaloides, 78
Striithiocephalus, 79
Struthionops, 79
Styracocephahis, 54
Sycocephalus, 59
Sycosaurus, 60
Syodon, 49
Syspliinctostoma, 63
Tamboeria, 87
Tangagorgon, 58
Tapinoceplmlus, 79
Taiirocephalns, 79
Taurops, 79
Tctracynodon, 74
Tetraodon, 61
Theriodenmus, 72
Theriognathus, 70
Therioides, 69
Therornits, 87
Theropsis, 66
Theropsodon, 65
Thomasw, 67
Thrinaxodon, 63
Tigriceplialus, 56
Tigrisaurus, 58
TigrisucJivs, 61
TUanognathus, 49
Titanophoneus, 49
Titanosuchus, 50
Traversodon, 65
Tribolodon, 64
Tricuspes, 66
Triglyphus, 67
Trirachodon, 65
Trirachodontoides, 65
Tritheledon, 67
Tritylodon, 67
Trochorhinus, 69
Trochosaurus, 69
Trodhosuchus, 69
Tytthoconus, 66
Ulemosaurus, 79
Uniserium, 67
I'raniscosaurus, 78
Venjulcovia, 80
FenyiiJiovm, 80
Walteria, 69
Watsoniella, 76
Whaitsia, 70

Bulletin of the Museum of Comparative Zoology

AT HA EVA ED COLLEGE
Vol. 114, No. 3

THE PLACENTA OF THE WOLVERINE {GVLO
GULO LUSCUS (LINNAEUS))

By George B. Wislocki

and
Emmanuel C. Amoroso

Department of Anatomy, Harvard Medical School,

Boston, Massachusetts, U.S.A. and the Eoyal Veterinary College,

University of London, England

With Eight Plates

CAMBEIDGE, MASS., U.S.A.
FEINTED FOE THE MUSEUM

February, 1956

Publications Issued by or in Connection

WITH THE

MUSEUM OF COMPARATIVE ZOOLOGY
AT HARVARD COLLEGE

Bulletin (octavo) 1863 — The current volume is Vol. 114.

Breviora (octavo) 1952 — No. 51 is current.

Memoirs (quarto) 1864-1938 — Publication was terminated with Vol. 55.

Johnsonia (quarto) 1941 — A publication of the Department of Mollusks.
Vol. 3, no. 34 is current.

Occasional Papers of the Department of Mollusks (octavo) 1945 — -
Vol. 1, no. 18 is current.

Proceedings of the New England Zoological Club (octavo) 1899-
1948 — Published in connection with the Museum. Publication terminated
with Vol. 24.

Bulletin of the Museum of Comparative Zoology

AT HAEVAED COLLEGE
Vol. 114, No. 3

THE PLACENTA OF THE WOLVERIXE {GVLO
GULO LUSCUS (LINNAEUS))

By George B. Wislocki

and
Emmanuel C. Amoroso

Department of Anatomy, Harvard Medical School,

Boston, Massachusetts, LT.S.A. and the Eoval Veterinary College,

University of London, England

With Eight Plates

CAMBEIDGE, MASS., U. S. A.
P E I N T E I) FOR T II E :M U S E U M

February, 1956

No. 3 — The Placenta of the Wolverine (Giilo finlo lusc^ts
(Linnaeus) )^

By George B. Wislocki and Emmanuel C. Amoroso

INTRODUCTION

Of the numerous fissiped carnivores the plaeentation of per-
haps fifteen species has been to some degree studied. It is of
interest when a specimen becomes available of a genus or species
in which the placenta has not been described before, and thus we
were fortunate in obtaining the placenta of a wolverine. This
animal is a member of the family of Mustelidae of the genus
Gulo, of which, according to Weber ('28), there are two species,
the Eurasian glutton, G. gulo L. and the North American wol-
verine, G. luscus L. ; others, however, regard the animals of the
two continents as conspecific (cf. Rausch '58). The placenta of
neither of these species has l)een described before.

The plaeentation of various other members of the Mustelidae
is to some extent known. Nearly a c(Mitury ago Bischoff (18(i5 a,
1)), reported on tlic plaeentation of martins {MusfeJa foina,
Musfela martes), the weasel (Mustehi vulgaru), and the otter
(Lutra vulgaris), while more recently the placenta of the ferret
(Puforins furo) has been described by Robinson ( '0-1) and Strahl
and Ballmann C15). and that of Zorilln by Ran ('25).

MATERIAL AND METHODS

The present specimen consists of a pregnant uterus (with the
ovaries removed) of a wolverine {Gulo gulo luscus (Linnaeus))
collected on February 9, 1954, at Mt. Susitna, near Anchorage,
Alaska. The specimen was presented to us for study through
the interest and courtesy of Dr. Robert Rausch, Chief, Animal-
borne Disease Branch. Arctic Health Research Center, Anchor-
age, Alaska, whose taxonomic terminology we have accepted.
The uterus with one of the gestation sacs laid open to ensure
better fixation was preserved in a lO^o solution of formaldehyde.

The specimen was first studied grossly by inspection and dis-
section. Then representative pieces of the placentas, fetal mem-

1 .\i(led b.v a jrraiit (G-220S(C3)) from the U. S. Public Health Service. National
Institutes of Health, and the Eugene Higgins Tru.st of Harvard University.

92 BULLETIN: MUSEUM OF COMPAEATIVE ZOOLOGY

braiies and iiniltilieal curds were removed. These were imbedded
in paraffin, sectioned and stained. The sections were stained by
means of Masson's trichrome stain, Gomori's ( '41) chrome ahim-
hematoxylin and phloxine stain and McManns' periodic acid-
Sehiff stain. The stained sections were cleared and monnted in
clarite. We are indebted to Miss Etta Piotti for three excellent
drawings of parts of the specimen.

GROSS ANATOMY

The specimen consists of a bieornnate nterns containing one
gestation sac in the right cornn and two gestation sacs in the
left horn. The three sacs are practically identical, each contain-
ing a fetns of approximately 75 mm. crown-rump length. The
interior of one of the sacs, showing the fetus attached by the
umbilical cord to the zonary placenta, is illustrated in Figure 1.
The allantoic sac has been entered and the zonary placenta flat-
tened out. The amnion has been removed to reveal the fetus. The
cut edge of the amnion fused l)ack to back Avith the wall of the
allantois, is visible at the placental end of the short umbilical
cord. The umbilical vessels are seen emerging from the vortex
of the allantois and branching to form the placental blood ves-
sels. A yolk sac, of digitiform shape with a somewhat bulbous
tip, is seen beneath the allantoic membrane, extending from the
attachment of the lunbilical cord across the placental girdle to
end between the allantoic membrane and the membranous cho-
rion. Separate vitelline vessels from the umbilical cord supply
it with blood. A few fine vessels extend from the margins of the
placenta into the membranous chorio-allantois.

The placenta consists of a girdle whicli has a gap in it in the
vicinitj' of the attachment of the umbilical cord. In this gap there
are two smaller, ovoid i^lacental masses separated from one
another and from the main ])lacental l)()dy by intervening mem-
branous chorion.

On the interior of the placental girdle, opposite to the attach-
ment of tlie umbilical cord, there is a conspicuous, dark red multi-
locular sac which protrudes from the surface of the placenta into
the interior of the gestation sac (Pig. 1). This is evidently a
variety of the Avell-known structures designated as hematomas,
which characterize the placentas of carnivores. In this case the

WISLOCKI AND AMOROSO : WOLVEEIXE PLACENTA 03

sac is a '•'central'" hematoma. Views of it cut in perpendicular
planes are shown in Figures 2 and 3. There are no grossly A'is-
ible "border" hematomas along the margins of the zonular
placenta, corresponding to the "brown'' and "green" borders of
cats and dogs respectively. In addition to the large central
hematoma, there are about half a dozen minute protuberances
on the fetal surface of the placenta (Fig. 1). The largest of
these is a pale brown elevated nodule close to the edge of the
central hematoma. Two other nodules, close together in the
neighborhood of the placental insertion of the umbilical cord,
are dark red. Another similar minute red fleck is present on the
opposite side of the placenta (near the animars snout in the
drawing) and tAvo further elevations, of brownish color, in other
places are scarcely discernible.

As revealed in cross sections of the zonary placenta, the
labyrinth has a perpendicularly striated texture (Fig. 3). The
thick placenta is loosely attached to the relatively thin wall of the
uterus from which it separates readily. At the lateral margins of
the placenta, the membranous chorion is visible.

Figure 4 is a diagram illustrating a gestation sac which has
been separated and removed from the uterus. The central hema-
toma is shown as it appears on the basal or maternal surface of
the placenta, with its orifice indicated by heavy black stippling.
The relation of this orifice to the placental labyrinth will be
further appreciated liy reference to Figure 2.

MICROSCOPIC ANATOMY

The placental Jahyrinih. Seen under the microscope, the
labyrinth of the chorio-allantoic placenta is typically of the
carnivore type consisting of rather indistinct twisting lamellae
of trophoblast which contain sinusoidal maternal capillaries and
are separated hy bands of pale-staining fetal stroma (Figs. 5. 7.
8). The lamellae are tortuous and apparently much branched,
in these respects resembling those of the dog and ferret rather
than the slender ones typical of the cat's placenta.

The individual lamellae show an arrangement of the tissue
layers characteristic of an endotheliochorial placenta according
to Grosser 's classification. Tortuous maternal blood vessels of
sinusoidal dimensions are lodged in lamellar sheets of tropho-

94 BULLETIN: MUSEUM OF COMPAEATIYE ZOOLOGY

blast which alternate with loose-meshed fetal mesenchyme which
contains delicate fetal capillaries (Figs. 7, 8). The endothelium
of the maternal sinusoidal capillaries consists of greatly enlarged
cells of nearly cuboidal form (Fig. 8), resembling closely
the ferret and Z or ill a and differing from the dog, bear and
cat which have lower, more flattened endothelium. The endo-
thelial cells rest upon a conspicuous basement membrane which
is intensely differentiated l)y the ])eriodic acid-Schiff stain
(Figs. 7, 9). Unlike the cat and resembling the dog, bear,
ferret and Zorilla, giant decidual cells are lacking between the
maternal vessels. At this stage, the maternal blood vessels ap-
pear to be surrounded solely by a relatively thin layer of syn-
cytial trophoblast, whereas cellular trophoblast is lacking (Figs.
7, 8), The trophoblast rests upon the fetal mesenchymal stroma.
The fetal capillaries indent the syncytium and in numerous
places follow an "intra-epithelial" course (Figs. 7. 8). The
basement membranes between the trophoblast and fetal stroma
and surrounding the capillaries, are so delicate that they are
scarcely differentiated by the periodic acid-Scliiff reagent.

From the nature of the single specimen at our disposal, the
exact pattern of the maternal vascular tree cannot be deter-
mined. Nevertheless, some information can be got from the
appearance of the blood A'e.ssels in the histological sections. Ap-
parently the maternal arterioles enter, and the maternal venules
leave, the placenta through the septa forming the walls of
glandular compartments of the junctional zone (see below). The
maternal arterioles traverse the placental labyrinth and finally
branch repeatedly beneath the fetal surface (Fig. 5, arrows).
The branches immediately assume the form of dilated sinusoidal
channels (Fig. 5) which are the equivalent of capillaries. As
these branches arise, they curve outward from the parent blood
vessel and become recurrent, following a course in the opjiosite
direction toward the base of the placenta where they join to
form the efferent veins. As a result of their recurrent course
near the fetal surface, many of the sinusoids encountered there
are cut tangentially. The appearance of these vessels is peculiar
in that the tangentially-cut endothelium lining them resembles
multinucleate giant cells or small masses of syncytium (Fig. 9).
The periodic acid-Schiff' stain, however, reveals the true nature
of these structures bv demonstrating that each of them is bounded

WISLOCKI AND AMOROSO : WOLVERINE PLACENTA 95

or enclosed by a basement membrane (Fig. 7) identical and con-
tiniTons Avith the basement membrane which is typical of all the
maternal capillary sinusoids (Fig. 9).

The junctional and glandular zones. The junctional zone (Fig.
5, jz) at the placental base consists of a series of dilated glandu-
lar bays or compartments which are lined by epithelium and
contain histotrophe (Fig. 10). Long tongues of fetal tissue, each
consisting of a core of vascularized mesenchyme which is cov-
ered by a single layer of columnar trophoblastic cells, project
into the glandular compartments (Fig. 12 j. Each of the bays
seen in histological sections has a floor and lateral walls clothed
by uterine epithelium (Figs. 10, 11). The floor of the compart-
ment is usually smooth, whereas the partitions have complex
outlines produced hy numerous irregularly-shaped, mucosal
leaves or folds, more easily illustrated than described (Fig. 11).
The proximal parts of the partitions, which extend into the
placenta for a considerable distance (Fig. 5), are lined by
cuboidal or low columnar cells, many of which have undergone
sj'mplasmic fusion and frequently contain conspicuous vacuoles
of variable sizes (Figs. 11, 12). The distal portions of the parti-
tions exhibit an increasing amount of symplasmic transforma-
tion of the epithelium, and the ultimate branching ends of the
septa appear to be disintegrating, resulting in the formation of
acellular, amorphous histotrophe -which fills the glandular lumens
(Fig. 12).

The trophoblast covering the processes of fetal tissue project-
ing into the glandular compartments, consists of columnar cells
which vary somewhat in height and appearance (Fig. 12). The
cytoplasm contains granular material and is frequently delicately
vacuolated. These cells are quite eviclenth^ engaged in absorbing
and phagocytizing the histotrophe which bathes their surfaces.
At the bases of the projections the columnar cells become lower
and change by transition into the syncytial trophoblast of the
placental labyrinth.

A decidual response Avith the formation of typical large de-
cidual cells is not evident in the stroma of the junctional zone.
It Avill be recalled that decidual cells Avere also absent in the
labyrinth.

In the present specimen Avhieli very prubaldy represents a stage
from the second half of gestation, the glandular zone comprises

96 BULLETIN: MUSEUM OF COMPAEATIVE ZOOLOGY

the dilated, superficial gland segments of the junctional zone
(spongy zone), which have just been described, and deeper,
relatively unaltered parts of the glands of the residual mucosa
^vhich extends down to the muscularis (Fig. 5, gz; and 10). The
glandular septa penetrating the junctional zone contain the
trunks of the maternal arteries and veins supplying the pla-
cental labyrinth.

Hematouias. One of the most conspicuous features of the
wolverine placenta is the large central hematoma (Figs. 1, 2, 3),
which consists of a multilocular sac formed by the chorionic
membrane and containing extravasated maternal blood. It
communicates through a gap in the placenta with the sub-
placental uterine mucosa (Figs. 2, 4). Its location, in relation to
the placenta, most nearly resembles that in the martin, otter and
badger (cf. Mossman. "37. Plate 20; Amoro.so, '52, Fig. 15.12).

The several minute elevations varying from brown to dee])
red, visible on the fetal surface of the placenta, are. on a minia-
ture scale, similar in nature to the central hematoma. Thev con-
sist of small sacs of folded chorionic membrane surrounding a
cavitv filled with varving amounts of extravasated maternal
blood (Figs. 13, 14, 15, 16). Some contain deeply stained, recently
extravasated blood (Fig. 16), others are somewhat older and
less deeply colored (Figs. 1, 13), while the oldest are quite pale,
both macroscopically and microscopically, and the blood within
them has been for the most pai't resorbed (Fig. 14).

Although seldom evident to the naked eye, circumscribed mar-
ginal hematomas of microscopic size are occasionally present
along the l)orders of the zonary placenta (Fig. 6). In addition,
microscopic examination reveals considerable, diffuse, parapla-
cental bleeding, and blood resorption by the chorion (Fig. 19).
These microscopic extravasations are equivalent to the macro-
scopically visible brown and green borders of other carnivores
(cat, dog). In all of these sites where maternal blood extrava-
sates, whether it be in the large central hematoma or in the
small macroscopic and microscopic areas, the pictures of the
chorionic epithelium are essentially similar. The variously in-
folded chorionic membrane is lined by tall, columnar, phago-
cytic cells which possess cytoplasm variably tilled with minute
vesicles, granules, disintegrating erythrocytes and pigment
(Figs. 15, 18, 19). '•Chorionic vesicles" of th(^ type depicted in

WISLOCKI AND AMOROSO : WOLVERINE PLACENTA 97

the cat (Amoroso, '52) have not Ijeen eueouiitered in the repre-
sentative, but random, sections of the wolverine placenta.

The chorio-aUanfoic )iie))ihranc. Close to the margins of the
placenta, the chorio-allantoic membrane is covered by tall, col-
nmnar, phagocytic epithelium (Fig. 19). On leaving the vicinity
of the placental borders, the epithelinm soon passes over into a
low columnar form characteristic of the membranous chorion in
general (Figs. 25, 26). Membranous chorion of the latter char-
acter separates the two lesser cotyledonary placental masses from
one another and from the main zonary placenta (Fig. 1).

The "rosettes" of large clear cells in the chorionic membrane
of the bitch, interpreted by Amoroso ('52) as being the equiva-
lent of the areolae of some ungulates, have not been encountered
in the random sections made of the wolverine's chorion.

The yolk soc. Under the low power of the microscope, the yolk
sac appears as a thin-walled cavity (Fig. 23) lined by deeply
stained, possibly pseudostratified cells, containing large empty
vesicles (Fig. 24). The latter perhaps represent fat which has
been removed.

The umhiUcaJ cord. The umbilical cord from the umbilical
ring to its placental insertion is extremely short (Fig. 1). Close
to the umbilical ring it contains two umbilical arteries and one
umbilical vein, an allantoic duct and a A'itelline artery and A'ein
but no vitelline duct. The allantoic duct forms a cleft lined by
flattened epithelial cells. Several millimeters farther out. the
umbilical vein divides into two vessels of about equal caliber
(Fig. 21). and still farther out, the vitelline vessels undergo
further division. The allantoic cleft increases progressively in
size. The mesenchyme in the center of the cord, in the region
around the umbilical vessels and the allantoic duct, is denser
than the more loose-textured stroma on the periphery of the
cord. The denser, central field of mesenchyme is richly vascular-
ized by small blood vessels (Fig. 22), a condition not previously
observed in a carnivore but of common occurrence in the umbili-
cal cords of cetaceans and various ungulates (cf. AVislocki, '33).

DISCUSSION

Comparison of the placenta of the wolverine with those of
other carnivores, with respect to its gross form and the presence

98 BULLETIX : MUSEU.M OF COMPAEATIVE ZOOLOGY

of the large, central hematoma, indicates that its closest placental
affinities are Avith the martin, otter and liadger (cf. Mossman.
'37, pi. 20; Amoroso, '52, figs. 12, 15). According to Weber
('28) and Simpson ('45), these fonr animals all belong to the
Mustelidae, but both authors believe that the wolverine and
martin are more closely related and consequently they place
these two members together in one of several subfamilies.

Comparison of the microscopic structure of the wolverine's
placenta with that of other carnivores again reveals a closer
affinity to other mustelids than to other fissipeds for which data
are available. The structure of the chorio-allantoic placenta of
the wolverine is almost identical with that of the ferret and
Zorilla. Common to these forms is the characteristic enlarge-
ment or hypertrophy of the endothelium which lines the maternal
placental capillaries, and the prominence of the basement mem-
brane upon which the endothelial cells rest. Comparisons of the
microscopic structure of other parts of the placenta and of the
fetal membranes are not feasible because of the paucity of de-
tailed information on other mustelids.

The placenta of the wolverine is an endotheliochorial one ac-
cording to Grosser 's classification. Actually, however, the ma-
ternal vascular channels of the labyrinth are composed of both
endothelial cells and a prominent basement membrane upon
which the endothelium rests. This membrane is probably a
secretion of the endothelium which serves to hold the cells in
place to constitute blood vessels. The unusual thickness of the
basement memln-ane is perhaps correlated Avith the exceptional
size and activity of the hypertrophied endothelial cells. In view
of the fact that the maternal capillary sinusoids of the placental
labyrinth of all carnivores possess a relatively heavy basement
membrane and some also are accompanieil by hypertrophied
decidual cells (e.g. cat), the term '"vasochorial" might Avell be
substituted for " endotheliochorial" in designating this placental
type (Wislocki, '55).

SUMMARY

The placenta of a wolverine {(riilo gulo lasctis (Linnaeus)) is
described. It is a zonary placenta Avith a large central hematoma
which has its closest affinities to the placentas of other mustelids,

WISLOCKI AX[) AMOROSO: WOLVEKIXE PLACENTA 99

particularly the martin, otter and badger. The placental lab-
yrinth is characterized by greatly hypertrophied endothelial
cells which line the large maternal capillaries, and by the prom-
inence of a basement membrane which completes the wall of the
maternal vessels. In two other mustelids, namely the ferret and
Zorilla, the maternal vessels are similarlv constructed.

REFERENCES CITED

Amoroso, E. C.

1932. Placentation in Marshall's Physiology of Repioduetiou, ed. by
A. S. Parkes. Longmans, Green & Co., London, vol. 2, pp. 127-
134.
BiSCHOFF, T. L. W.

1865a. Ueber das Vorkomnien eines eigenthiimlichen Blut luul Hama-
toidin enthaltenden Beutels an der Placenta der Fisehotter
(Lutra vulgaris). Sitzungsb. Akad. Wissenseh., Miinc-hen, 1:
214-225.
BiSCHOFF, T. L. W.

1865b. Ueber die Ei-und Phuentabildung des Hein und Edel Marders
(Mustela foina und martes) und des Wiesels {M. vulgaris).
Sitzungsb. Akad. Wissenseh., Miinc-lien, 1: .339 348.
GOMORI, E.

1941. Observations with differential stains on human islets of Langer-
hans. Am. J. Path., 17: 395-406.
MOSSMAN, H. W.

1937. Comparative morphogenesis of the fetal memliranes and acces-
sory uterine structures. Contrib. Embryol. Carnegie Inst., 2G:
129-246.
Rau, a. Subba

1925. ContrihutLons to our knowledge of the structure' of the phicenta
of Mustelidae, Ursidae and Sciuridae. Proc. Zool. Soc. London,
B: 1027-1070.
Rausch, Robert

1953. On the status of some Arctic mammals. "Arctic," J. Arct.
Inst, of North America, 6: 91-148.
Robinson, A.

1904. Lectures on the early stages in the development of mammalian
ova, and on the differentiation of the placenta in the different
groups of mammals. J. Anat.. 38: 186-204, 325-340, 485-502.
Simpson, G. G.

1945. The principles of classification and a classification of mammals.
Bull. Amer. Mus. Nat. Hist., 85: 1-350.

100 BULLETIN : MUSEUM OF COMPAKATIVE ZOOLOGY

Strahl, H. and E. Ballmann

1915. Embryonalhiillen mid Placenta von Putorius furn. Abli. preuss
Akad. Wissenseh., Berlin, 1-69.
Weber, Max

1928. Die Saugetiere. Gustav Fisc-her, Jena. Vol. 2. Carnivora, pp.
299 354.
WiSLOCKI, G. B.

1933. On the placentation of the harbor porpoise (Phocaena pho-
coena (Linnaeus)). Biol. Bull.. 65: 80-98.
WiSLOCKI, G. B.

1955. Gestation: Transactions of First Conference. The Josiah Maey
Jr. Foundation. New York.

PLATES

PLATE 1

Fig. 1. Drawing of one of the gestation sacs of the wolverine, which has
Iteen opened to show the character of the zonary placenta nnrl the fetal
membranes. The drawing is described in the text. X 1.20.

^-^r^,/

PLATE 1

PLATE 2

Figs. 2 and 3. Perpendicular, gross sections through the center and periph-
ery of the central hematoma of the wolverine placenta. In the former, the
uterine wall had separated and was removed. The orifice of the hematoma
in tlie center of the placenta is visible in Figure 2. X l.'2o.

Fig. 4. A diagram of an unopened gestation sac illustrating the position
and shape of the orifice of the central hematoma. It is interesting to com-
pare this pattern with different ones encountered in other carnivores (cf.
Mossman '37. Plate i!0 ; Amoroso '52. Fig. 15. 12).

■ A'

mmmmsm

.'»•*• ^•*. •

•.V:/^

\-: :'■ :■:■'.' '• ■ ■' :'. :\ : ■: ■::'■ '■■•'< !•> • •

PLATE 2

PLATE 3

Fig. 5. A low power, mk-roscopic view of the tyincal appeaianee of the
zonarv placenta of the wolverine. A maternal artery is visil)le (arrows)
jnirsuino- its way toward the fetal surface of the placental hJiyrinth where
it liranclies. J~, junctional zone; gs, deep glandular zone. X 9.

Fig. 6. The border of the wolverine's placenta, at a point where there is
a small, round hematoma as well as hemotrophe and histotrophe between
the nu'mbranous chorion on the one hand, and the placental labyrinth and
tlie parajdacental junctional zone on the otlier hand. The field enclosed in
a rectangle is shown at higher magnification in Figure 17. The uterine
lumen, mucosa and musculature are visible at the very bottom of the picture.
X 11.25.

PLATE 3

PLATE 4

Fig. 7. The placental labyrinth shown at high magnification, illnstrating
the maternal sinusoidal capillaries lined by hypertrophied endothelial cells
which rest upon a liasement membrane which is intensely stained by the
periodic acid-Schiff reagent. Surrounding the basement membrane the
trophoblast is visible; it is syncytial in chaiacter and contains nunun'ous
■■intraepithelial" fetal capillaries. Pale, unstained, relatively acellular
stroma without any basement meml)rane, abuts the trophoblast. X 200.

Fig. 8. A portion of the laliyrinth stained by Masson's trichrome mix-
ture. With this staining method, the endothelium lining the maternal vessels
and the syncytial tropholilast is well differentiated, but the intervening
basement membrane which is selectively stained by the ]ierio(lic acid-Schiff
reagent (Fig. 7) is entirely unstained. X '200.

Fig. 9. A portion of the laljyrinth close to the fetal surface, illustrating
the spurious presence of giant cells or syncytial tags, resulting from the
tangential planes in which the endothelium of the recurrent maternal ves-
sels in this region is cut. X 152.

PLATE 4

PLATE 5

Fig. 10. Higher power niicroseoijit- view of the junctional and glandular
zones of the wolverine's placenta. Compare with Figure 5 (./~, gz). Two
entire glandular compartments and parts of two others are visible; their
lumens are filled witli histotro]ih(' and tlie i)artitions between them carry tlie
afferent and efferent i)lacental maternal blood vessels. X 32.

Fig. 11. Illustrates a single glandular compartnieut with a simple tioor
and elaborate side Avails; the walls consist of folds or leaves of stroma, cov-
ered by uterine epithelium some of which is syncytially transformed and
contains vesicles. X 144.

Fig. 12. Another compartment, showing a tongue of fetal stronui covered
l)y columnar trophoblastic cells projecting into a mass of amorphous histo-
trophe in the chamber. The fetal projection is flanked on its sides by folds
of vesiculated nuiternal mucosa, some which is visibly disintegrating to
form histotrophe. X 144.

PLATE 5

PLATE 6

Fig. 13. A section through the larger of the two minute hematomas
on the internal surface of the placenta near the insertion of the umbilical
cord in Figure 1. In the upper corner of the ijhotograph there is a similar
lieniatoma of microscopic dimensions. X 40.

Fig. 14. A section through the small, pale nodule on the internal surface
of the placenta near the attachment of the large central hematoma (cf.
Fig. I). This hematoma consists primarily of elaborate infolded masses
of the chorion with very little actual extravasated blood in its interior. It
is interpreted as being an older hematoma in which resorption is nearly
completed. X 32.

Fig. 1.5. A higher jjower view of part of the wall of the hematoma seen in
Figure 13, showing folds of chorion dipping into the interior of the extrava-
sated maternal l)lood. The low columnar trophoblastic cells covering the
clKjriduic folds are phagocytic and are actively engaged in resorbing the
blood. X 144.

Fig. l(i. A view of an extremely minute, very recent hematoma in con-
tact with the placental labyrinth at a site where the extravasation of blood
occuired probably fiom neighboring maternal sinusoidal capillaries. X 100.

^^l

^v?!?<^. !?»^

7 '^ ■« i " ■ ■' ^

" , ■■' --■ ■"■'.» ■ • ' -v' ■' ■

"^^*r

^^ *t:"';>^;-^- ';''

Sfo^^^' 16:

PLATE 6

PLATE 7

Fig. 17. A field at the border of the placenta, magnified from the area
shown in the rectangle in Figure 6. The area in this figure delimited by a
rectangle is similar to the one shown at higher magnification in Figure 18.
The two arrows in this pliotograph indicate parts of the paraplaeental mem-
branous chorion similar to that shown at higher magnification in Figure 19.
X -iO.

Fig. 18. A field, similar to the one contained in the rectangle in the pre-
ceding figure, still further magnified, to show a fold of chorion dipping
into a mass of extravasated maternal lilood. The chorionic epithelial cells
are variously laden with disintegrating blood cells, pigment, granules and
vacuoles. X 144.

Fig. 19. Paraplaeental membranous chorion from a site similar to the
ones indicated by arrows in Figure 17, again showing the columnar chorionic
epithelial cells engaged in the resorption of extravasated maternal blood.
X 144.

Fig. 20. Glands in the paraplaeental uterine mucosa. The low surface
epitheliimi lining the uterine cavity is apparent above the glands. The
mucosa is visil)le at nuu-h lower magnification in Figure 6. X 144.

*^%V'''*^

«>*^*^

r.- ■ '

PLATE 7

PLATE 8

Fig. 21. Umbilical cord of the wolverine, hnlf n centimeter from the
umbilicus, showing paii'ed umbilical arteries and veins with the slit like
lumen of the allantois between them and. below, the vitelline artery and
vein. X 16.

Fig. 2'2. A portion of the denser connective tissue in tlie neighborhood of
the umbilical arteries, showing a rich supply of small, interstitial blood
vessels which vascularize the dense, central stroma of the cord. X 160.

Fig. 23. A longitudinal section through the tip of the yolk sac, the
chorionic membrane and the wall of the allantois. Abbreviations: ys, yolk
sac; e.r, exocoelome; cli, membranous chorion; al, allantoic cavity. X 40.

Fig. 24. Th? detailed appearance of a portion of the wall of the yolk sac,
at high magnification. X 260.

Fig. 25. A portion of the membranous (du)rion t)verlying the yolk sac.
The chorionic epithelium is hiwei- and more inactive here than in the para-
placental region. X 2(')0.

Fig. 26. A ptirtion of the membranous chorion at a distance from the
liorder of the i)lacenta. The cells, lu'i-e, are (piite low and inactive. X 260.

\ ■ V

*; - — ' ■ ',/ ■''/' /' , •.'.*»>.? #^--*^

PLATE 8

Bulletin of the Museum of Comparative Zoology

AT HARVARD COLLEGE
Vol. 1 14, No. 4

THE FUNCTIONAL ANATOMY OF THE
DELPHINID NOSE

by Barbara Lawrence and William E. Schevill

With Thirty Figttres

CAMBRIDGE, MASS., U. S. A.
PRINTED FOR THE MUSEUM

February, 1956

Publications Issued by or in Connection

WITH THE

MUSEUM OF COMPARATIVE ZOOLOGY
AT HARVARD COLLEGE

Bulletin (octavo) 1863 — The current volume is Vol. 114.

Breviora (octavo) 1952 — No. 51 is current.

Memoirs (quarto) 1864-1938 — Publication was terminated with Vol. 5').

JoHNSONiA (quarto) 1941 — A publication of the Department of Mollusks.
Vol. 3, no. 35 is current.

Occasional Papers of the Department of Mollusks (octavo) 1945 —
Vol. 1, no. 18 is current.

Proceedings of the New England Zoological Club (octavo) 1899-
1948 — Published in connection with the ^Museum. Publication terminated
with Vol. 24.

BiiUetin of the Museum of Comparative Zoology

AT HARVARD COLLEGE
Vol. 114, No. 4

THE FUNCTIONAL ANATOMY OF THE
DELPHINID NOSE

by Barbara Lawrence and William E. Schevill

With Thirty Figures

CAMBRIDGE, MASS., U. S. A.
PRINTED FOR THE MUSEUM

February, 1956

No. 4 — The Funcfional Anafomxi of the Delphinid Nose'^
By Barbara Lawrence and William E. Schevill

TABLE OF CONTENTS

Page

Introduction 103

Materials and methods . 106

General description of the nasal passage 108

Detailed description of the nasal passage . . Ill

Outer part: blowhole and vestibular sacs Ill

Constriction: tubular and connecting sacs 113

Inner part: nasal plugs and premaxillary sacs 116

Musculature of the blowhole region in Tursiops, Stenella, and Belphinus 118

Introduction ........ 118

Pars nasalis of the maxillonasolabialis 119

Posteroexteruus layer .... 119

Intermedins layer 121

Anteroexternus layer 122

Posterointernus layer .... 123

Anterointernus layer . . 124

Profundus layer ... 126

Nasal plug muscle .... . 126

Pars labialis of the maxillonasolabialis 127

Major intrinsic muscle of the tubular sac 128

Ligament muscle 129

Minor intrinsic muscle of the tubular sac 129

The blowhole mechanism ...... 130

Observations on its functioning . . . . 130

Anatomical explanation of function . . . . 136

Contrast with mysticetes 138

Discussion of previous work 140

Effect of hydrostatic pressure on the blowhole mechanism 147

Summary 148

Acknowledgments . 149

References . . 150

INTRODUCTION

No description or anatomical examination can possibly give an
adequate picture of a free-swimming porpoise's breathing cycle.

Nor can they convey a sense of the urgent purpose behind the

iContribution uo. 79.3 from the Woods Hole Oceanographic Institution

104 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

grace and power with which it rolls to the surface and the split-
second timing when it goes down. The strong rush of air as it is
expelled is distinctly audible, the slight gasp of the far briefer
inhalation much less so. There is a definite feeling of force be-
hind this swift exchange of air that is never apparent in land
mammals even during hurried or labored breathing. A thin puff
of moisture more often than not accompanies a blow, especially
when the porpoises are swimming fast and so speeding up the
breathing cycle. In Tur slops these puffs are sometimes visible
for quite a distance and frequently betray the presence of in-
dividuals that are rolling low and otherwise quite difficult to see.
Occasionally the spouts are larger, spreading rapidly and mush-
rooming at the top, suggesting those of a humpback whale in
miniature.

To a casual observer all this seems something to admire rather
than to marvel at, and yet here is a mammal which is keeping its
ancestral dependence on air while living its entire life in water.
No wonder zoologists have long been fascinated by the adapta-
tions arising from this most successful compromise. Of these,
one of the most obvious and yet the most baffling is the breathing
of the porpoise. How does it keep the air in or the water out?
Investigation of this problem was begun a long time ago ; as early
as 1787 Plunter considered it in his "Observations on the struc-
ture and oeconomy of whales." Since then numerous workers
have reported on different aspects of the blowhole region, but
none of them with sufficient accuracy of detail about the structure
and interrelationship of the parts involved to make possible a
correct interpretation of their function. More particularly, de-
scriptions that deal with muscle layers do not give sufficient
information about the thickness of the different layers, the way
in which they merge at origin or insertion, or the actual eft'ect
of their contraction. Furthermore, descriptions of the nasal
sacs lack adequate detail on their position and that of their
openings relative to the rest of the nasal passage. This is no
reflection on the diligence of these early workers ; rather it is an
expression of the difficulty of distinguishing these details in any
l)ut fresh (or recently thawed) material, and of the importance
of having many specimens available.

It should also be borne in mind that the focus of much of this

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 105

earlier work has been on the possible homology of structures of
the blowhole region with the nasal cartilages, bones of the nose,
and facial muscles of land mammals. As a result, function has,
in a sense, been speculated on rather than investigated. Such
speculation, not checked on living animals, has produced a con-
flicting mixture of fact and fancy, and even the most careful
workers have failed to grasp some of the most important points.

The purpose of our work has been to unravel some of this
confusion. In this paper we have attempted to explain the move-
ment of air in the outer nasal passages, how the blowhole is
opened, how it closes, and how the seal is tightened. Since such
explanation will be easier to follow if the myology of the blow-
hole region and the interrelationships of the nasal sacs and main
air passage are understood, these will be described first. We
have accepted Huber's (1934, p. 131) opinion that the rostral
muscles are the pa7's JabiaUs of the maxillonasolahialis and that
all the layers of blowhole muscle are derived from the pars nasalis
of the maxillonasolabialis. We recognize six such layers, all
functionally distinct, which we have called : posteroexternus
ipe), intermedins {i), ant eroext emus {ae), poster oint emus (pi).
anterointcrnus (ai), and profundus (pr) (Figs. 10-17). This
recognition implies a more clearcut anatomical separation than
actually exists. In our cross sections the only obvious division is
between pi and ai (Figs. 7-9) ; also various layers, notably pr, in-
clude more than one section, but where there is no sharp distinc-
tion either in origin, insertion, or direction of fibers, such
subdivisions within a layer have not been considered separately.

In studying the nasal sacs we have found that only the pre-
maxillary sac has been consistently so identified. This term we
have continued to use; otherwise we have selected appropriate
descriptive names. For the sake of simplicity the descriptions
are usually of the right side only ; hence the sacs, although paired,
are usually referred to in the singular. In keeping with the
characteristic asymmetry of the skull (Fig. 1), the passage and
associated sacs on the left are all less developed than those on
the right.

The part of the nasal passage here considered is the distal part
which lies superficial to the openings of the paired bony nares,
in other words that part which lies in the soft tissues of the head
between the bones of the skull and the protuberance of dense,

106 BL'LLETIX : MUSEUM OF COMPARATIVE ZOOLOGY

fatty, connective tissue above the rostrum, commonly called the
melon.

MATERIALS AND METHODS

This paper is based on a stud}^ of both dead and living ma-
terial. Detailed dissections were made of two Tur slops truncatus,
two Stenella plagiodon, and one Delphinus delphis, all preserved
by freezing'. This kept the material in excellent condition even
when partially thawed for dissection and then refrozen repeat-
edly. Cross and sagittal sections were made of two fresh adult
Tursiops, and sagittal sections and dissections of a fetal one
preserved in formalin. Dissections and comparative observations
were also made on a number of Glohicephala melaena and a
single specimen of Lagenorhynchus acutus. Latex casts of the
passage of Stenella and Delphinus were generously supplied by
Dr. Henrj^ Kritzler. In addition, the movement of air in the
passage was studied l)y pumping air up the trachea of specimens
of Tursiops, Delphinus, and Stenella. Most of the observations
on live animals were made during the course of four visits to
Marine Studios. Marineland, Florida. Here there was ample
opportunity to watch, from both above and below the surface of
the water, the breathing cycles of captive Tursiops, and less
frequently of Stenella. Motion pictures taken at 64 frames per
second show the relation of closing of the blowhole to submerging,
both in groups of animals swimming at various speeds and in
two different individuals idling near the surface. Other close-up
pictures taken at the same speed show, in detail that the unaided
eye cannot catch, how various parts of the passage move at such
times.

Finally, and most important, investigations were made inside
the blowhole of a live bull Tursiop.s which was kept under such
observation for three days. By holding a finger in various parts
of the passage (on occasion as far down as the bony nares) for
extended periods between and during blows it was possible to
feel both muscular and pneumatic motion. It was also possible
to introduce a rubber tube and take pressure readings with a
mercury manometer in different parts of the passage.

The behavior of our live animal during the three days when
we had him under ol)servation is of considerable interest in view

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 107

of a rather widely held opinion that porpoises will not tolerate
tampering with the blowhole. This is probably based in part on
Triie's observation (1891, p. 202) that stranded Tursiops at
Hatteras "invariably showed signs of discomfort by lashing the
tail violently" when he touched the blowhole. In contrast, our
bull never thrashed around until in his death flurry, when he was
apparently unconscious. The rest of the time he was remarkably
passive, whether afloat or aground. This does not mean that he
submitted with complete tranquillity to our investigations. His
efforts to hold his nasal passage closed will be described later.
In addition, as he lay on his side, he occasionally bent his head
down away from the investigator, at the same time lifting it
slightly from the deck, and sometimes opened his mouth a little
as if gagging. This occurred particularly w^hen the plugs were
pulled open by hand, allowing air to escape, or when he was
probed near the entrances to the tubular and connecting sacs.
These occasional efforts seemed to be to aA^oid rather than to
attack the investigator. The only time he snapped was once when
we attempted to muzzle him with a rope, a precaution we gave
up as unnecessary.

Irving, Scholander, and Grinnell (1941, p. 158), when working
on the respiration of Tursiops, found them so sensitive to inter-
ference Avith their breathing that on occasion individuals as-
phyxiated themselves. These authors stated that "if a little
water enters the nostril during inspiration, the animal often will
not breathe again." This certainly was not the case with our
animal, even when we ourselves opened his blowhole under
water. Apparently neither the presence of water in the outer
nasal j^assage nor interference with the mechanics of l)lowhole
action are by themselves sufficiently disturbing to inhibit l)reath-
ing. Some other factors must also be involved.

This individual (length 262 cm., weight about 260 kg.) was
not in perfect health when caught, as shown hj an extraordinary
amount of flatulence and a rather bloated appearance. For this
reason the mild behavior described above raav not be entirelv
typical ; it is, however, in keeping with our previous observations
on these creatures.

108 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

GENERAL DESCRIPTION OF THE NASAL PASSAGE

A brief general description of the nasal passage, together with
the accompanying diagrams, will help to orient the reader and
enable him to understand the subsequent detailed discussion of
the structure and functions of the various parts.

The 'inner or bony nares pass upward through the skull as a
pair of simple, somewhat rounded tubes between the cranium
and the rostrum (Fig. 5). Being surrounded by bone, they are
not collapsible, and they lack diverticula. Paired plugs occlude
their upper ends, shutting them off from the collapsible and far
more complicated superficial portion of the passage which slopes
more anteriorly (Fig. 2). In this paper the term "plug" refers
to the mass of tissue which occludes the upper end of the bony
nares, is moved by the nasal plug muscle, and has a lateral ex-
tension on the premaxillary bone.

For most of its extent the distal nasal passage is paired, being
divided by a cartilaginous extension of the bony septum which
separates the internal nares (Figs. 1, 5, 19, 23). The blowhole,
or opening through Avhieh the passage communicates with the
exterior, is single, smaller than the rest of the passage, approxi-
mately crescentic in outline when closed, and nearly round when
wide open (Figs. 26-28). Essentially, the main air passage may
be considered as made up of (1) a relatively fixed posterior wall
sloping downward and backward from the posterior lip of the
blowhole, and (2) a highly movable anterior wall comprised of
the valve and the plugs, which are closely coapted to the posterior
wall (Fig. 2). Bisecting the nasal passage in the midline a short
distance within the blowhole is the cartilaginous septum men-
tioned above. Anteriorly this attaches to the wall of the passage
between the plugs (Figs. 5, 19) ; posteriorly it attaches to the
bony wall of the passage and thence curves upward and forward
beneath the posterior lip (Fig. 28). The plugs are freely movable
on each side of this septum, which itself limits the extent to
which the passage can he drawn open anteroposteriorly.

That the apparent simplicity of a view down the open blow^-
hole is very misleading is immediately discovered when the closed
passage is in^•estigated. Then one finds that the overlapping
posterior wall fits so snugly on top of, as well as behind the
anterior wall (Figs. 2, 3) that a probe thrust down must follow

LAWRENCE AND SCIIEVILL : THE DELPHINID NOSE 109

a tortuous course to reach the bony nares. As a further compli-
cation there are four pairs of diverticula (Figs. 1, 20), two of
which, the premaxillary and the vestibular, are broadly con-
tinuous with the passage and so widen it to a diameter far greater
than that of the nares or blowhole.

The paired premaxillary sacs are the largest (Fig. 1) and lie
immediately on top of the smooth portion of the premaxillary
bone around the opening of the bony nares. They communicate
with the nares internal to the plugs along the entire narial border
of the premaxillae, and their superficial wall appears as an
extension of the integument covering the nasal plugs (Figs. 2-
4). For the most part, the premaxillary sacs are internal to the
plugs, but each has a small opening at its posterolateral corner
external to the plug (Fig. 22) through which air may pass into
the outer passage even when the plugs are seated.

The outermost pair of sacs, the vestibular, are posterolateral
to, and lie immediately within the lips of the blowhole (Figs.
10, 12) . Thej^ are highly distensible and, when uninflated, roughly
half the size of the premaxillary sacs. Like the latter they are
continuous with the main passage, so that it is difficult to deter-
mine where the one begins and the other ends. Together, passage
and sacs form a widened vestibular area whose inner limit is the
transverse slit where the plugs pass beneath the overhanging
posterior wall (Figs. 21, 22). This slit or constriction is of par-
ticular importance, as it is here that the porpoise can volun-
tarily tighten the closure of the passage. There is a considerable
mass of tissue lying between the vestibular area and the dorsal
surface of the plugs (that is to say, beneath the ventral surface
of the vestibular sac) and it is here that the last two pairs of
sacs, the iuhular and the connecting, lie embedded. These are
both true sacs in the sense that they are perfectly distinct from
the main air passage and connect with it only via rather small,
well-marked openings. The tubular sacs are U-shaped, lie almost
horizontally in the head, and on each side surround the slitlike
opening between the top of the plugs and the posterior wall (Figs.
1-3). A transverse ligament beneath the posterior limb of the
tubular sac helps to stiffen the posterior wall of the passage here
(Fig. 20). The connecting sacs appear as small appendices
lateral and internal to the tubular sacs, into which they open
(Figs. 1, 20, 22). They also lie directly over the external opening

110 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

of the premaxillary sac described above (Fig. 22), so that there
is a direct passage for air from the outer margin of the premaxil-
lary sac via the connecting sac into the tubular sac. When the
tubular sac is thus inflated it pneumatically tightens the seal of
the inner passage.

Surrounding the passage and sacs are six layers of blowhole
muscle. Omitting the small and often poorly developed inter-
medins layer, the other five take origin over a wide area in the
fronto-maxillary region (Fig. 23) and are found to be in two
groups which insert respectively posterior and anterior to the
passage, but alternate, and overlap broadly at their origin. Thus
the anteroexternus partially separates the two posterior layers,
while the anterior portion of the posterointernus lies between the
anteroexternus and the anterointernus (Fig. 11). Where the two
posterior layers are not separated by an anterior layer they tend
to merge. Thus while these four layers are functionally distinct,
they are by no means anatomically so throughout their entire
extent. The remaining, or deepest layer, the profundus, is dis-
tinguished from the anterointernus by its more intimate connec-
tion with the rostral muscles and its insertion, chiefly in the
melon (Fig. 16). The effect of these muscles on the sacs is inci-
dental to their control of the passage ; the vestibular and pre-
maxillary sacs lack intrinsic muscles. The tubular sac is unique
in being provided with a weak l)ut rather extensive intrinsic
musculature and the connecting sac is partially affected by a
portion of this. The nasal plugs have their own distinct muscles.

In summary, then, the distal passage maj' be considered to be
divided into two rather wide spaces separated by a slitlike con-
striction. For a more detailed description it will be convenient
to deal with these three regions separately, the blowhole and
vestibular sacs forming the outer part, the tubular and con-
necting sacs and associated parts forming the constriction, and
the nasal plugs and premaxillary sacs forming the inner part.
In fundamental plan the distal nasal passage is very similar
in all three forms studied, so our description of Tnrsiops trunca-
tus is equally applicable to Stenella plagiodon and Delphinus
delphis. The same is true of the rather general remarks on the
musculature of this region. A more thorough description of the
muscle layers, as well as an account of certain slight differences
between the three forms, is given below (pp. 118-129).

LAWRENCE AND SCIIEVILL : THE DELPIIINID NOSE 111

DETAILED DESCRIPTION OF THE NASAL PASSAGE
Outer Part : Blowhole and Vestibular Sacs

The blowhole is itself somewhat asymmetrically curved with
its convexity directed backward. The anterior lip, or valve,
slopes back to pass under the posterior, thus forming a slight de-
pression which often holds a little water when the animal sur-
faces. The lips are thick and non-muscular, being composed of
what appears to be very fibrous, less oily blubber than that w'hich
covers the rest of the animal. The posterior lip in particular is
stiff and densely fibrous ; in a live animal only the margin along
the middle third was found to l)e of softer, freely movable tissue.
Since the lips have no muscle attachments in them, they can
move only as the underlying tissue is moved. Thus the thick,
fibrous mass of the posterior lip, projecting as it does well
l)eyond the subjacent muscles, moves relatively little, while the
anterior lij) or valve moves freely as part of the mass of connec-
tive tissue on which it lies. The valve in the living animal is soft
and pliable to the touch, and has a well-marked margin where it
passes beneath the posterior lip. It slopes rather steeply in the
midline, and is continuous below and laterally with the two
j)lugs, which are separated by the septum. On either side the
passage widens into a large sac, the vestibular, which when un-
inflated lies lateral and somewhat posterior to the blowhole (Fig.
10). At the lips of the blowhole the smooth wall of the passage
passes gradually into the more wrinkled lining of the sac with
no real division between sac and passage (Fig. 13).

On the under or forward side of the passage the valve passes
with no sharp boundary into the anterior fold of the vestibular
sac (Figs. 2, 13, 21). This fold divides the anterior part of the
sac into two unequal pockets, an inner one which is shallower
anteroposteriorly, and an outer one which is much deeper and
communicates directly with the blowhole opening. The fold is
thickest at its medial end, and here takes origin midway between
th(^ margin of the valve and the jdug. Laterally it tapers to
nothing just beyond the end of the slitlike opening ])etween the
plug and the posterior w^all. When the passage is closed and
the sac uninflated, this fold covers the slit, thrusting against the
less well developed posterior fold which forms a soft yielding
mass Ix^tweeu the stiff overhanging posterior lip and the almost

112 BILLETIX : MUSEUM OF COMPARATIVE ZOOLOGY

equally stiff convex inner mass that interlocks with the plug
(Figs. 13, 19). This posterior fold has no real pocket beneath
it and appears merely as a puckering of the wall of the uniu-
flated sac. Various degrees of inflation change the shape of this
sac in various ways, as is shown by latex casts. When it was
slightly inflated, the latex reached only that part of the sac
beneath the anterior fold; when strongly inflated both folds are
obliterated. The vestibular sacs lie close beneath the blubber
and are separated externally from the thick integument around
the blowhole by the thin aponeuroses which form part of the
insertion of the outermost layers of blowhole muscle, pe, i, and
ae (Figs. 10, 11). Posteriorly these sacs are tightly bound to
the insertion of the underlying muscles, so that here they appear
thick walled, while anteriorly they are thinner walled and more
loosely held in the investing tissue.

As well as enclosing the sac, the outermost layers of blowhole
muscle insert internal to the dense collagenous connective tissue
mass comprising the outer lips. As with the rest of the blowhole
muscles, these layers are not entirely distinct from the surround-
ing layers, but usually merge with them at either origin or in-
sertion so that it is extremely difficult to separate them on
dissection. For this reason the pull that they exert is on an
area, not a particular spot, and their contraction has a general
rather than a local effect. The four outermost layers, pe, i, ae,
and pi (Figs. 10-13), are chiefly concerned with opening the
blowhole by drawing the valve down and forward, and by pulling
upward and backward on the walls of the nasal passage which
lie immediately inside the rigid posterior lip. These layers fur-
ther act to hold the vestibular sac in place. Pulling on them does
not stretch this sac, neither does it keep it from expanding. They
do however fix this sac, determining the direction in which it
expands when air is forced into it from below. Furthermore,
when contracting to open the outer part of the blowhole, the
outermost layers of muscle exert pressure which tends to flatten
the sac, thus preventing air from being impounded there as the
l)lowhole closes.

In opening the passage, pe and pi act chiefly on the posterior
part. Pe draws up and back the outer wall of the vestibular sac
and that portion of the passage which lies just inside the posterior

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 113

lip. The tendinous insertion of pi (Fig. 14) curves over the
nasal bones to a deeper attachment, chiefly in the slightly de-
veloped posterior fold of the vestibular sac and the tissue ventral
to this, including the outer side of the tubular sac. Posteriorly
pe and pi are not separated by other layers, and at their insertion
are not truly distinct. Here their thick tendinous mass fills the
space between the nasal bones and the nasal passage, and, where
the vestibular sac lies on top of this mass, it is closely attached
to it.

While pe and pi have their main mass posteriorly and control
the posterior part of the nasal passage, ae has its main mass
anteriorly and acts on the anterior part of the passage (Figs. 11,
13). It is largely responsible for depressing the valve by pulling
the tissue beneath it anteroventrally, and further helps to open
the nasal passage by pulling the well-developed anterior fold of
the vestibular sac anterolaterally as well as ventrally, away from
its position on the top of the slitlike opening of the passage
above the plugs. In addition, ae anchors the underside of the
vestibular sac, so that when inflated it has to expand on top of
ae. Pe, ae, and pi are all separated from the melon and the pars
labialis of the maxillonasolabialis by the collagenous connective
tissue in which the posterior, external fibers of this latter muscle
end (Figs. 11, 13). Layer i is very thin with no bony attach-
ments, and makes a connection, otherwise lacking, between pe
and ae and the melon (Fig. 11).

Constriction : Tubular and Connecting Sacs

Although, for convenience of description, the nasal passage is
here divided into three, there is no real separation between these
regions. As noted above the innermost portion of pi attaches on
the tubular sac, and the slitlike opening at the inner end of the
vestibular area also marks the outer end of the constricted area.
Surrounding the slitlike opening on each side and more or less
embedded in fatty connective tissue are the tubular sacs, which
are paired, asymmetrical, and U-shaped (Figs. 15, 22). The
posterior portion of each of these lies between the nasal bones
and this opening. Together with its surrounding muscles, the
blowhole ligament, and connective tissue, this portion of the
sac forms a transverse, convex band in the posterior wall of the

114 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

passage, which, when the passage is closed, fits into a concavity
on the superficial surface of each plug (Fig. 3). The smaller
anterior portion of the sac lies beneath the valve. The tubular
sacs each open downwards into the nasal passage through a more
or less horizontal slot in the posterior part near the bend of
the U. This slot slopes postero ventral ly for a very short distance
to open immediately dorsal to each plug and slightly, anterior
to where the plug lies against the posterior wall of the bony
nares (Figs. 1, 3, 17, 20, 22). Thus the only entrance to the sac
is from below, and when the passage is closed this opening is
external to the plugs. Mediallj^ the opening lies very close to
the bony wall of the naris; laterally, owing to the slope of the
skull, the opening lies a slight distance in front of this wall.

Lateral to the tubular sac and deeper, lies a small diverticulum,
the connecting sac (Figs. 1, 16, 20), which is separated from the
former by a deep section of the major intrinsic muscle of the
tubular sac, which is here poorly developed and liberally streaked
with collagenous connective tissue ; the connecting sac opens
downward into the nasal passage near the posterolateral corner
of the na-^al plug, and upward into the end of the ventral opening
of the tubular sac (Figs. 20, 22). The opening of the connecting
sac into the nasal passage lies over the lateral opening of the
premaxillary sac. It is Ijounded anterior^ by the blowhole liga-
ment (Fig. 20), to which its ventral surface is closely attached,
so that its entrance does not collapse when the passage is closed.
This makes it easily possible, even when the passage is closed,
for air passing out of the posterolateral corner of the premaxil-
lary sac external to the plug to enter the connecting sac, and
difficult for it to go elsewhere. The blowhole ligament passes
from the premaxillary bone to the lateral extremity of the slit-
like opening above the plugs, where in some specimens it is
stiffened by a small band of cartilage ; from here a few fibers form
a transverse band immediately posterior to this opening and end
in the tissue beside the septum. This ligament is a very important
structure that strongly anchors the commissure of the opening
and makes taut the posterior wall above the plugs. Moreover
it is slightly elastic, so that contraction of pi can lift it away from
the dorsal surface of the plug, but when this layer is relaxed the
ligament snaps back into place. Immediately beneath it and with
similar orientation is a poorly developed slip of the major in-

LAWRENCE AND SCHEVILL : THE DELPIIINID NOSE 115

trinsic muscle, the lio-araent muscle ; together they separate the
connecting sac from the lateral edge (Figs. ]6, 20) of the pre-
maxillary sac.

Of the layers of blowhole muscle, three {j)i, ai, and pr) are
more or less associated with this middle region of the distal nasal
passage (Figs. 13-16). Pi, as described earlier, acts to some de-
gree on the outer part of the passage and has, externally, a
strongly tendinous insertion with anteriorly directed fibers (Fig.
13). Deeper dissection shows the direction of the fibers to be
progressively less horizontal, more anteroventral, and where the
innermost layers attach on the tubular sac there is no intervening
mass of tendon (Fig. 14).

Ai and pr have in common a rather coarse and tendinous struc-
ture, attachments in front of the nasal passage, and an intimate
connection with the pars labialis. Ai has as its primary function
to help ae depress the valve by drawing the tis.sue beneath it
downward and forward. The main mass of ai is anterior, and
the thin aponeuroses which form the insertion of the outermost
layers lie in part on top of the anterior part of the tubular sac
(Fig. 14). Insertion of the deeper portion is more via slender
strands than aponeuroses, and is largely anterior to the tiibular
sac, which is firmly attached to it (Fig. 15). As the posterodorsal
border of this muscle tends to follow the edge of the premaxilla,
and as in this region its fibers course anteroventrally as well as
medially, its contraction would not put pressure on the lateral
corner of the nasal plug, although it would do so on the anterior
part of the connecting sac.

Pr is made up of many subdivisions which arc impossible to
separate fully, but Avliich show a progressive change in the
direction of the fibers (Figs. 16, 17) from the outermost, which
take about the same direction as those of ai, to the innermost,
which are nearly horizontal. At their origin and. except for the
innermost layers, at their insertion, the fibers of pr are continuous
with those of the pars labialis of the rostrum. The innermost
layers curve inward anterior to the nasal plug muscle and be-
tween it and the melon (Fig. 18). Functionally also, the two
muscles work together to draw the mass of the melon down to-
wards the skull and back against the premaxillary sacs, thus
providing a relatively i-igid backing for the nasal ])liig muscles,
and putting pressure on these sacs.

116 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

lu addition to these layers of blowhole muscle, there are also
associated \^dth the tubular sac, and to a certain extent also with
the connecting sac, two pale, ill-defined muscles, the major and
minor intrinsic muscles of the tubular sac (Pigs. 15, 16) and a
small slip of the former here called the ligament muscle ; these
are exceedingly hard to separate from the surrounding connective
tissue. The major intrinsic muscle attaches to the posterior and
underside of the posterior part of the tubular sac, and runs
posterolaterally to end in the fatty connective tissue between
this sac and the skull ; near the bend of the tubular sac the fibers
change direction as shown (Fig. 15), and pass to the underside of
the sac. Anteriorly the muscle fibers become much more sparse ;
the outer ones insert along the anterior margin of the tubular sac
and at their origin are hard to distinguish from the long slender
threads of the tendinous insertion of ai (Fig. 15). This combi-
nation of scattered muscle fibers and fatty collagenous connective
tissue is rather thick in this region, and, as well as attaching to
the underside of the tubular sac, attaches also to the top of the
connecting sac. Closely connected with the anterior part of the
major intrinsic muscle is the ligament muscle, which has its origin
in the fascia beneath the blowhole ligament and inserts in part in
the tissue at the commissure of the passage and in part in the wall
of the premaxillary sac where this attaches external to the plugs.
Wrapping around the anterior part of the posterior half of the
tubular sac is a similarly pale and even less well developed
muscle, the minor intrinsic (Fig. 15), which arises in the fatty
collagenous tissue between the sac and the nasal passage, and
passes posterodorsally to insert beneath, on the anterior edge,
and on top of this part of the sac. Apparently the function of
both the major and minor intrinsic muscles is to help anchor and
hold open the tubular sac. The ligament muscle probabh" helps
also to hold open the external opening of the premaxillary sac.

Inner Part : Nasal Plugs and Premaxillary Sacs

The above described, non-rigid parts of the passage are shut off
from the permanently open Ijony nares by paired plugs attached
in the midline to a cartilaginous septum which divides the passage
and limits the extent to which it can be pulled open (Figs. 19,
27, 28). The openings of the nares face anterosuperiorly so that

LAWRENCE AXD SCHEVILL : THE DELPHIXID XOSE 117

the plugs lie obliquely across the passage on each side of the
septum (Figs. 2, 3). This septum is roughly triangular, with the
hind margin extending along the upper end of the bony septum,
then up the bony wall of the nares and finally curving forward
under the posterior lip; and with the front margin rising more
nearly vertically between the two plugs (Figs. 5, 19). At the
anterior edge of the bony nares the lower margin of each plug is
closely attached to the septum; this then is the pivot on which
the rest of the mass works as it is drawn anteroventrally to open
the passage. At this point the plugs meet the septum almost at
right angles and form a very small posterior projection on each
side of it. As they slope back across the passage they project in-
creasingly far, till at their upper edge septum and plugs meet
at the apex of a Y (Figs. 5, 22). This allows the posterior part
of the plugs to be freely movable along each side of the septum.

The position and extent of the premaxillary sacs is marked on
the skull by a characteristic smooth area on the premaxillae
surrounding the bony nares. Lying as they do beneath the plugs,
these sacs communicate with the inner passage along the narial
borders of the premaxillae (Figs. 1, 23). Their anterior portion
forms a well-defined pocket with no other openings. Along its
peripheral margin the sac attaches to the premaxilla and then
along the maxillo-premaxillary suture where its black wall is
visible when pr is cut away (Figs. 16, 18. 19). Medially the sac
wall is continuous with the integument of the under surface of the
plugs, except posteriorly where it attaches on top of the liplike ex-
tension of the plug instead of being continuous with its margin.
This forms an opening to the sac external to the plugs (Fig. 22)
where air which has passed into the premaxillary sac along the
hony margins of the nares can pass outward into the upper nasal
passage without unseating the plugs. This opening lies close
beneath the opening of the connecting sac, so that when the main
passage is closed, air from here reaches first the connecting sac
and then the tubular sac (Fig. 20).

The dorsal surface of each plug is not horizontal, but slopes
somewhat ventrolaterally and is characteristically smooth-skinned
and concave. Fitting into this concavity is the above-described
convexity foi"med by the posterior limb of the tubular sac, the
blowhole ligament, and the mass of tissue associated with them
(Fig. 3). On their under surface, posteriorly, the plugs are

118 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

roughly oval and somewhat protuberant. The mass of collagenous
connective tissue of which they are formed is thickest here
where the plugs lie snugly against the margins of the opening of
the bony nares. Anteroventrally this cushion becomes thinner
and passes without a well-defined boundary into the membranous
external wall of the premaxillary sac (Fig. 19). Laterally the
cushion of the plugs is more sharply set off from the liplike
extension which lies on top of the premaxillary bone beneath the
connecting sac (Figs. 18, 19). The paired openings of the bony
nares are made oval by a tough membrane (Fig. 19) stretched
across the angle where the maxilla and mesethmoid meet, that is
to say, the posterolateral corner. This membrane is so sitviated
that air from below cannot lift the liplike extension of the plug
up against the opening of the tubular sac, while it can press the
medial part of the plug more tightly against the interlocking
mass of the posterior lip.

The paired nasal plug muscles themselves are distinct from
the layers of the pars nasalis and from the pars labialis. They
take origin in part lateral but chiefly anterior to the premaxillary
sacs and course posteromedially to a diffuse insertion in the col-
lagenous connective tissue of the cushion of the plug, and are
themselves heavily streaked with collagenous connective tissue
(Figs. 3, 19). Their contraction draws the plugs down and
forward.

MUSCULATURE OF THE BLOAVHOLE REGION IN
TURSJOPS, STENELLA, AND DELPHINUS

Introduction

Before describing in detail the various layers of blowhole
muscle (Iluber's pars nasalis of the maxillonasolabialis), a few
general remarks may be helpful. The layers are in thin sheets.
Those controlling the posterior part of the passage are thickest
posteriorly, hard to distinguish from each other in this region,
and, at their origin, extend less far forward. Obviously, also,
because of the shape of the passage they are the more external
layers. Those controlling the anterior part of the passage are
correspondingly thickest anteriorly Avhere they tend to merge
with one another, and extend less far posteriorly. Thin layers of

LAWREXCE AND SCIIEVILL : THE .DELPIIIXID NOSE 119

loose connective tissue separate more layers of muscle fibers than
seem to be functionally distinct. Of the distinct layers the outer-
most are inserted via aponeuroses of varj'ing thickness while the
inner ones have a more tendinous insertion, the individual strands
often being discrete instead of in bundles. It will also be seen
from the figures that while the actual muscle fibers of the super-
ficial layers extend at least as far as the vestibular sac before
being replaced by aponeuroses, the muscle fibers of the deep
layers become increasingly shorter as the passage widens. Other
interesting features are the way in which the anterior and pos-
terior layers alternate with each other, as well as the fanning out
of certain layers. The effect of this rather intricate arrangement
is to balance the forces pulling in various directions, so that while
the passage is being opened it is also anchored against too strong
a pull in any one direction.

The three forms studied resembled each other closely. Such
differences as were found were chiefly in the position and degree
of development of minor subdivisions of the various layers. For
this reason Tursiops is discussed in detail and only the ways in
which the others differ are noted. Some of these differences may
be merely individual variation; the number of specimens dis-
sected is not sufficient to show how great this is ; and the bilateral
asymmetry is considerable.

It should also be noted that the way in which the bones in this
region overlap varies within a species so that, while the region
of attachment of each layer remains fairly constant, the actual
bones involved may vary. For instance, in the specimen figured
the frontal does not intervene broadly between the occipital and
the nasals, while in others it does.

Pars Nasalis of the Maxillonasolabialis

Posteroexternus layer of the pars nasalis of the
maxillonasolabialis

Tursiops {pe; Figs. 10-12, 23)

This is the outermost of the blowhole muscles and takes origin in
a broadly continuous sheet over a wide area. Posteriorly it is
thick, and at its posteromedial edge along the raised margin of
the naso-frontal bones it is hard to distinguish from pi which lies

120 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

beneath it. Anteriorly it tapers to a very thin sheet which is diffi-
cult to separate from i. At its insertion the converging fibers
overlap in a way that makes the muscle appear double. A very
thin inner section is separable onty with difficulty from the main
mass of the muscle and intimately connected with ae where the
two meet.

Origin : On the maxilla, frontal, and occipital ; where the occipi-
tal forms a crest along the occipitofrontal suture, pe attaches
beneath it. Anteriorly pe is difficult to separate from ae.
Insertion: Posterior to the blowhole opening. The anterior part
of the muscle has dorsally directed fibers which pass in a very
thin aponeurosis across the vestibular sac and then continue su-
perficial to the posterior part of the muscle to attach in the denser
collagenous connective tissue of the wall of the nasal passage
just internal to the stiffened, crescentic posterior lip. The pos-
terior part of the muscle has fibers that run anteromedially,
passing deep to the anterior part to insert in the collagenous
connective tissue of the posterior Avail of the nasal passage. It
also inserts on the vestibular sac. When this is flattened and
uninflated these attachments are along the posteromedial mai'gin
and merge with similar attachments of the posterior part of ae.
The anterior part of pe is intimately connected with the loose
connective tissue investing the vestibular sac, chiefly over its
superficial surface. These attachments of pe on the vestibular
sac merge anteriorly with similar attachments of i which here
lies internal to it.

Funciion: The posterior part when contracted pulls back the
posterior wall of the blowhole and the movable, medial part of
the posterior lip ; more laterally it draws up and back the pos-
terior wall of the vestibular sac, including its posterior fold.
The attachments on top of and along the posteromedial margin
of the vestibular sac are such that contraction of this layer does
not stretch the sac, while the aponeuroses which lie on top of
the sac help to flatten it when the blowhole opens.

Delphinus

Resembles Tursiops closely except that the anterior portion
of pe has its origin slightly more posteriorly and the insertion of
this part forms a more conspicuous transverse band across the
insertion of the posterior part. Also the anterior margin of pe

LAWRENCE AXD SCHEVILL : THE DELPPIIXID XOSE 121

and posterior margin of i are more distinct.
Stenella

The fibers of the anterior and posterior sections do not cross
at their insertions as in Tursiops and Delpliinns, bnt converge
evenly. Tn StcneUa, pe differs further in having a more distinct
inner layer which has, however, the same origin and insertion as
the outer layer and so is here considered as part of pe.

Intermedins layer of the pars nasalis of the
raaxillonasolabialis

Tursiops (i; Figs. 10, 11)

This is a small, not very distinct, anterodorsally directed muscle
which lies between pe and ae.

Origin: Beneath pe from the connective tissue sheath of ae.
Insertion: In the fatty, fibrous mass of the melon and on the an-
terior part of the superior surface of the vestibular sac (Figs.
10, 11). Its insertion, via a thin aponeurosis, on the vestibular
sac lies beneath and merges with that of the most anterior part
of pe. In front of the vestibular sac its fibers merge with those
of ae to attach via strong aponeuroses in the connective tissue
of the valve at the commissure of the blowhole, as well as beneath
the valve more medially. Those anteriormo.st fibers which insert
in the melon are quite distinct from the underlying layers, and
it is this distinctness which is the principal reason for recognizing
this layer as a separate one.

Function: Forms the main connection between the outer layers
of blowhole muscle and the melon.
Delphinus

Layer / is broader and better developed than in the other two
genera, with fibers directed more anteriorlv. Posteriorlv a few
fibers originate directly from the maxilla, and the attach-
ments of this layer on the superficial aspect of the A'estibular sac
extend further posteriorh' and nearer the ventral margin of
the uninflated sac.
Stevella

Differs chiefly in having the fibers directed slightly more dor-
sally and less anteriorly than in Tursiops, so that their insertion
in the melon is nearer the bloAvhole ; further, part of the origin
is from the maxilla via a very thin aponeurosis.

122 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Anteroexternus layer of the pars nasalis of the
maxilloiiasolabialis

Tursiops {ae; Figs. 10-13, 23)

While the main mass of this muscle is anterior, it is entirely
distinct from the melon and the pars labialis of the maxillonaso-
labialis, from which it is separated by a thin fibrous sheet.
Origin: Chiefly on maxilla, with a thin extension onto the frontal.
While at its origin it is largely distinct from pe, it tapers pos-
teriorly to a thin sheet of connective tissue which is difficult to
separate from this layer. Anteriorly its connections with the two
layers internal to it are equally close. Here at its insertion on
the maxilla it merges with pi and anterior to this a few fibers are
inseparable from ai.

Insertion: Via strong aponeuroses chiefly in the dense, colla-
genous connective tissue beneath the valve from the commissure
of the blowhole almost to the midline, and continuously from
this onto the anterior fold of the vestibular sac. In addition, thin
aponeuroses attach over the vestibular sac. Anteriorly these are
distinct; posteriorly they merge with similar layers of pe. In
effect, pe, i, and ae enclose the vestibular sac in thin aponeu-
roses, pe and i attaching more to the margin with only a few
sheets passing to the outer and inner sides. As described above,
the insertion of i merges in part with that of ae; the insertions
of ae and pi are, however, entirely distinct.

Function: This layer controls quite a large area. Contraction
depresses the valve by pulling the tissue beneath it and as far
internally as the beginning of the separation of the plugs ; it also
draws the anterior fold of the vestibular sac forward, laterallv,
and down. Counteracting this forward pull on the vestibular
sac, the posterior, less elastic part of the layer helps to hold the
sac in place. The internal attachments of this muscle are such
that the sac must expand on top of, rather than underneath it.

Dclphinns

In general ae resembles that of Tursiops. It differs in having
a very thin l)ut distinct inner layer attaching to the underside
of the vestibular sac and merging with pi posteriorly. This in
Tursiops is probably represented liy a tenuous layer of con-
nective tissue.

LAWRENCE AND SCHEVILL : THE DELPIIIXID NOSE 123

Stenella

Here there seems to be a more obvious separation of ae into
two parts, the second being posterior as well as internal to the
first. Together they perform the function of ae in Tursiops and
their joint origins and insertions differ only in a more extensive
attachment of the posterior part on the posterior margin of the
vestibular sac and in a closer connection with pi at the maxil-
lary origin.

Posterointernus layer of the pars nasalis of the
maxillonasolabialis

Tursiops (pi; Figs. 11-14, 23)

Although this layer has a very wide origin, its greatest de-
velopment is posterior. The actual muscle fibers are peripheral;
beneath the vestibular sac and superficial to the plugs these fibers
end in and are replaced by thin aponeuroses.

Origin: On the maxilla and frontal with a small area on the
occipital. At its origin it is thickest posteriorly, covering a much
wider area than the outer layers. In the region where the maxillo-
frontal suture bends upward it tapers to a thin layer without
muscle fibers. Anteriorly, it l)ecomes more muscular and again
somewhat thicker. At its anteriormost attachment on the maxilla
it is hard to separate from ai: it also merges with ae. From
this margin thin layers of tissue merge with the strands at the
end of the pars labialis of the maxillonasolabialis. Posteriorly
where it arises beneath pe it is distinct from this layer.

Insertion: Its principal insertion is internal to pe in the posterior
wall of the nasal passage beneath the vestibular sac, as well as
in the shallow posterior fold of the sac. This insertion is via a
thick, shiny, tendinous mass, whose fibers trend anteroventrally
to fill the space between the nasal bones and the passage, and
which is closely attached to the underside of the vestibular sac,
particularly posterior to the slitlike opening of the main passage
between the dorsal surface of the plugs and the posterior wall.
Where the tendinous insertion of pe lies on top of that of pi, the
two are ditfieult to separate. Deeper dissection shows the fibers
of pi to slope progressively more ventrally, less anteriorly, to a
less tendinous insertion on the tubular sac. Additional lesser
attachments are via aponeuroses to the underside of the vestibular

124 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

sac in the region lateral and slightly anterior to the slitlike open-
ing between the dorsal surface of the plugs and the posterior wall ;
other similar sheets extend forward between ae and ai to merge
in the melon with the insertion of the pars labialis of the maxil-
lonasolabialis. These attachments are of very secondary im-
portance, and most of the muscle, even the anterior part,
converges to insert behind the nasal passage, as described above.

Function: Contraction of this muscle helps pe to draw back the
posterior fold of the vestibular sac; it also (and this is very
important) helps to open the slitlike passage between the pos-
terior wall and the dorsal surface of the plugs. This is accom-
plished by pulling up and back the convex mass including the
blowhole ligament and the tubular sac which lies in the con-
cavity of the plugs. The anterior insertions of this layer help
to counteract this powerful backward pull and, together with ae,
prevent the contraction of the very strong posterior section from
distorting the entire passage. A further effect of the contraction
of pi is to put pressure on the posterior part of the tubular sac,
which would force out air trapped here.

Delphimis

Differs from Tursiops and SteneUa in having the anterior
attachments reduced to a very few poorly differentiated strands
which pass to the posterolateral commissure of the nasal passage
beneath the vestibular sac. Further, at its origin it tapers
gradually from a thick posterior portion to a thin anterior, but
lacks a pronounced thinning midway. At its origin anteriorly
it is also distinct from ae, though it merges with ai.

Stenella

Two very distinct parts of this layer were found, the external
of which differs from Tursiops and Delphinus in having the
anterior fibers directed more dorsally and with better developed
attachments anterior to the nasal passage. The thicker internal
layer controls the posterior wall of the passage beneath the ves-
tibular sac.

Antei-ointeruus layer of the pars nasalis of the
maxillonasolabialis

Tursiops {ai; Figs. 11, 13-15, 23)

This is a very thin tendinous layer, slightly thicker anteriorly,

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 125

which could possibly be considered as merely a very distinct part
of pr. Along its anterior and posterior margins it is closely
connected with pr, but otherwise the two layers of muscle fibers
are quite distinct, the boundary being marked by strong shining
tendons connected by thinner sheets of connective tissue. The
actual muscle fibers are chiefly lateral to the maxillo-premaxil-
lary suture, the posterodorsal ones following around this rim.
Origin: On the maxilla and frontal internal to pi and anteriorly
internal to ae as well. A few of the anteriormost fibers merge
with pi. At its origin anteriorly it merges with the pars labialis
of the maxillonasolabialis so that the two are impossible to
separate.

Insertion: Anterior to the nasal passage in the dense collagenous
connective tissue beneath and lateral to the valve. The strong
transverse sheet of its insertion shows many slender tendons
which branch and vanish and are hard to separate from similar
insertions of pr. In relation to the tubular sac, to which it is
firmly attached, the insertion is in tissue which lies in part on
top of this sac, in part anterior to it. The fibers converge towards
this area from their rather wide origin, the most posterior ones
assuming an anteroventral direction to do so. As the main mass
passes dorsally anterior to the slitlike opening between the dorsal
surface of the plugs and the posterior wall, it lies on top of the
tubular sac; laterally it only partly covers this sac. At its in-
sertion anterior to the tubular sac, some of its fibers are hard to
distinguish from those of the major intrinsic muscle of the
tubular sac.

Function: Helps depi-ess the valve by pulling downward and
forward on the tissue beneath it, and puts pressure on the an-
terior part of the tubular sac.
Delphinus

Differs in having a better developed posterior section, with
origin extending as far as the nasal bones, and some of its in-
sertion via thin aponeuroses in the area posterolateral to the
slitlike opening of the passage between the plugs and the posterior
wall. Anteriorly it also differs in being more closely connected
with ae at its origin.
Stenella

This layer does not differ significantly from that of Tursiops.

1 26 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Profundus layer of the pars nasalis of the
maxillonasolabialis

Tursiops {pr; Figs. 16, 17, 18, 23)

This muscle is composed of a number of sections which are
separated with difficulty. The fibers of the outermost ones have
approximately the same direction as those of ai. The deeper
ones are directed progressively more anteriorly (less dorsally).
The outer layers of pr are inseparable both at their origin and
their insertion from the pars labialis of the maxillonasolabialis.
The innermost layers are distinct at their insertions only. This
muscle is the most tendinous of all the blowhole muscles with
numerous small bipinnate sections. The tendons of these sections
converge to form strong strands which divide again to insert in
the melon.

Origin: Beneath ai. The inner layers arise on the maxilla and
from the strong, thick, lateral margin of the premaxillary sac.

Insertion: In the fibrous fatty mass of the melon anterior to the
tubular sac and its major intrinsic muscle. The innermost layer
forms a more or less transverse band across the attachment of
the nasal plug muscle, separating it from the pars labialis of
the maxillonasolabialis.

Function: Draws the melon stronglj' down towards the premax-
illa. This flattens the premaxillary sac and also makes a firm mass
towards which the nasal plug muscles can pull. The band across
the base of the nasal plug muscle helps to limit its motion here.

Delphinus

Does not differ significantly from Tursiops.

Stenella

Does not differ significantly from Tursiops.

Nasal Plug Muscle

Tursiops (Figs. 2-4, 8, 9, 19, 23)

This muscle is entirely independent of the six layers of blow-
hole muscle. It is abundantly streaked with collagenous con-
nective tissue which becomes more and more concentrated until
it finally replaces the muscle entirely in the pad of the plugs.
From its origin, it takes a posteromedial direction.

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 127

Origin: Chiefly on the premaxilla anterior to the premaxillary
sae, with a few fibers arising in the connective tissue band along
the margin of the premaxilla lateral to the sac.
Insertion: In the dense collagenous connective tissue of the plugs.
Function: To withdraw the plugs from the openings of the bony
nares. The diagonal direction of the fibers and their strong
medial insertion makes the dorsomedial parts of the plugs the
most movable. This motion can be seen in photographs taken at
64 frames per second.
Delphinus

As in Tursiops.
Stenella

As in Tursiops.

Pars Labialis of the Maxillonasolabialis

Tursiops (Figs. 11-13, 16-17, 18, 23, 25)

This large muscle lies anterior to the layers of blowhole
muscle and the nasal plug muscle. It is very coarse and mixed
with numerous heavy collagenous connective tissue fibers, espe-
cially externally. Here the muscle fibers are directed laterally
and posterolaterally ; deeper dissection shows a gradual shift in
direction, first to more vertical and then to anteromedial. The
muscles on each side, although for the most part separate, are
connected by a few fibers anteriorly, and behind this by a poorly
developed transverse band of connective tissue beneath the melon.
Origin: From the maxilla ; on the premaxilla it merges gradually
with the tissue beneath the melon. The origin of the deeper por-
tions is not distinct posteriorly from that of the deeper layers
of blowhole muscle.

Insertion: Externally the more ventral fibers end in the fibrous
connective tissue between the upper lip and the melon, the more
dorsal in the melon itself and in the fibrous sheet of tissue
separating the melon from the outer layers of blowhole muscle.
Internally the fibers insert in the melon.

Function: Helps keep the melon under tension and makes a
strong connection between it and the rostrum.
Delphinus

As in Tursiops.

128 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Stenella

As in Tur slops.

Major Intrinsic Muscle of the Tubular Sac

Tursiops (Figs. 15, 16)

This is a poorly developed muscle with a strong mixture of
connective tissue, particularly in its anterior part. It merges
with and vanishes in the adjacent tissue to such a degree that it
is extremely difficult to determine its limits. Anteriorly it is
thick, with the muscle fibers becoming sparser as one dissects
deeper. The direction of the fibers is difficult to determine ; pos-
teriorly they run anteromedially and gradually change direction
to run posteromedially in their anterior part.

Origin: In the fatty, collagenous connective tissue between the
tubular sac and the skull posteriorly and laterally. It also at-
taches in a thick mass of this same tissue on top of the connecting
sac. Anteriorly it intermingles with the connective tissue in-
sertion of ai.

Insertion: On the margin and underside of the tubular sac, the
anterior part being attached more on the underside and less on
the margin.

Funciion : The posterior part helps to anchor and hold open the
tubular sac ; the part lying between the tubular and connecting
sacs may also help to separate the walls of the latter.

Belphinus

The structure of this muscle was exceedingly hard to deter-
mine here. The anterior part in particular was not clearly
separable from the surrounding layers.

S^tenella

As in Delphinus, this muscle seems to be largely posterior.
It diff'ers from that in Tursiops in having the fibers run antero-
medially to insert on the tubular sac. Thus while the anchoring
effect is apparently the same, the direction of the fibers is dif-
ferent. It should be emphasized again that this muscle is poorly
developed and hard to trace.

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 129

Ligament Muscle (not figured)

Tursiops

This poorly developed band of muscle fibers is apparently an
internal slip of the major intrinsic muscle, with which it is closely
connected. It lies beneath and in front of the blowhole ligament,
on top of the connecting sac.
Origin: Beneath the blowhole ligament.

Insertion: In the tissue near the commissure of the slitlike open-
ing above the plugs.

Comments: Failure to identify this slip in Stenella and Del-
phinus could easily have been due to the great difficulty of
separating it from the rather formless mass of connective tissue
and muscle fibers in this area. A special effort to determine its
extent in Lagenorhynchus showed it to be distinguishable from
the major intrinsic muscle only by the direction of the fibers.
In Lagenorhynchus the ligament muscle, as well as inserting
near the commissure, has a few fibers that insert in the outer
wall of the premaxillary sac where this latter overlaps or lies
on top of the lip of the plug. It is possible that this helps open
the passage here.

Minor Intrinsic Muscle op the Tubular Sac

Tursiojjs (Fig. 15)

A very poorly developed muscle wrapping around the anterior
part of the posterior half of the tubular sac.
Origin: In the fatty collagenous connective tissue between the
tubular sac and the nasal passage both internal and anterior
to the tubular sac, so that the fibers trend dorsoposteriorty.
Inserfion: Weakly around the anterior part of the posterior half
of the tubular sac.

Function: As with the major intrinsic muscle of the tubular
sac, this helps to anchor and hold open the tubular sac.
Delphinus

Not found here, possibly due to state of preservation.
Stenella

Relatively better developed here than in Tursiops, and passing
across the tubular sac near the bend to merge with the major
intrinsic muscle of the tubular sac.

130 BTTLLETIN : MUSEUM OF COMPARATTV^E ZOOLOGY

THE BLOWHOLE MECHANISM

Structure must be understood in order to understand func-
tion, but the detailed dissections described above do not by
themselves suffice to explain the blowhole mechanism of por-
poises. The additional information obtained from experiments
on living and freshly killed animals has solved a number of
controversial problems which have baffled zoologists in the past.

Observations on its Functioning

All of the evidence agrees in showing that the passage is opened
l)y muscle action, closed by the natural elasticity of the tissues,
and the seal is tightened not by muscle action but by forcing the
air back up the passage into the tubular sac. Inflation of this
sac makes a very efficient seal where the passage narrows to a

slit beneath the vestibular sac so efficient that internal to

this the porpoise can keep its bubble, or internal air, at greater
than ambient pressure. It makes a less efficient seal at the blow-
hole entrance, where it is possible with small effort to separate
the lips. When the passage opens, the plugs, the anterior fold
of the vestibular sac, and the valve move so nearly simultaneously
that pictures at 64 frames per second fail to show any differential
sequence of events (Fig. 26), and we could not detect any per-
ceptible lag between the first movement and the outward rush
of air. When the passage closes it is possible, again in slow
motion pictures, to see the plugs falling into place first, next
the anterior fold of the vestibular sac, and finally the valve
(Figs. 27, 28).

This opening of the passage is in no sense the separating of
the lips of an orifice ; it is the drawing apart of masses of tissue
which, when undisturbed, lie pressed closely together. Most of
the motion is in the anterior part of the passage, the great wedge
of tissue formed by the valve, the folds of the vestibular sac, and
the plugs being drawn down and forward away from their
])osition beneath and against the posterior wall. Although the
posterior part moves relatively little, it is possible to see the
mid-part of the crescent drawn back to widen the blowhole
aperture (Fig. 26) and to feel the wall inside lifted up and back
from its position on top of valve and plugs.

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 131

Within the vestibular sac and well behind the opening of the
main passage, it is possible to feel a very strong contraction of
muscle layer pe between the outer wall of the vestibular sac and
the blubber as the blowhole opens. Deeper than this, individual
action of the muscle layers cannot be felt, although it is easy to
feel the swift separating of the actual walls of the passage. It is,
however, apparent from their arrangement and the relative
strength of their parts as well as from pulling on them by hand,
that pe and pi control the posterior wall of the passage, drawing
the lip slightly back and lifting the wall of the passage, including
the convex mass, away from the valve and dorsal surface of the
plugs. Layers ae and ai act on the anterior mass external to the
l)lugs, drawing the valve and anterior fold of the vestibular sac
forward and down. Finally, the nasal plug muscle withdraws
the plugs from their position across the bony nares.

The nasal plug muscle attaches in the cushion of the plugs, and
of necessity it is that part of each plug which lies against the
posterior wall beside the septum which must move the most to
open the passage. The liplike, lateral extension of the plugs is
moved only indirectly, and very little, as is shoAvn in motion
pictures. This shows less clearly in the photographs, but in
Figure 26 this part of the plug shows near the back corner of
the otherwise widely opened passage. There is also very little
motion at the lower end of the plugs where they attach on the
cartilaginous septum close to the anterior margin of the bony
nares (Fig. 19). For this reason, withdrawing them does not
open the premaxillary sac, nor can the plugs slide forward into
the sac. Instead, the plugs hinge here and their forward move-
ment makes a distinct and perceptible midline bulge in the
melon, which at the same time seems to flatten on the sides.
Apart from this bulging of the melon when the blowhole opened,
no other muscle action was seen or felt through the thick layer
of blubber which coA'ered the Tursiops head to a depth of about
two-thirds of an inch.

All of this happens very quickly, and during a normal blow
the main column of air passes directly and with considerable
force to the outside. There is no preliminary inflation of the
vestibular sac, as determined by manometer readings, though
there is here a verv slight rise from 2 mm. of mercurv to about

132 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

8 mm. during a blow, with readings in the bony nares going up
to at least 30 mm. at such times. This rise is part of the general
rise in pressure at exhalation, and one sequence of pictures
showed a slight bulge over the vestibular sac at this time. On
inhalation the pressure drops, and the same sequence then showed
a flattening of this area. The pattern of a blow is for the exhala-
tion to take longer than the inhalation. This can best be told by
ear, as there is no visible change in the widely opened passage
when exhalation stops and inhalation begins. The timing of
events during the period between the first opening of the blow-
hole and its complete closing varies. In eight blows of a single
individual loafing near the surface, it was found that the per-
centage of time for opening the blowhole averaged about the
same as that for closing, though sometimes the one was swifter,
sometimes the other, and on two occasions the blowhole seemed
to widen slightly after it was apparently fully open. These blows
were analyzed as follows :

Total iloic
.63 sees.
.77
.69
.77
.75
.74
.78
.47

That the opening and closing took so nearly the same time is
especially interesting in view of the fact that only in closing is
there a clearly visible sequence of events (Figs. 27, 28). At this
time the plugs fall in place first, then the folds of the vestibular
sac, and last the valve. More often than not the closing appears
asymmetrical with, in two porpoises at least, a lag on the left side.
As is often the case with aquatic mammals (HoAvell 1930, p.
87), the closed position is the involuntary one. In the forms
studied this is dependent on the great elasticity of the tissues.
The posterior lip presses on top of as well as behind the anterior,
so that AA'hen it is pulled up by hand and released it falls back
quickly in spite of the absence of any muscles to draw it in this

Opening

Closing

17.5%

12.5%

16.3

20.0

19.0

18.2

14.2

14.3

16.6

14.6

14.9

19.2

14.0

20.0

16.6

LAWREXCE AND SCHEVILL : THE DELPHIXID XOSE 133

direction. Similarly, the valve when depressed by finger im-
mediately lifts into place when the finger is removed. A finger
inside the vestibular sac during repeated blows detected no
muscle action between the sac and the blubber to pull the lips
of the blowhole together, nor does anatomical investigation show
any muscle capable of this action. Air is not retained in the
vestibular sac at the end of a blow, as repeated investigation of
the live animal showed. ^Manometer readings showed no pressure
here at such times, and opening only the vestibular sac by sepa-
rating the lips of the blowhole right after an inhalation never
resulted in the escape of air, even when this was done under
water. Beneath the vestibular sac, the passage narrows, as pre-
viously described, to a transverse slit between the top of the
plugs and the posterior wall of the passage. The soft mass of
the anterior fold of the vestibular sac covers this slit so com-
pletely that considerable probing is needed to get into the
passage beneath it. In the posterior wall internal to this slit,
the elastic blowhole ligament and associated tissues form a taut
convex mass which is difficult for the finger to lift from the top
of the plug, and which immediately snaps back into place when
released. Deeper than this it is possible further to open the
passage by unseating the plugs manually. This does not neces-
sarily result in the escape of air, indicating that at such times,
anyhow, the animal's internal air is at no greater pressure than
his surroundings.

Attempts to open forcibly different parts of the nasal passage
showed that when the Tursiops resisted it did so by inflating the
tubular sac. This resistance could be felt strongly internal to
the vestibular sac and weakly at the lips of the blowhole. Within
the vestibular sac over the posterior part of the tubular sac, an
obvious swelling could be felt at such times, and when a finger
was forced down the slit between the plugs and the posterior wall
a tightening could be felt on each side of the finger. Further,
when the pOrpoise was thus resisting, manometer readings in
the lateral margin of the premaxillary sac showed sharp rises in
pressure, measured in millimeters of mercury above ambient
(^ 0 mm.). At these times the pressure fluctuated irregularly,
the usual pattern being for the animal to hold pressures of about
6-10 mm. and sometimes as high as 20 mm. between blows, Avith

134 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

occasional rapid excursions higher. The greatest pressure re-
corded at such a time was 124 mm., though peaks were more
often between 40 and 60. On the inhalation the pressure dropped
to about 0. Also at times of resistance there was often a slight
but appreciable bulging of the melon somewhat in front of the
valve, as might be caused by inflation of the premaxillary sac.
With the manometer tube at the entrance to the tubular sac,
similar changes in pressure were recorded. These were somewhat
more erratic, probably because of leaks around the rubber tube,
and reached as high as 86 mm. ; as with the premaxillary sac,
the pressure dropped at the end of each blow, seemingly on the
inhalation. Clearly, efforts to hold the blowhole closed involved
forcing air back up the passage under considerable pressure ;
unless the plugs were unseated by hand, or the walls of the pas-
sage near the entrance to the connecting sac were separated, this
air did not escape but was retained internal to the vestibular sac.

More often than not the porpoise made no perceptible re-
sistance to our explorations of the vestibular sac, as he almost
always did when we tried to push further down the passage.
On the not very frequent occasions when the porpoise resisted
separating the lips of the blowhole, a lifting of the cushion of
the valve as by inflation of the tissue beneath it could be felt,
sometimes accompanied by what seemed to be an inflation of the
premaxillary sac. The lifting was immediately external to the
anterior part of the tubular sac and made a visible depression
along the hinge line of the valve. This, however, was not strong
enough to prevent the valve from being forced do\^'n by hand
with very little effort. When this was done with the blowhole
submerged, Avater ran into the vestibular sac. At the next blow,
the water was ejected.

That this use of the vestibular sac as a water trap is not acci-
dental was suggested by many pictures of swimming porpoises
in which one frame shows the blowhole well open with the water
at the margin, and the next frame shows the blowhole submerged.
This evidence was later confirmed bv a long series of similar
observations of another porpoise floating at the surface in a
confined space. Here sequence after sequence showed the animal
submerging with the blowhole open (Fig. 29). As it went under,
the water trickled in at the forward corner of the blowhole, on
either side ; if it delayed closing for .05 to .06 second, the trickle

LAWREXCE AXD SCHEVILL : THE DELPHIXID XO^E 135

spread to a stream extending around the rim of the opening to
the midline in back. At such times there was often a visible
spurt of air and water as the animal submerged while closing the
blowhole (Fig. 29), suggesting that if there is any danger of
the animal flooding itself, it forces air back up the passage at
the end of an inhalation. The spray was particularly conspicuous
if water had l)een allowed to trickle in along the midline, that
is to say, posteromedial to the entrance to the vestibular sac.
Water leaking in here could not be trapped in this sac, and the
only way to get rid of it would be to blow it out. Only part of
the water was ejected in this way. On the subsequent blow, a
thick spray was apparent before pictures showed a visible opening
of the passage (Fig. 30). When the spray was fairly well dissi-
pated the blowhole began to open visibly ; in seven sequences this
delay varied from .05 to .12 second, depending apparently on
the amount of water blown out. This seems to indicate that the
animal may force air up the passage, as explained below, Avithout
retracting the plugs, while there is water in the vestibule. This
would eliminate the danger of water making its way down into
the bony nares past the plugs when the animal pulls them. If this
should happen it could endanger the larjmx or the .Eustachian
tube, as the porpoise has no mechanism for getting rid of water
once it is in the inner nares. The larynx projects into the back
of the bony nares and is held in place by a sphincter muscle in
such a way that a shallow trough is left around the two, near
the bottom of which is the opening of the Eustachian tube. We
do not see how water accumulating here, ventral to the opening
at the end of the larynx, could either be blo^vn out or ejected by
muscle action.

The mechanism for closing the blowhole is not strong enough
to retain any volume of air in the vestibular sac under much
greater than ambient pressure. Small (|uantities of air can, how-
ever, be trapped in this sac beneath the folds. This was demon-
strated in two Avays. A small amount of latex blown into tho
vestibular sac from the inner passage will fil] the space beneath
the folds, but not reach the outer part. Experiments in pumping
air up the nasal passage of a Delphl)U!s soon after death show-
that after the tuliular sac has been inflated, slightly more pressure
will inflate the vestibular sac with scarcely anj^ air escaping from
the blowhole. AiDparently, undei- moderate pressin-(\ air from

136 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

below will lift the anterior fold up against the posterior lip, thus
tightening the seal, which helps to trap air in the inner part of
the vestibular sac. "When internal pressure is released, the sac
deflates slowlv, and it is difficult to force out the residual air bv
hand pressure. Similarly in a dead Lagenorhxjnclius where air
was found in the tubular sac, this air could be forced into the
vestibular sac when the anterior part of the tubular sac was
squeezed by pulling ai. When this tension was relaxed the air
moved back into the tubular sac. It did not escape at the blow-
hole nor did it find its way past the plugs and out at the cut end
of the trachea.

These two experiments seem to show that once the tubular sac
is inflated, as in the first case, it is difficult to force air past the
slitlike opening which it constricts. If, on the other hand, the
tubular sac is squeezed more or less empty, as in the second
experiment, air passes more easily through this slit in either
direction. The fact that air returned to the tubular sac when
pressure was released would seem to indicate that this sac ordi-
narily remains partly open. Anatomically, particularly at the
bend, this is possible. Of further interest is the presence of
observable amounts of air in the outer passage of this individual
only, among the various ones dissected. Since the animal was
harpooned and lanced, and died after considerable flurry, it
seems very likely that this tightening of the seal of its air passage
was part of its death struggle.

Anatomical Explanation of Function

The observations on the various changes that take place in the
outer nasal passage during and betAveen blows are both con-
firmed and explained by the accumulated anatomical data. The
absence of muscles for closing the blowhole was striking. No
muscles were found on top of the vestibular sac with attachments
near the corner of the crescent which could possibly draw the
posterior lip forward. Instead, the two large layers, pe and pi.
attaching posterior to the passage, have medial attachments and
exert a strong posterior pull. Similarly the wide spi-ead of the
vestibular sac and the position of the plugs make it impossible
for muscle action to lift the valve up and back against the pos-
terior lip. The two layers, ae and ai, which attach in front of the

LAWRENCE AND SCHEVILL : THE DELPHIXID NOSE 137

passage beneath the valve, pull strongly downwards and for-
wards. In each instance the outermost layers attach well beneath
the thick layer of blubber and connective tissue which surrounds
the orifice, far too deep to draw the edges together in any way.

While elasticity of the tissues closes the passage, the normal
watertight fitting together of its parts is ensured by the texture
of the surfaces and consistency of the masses involved. In the
posterior wall there are two stiffened and very smooth surfaces.
The more external is the underside of the crescentic posterior lip,
which closes against the equally smooth anterior wall where it
slopes posteroventrally from the valve. The more internal is the
convexity formed by the posterior part of the tubular sac and
the associated muscles and connective tissue, and stiffened by
the blowhole ligament. This, too, has a very smooth surface which
lies in the no less smooth concavity of the external surface of
the plugs. Between these two stiffened surfaces is the soft
wrinkled mass of the two folds of the vestibular sac. Thus be-
tween two pairs of beautifully conforming surfaces there is
a gasket-like cushion that can be distorted to fill any crack.

The pneumatic closing action of the tubular sac is also easily
understood if the structure of this I'egion is examined in detail.
As is most readily seen in sagittal section, the plugs lie diagonally
across the upper end of the bony nares in such a waj' that air
under pressure from below will push the posteromedial part of
the plugs more tightly against the surface dorsal to them, thus
hindering the direct upward passage of air while at the same time
lifting the lateroventral part of the plugs to open the entrance
to the premaxillary sac so that air can pass in this direction.
While the cushion of the plug is thus lifted by air forced up
the bony nares, the lateral extension is not so affected, lying as
it does on top of the protecting diagonal membrane and the
premaxillary bone. This means that this part of the plug is not
pushed up against the entrance to the tubular sac, which thus
remains open while the main passage closes more tightly. From
the premaxillary sac air passes into the connecting sac, whose
opening lies immediately above the lateral external opening of
the premaxillary sac. From here the direct upward passage of
air takes it into the tubular sac, whose intrinsic muscles are suited
to open it, facilitating the passage of air in this direction. As
described above, this inflation of the tubular sac tightens the

138 BULLETIN' : MUSEUM OF COMPARATIVE ZOOLOGY

seal of the passage. When latex was blown up the passage of a
Delphinus, a concentration of it was found in the lateral edge
of the premaxillary sac continuous with the latex in the bony
nares and that lying external to the lip of the plug. This shows
how air from the premaxillar}' sac thus passes externally to
the plugs into the connecting sac.

There is no evidence that hydrostatic pressure is important in
sealing the blowhole. Animal tissue differs little from water in
its compressibility; so, once the outer passages are emptied of
air, sea pressure could be used in this way only if the air in the
bony nares were at less than ambient pressure, and the por-
poise's voluntary efforts to tighten closure of the passage involve,
as we have seen (pp. 133-134), a rise to greater than ambient.
This is surely the easiest arrangement, for achieving an intrapul-
raonary pressure less than ambient would involve expanding the
thorax, which would be difficult at any but the shallowest sub-
mergence. To put it differently, this postulates the animal
working against its medium (the water) to tighten the closure
of its nasal passage when it is organized to do this much more
easily by working with the medium. It is further significant that
in relying on an increase in internal pressure to tighten its seal,
the porpoise has perfected an arrangement which allows it to
retain its internal air under greater than ambient pressure.

Contrast with Mysticetes

The reasons for these complications in the porpoise's outer
nasal passage are not completely understood. Aquatic mammals
in general develop some mechanism for closing the nostrils. In
delphinids this is particularly elaborate, far more so, for in-
stance, than in their relatives the baleen whales. Both groups
are faced with many of the same problems, and the different ways
in which they have solved these must inevitably depend on the
variation possible in each group. In the mysticetes. or baleen
whales, the nasal passage is much less modified than in the odon-
tocetes. The nostrils are still paired, the passage is more nearly
horizontal, and part]y roofed over by the nasal bones. All this
has made possible a comparatively simple method of closing. In
one form dissected. Balaenoptera acuto-rostrata. which is prob-
ably fundamentally similar to the others, the nares are closed

LAWRENCE AND SCHE^^LL : THE DELPHINID NOSE 139

by two large plugs moA'ing on each side of a stiff median septum.
These pass beneath the posterior wall which is stiffened by carti-
lage. Posterolateral to this opening of the inner passage and
extending anteriorly are shallow diverticula which probably form
the basis for Kernan's somewhat misleading and oft-quoted
reference to "spiral folds" as a closing device (Kernan in
Schulte, 1916, p. 435). The muscles opening the nostrils and
moving the plugs are simple, and the elasticity of the tissue as
felt when one thrusts an arm down the nasal passage of a dead
whale is surely great enough to close it. There has been no
development of an elaborate pneumatic mechanism to tighten
closure.

Both groups of cetaceans have in common their need for some
way to keep water out of the inner passage. Stormy weather and
fast swimming with a minimum of time at the surface make this
a real problem for each, but the solution is different. The odon-
tocetes. as we have discovered, have a water trap inside the
blowhole. The mysticetes have a conspicuous V-shaped pair of
ridges which elevate in front of the nostrils as these open. Ob-
servations at sea of the humpback, Megaptera novae-angliae, and
the little piked whale, Balaenopfera acuto-rostrata, show these
to be very efficient water deflectors, at least in moderate weather.

The use of buoyancy changes may also be common to both
groups, although we have not observed it in the mysticetes. We
have frequently seen our captive Tursiops sink verticalh^ This
sudden sinking without exhaling and without any horizontal
motion (thus excluding the possibility of planing forces driving
the porpoise down) seems best explained by a buoyancy reduc-
tion from making the body smaller. This could be done only by
compressing the internal air, which requires a tight seal in the
i-espiratory passage ; since human swimmers can do this in slight
degree, it is not surprising to find a cetacean adept at it. Such
a maneuver differs from the sinking after exhaling often observed
by us in Stenella plagiodon and reported by "Woodcock and Mc-
Bride (1951, p. 215). One is reminded of the old whalers' ac-
counts of "settling" by the sperm whale, supposedly as an
emergency maneuver when surprised by a whale boat (e.g.,
Davis, 1874. p. 187).

Different solutions to similar problems based on fundamental

140 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

dissimilarities in structure only partly explain the elaborate
pneumatic arrangement of the odontocete passage. Other dif-
ferences based on different habits and activities must also con-
tribute. One of the most conspicuous features of the odontocetes,
and one in which they differ widel}^ from the mysticetes, is their
habitual production of a wide variety of underwater sounds.
Experiments have led us to believe that these sounds are laryn-
geal, and we have found that thev all can be made without
necessarily allowing air to escape from the blowhole. Undoubt-
edly the ability to subject the internal air to considerably greater
than ambient pressure is important in this connection. Regula-
tion of the air pressure in the sinuses connected with the middle
ear may also be affected by this arrangement, much as a noseclip
aids a human diver, especially if he goes deep.

DISCUSSION OF PREVIOUS WORK

Much of the early work on the blowhole region does not deal
with function, and since this is our main interest an historical
account of these investigations is not attempted here. The reader
who wishes to pursue this can find fairly detailed references in
Murie (1873) and Gruhl (1911). It is, however, opportune to
summarize the varying opinions on the manner in which the
outer nasal passage is opened and closed and the anatomical
evidence on which these opinions rest.

Since the early 19th century there have been two general
theories ; the one held by such workers as von Baer, Kiikenthal,
Boenninghaus, and Gruhl, that the blowhole closes when its
muscles relax and is sealed tighter by increasing hydrostatic
pressure, and the other held by Cuvier, Sibson, Murie, and more
recently by Huber and others, that certain of the layers of blow-
hole muscle pull the blowhole closed while others open it. Fur-
ther, there are differences of opinion as to the number of layers
involved. The first group of workers tends to emphasize the
essential singleness of the blowhole muscle in structure as well
as in function. The second group, in accordance with the double
function they assign to this musculature, defines numerous
layers. Thus, those workers who believed the blowhole to be
opened, not closed, by muscle action felt it unnecessary to recog-
nize more than a very few laj^ers, while those whose approach is

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 141

through dissections (accurately showing more numerous muscle
layers) have felt it necessary to assign a double function to these.
Neither group has adequately described or shown the use of
the nasal sacs.

A considerable part of these investigations was carried out on
Phocoena. Piecing together the accounts of phocoenids by von
Baer, Kiikenthal, Sibson, Boenninghaus, Gruhl, and Howell, one
finds a certain similarity in structure between these and the
delphinids, but the differences are such that more anatomical
detail and an actual examination of live specimens is necessary
before function can be compared in the two groups. However,
we see no reason to doubt von Baer's assertion of 1826 (col. 820)
that the lips of the blowhole are closed except when opened by
muscles. Turning to the delphinids, we find that the earliest good
descriptions and illustrations of the blowhole region and asso-
ciated muscles in this group are those of Murie, who published on
Grampus rissoanus {= g risen s) (1870), Lagenorhynchus al-
hirostris (1871), and Glohicephalus nielas (^ Glohicephala
melaena) (1873), stating in the latter paper (p. 247) that the
function of the layers is "nearly identical in the several forms."
Of these genera the second is the most closely related to the
three discussed in this paper, and dissection by us of Lagenorhyn-
chus acutus shows this species, at least, to resemble our three
closely. In this it disagrees with Murie 's figures of L. alhirostris.
which differ chiefly in the wide areas of insertion of the outer
layers of blowhole muscle, structure and musculature of the
tubular sac, and the arrangement of the rostral muscles. In spite
of this, a comparison of his text and figures leads us to believe
that his layers (1871, pp. 147-148 and pi. 5) may be homologized
with ours as follows : first layer, occipito-frontalis, with pe, and
probably also i; second layer, levator labii superioris alaeque
nasi, with ae (the posterior part is probably homologous with
our inner layer of pe which we found merging with ae) ; third
layer, levator superioris proprius, with pi: fourth layer, zygo-
maticus, and the posterior part of the fifth layer, depressor alae
nasi, jointly with ae and pr; the anterior part of the fifth layer,
which he calls naso-labialis, with the pars labialis of the maxil-
Icmasolabialis. The nasal plug muscle is described (1871, p. 149)
l)ut not named or figured. Finally pyramidalis, in spite of its
difference in size and orientation, is probably equivalent to our

142 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

major intrinsic muscle of the tubular sac. Murie's sacs also maj^
be homologized as follows : the maxillary with our vestibular sac,
the naso-frontal with the posterior part of our tubular sac, and
the anterior or premaxillary with our premaxillary sac. While
our examination of Lagenorhynchus was not as thorough as that
of the other three forms, we found nothing to suggest that it
functions differently. In spite of this strong similarity, we also
found that as a group Tursiops, Stenella, and Delphinus resemble
each other more closely than any of them does Lagenorhynclius.
In some features this latter, a shorter beaked porpoise, tends to
approach Glohicephala. Although the evidence is not conclusive,
the resemblance (in such detail as Murie published) of Grampus
and Globicephala to our material suggests very persuasively that
here also function will be found to be very similar. Our own
superficial examination of the passage and sacs in Glohicephala
tends to confirm this. Murie himself felt this to be true of his
forms and so stated (see below) in the last of his three accounts.
His work on Glohicejjhala, although completed first, was not
published till after the other papers had come out. His work
on Grampus was completed next and published first, while that
on Lagenorliyncliiis refers to both of the previous dissections.
Homologous outer layers are given the same names in all three
although they are sometimes numbered differently. Identification
of the inner layers varies, and one more layer is recognized in
Lagenorhynchus than in the other two. This inconsistency is
based not so much on an apparent difference in the layers as on
attempts to homologize them with facial muscles of other mam-
mals, and so does not concern us here. What is of interest is
Murie's inability to come to any single, final conclusion as to
the functions of these layers in spite of his statement (1873, p.
247) : " As regards the action of the diff'erent layers in Glohiceps
etc., they are nearly identical in the several forms." In the paper
on Grampus he says that the three outer layers close the blowhole
while the fourth layer opens it, as does also the nasolabialis.
In the paper on Lagenorhynchus he claims the opposite function
for these same three, saying they "dilate'' the blowhole; the use
of the fourth is more obscure : he appears to feel that primarily
it helps close the passage by compressing the nasal blubber, while
it could also have a tendency to open the blowhole ; the fifth layer
he feels has a divided function: the anterior part (1871, p.

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 143

149) "drags forward and everts the anterior lip," the posterior
part retracts it. Evidently in Glohicepliala the various layers
function as in Lagenorhynchus. This is apparent when on com-
paring text and figures (1873, pp. 247-248, figs. 63-65), it be-
comes clear that the "superior layer" of Glohicepliala equals oe-
eipito-frontalis and levator labii superioris alaeque nasi (the first
and second layers of Lagenorhynchus) , while the "second sheet"
equals levator superioris proprius (or the "third layer" of
Lagenorhynchus) . Clearly, he frequently had in mind that what
applied to the blowhole mechanism of one genus was equally
true of another, as witness his numerous cross references in lieu
of restatement. Nevertheless, he never summarized or clarified
these conflicting remarks, and his sequence of dissections and
publication is such that it is not possible to decide what his final
conclusions were. Although this might make his work seem
confused, one has only to consider how difficult his dissections
must have been to realize how remarkable it is that he was able
to find out so much. As regards the anterior part of the naso-
labialis, he seems to understand that this has an opening func-
tion, though, as in the case of the other layers, the actual region
of the nasal passage affected is not made clear.

Kiikenthal's (1893) discussion of the odontocete nose in his
monumental whale monograph is chiefly concerned with homolo-
gies and phylogeny. He adds little to what von Baer had long
before said of function, and his anatomical work bearing on
these matters is adequately built on by Gruhl (1911). Gruhl's
important comparative account of the nasal passage in the Ceta-
cea deals primarily with odontocetes, of which he dissected four
genera, and includes a detailed discussion of previous work. His
comments on function and such anatomical details as he gives
for Tursiops and Delphinus need to be considered here. Very
properly he realizes that elasticity of the walls closes the passage,
but he dismisses the musculature as essentially a large paired
muscle which embraces the outer sacs and is suited to widen them.
As with previous workers who have understood the basic single
function of the blowhole muscle layers, there is no apparent
grasping of the fact that the valve moves in one way. the pos-
terior wall of the blowhole in another, and the folds of the
vestibular sac still differently. It is this difference in the way
particular parts of the nasal passage have to move in order for

144 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

it to open which has made it necessary for the musculature to be
organized in alternating layers with differently directed fibers
attaching in specific places, not all around the blowhole. In
describing the sacs in Tursiops and Delphinus he calls attention
to considerable variation in them, particularly in the connecting
sac, which he found to be smaller in Delphinus than in Tursiops.
This is in agreement with our findings. However, his lack of live
or fresh material has led him to believe that the small size of
the tubular sac means that it is a mere vestige of the "Regio
olfactoria, " and hence insignificant. He also thinks that the
function of the premaxillary sacs is to facilitate the action of
what we call the nasal plugs. His detailed and careful descrip-
tion of what we have called the blowhole ligament is of particular
interest in that he refers to it as a cartilage. This suggests that
chondrification may be more usual than we found it.

Apparently disregarding Gruhl's work, Kellogg (1928, pp.
194-195; 1938, pp. 659) refers to a muscular closing of the nose
which he describes (1928) as follows: "... the narial cavity is
divided into a series of pockets into which the narial passages
open, and between these pockets are a series of plugs controlled
by muscles that apparently contract when pressure is exerted
on the external flap of fibro-elastic cartilage that covers the
external nostril." There are no further anatomical details nor
explanations, and since this description does not agree with what
we have seen, we are at a loss to reconcile it with our findings.

The most recent investigations of the blowhole mechanism in
odontocetes are those begun by Ernst Huber as part of his careful
series of papers on the evolution of the facial musculature of
vertebrates. Through studying the innervation in both Tursiops
and Monodon he came to the, conclusion that only one muscle,
the maxillonasolabialis, is involved. The blowhole laj^ers and
nasal plug muscle are derived from the pars nasalis of this muscle
and the rostral muscles from the pars labialis. In addition to
studying homologies, he also considers the mechanism of the
blowhole, returning to the earlier theory that muscle action helps
to close it. A posthumously published paper, assembled from
iiis notes and sketches, is very explicit as to how this might
be done.

T"'^nfortunately he himself never lived to complete this work,
so the anatomica] basis for this theorv is not sufficientlv detailed ;

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 145

indeed the publiished schematic diagrams of muscle action do not
agree with the figures of dissections. The latter show in more
detail than had previously been done the layering of the blowhole
muscle, how the outermost layers are more posterior, the inner
ones more anterior, and the increasing horizontality of the
deeper layers. The direction of the fibers of the pars labialis
of the maxillonasolabialis and their continuity with the inner
layers of the blowhole muscle are also clearly shown. What are
lacking are complete descriptions of the individual layers and
their subsidiary parts, their relative thickness, fiber direction,
and the minutiae of their insertion, as well as detailed descrip-
tions of the sacs and their relation to the main passage. Probably
it is in part lack of this evidence which has led him to con-
clusions so at variance with what our studies have demonstrated.
The assumption that the outer layers of muscle close the blow-
hole is based largely on their attachment in the tissue on each
side of the passage. This is easily seen, and it is interesting to
note that while Huber interprets this to mean that they pull
the blowhole closed, earlier workers interpreted this as evidence
of a dilating action. Actually, if thickness of the layers and
insertion of their strongest parts are known, it is perfectly clear
that the layers which attach around the passage in this way
have a powerful portion which moves one part of the passage
so as to open it. and a weaker, thin portion which helps anchor
an opposite part against distortion. The deeper layers he be-
lieved assisted in closing the blowhole by pulling the "elastic
cushion" beneath the valve backward, thus squeezing the valve
back and up against the posterior lip. We found no layers
which could do this. Either they attach posterior to the passage
or too deep for their contraction to have this effect.

Though he thus postulates a closing action for layers which we
find open the passage, he also refers (1934, p. 124) to "the
dilator musculature of the blowhole orifice. ' ' This is not further
described or figured so it is difficult to know Avhat muscles or
layers he had in mind.

As well as ])ostulating muscle action to close the blowhole,
Huber also said that air trapped in the vestibular sac helps to
keep the plugs seated ; but after a blow we always found this sac
empty of air. The pneumatic action which he felt around the
outside of the blowhole was probably deeper than he suspected.

146 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

His forcing the animal to breathe by pressing here might have
been due, as he suggests, to unseating the plugs. We ourselves
were only once able to make a porpoise breathe by squeezing
lateral to the blowhole, though we tried repeatedly. Unseating
the plugs by means of a finger inserted down the passage often
made it blow. Sometimes, also, when it was resisting our opening
the blowhole it was possible to make it blow by pressing down
on the valve.

In respect of these remarks about cetaceans, it is interesting
to note that of seals Howell (1930, p. 88) says that two muscles
control the nose : the maxillonasolabialis opens it and the naso-
labialis closes it. Now we find in cetaceans, where the closing is
not by muscle action, that the nasolabialis is rudimentary.

Huber's paper was prepared for publication by A. B. Howell,
who gives no indication of disagreeing with the theories therein.
However, in an earlier work while speculating on the origins of
the sacs, he says (1930, pp. 98-99), "The intricate and laminated
muscle layers which converge to operate its opening cannot func-
tion as dilators with precisely equal force at each and everv'
point, however, but their pull must be very unequal." This ex-
cellent observation is subsequently confused by a statement that
the nasal valve (which etiuals our anterior lip and may also be
used here to include the plugs) "is opened by contraction of
certain of the rostral muscles, and that in closure there is tension
from laterad of the blowhole, but whether [it] is voluntary . . .
is not clear." As the rostral muscles, those beneath the melon,
are obviously not the same as the intricate muscle layers converg-
ing on the blowhole and as he refers the reader to Huber for
further details, one gathers that he is in fundamental agreement
with the latter. Howell's published diagram (1930, p. 98, fig. 15)
is further misleading in that the interlocking arrangements of
the supracranial part of the passage are not shown, nor is the
attachment of the nasal plug (valvular plug of Huber) to the
membranous septum at the margin of the premaxillary sacs. This
is what makes it impossible for a forward movement of the plug
to open the premaxillary sac as Howell's diagram shows.

"The handbook of R. H. Burne's cetacean dissections" (Fraser
1952), although it does not deal with function, is of interest in
that the sagittal sections of Delphiniis resemble closely ours of
Tiirsiops. This is to be expected in view of the close resemblance

LAWRENCE AXD SCHEVILL : THE DELPHINID NOSE 147

we found between the two forms in the arrangement of sacs and
muscle layers. Comparing the terms used for the various struc-
tures which he found, his "boss" matches well our nasal plugs,
while his vestibular and premaxillary sacs are the same as ours.
The intermediate areas as shown in his Figure 19 are harder
to identity with the structures found in our dissection of Del-
phi nus. Apparently the "tongue-like" process equals both our
cartilaginous septum and the concave mass of the posterior wall ;
the "subsidiary sac" seems closest to our connecting sac, while
the space above it could easily be the tubular sac. The dissection
shown is exceedingly difficult to make without distorting the
region, and this probably accounts for the discrepancies between
our findings and his.

EFFECT OF HYDROSTATIC PRESSURE
OX THE BLOWHOLE MECHANISM

The failure of earlier workers properly to explain the blow-
hole mechanism has been partly due to lack of observations on
living or recently-killed animals. Further, the anatomical basis
for these theories has been incomplete. The arrangement of the
muscles had been only partly worked out and too little emphasis
put on the actual shape of the air spaces. In addition, the relation
of pressure to closing the passage had not been properly analyzed.
The effect of increasing sea (hydrostatic) pressure on a closed,
collapsible container full of air is to reduce the volume of that
air so that pressure inside the container is the same as pressure
outside. This means that two very common, essentially opposite
theories about pressure in relation to closing the passage are false.
The first holds that pressure tends to expel air from the animal,
thus necessitating a very strong closing mechanism. The second
supposes that pressure itself is the closing mechanism — on the
assumption that increasing hydrostatic pressure squeezes the lips
of the blowhole ever more tightly together.

As far as the first is concerned, the OA'erlapping and close-
fitting-together of the lips and walls of the passage form a series
of check valves which trap the air so that its relative density
cannot make it bubble out. Further, the cetacean diaphragm
slopes in such a way as to allow far more compression of the
lungs than in land mammals. Thus the porpoise as it submerges

148 BULLETIN : MITSEUM OF COMPARATIVE ZOOLOGY

is effectively a compressible closed container of gas, and subject
as such to Boyle's law.

The second assumption is equally wrong, since the compressi-
bility of the air bubble results in the animal maintaining an
internal pressure equivalent to the hydrostatic pressure around
it. If the pressure on both sides of his valves remains the same,
any increase in outside pressure squeezing the passage shut would
be acted against by an equivalent increase in internal pressure
forcing the walls apart. Outside pressure could tighten the seal
only if it acted against a lesser inside pressure. This would be
possible only if the porpoise could keep its pulmonary air at less
than sea pressure, which, as we have said (p. 138), would be
a difficult feat involving holding the ribs lifted against very
considerable outside pressure. It is far easier for the porpoise
to increase than to decrease its internal pressure, and it is just
such an increase, as we haA'e shown above, which pneumatically
tightens the seal of the passage.

SUMMARY

Investigations have been made on the blowhole and related
nasal structures in the delphinid genera Tur slops, Stenella, Del-
phinus, Lagcnorhynchus, and Glohicephala. These have included
dissections and pneumatic studies of fresh and frozen carcasses,
slow-motion moving pictures, and observations on live animals,
both in captivity and at large. We have demonstrated that the
nasal passage is opened by muscle action and closed by the
elasticity of the tissues, and that this closure may be voluntarily
tightened by a pneumatic seal.

The blowhole musculature which opens the passage is incom-
pletely divided into six different layers, of which five are func-
tionally important. Of these, two insert posterior to the passage
and three anterior, alternating with each other to do so and
overlapping broadly at their origin. Those layers which lift the
posterior part of the passage up and back are in general more
external and are thickest posterior to the passage ; those which
draw the anterior part down and forward are more internal and
have their main mass anterior and lateral to the passage. The
passage is anchored against distortion by thinner portions of
these various lavers attaching in such a wav as to counteract the

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 149

pull of each major portion. As the walls of the passage are
drawn apart, the plugs covering the opening into the bony nares
are withdrawn by an independent muscle. When the passage
is closed the posterior wall presses closely on top of and behind
the anterior. The outer j^art of the passage, including the ves-
tibular sac, may be considered as a vestibule. This is used as a
water trap when the animal submerges with the blowhole open.
Beneath this is a transverse slit which marks the entrance to
the deeper part of the passage. Lying around this slit on each
side is the U-shaped tubular sac, which can be inflated without
withdrawing the plugs and whose inflation tightens the seal at
this slit. This control allows the porpoise to keep its internal air
at greater than ambient pressure. Beyond a certain limit an
increase in pulmonary pressure forces the air past this point
and into the vestibular sac, where a small amount can be trapped
beneath the anterior fold, any excess escaping at the blowhole,
where there are no constricting muscles.

The blowhole mechanism is efficient in protecting the inner
passages from flooding and also in allowing the porpoise a certain
independence from its medium in controlling its pulmonary
pressure. The odontocete nose is so much more complicated than
that of the mysticetes that it suggests other non-re.spiratory
functions. Of these, buoyancy reduction, vocalization, and pres-
sure equalization in the sinuses connected with the middle ear
are probably all important.

ACKNOAVLEDGMENTS

This work was supported chiefly by the Office of Naval Re-
search through a contract with the Woods Hole Oceanographic
Tn.stitution.

As u.sual, we are grateful to many people for help. In par-
ticular we wish to thank Marine Studios, Marineland, Florida,
for their hospitality over a number of years. Not only have they
repeatedly granted us the use of their unparalleled facilities, but
they have also provided and maintained experimental animals
for us. and their staff has helped enormously with unflagging co-
operation and interest. AVe would like to mention especially
Messrs. W. P. Rolleston, F. G. Wood, Jr., and Frank S. Essapian.
as Mell as Captain R. Capo and INIr. L. Andreu.

150 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

We are indebted also to Dr. Henry Kritzler for his help in the
early stages of this work and for the two latex injected specimens
used in this study.

The opportunity to study Globicephala in Trinity Bay, New-
foundland, was due to the kindness of Arctic Fisheries Products,
Ltd., Mr. Bernard Andrews, Mgr., and the help of Dr. David E.
Sergeant, Fisheries Board of Canada.

We wish also to thank Drs. H. B. Bigelow and R. H. Backus
for reading the manuscript, and especially Drs. C. P. LjTuan and
D. C. Fawcett for giving us a great deal of time and for their
constructive criticism.

The drawings were made by Miss Jessie H. Sawj'er. Photo-
graphs of transverse sections were taken by Mr. Frank S. Essa-
pian, and the motion pictures by the authors.

EEFERENCES

VON Baee, K. E.

1826. Die Nase der Cetaceen, erlautert dnrch Untersuchung der Nase
des Braunfisches (Delphinus Phocaena) . Tsis von Oken, Jg. 1826,
Heft 8, columns 811-847. pis. 5-6.

BOENNIXGHAUS, GeORG

1902. Der Eachen von Phocaena coininunis Less. Zoologische Jahrbii-
cher, Abt. Anat., 17 (1-2) : 1-98, 20 text figs., 1 pi.

Davis, William M.

1874. Nimrod of the sea. New York, Harper & Brothers, 403 pp.. 30
text figures.

[Eraser, Francis C]

1952. Handbook of R. H. Burne's cetacean dissections. London, British
-Museum (Natural History), 70 pp., 41 figs.

Gruhl, Kurt

1911. Beitrage zur Anatomie und Physiologie der Cetaceennase. Jena.
Zeitschr. Xaturw., 47(3) : 367-414, 21 text figs.

Howell, A. Brazier

1927. Contribution to the anatomy of the Chinese finless porpoise,
Xeomeris phocaenoides. Proc. U. S. Nat. Mus., 70 (art. 13") : 1-43.
14 text figs., 1 pi.

1930. Aquatic mammals. C. C. Thomas, Springfield (Dlinois) and
Baltimore (Maryland), xii -|- 338 pp., 1 pi., 53 text figs.

LAWRENCE AND SCHEVILL : THE DELPHINID NOSE 151

HuBER, Ernst

1934. Anatomical notes ou Pinnipedia and Cetacea. Carnegie Instit.
Washington, publ. no. 447: 105-1.36, 12 text figs.

Hunter, John

1787. Observations on the structure and oeconomy of whales. Phil.
Trans. Eoy. Soe. London, 77 (pt. 2): 371-450, pis. 16-23.

Irving, Laurence, P. F. Scholander, and S. W. Grinnell

1941. The respiration of the porpoise, Tursiops truncal us. Jour. Cell.
Comp. PhysioL, 17(2) : 145-168, 15 text figs.

Kellogg, Remington

1928. The history of whales — their adaptation to life in the water

(concluded). Quart. Rev. Biol., 3(2): 174-208, figs. 12-24.
1938. Adaptation of structure to function in whales. Cooperation in
Research. Carnegie Instit. Washington, publ. no. 501: 649-682.

Kukenthal, Willy

1893. Vergleichend-anatomische uud entwickelungsgeschichtliche Unter-
suchungen an Walthieren: Bau und Entwickelung der Cetaceen-
nase. Denkschr. Med. Naturw. Ges. Jena. 3(2): 322-349, text
fig. 51, pis. 21-24.

MuRiE, James

1870. On Risso's grampus, G. rissoanus (Desm.). Jour. Auat. Physiol.,
5: 118 138, 2 text figs., pi. 5.

1871. Notes on the white-beaked bottleuose, Lagenorhynclius albirostris,
Gray. Jour. Linn. Soc, Zool., 11: 141-153, pi. 5.

1873. On the organization of the eaaing whale, Globicephalus melas.
Trans. Zool. Soc. London, 8(4): 235-301. 6 text figs., pis. 30-38.

Schulte, H. von W.

1916. Anatomy of a foetus of Balaenoptera borealis. Mem. Am. Mus.
Nat. Hist., n. s., 1(6) : 389-502, 10 text figs., pis. 43-57.

SiBSON, Francis

1848. On the blowhole of the porpoise. Phil. Trans. Roy. Soe. London,
1848: 117123, pi. 12.

True, Frederick W.

1891. Observations on the life history of the bottlenose porpoise. Proc.
U. S. Xat. Mus., 13: 197-203, 1 text fig.

Woodcock, A. H., and A. F. McBride

1951. Wave-riding dolphins. Jour. Exp. Biol., 28(2'): 215-217.

EXPLANATION OF FIGURES

All of the figures are of Tursiops truncatus (Montagu).

The dissections are of a young female (skull length about 150 mm.), the
sagittal sections of an old female, and the transverse sections of an old bull.

In Figures 1, 21-23, and 26-30 the anterior end is at the bottom of the
picture; in all others except the transverse sections it is at the right. Orien-
tation of the head is the same in Figures 10, 11, 13-17, and 19.

Abbreviations used:
Layers of the pars nasalis of the maxillonasolabialis muscle
pe, posteroexternus
ae, anteroexternus
t, intermedins
pi, posterointernus
ai, anterointernus
pr, profundus

MAJ. I\T., major intrinsic muscle of tubular sac
MIN. IXT., minor intrinsic muscle of tubular sac
VEST. S., vestibular sac
TUB. S., tubular sac
CONX. S., connecting sac
PBEMAX. S., premaxillary sac

Fig. 1. Plan of nasal sacs in dorsal view, and positions of sagittal sec-
tions, Figs. 2-4. In vertical sequence the vestibular sac is the outermost,
then the tubular, the connecting, and the premaxillary. The posterior wall
of the bony nares is stippled lightly; the margin of the premaxillary sac
is shown by a broken line, the posterior margin of the plug by a solid line.
The left vestibular sac is omitted.

BULL. MUS. COMP. ZOOL

OFL'NING- BEriAiELN DORSAL'
SURFACE OF PLU&S L

PCST ^MALL

Pkfmax. S.

Fig. 2. Sagittal section through blowhole made at a, Fig. 1.

BULL. MUS. COMP. ZOOL.

Fig. 2

^ ^

° ^

^ <

^ >

n ^

t— "

LLJ

/
/

O H-
CO

I- -^ >

, CO CD

> O L._

1 ^^ CO .

1 q: h- '

1 O

cb \

, 2

^ \

< 5

\ err

r ~i

^ ^

\ <

o CQ

' CO

c!^S

<; u_j

^c:>

Fig. 3. Sagittal section lateral to blowhole made at 6, Fig. 1.

BULL. MUS. COMP. ZOOL.

Fig. 3

$,-^

'-^ ,^

_j
<

Fig. 4. Sagittal section near lateral edge of passage made at c, Fig. 1.

BULL. MUS. COMP. ZOOL.

Pig. 4

cr
I —

Fig. 5. Diagram of sagittal section through right side of nasal passage
of fetus with valve and right plug drawn forward to show attachment of
cartilaginous septum between plugs. In older animals the larynx lies farther
lip in the posterior bony nares.

BULL. MUS. COMP. ZOOL.

Fig. 5

Tub. s.

Left plug-
^^^. behind septum

Post lip

Valve

"f\l&HT PLL'(j

DKAIMN rOR-

HARD

PrEMAvX.

Sac •

Tip of larynx

Trache/A

Fig. 6. Oblique side view of skull showing positions of transverse sec-
tions, Figs. 7-9, 24, 25. The nares are shaded.

Fig. 7. Rostral face of approximately transverse section, posterior to
bony nares, made at a, Fig. 6, showing relative position of sacs and passage
as well as principal separation Itetween layers of blowhole muscle. See Fig.
24 for photograph of this same section with passage cut open along lines
shown here.

Maxilla

Nasal

BULL. MUS. COMP. ZOOL.

M&StTHMOlD

; ,B LOIN HOLE

I /

*.'^ I Lacrymal

Occipital-.

Frontal-

Parietal-''
Temporal''

Palatine
/ Pterygoid

Bony nares

Fig. 6

HEAD

MAXILLA

Fig. 7

Separation

Blubber

Tu B^ s.

Passa&e between

PLU& L POST I^ALL

-Ant fold vests

-'^Cranial

CAVITY

.' In Fig. 24 passa&e cut here
f^assaoe beneath dia&onal membrane

Passage above diagonal
membrane

Fig. 8. Eostral face of approximately transverse section through bony
nares made at b, Fig. 6. The heavy black line shows the extent of the
paired premaxillary sacs lateral and anterior to the nares. On the left of
the figure the lining of the naris has been removed.

Fig. 9. Eostral face of transverse section anterior to bony nares, made
at c, Fig. 6. The premaxillary sacs are here separated bj' a thin partition,
and the separation l)etween the anterior and posterior layers of the blowhole
muscle is well marked.

BULL. MUS. COMP. ZOOL.

Nasal plu'j muscle
Bony septum^

Fig. 8

pRfMA

XILLA

B^lUBBER

Frontal

Premaxillary s.

.-Separation be-
^ ^n\.i.H pLiol

'Naris

lntranceto
eustachian

TUBE

Sphincter

MUSCLE

Melon
Nasal plu& muscle

pREMAXILLARY
""--------. SAG

Mesethmoid cartila&e ^'

Vomer'' ]_
Fnd of ptery&oid sinus

Premaxilla

-Sinus con-
nected lAllTH
eustachian
tube

""""""Jaiai

Fig. 9

RATION BE-

vi'im pu Lac

Palatine

""Jaiaj
Maxillary sinus

Fig. 10. Outermost layers of blowhole muscle; vestibular sac beneath
posterior lip dotted in.

Fig. 11. Posteroexternus layer partly cut away to show attachments of
outer layers of blowhole muscle on and around vestibular sac, and orienta-
tion of deeper layers. The fibrous sheet separating anteroexternus
and pars labialis has been reflected.

BULL. MUS. COMP. ZOOL.

Fig. 10

pe POST. PART

Vest, s

^.Blo/m hole

.pe ANT PART
-^?C -OOLLA&ENOUS CONNECTIVE TISSUE INSERTION OF PARTS ^^f 6LCLe

/Melon, LESS DENSELY fibrous

,MelON, MORE DENSELY FIBROUS

m^MemMirgc

O'^l

Fig. 11

/7^,THIN INNER LAYER

^DOE Ofae MERGING lAllTH /?(?

'ED&E OF 6 MERGING lAIITH /7C

^ y' ^'^""-'^.^--Vest. s.

^,6, REFLECTED

-Pars labialis

Layer of fat between
, blubber l muscle

aaS^^^MlS?

Fig. 12. Dorsal view of outer layers of blowhole muscle

Fig. 13. Posterior lip of blowhole partly cut away and folded back, and
vestibular sac cut open to show its inner boundary where the plugs pass
beneath the posterior wall. See Figs. 21 and 22 for transverse extent of
this opening.

BULL. MUS. COMP. ZOOL.

Fig. 12

Nasal

BONE

/Vest. s.

BL0^^HOLE

^-Pars labialis

^(f PARTLY CUT AWAr TO SHOlAj^/,

Blubber

Fig. 18

ST. S-

Cut ed&e: of vest s.
\\post fold of vest s.

Smooth unin& cr post lip passing into wrinkled linins ofvesi s.
,,'Mar&in OF valve against i^hich post, lip closes
'Fold markino edge or tougher external tissue

-'-Ant fold of vest s. pulled forward

•Opening or passage betvneen

PLUa & post lAJALL

nsektion 0f'^(?

-Deeper fibers of paks labiaus

MERGING lAllTh aL

.-''"Pars labialis, post
part reflected

Fig. 14. Vestibular sac and folds cut away to show insertion of an-
terointernus and posterointernus.

Fig. 15. Posterointernus and outer part of anterointernus cut away to
show tubular sac and its intrinsic muscles.

BULL. MUS. COMP. ZOOL.

Insertion cf pc

Post, lip
.Valve

Fig. 14

'Opening of passa&e be-

^NEEN FLU G-&. POST. l^ALL

I
_U-CUT MARGINS
OF VESTS.

~ ~ at

Post lip or blcai hole

Fig. 15

/Valve

Openin(j of passag-e be-

TIAIEEN PLUG- h POST l^ALL

Inslrtiom [)^aL

>-J.-

MaJ. INT

Fatty connective tissue

~ ~ ~~~MlN. INT

Fig. 16. Tubular sac partly cut away to show connecting sac, edge of
premaxillary sac, and inner part of major intrinsic muscle.

Fig. 17. Connecting sac and blowhole ligament cut away. The tubular
sac is partly cut away, and its entrance pulled open to show the plug in-
ternal to it. The premaxillary sac is cut along the margin of the premaxilla
and pulled open to show its relation to the tubular sac and the main passage.
Two probes show respectively the opening of the tubular sac into the
passage (solid line) and the lateral extent of the passage between the plug
and the posterior wall (dotted line).

BULL. MUS. COMP. ZOOL.

Fiff. 16

■'^ffwwwi^jyay

Tub. S. DRAt^N BACK TO SHOl/^ CONN. S.

Opening or tub. s. into nasal PASSAO-f

WALL

LaT IDQl OF PREMAK. S

P^OBt IN OPENING OrTUB.S. INTO PASSAGE

Post part or tub. s. i/nith entrance pulled open

Fig. 17

lateral cor-

PASSAGE BETl^EEN

PLUO «oPOST/yALL

vmiwr

,tiltf»"'''^'''^^

Fig. 18. Eight half of melon cut away to show inner edge of insertion
of pars labialis and its relation to profundus. The premaxillary sac lateral
to the naris has been pulled open to show the diagonal membrane.

BULL. MUS. COMP. ZOOL.

Fig. 18

K-

c/j

~ —

\—

u—

U-J

-z.

—1
<

\
\

V—

ct^

U_J

-z.

C/O

cr:^

Li_l

-ZL

(>i

Fig. 19. Valve and plug drawn forward to show anterior and posterior
limits of attachment of plug on septum and position of anterior and posterior
folds of vestibular sac on left side.

BULL. MUS. COMP. ZOOL.

-^ (±3 j4^ Q

Fig. 19

1 1

^—

Li_

Li-

U_J

ce:

—

U_I

Ll_)

U-J

•z.

a::

Fig. 20. Exploded diagram of the sacs cut open and distorted, to show
their relation to each other and to the main passage. The outer part, con-
striction, and inner part have been separated as shown by arrows. Jagged
lines show where the sacs have been cut open. Details of the plug and main
passage are omitted, and the folds of the vestibular sae are not shown.

Outer part, A. This shows the vestibular sac (vs) as an expanded part
of the passage between the valve (v) and the slit (s), Avhere the plug (p)
passes beneath the posterior wall (piv)- This slit is at once part of the
main passage and the inner opening of the vestibular sac. The outer open-
ing (o) of the vestibular sac is at the lips of the blowhole.

• Constriction, B. The tubular sac (ts) lies on each side of the slit (s),
and its opening (e) is downward into the passage external to the plug.
The entrance to the connecting sae (cs), which has been stretched open (ef.
Figs. 1, 22), lies external to the liplike extension (I) of the plug and passes
upward (u) to join the entrance of the tubular sac.

Inner part, C. The plug has been cut away medially to show the naris
(n) and the inner opening of the premaxillary sac (ps) around the mar-
gin (m) of the premaxillarj- l)one. Laterally the premaxillary sac has been
cut, and part of the blowhole ligament (t) removed to show the second
opening (h) of this sae which is into the passage external to the liplike
extension (/) of the itlug. The opening of the connecting sae (cs) downward
into the imssage in the region marked i has been mm-li widened.

BULL. MUS. COMP. ZOOL

/ ^ Y (cut ends)

Fig. 21. Dorsal view of passage cut open at commissure of blowhole.
Anterior fold of vestibular sac drawn forward to show transverse extent
of passage at point where plugs pass beneath posterior wall.

Fig. 22. Diagram of dorsal view of passage with outer part including
vestibular sac cut away. On the left side of the figure the opening where
the plug passes beneath the posterior wall marks the inner limit of the
vestibular area; the position of the tubular sac around this slit and the
outline of the plug beneath this sac are indicated by broken lines; the con-
necting sac is omitted. On the right side of the figure, the posterior wall of
the passage has been cut away from the skull (along irregular line) and
folded forward to show the lateral entrance to the premaxillary sac. The
dotted line shows the attachment of the superficial wall of the premaxillary
sac external to the liplike extension of the plug. The broken line shows
the margin of the plug continuous with that of the premaxillary sac. The
entrance to the connecting sac (as well as the sac, which is not shown) is
very much smaller on this side than on the right (cf. Fig. 1).

BULL. MUS. COMP. ZOOL.

Fig. 21

Vesi s..

Ofenin(j of ^-^

PASSAGE BETl/^EEN
DORSAL SURFACE
or PLUCj h

posi nMi

Ant fold of
rio-ht vest s.

/Post fold of

VESI s.
/Post, lip

^'DRAIAIN BACK/

Valve

Opening of passage

\ BETlAjEEN dorsal

/ v
/ \

SURFACE OF PLUG / \

\ Sb POST l^ALL /^

//Right \
// plug \

Fig. 22

Transverse extent of passage
beneath vests.

Bony naris

PUKE
EXTENSION
OF PLUG

?0 ST PART- - ~%y'

OF TUB.S

Premaa. s.

/ Concave dorsal sur-
face DELEFT PLUG

.lAT ENTRANCE TO
/ PREMAX. S. EX-
TERNAL TO PLUGS

Entrance TO

CONN. s.

Entrance to

TUB. s

Ventral surface
of post imall

Fig. 23. Dorsal view of skull showing muscle attachments (on left of
figure), and arrangement of bones around nares.

BULL. MUS. COMP. ZOOL.

Occipital

Fig. 23

^^/^

WW&;. ae

TM/AP^

lligg] at

REMAXILLA

pars lab/cbJis
nasa) p)ua m

Fig. '24:. Photograph of rostral face of section at a, Fig. (i. The cuts iiidi
eated in Fig. 7 have been made and pulled open to show liplike extension
of plug and entrance to connecting sac as well as the interior of the posterior
part of the tubular sac. The string following the passage outward is hidden
as it passes between the dorsal surface of the plug and the concave mass of
the posterior wall, visible as it passes around the anteiior fold of the
vestiljular sac, and hidden again till it emerges from lilowhole. A1)brevia-
tions: a, pajier marker at entrance to connecting sac; //, lateral end of
opening of tuliulai' sac into passage; c, paper marker in bend leading to
anterior jiart of tubular sac; d, interior of posterior i^art of tubular sac; e,
valve, or anterior lip of blowhole; /, diagonal membrane in left side of
passage; r/. anterior fold of vestibular sac; //, passage beneath plug; i,
string entering passage superficial to plug; ,/, anterior nutrgin of opening
of tubular sac; 7;, liplike extension of plug.

Fig. l^.l. Rostral face of section through melon made at ^7, Fig. fi. showing
insertions of pars labialis and the gradual transition from the more oilv
melon to the more fibrous blubber. Abbreviations: ))t, melon; n, blubber;
o, rostrum ; /), pars labialis.

BULL. MTS. COMP. ZOOL.

Fi-. 24

BULL. MIS. (■0.\li\ ZOOL.

Fig'. 25

Fig. 26. Beginning of blow. Alternate frames from pictures taken at
64 frames per second looking into left side of passage as it opens. In this
and Figures 27 and 28 the following alibreviations are used:

b, bony wall of naris; r, convex mass of posterior wall;

(1, diagonal membrane; /, anterior fold of vestibular sac;

/, liplike extension of plug ; p, plug ; s, septum ; r, valve.

Here as well as in the following sequences (Figs. 27-.30), anterior is at
the bottom of the picture, posterior at the top. In this sequence in the
fifth frame the edge of the diagonal memlirane shows internal to the
convex mass; as the blowhole aperture widens, the liplike extension of the
plug appears as a pale mass external to the diagonal membrane in the lower
left hand corner of the open passage (frames 11, 13, and 15).

BULL. MUS. COMP. ZOOL.

Fiff. 26

Fig. 27. End of blow l)egun in Figure 26. Uninterrupted sequence of
frames from pictures taken at <i4 frames per second looking into passage
;l little to left of midline. As it closes, first the plugs (frame 3), then the
anterior fold of the vestibular sac (frame -4), and last the valve are seen
falling into place one after the other. The asymmetrical closing as well as
the ])Ositions of the sejitum and concav(» mass of the jiosterioi- wall show
clearly.

BULL. Ml^S. COMP. ZOOL.

Fig. 27

Fig. 28. Slightly different view of blowhole closing, showing more of
right half of passage. The frames are in uninterrupted sequence and were
taken at 6-i frames per second. The outer shadow in the left side of the
passage (see especially frames 1-4) marks the entrance to the vestibular
sac; the inner shadow is in the naris internal to the convex mass of the
l)Osterior wall. The fourth frame shows the plug closing beneath this mass.
The sixth, seventh, and eighth frames show the anterior fold of the vestibu-
lar sac closing on top of this mass.

BULL. MUS. COMP. ZOOL.

Fig. 28

Fis. -^. rniiitiMrupted sequence of frames from pietiues taken at 64
flames i>ci- second slio\vin«- water running into open blowhole as porpoise
sulinierjics, ami low spray as some water is forced liack out.

BULL. MUS. COMP. ZOOL.

Fig. 29

Fig. 30. Uninterrupteil sequeiu-e of frames from pit-tures taken at 64
frames per second showing wet Idow following submergence with blow-
hole open.

BULL. MLTS. COMP. ZOOL.

Fig. 30

Bulletin of the Museum of Comparative Zoology

AT HARVARD COLLEGE

Vol. 114, No. 5

STLICIFTED MIDDLE ORDOVICTAN TRILOBTTES
THE ODONTOPLEURTDAE

By H. B. Whittington

With Twenty-pour Plates

CAMBRIDGE, MASS., U. S. A.
PRINTED FOR THE MUSEUM

May, 1956

Publications Issued by or in Connection

WITH THE

MUSEUM OF COMPARATIVE ZOOLOGY
AT HARVARD COLLEGE

Bulletin (octavo) 1863 — The current volume is Vol. 1 14.

Breviora (octavo) 1952 — No. 53 is current.

Memoirs (quarto) 1864-1938 — Publication was terminated with Vol. 55.

Johnsonia (quarto) 1941 — A publication of the Department of Mollusks.
Vol. 3, no. 35 is current.

Occasional Papers of the Department of Mollusks (octavo) 1945 —
Vol. 1, no. 18 is current.

Proceedings of the New England Zoological Club (octavo) 1899
1948 — Published in connection with the Museum. Publication terminated
with Vol. 24.

Bulletin of the Museum of Comparative Zoology

AT HARVARD COLLEGE

Vol. 114, No. 5

SILICIFIED MIDDLE ORDOVICIAN TRILOBITES
THE ODONTOPLEURIDAE

By H. B. Whittington

With Twenty-four Plates

CAMBRIDGE, MASS., U. S. A.
PRINTED FOR THE MUSEUM

May, 1956

No. 5 — Silicificd Middle Ordovician trilohites.
The Odontopleuridae

H, B. Whittington

CONTENTS

Introduction and acknowledgements .

Terminology

Stratigraphical occurrence and localities
Part I : Morphology, Ontogeny and Evolution
Morphology of holaspid exoskeleton
Cephalon ....

Thorax .....

Pygidium ....

External surface
Abnormal specimens
Ontogeny .....

Locus of segmental divisions
Mode of life ....

Origin and evolution .
Part II : Systematic Descriptions
Famity Odontopleuridae Burmeister
Discussion of svstematics of Odontopleuridae
Subfamily Odontopleurinae Burmeister
Odoniopleura Emmrich ....

Primaspis K. and E. Richter .

P. ascitus n. sp. . . . . .

Leonaspis R. and E. Richter .
Diacanthaspis AYhittingtou
D. coo peri Whittington

D. lepidus n. sp

D. secretus n. sp. .....

Discussion of ontogeny of Diacanthaspis

D. ulrichi n. sp

D. aff. ulrichi n. sp. .
Z>. orandensis n. sp. .
D. scitidus n. sp. .

Page
159
160
162
166
166
166
173
175
175
178
178
182
185
186
193
193
194
195
196
198
199
205
207
211
216
220
224
225
228
228
230

156

BULLETIN : MUSEUM OF COMPARATm: ZOOLOGY

Acidaspis Murchison ....
Dudleyaspis Prantl and Pfibyl
Radiaspis R. and E. Richter .
Subfamily Miraspinae R. and E. Richter .
Miraspis R. and E. Richter .

fM. sp. ind

Ceratocephala Warder ....

C. laciniata Whittington and Evitt

C. rarispina n. sp.

C. (Ceratocephalina) n. subgen. .

C. (C.) tridens n. sp. ....

Proceratocephala Prantl and Pfibyl
Whittingtonia Prantl and Pfibyl .
Dicranurus Conrad ....

Notes on other miraspinid genera .
?Subfamily Miraspinae R. and E. Richter .
Odontopleurid protaspis ....

Subfamily Apianurinae n. subfara. .
Apianurus n. gen. .....

A. harhatus n. sp. .

Specimens at localities 2 and 3 differing

A. 'barhatus n. gen., n. sp.,
A. glaber n. sp. .
A. sp. ind. ......

A. aff. fiircata Linnarsson .
Calipernurus n. gen. ....

C. insoUtns n. sp.
Subfamily Selenopeltinae Corda

Selenopeltis Corda .....

Other genera, subgenera and species sometimes
ferred to Odontopleuridae

References

Explanation of Plates 1-24

from

Page

234
235
235
236
238
238
239
242
243
244
247
248
248
248
250
250
252
252
254

205
268
269
270
271
273
279
279

282
283
289

ILLUSTRATIONS
Plates

Plate
1.

Primaspis ascitus n. sp.

WHITTINGTON : SILICIFIED ODONTOPLEUBIDAE

157

Plate

2.
3.
4.
5.
6.
7.

8.
9.

10.
11.

12.
13.
14.

15.

16.

17.
18.
19.
20

21.

22.
23.
24.

Primaspis ascitus n. sp. .
Diacanthaspis cooperi Whittington .
Diacanthaspis lepidus n. sp. .
Diacanthaspis lepidus n. sp.
Diacanthaspis secretus n. sp. .
Diacanthaspis secretus n. sp. and D. lepidus

n. sp

Diacanthaspis ulrichi n. sp.

Diacanthaspis aff. ulrichi n. sp. and D. cooperi

Whittington

Diacanthaspis orandensis n. sp.
Diacanthaspis orandensis n. sp. and D. cooperi

Whittington

Diacanthaspis scitulus n. sp.

Diacanthaspis scitulus n. sp. .

Miraspis sp. ind. and Ceratocephala laciniata

Whittington and Evitt ....
Ceratocephala rarispina n. sp. and C. {Cerato

cephalina) n. subgen., sp. ind.
Ceratocephala (Ceratocephalina) tridens n

subgen., n. sp.
Apianurus harhatus n. gen., n. sp. .
Apianurus harhatus n. gen., n. sp. .
Apianurus harhatus n. gen., n. sp. .
Apianurus harhatus n. gen., n. sp., and A. aff

furcata Linnarsson ....
Apianurus glaher n. gen., n. sp., Apianurus n

gen., sp. ind., and miraspinid ? protaspis
Calipernurus insolitus n. gen., n. sp.
Calipernurus insolitus n. gen., n. sp.
Calipernurus insolitus n. gen., n. sp. .

Figure
1.

Figures

Notation for spines of Stage 0 cranidium .
Ilypostomes characteristic of subfamilies .
Classification, relationships, range in time of
Odontopleuridae ......

Page
161
171

: 88-1 89

158

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Figure Page

4. Odontopleura ovata Emmrich .... 197

5. Primaspis primordialis (Barrande) . . . 198

6. Development of Primaspis ascitus n. sp. . . 204

7. Leonaspis n. sp. . . . • • ■ 206

8. Age and comparison of species of Diacanthaspis 208-209

9. Development of Diacanthaspis cooperi Whit-

tington ........ 213

10. Keconstruction of Diacanthaspis lepicius n. sp. g^g

11. Development of Diacanthaspis secretus n. sp. 222

12. Development of Diacanthaspis ulrichi n. sp. . 227

13. Acidas2:)is hrightii Murchison .... 233

14. Dudley aspis quinquespinosa (Lake) . . 234

15. Miraspis mira (Barrande) .... 237

16. Ceratocephala laciniata Whittington and Evitt 239

17. Whittingtonia Mspinosa (M'Coy) . . . 247

18. Dicranurus monstrosus (Barrande) . . . 249

19. Reconstruction of Apianurus barhatus n. gen.,

n. sp 253

20. Cephalon of Apianurus barbatas n. gen., u. sp. 256

21. Triangular graph of dimensions of Apianurus

harbatus n. gen., n. sp. . . . . . 259

22. Development of Apianurus barbatus n. gen.,

n. sp 261

23. Triangular graph of dimensions of Apianurus

barbatus n. gen., n. sp., and Calipernurus

insolitus n. gen., n. sp. . . . . . 272

24. Reconstruction of Calipernurus insolitus n. gen.,

n. sp 274

25. Selenopeltis buchi (Barrande) .... 281

Tables

Table Page

1. Numbers of exoskeletal parts of two species

of Diacanthaspis at localities 2-4 . . . 217

2. Numbers of exoskeletal parts of three species

of Diacanthaspis at locality 8 . . . 229

3. Numbers of odontopleurid cranidia at local-

ity 8 242

4. Numbers of exoskeletal parts of all sizes of

Apianurus n. gen., and Calipernurus n. gen. 267

WHITTINGTON : SILICIFTED ODONTOPLEURIDAE 159

INTRODUCTION
AND ACKNOWLEDGEMENTS

The photographs show the remarkable nature of the silicified
material from Virginia — superbly preserved, free from the
matrix, undistorted, and including adults and growth stages.
These specimens afford a detailed picture of an early part of the
history of odontopleurid trilobites, and thus throw new light
on the morphology and evolution of the family. The best odon-
topleurid material previously known was that described by
Barrande (1852, 1872) and recently re-studied by Prantl and
Pi^'ibyl (1949). Although these Bohemian odontopleurids range
in age from Ordovician to Devonian, they are not as well pre-
served, nor are growth stages known. It is my good fortune that
the Museum of Comparative Zoology houses the Schary Collec-
tion (to which Barrande originally had access), for this circum-
stance enabled me to study the Bohemian species at first hand.
A visit to Britain in 1953, and generous loans from American
museums, have permitted me to see additional specimens. The
type and certain other species have been described elsewhere
(Whittington, 1956b), and two silicified species from Virginia
were treated of by Whittington and Evitt (1954). In Part II
of this paper the remainder of the silicified odontopleurids are
described in detail. In Part I, the morphology, ontogeny, and
evolution of the family are discussed in general terms. Part II
embodies a classification arising out of the new data, and my
views on evolution, and includes notes on all the genera which
have been referred to the Odontopleuridae.

The amount of morphological detail which the silicified exo-
skeletons reveal enables fine distinctions to be drawn between
genera and between species, i.e. it encourages "splitting." Ex-
amples are the new subgenus Ceratocephala (Ceratocephalina),
the differences between Apianurus n.gen. and Calipernurus
n.gen., and between some of the species of Diacanthaspls. On
the other hand, the seven species of Diacanthaspis seem to form
part of a natural, related group, and I have widened the original
diagnosis of Diacanthaspis rather than split it into subgeneric
groups. In dealing with the less well-preserved and often in-
complete material on which some Ordovician and all later genera
have been established, I have tended to "lump" rather than

160 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

"split." Thus, widely varying degrees of morphological dif-
ference separate the genera shown in Text-figure 3, and these
differences are not of equal rank.

Dr. G. Arthur Cooper and his colleagues at the U. S. National
Museum first discovered silicified trilobites in the Middle Ordo-
vician limestones of Virginia about 1935. In succeeding years
a large collection was prepared, and in 1946 Dr. Cooper invited
me to study it, under project grant 491-46 from the Penrose
Bequest of the Geological Society of America. Meanwhile, Dr.
William R. Evitt, University of Rochester, independently dis-
covered these fossils, and since 1947 he and I haA^e collaborated
in the preparation of additional material and in studies of it.
The silicified trilobites here described are thus derived from
our own as well as IJ. S. National Museum collections. A special
debt of gratitude is owed to Mrs. Evitt for her painstaking sort-
ing of the finest residues, which brought to light many of the
tiny growth stages. I am also grateful to Mrs. Stanley J. Olsen,
Mrs. Robert E. Kay, and Mr. Ira B. Laby for preparing the
enlargements from my negatives and for aiding in mounting the
plates. Text-figures 10 and 19 were drawn from my sketches by
Mr. F. Y. Cheng, the remainder by Miss Pat Washer. Professor
L. Stormer, Paleontological Institute, Oslo, kindly permitted me
to describe a Norwegian species here.

TERMINOLOGY

The terminology used herein is the same as that employed by
Whittingtou and Evitt, 1954, pp. 11-14, with certain additions
and emendations : —

Cephalic spines: to simplify description of cephalic spines a
system of letters and numbers has been adopted (Text-fig. 1).
Median or paired axial spines of the glabella have been num-
bered 1-5, commencing with the median occipital and numbering
forwards; 2a is an additional pair appearing later in the on-
togeny than 1 and 2. These numbers do not correspond w4th
those used by Whittington and Evitt (1954, text-fig. 16), where
2a is numbered 2, 2 is 3, 3 is 4, etc. Certain paired cheek spines
have been lettered as shown, in the protaspis and Stage 0
cephala of Diacanthaspis (Text -figs. 9, 11, 12) and Apianurus
n. gen. (Text-fig. 22) there is a long, backwardly-directed spine

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 161

at the extremity of the posterior border of the fixed cheek, as
well as a long spine on the border of the free cheek. These
spines have been called fixigenal and lihrigenal respectively
(terms proposed by Richter, 1932), in preference to the older
terms metacranidial and parial of Raw (1925).

Antennular notch: a notch in the margin of the free cheek
immediately outside the anterior branch of the facial suture (a in
Text-fig. 20; see also PI. 1, figs. 2, 5; PI. 17, figs. 7, 10). As

Figure 1. Cranidium of Stage 0 odontopleurid based on that of Apianurus
harbatus n.gen., n.sp., with paired spines numbered and lettered as follows :
1, 2a, 2, 3, 4, 5, six pairs of axial glabellar spines numbered forward from
the occipital ring; Ai, A2, A3, on fixed cheek; B, C, on posterior border; D,
on fixed cheek behind palpebral lobe; Er, on eye ride; Fx, fixigenal; Op,
paired occipital spine; PI, on summit of palpebral lobe.

explained under "Cephalon: hypostome" and "Mode of Life,"
this notch may permit the antennule to protrude forward when
the animal is resting on the sea bottom.

Pleural spines: The pleura may be divided by a pleural furrow
into a convex anterior and posterior band, and each of these
bands may be prolonged distally into a spine (the anterior into
more than one spine), here called anterior pleural (not terminal

162 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

of Prantl and Pfibyl, 1949, p. 128) and posterior pleural spines
respectively.

Pleural region is used for that part of the pygidium outside
the axis, rather than pleural lobe.

Pygidial spines: Commonly one paired spine is longer and
stouter than the others, and is here called the major spine; it
may arise from the border or the surface of the pleural region,
and the base is always connected to the first axial ring by a low
ridge, here termed the pleural ridge.

Sagittal (sag.), Exsagittal (exs.), and Transverse (tr.) refer
respectively to the median line, a line parallel to, but outside of
the median, and a direction at right angles to the median. The
abbreviations used in the text are given in parentheses.

STRATiaRAPHICAL OCCURRENCE
AND LOCALITIES

The specimens described here were obtained from limestones
of the Edinburg, Oranda and Martinsburg formations of the
Shenandoah Valley, northern Virginia. The stratigraphy of
these formations has been described by Cooper and Cooper
(1946), and the classification and correlation is summarized in
Twenhofel et al. (1954). Trilobites from the Edinburg and
underlying limestones have been described by Evitt (1951),
Cooper (1953), and Whittington and Evitt (1954), and the
latter work also contains an account of the mode of occurrence,
preservation, and method of extraction of the silicified specimens.
Trilobites from the lower Martinsburg formation have been de-
scribed by Whittington (1941), Evitt (1953) and Evitt and
Whittington (1953). Odontopleuridae first appear in the lower
Lincolnshire limestone (Ceratocephala triacantheis Whitting-
ton and Evitt and fragments of Apianurus n. gen.), below the
Edinburg formation. In the lowest part of the latter formation
(localities 2-4) they are abundant and exhibit the greatest
variety. In the slightly higher horizon at locality 6 odonto-
pleurids are rare, but they become numerous and varied again
some 400 feet higher in the section in the Oranda formation. The
latter is not more than 50 ft. thick, and in the lowest part of
the succeeding Martinsburg formation silicified odontopleurids
again occur in fair abundance and varietv. Not all the blocks of

WHITTINGTOlSr : SILICIFIED ODONTOPLEURIDAE 163

limestone collected yield good material, and the same species do
not appear at different localities in the same formation — per-
haps because of slight differences in horizon. Thus Diacanthaspis
ulrichi n. sp. is the only odontopleurid at the lower Edinburg
locality 7, is rare at locality 4, and unknown at 2 and 3. Pri-
maspis ascifus n. sp. is only known at locality 10 in the lower
Martinsburg, and does not accompany t). cooperi at localities 9,
11 and 12.

The localities (cf. Whittington and Evitt, 1954, pp. 5-6) and
the odontopleurids obtained at each are listed below. Tables
1-4 give numbers of specimens recovered of particular species.

Edinburg limestone

Locality 2 — Lower part of Edinburg limestone (bed 18 of
Cooper and Cooper, 1946, geologic section 19, pp. 94-95), yellow-
ish-weathering argillaceous limestone forming edge of quarry
and along strike of same bed, in field between quarry and rail-
road ; just north of railroad tracks at switch a quarter of a mile
east of Strasburg Junction, just west of Strasburg, Shenandoah
County, Virginia. Some of the finest specimens illustrated here
came from blocks collected at this locality and prepared by Dr.
G. Arthur Cooper.

Diacanthaspis lepidus n. sp.

Diacanthaspis secretus n. sp.

Ceratocephala laciniata Whittington and Evitt, 1954

Apianiirus barhatns n. gen., n. sp.

Apianiirus glaher n. gen., n. sp.

Apianurus sp. ind.

Calipernurus insoUtus n. gen., n. sp.

Locality 3 — Lower part of Edinburg limestone, section in
field on south side of road, 0.2 mile east of Strasburg Junction,
just west of Strasburg, Shenandoah County, Virginia. The sec-
tion dips about 38° SE. The lowest beds were 6 feet of granular
limestones with GirvaneUa sp., about 91 feet from the east edge
of the quarry dump at the top of the field. These may be upper
Lincolnshire limestone. About 20 feet stratigraphically higher,
alternations of dark granular limestone and dark fine-grained
limestone with sponge spicules were seen. These beds seem to be

164 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

the basal Botetourt limestone member of the Edinburg formation
(Cooper and Cooper, 1946, p. 80), and blocks from them were
collected and prepared by Dr. G. Arthur Cooper, Dr. and Mrs.
W. R. Evitt, and Whittington. These yielded all the species
known from locality 2 except Apianurus sp. ind. In addition
Miraspis sp. ind. and Ceratocephala {Cerafocephalina) tridens
n. subgen., n. sp., occur.

Locality 4 — Botetourt member, lower part of Edinburg lime-
stone, in upper part of field northeast of Virginia State Highway
639, at a point 0.25 mile from its junction with U.S. Highway 11.
This junction is 0.7 mile southwest of Strasburg, Shenandoah
County, Virginia. The outcrop is approximately half a mile
southwest of locality 3 along the strike of the beds. Collected
and prepared by Dr. and Mrs. W. R. Evitt and by Whittington.
and notable for yielding some of the finest tiny specimens.

Diacanthaspis lepidus n. sp.

Diacaniliaspis sccretus n. sp.

Diacanthaspis ulrichi n. sp.

Ceratocephala laciniata Whittington and Evitt, 1954

Ceratocephala {Ceratocephalina) tridens, n. subgen., n. sp.

Apianurus harhatus n. gen., n. sp.

Apianurus glaher n. gen., n. sp.

Locality 6 — Edinburg limestone, lower part. Hupp Hill, at
entrance to Battlefield Crystal Caverns, and in field on opposite
(east) side of U. S. Highway 11, about ly^ miles north of Stras-
burg, Shenandoah County, Virginia. Discovered by Dr. G.
Arthur Cooper, later collections by Whittington. Odontopleurids
are rare, only one or two parts of exoskeletons of the following
being known: —

^te

Diacanthaspis secretus n. sp.

Ceratocephala laciniata Whittington and Evitt, 1954

Apianurus harhatus n. gen., n. sp.

Locality 7 — Lower part of Edinburg limestone, 300 feet ±
south 40° east of bridge, li/4 miles east of Edinburg, Shenandoah
County, Virginia. Originally collected by E. 0. Ulrich, later
(1931) by Charles Butts, and yielding only one odontopleurid,
Diacanthaspis ulrichi n. sp.

WHITTTNGTON" : SILTCIFIED ODONTOPLEURIDAE 165

Oranda formation

Locality 8 — Lower 5 feet of formation, eobbly limestone, in
bank and pasture on north side of Virginia secondary highway
777, just west of its junction with Virginia secondary highway
910, and circa 300 yards north of Greenmount church, five
miles north of Harrisonburg, Rockingham County, Virginia.
Discovered by G. Arthur Cooper, later collections by Cooper
and A. R. Loeblich, Jr., W. R. Evitt, and Whittington.

Diacanthaspis orandensis n. sp.
Diacanthaspis scitulus n. sp.
Diacanthaspis aff. ulrichi n. sp.
Ceratocephala rarispina n. sp.
Apianurus harhatus n. gen., n. sp.

Martinsburg shale

Locality 9 — Road cut, gutter, and loose blocks in pasture on
west side of Virginia secondary highway 910, about half a mile
north of Greenmount church, five miles north of Harrisonburg,
Rockingham County, Virginia. Same as locality 1 of Evitt and
Whittington, 1953, p. 55. Collected and prepared by Dr. and
Mrs. W. R. Evitt, Dr. G. Arthur Cooper, and Whittington.

DiacantJiaspis cooptri Whittington, 1941
Locality JO — Pasture on north side of Virginia secondary
highway 772, about 1 mile east of Greenmount church, five miles
north of Harrisonburg, Rockingham County, Virginia. Collected
and prepared by Dr. and Mrs. W. R. Evitt.

Primaspis ascitus n. sp.
. Diacanthaspis cooperi Whittington, 1941.

Locality 11 — Loose blocks in pasture on north side of Virginia
secondary highway 616, Vj mile east of intersection with Virginia
secondary highway 699, and 2i/2 miles north-northeast of Spring
Hill, 7 miles north of Staunton, Augusta County, Virginia. The
Oranda formation immediately underlies the lower Martinsburg
formation and outcrops a short distance to the west. The locality
is that from which all the material described by me (including
Diacanthaspis cooperi) in 19.41 came. At that time (Whitting-
ton, 1941, p. 492) tAvo localities were given, distant respectively
2^2 miles north-northeast, and 3 miles north-northeast, of Spring

166 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Hill and Long Glade. Spring Hill is a new name for the settle-
ment formerly called Long Glade, and the confusion probably
arose because the blocks of limestone were collected at different
times (cf. Evitt, 1953, p. 34). The first blocks were collected
by Dr. G. Arthur Cooper, later ones by Cooper and "Whittington.
Locality 12 — In field on south side of Virginia secondary
highway 753, 1 mile west of intersection with Virginia secondary
highway 732, and 31/2 miles north-northeast of Spring Hill,
Augusta County, Virginia. This locality, visited by Dr. G.
Arthur Cooper and Whittington, is one mile northeast of locality
11, and yields Diacanthaspis cooperi.

PART I: MORPHOLOGY, ONTOGENY
A^^) EVOLUTION

Morphology' of Holaspid Exoskeleton

The family diagnosis (p. 193) epitomises the morphology of
an odontopleurid, and in this section certain aspects are com-
mented on more fully, these being chiefly features on which I
have information additional to that available to Prantl and
Pfibyl (1949).

Cephalon

Attitude: The characteristic attitude of the strongly convex
cephalon is discussed below under "Mode of Life." The convex-
ity of the cephalon and its position relative to the thorax and
pygidium were recognized by some earlier authors (e.g. Weller,
1907, PL 23, figs. 1-4; Warburg, 1933, p. 9, footnote; see also
Whittington and Evitt, 1954, p. 54), but not, apparently, by
Barrande. Many of Barraude's drawings were made from speci-
mens flattened in shale or calcareous mudstones, and show an
exterior view of the cephalon combined with a dorsal view of
thorax and pygidium (for definitions of these terms see Whit-
tington and Evitt, 1954, p. 11). The present reconstructions
attempt to remedy this situation (compare Text-figs. 4, 15,
25 with Barrande, 1852, PI. 39, figs. 33, 1; PL 37, fig. 25, respec-
tively).

Occipital ring: Conspicuous in odontopleurids is the way in

WHITTINGTON- : SILICIFIED ODONTOPLEURIDAE 167

which the antero-lateral portion of the occipital ring merges
with the inner posterior corner of the cheek. Behind this region
the posterior border is distinctly separated from the remainder
of the occipital ring, and the axial furrow more or less well
developed, as it is beside the lateral glabellar lobes. The occipital
ring in most odontopleurid genera is strongly convex and com-
monly elongated so that it projects behind adjacent parts of the
posterior border. As Prantl and Pfibyl pointed out (1949, p.
127), a median occipital tubercle is always present, and a median
spine or paired spines may also be developed. The species of
Diacanthaspis here described (Text-fig. 8) show that the median
tubercle, or paired spines, or both may be greatly enlarged, or
none may be especially prominent, in a group of closely related
species. In genera such as Ceratocephala and Miraspis (Text-
fig. 15), where long, stout, paired occipital spines are developed,
a posterior occipital band, below and behind the bases of these
spines, is developed (cf. Warburg, 1933). Such a band is, how-
ever, not present in Acidaspis (Text-fig. 13) or Bicranurus
(Text-fig. 18), though the median (in the former) and paired (in
the latter) occipital spines are large. The view of Reed (1925.
p. 420) that the posterior occipital band represents the occipital
ring, and that the part in front belongs to the glabella, has been
adequately refuted bv Warburg (1933) and Opik (1937. pp.
45-47).

Where I have been able to observe both surfaces of the exo-
skeleton, as in Primaspis ascitus, n. sp. (PI. 1, figs. 1, 6), the
occipital lobes are small, gently convex, subcircular in outline,
situated immediately behind the deep outer part of the occipital
furrow. Their convexity, rather than any distinct furrow, sep-
arates them from the rest of the occipital ring inside and behind
them, and they are indistinctly bounded posteriorly. On the
inner surface they are seen to be in front of the outer part of
the doublure of the occipital ring. In structure and position
they are analagous to the lateral glabellar lobes, though they do
not extend back to the posterior margin of the segment, and
are much less distinct than the first and second glabellar lobes.
In Ceratocephala laciniata and C. triacantheis (Whittingtou and
Evitt, 1954, PL 6, fig. 4; PI. 8, figs. 1, 2; PI. 25, figs. IQ, 11)
occipital lobes are scarcely distinguishable, whereas in certain

168 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

younger species of the same genus (e.g. Warburg, 1933, text-
figs. 2, 3) the outer part of the occipital ring, between the occipi-
tal furrow and posterior band, is inflated. Similarly, in Dicra-
mirus hamatus (Whittington, 1956b, PI. 60, figs. 11, 12) the
inflation of the antero-lateral part of the occipital ring, though
vaguely defined, especially postero-laterally, is greater than in
D. mo7istrosus (Text-fig. 18). Species of other genera show the
same features, and indicate that, when occipital lobes are de-
veloped (they are absent altogether in some genera and absent
in certain species only of other genera), they are situated in the
antero-lateral corner of the occipital ring, probably outside the
margin of the doublure, and ill-defined on the posterior and inner
sides. Their variability suggests that presence or absence of such
lobes can scarcely be used as a generic character.

Glahellar lohes and furrows: Two pairs of lateral glabellar
lobes are always present, the first (basal) pair the larger, and,
except in a few genera and certain species of other genera, small,
variably-developed, third lateral lobes are present. Particularly
in Ceratocephala and Miraspis the third lobes are small and de-
pressed, and have been regarded as not developed by Prantl and
Pfibyl (1949, pp. 180, 194).- The convexity of the' lateral lobes
and degree of their separation from the median glabellar lob(^
are variable (contrast, for example, Leonaspis, Text-figure 7,
with Whittingtonia, Text-figure 17), and in Apian urns n.gen.
and Primaspis keyserlingi (Barrande, 1852, PI. 36, figs. 10, 12)
the lateral lobes are fused. The peculiarities of the glabellar
lobation of Selenopeltis (Text-fig. 25) set it apart. The frontal
lobe varies in width, sometimes being only as wide as the median
lobe {Acidaspis. Text-fig. 13) but usually it is expanded to
extend in front of the second or third lobes (Leonaspis n sp.,
Text-fig. 7 ; Dicranurus liamatiis, Whittington, 1956b, PI. 60,
fig. 15; Miraspis mira, Text-fig. 15). The eye ridge terminates
opposite the lateral extremity of the frontal lobe, and may merge
into it {Ceratocephala, see Whittington and Evitt, 1954, PI. 9.
fig. 2) or be separated from it by a low depression, the axial
furrow (Primaspis asciius n. sp., PL 1, figs. 2, 5). Hupe's dia-
gram (1953, p. 102, text-fig. 62; repeated in 1955, p. 236, text-
fig. 205) conveys a misleading impression, in that figure 62 (1),
.'howing 3 pairs of lateral glabellar lobes, is labelled "primitif"

WHITTINGTON : SILICIFIED ODOXTOPLEURIDAE 169

and ''normale," and figure 62 (3) with 2 pairs of lateral lobes,
is labelled as showing ''regression progressive et complete" of
the fourth segment (from the posterior margin). In fact no such
progressive reduction occurs, for genera with 2 and 3 pairs of
lateral lobes appear at about the same time {Diacanthaspis,
Primaspis, and Ceratocephala), and Ceratocepliala and Dicra-
nurus, both having 3 pairs of lateral lobes, are known in the
Middle Devonian.

The outer part of the occipital furrow, and each glabellar
furrow (especially at the inner end), is deepened and forms a
strong ridge on the inner surface (PI. 1, fig. 6; PI. 5, fig. 4;
PI. 7, fig. 5; PI. 12, fig. 7"; PI. 15, fig. 5; PI. 20, fig. 11; PI. 21,
fig. 6; PI. 22, fig. 6). The quartz along the crests of these ridges
may appear darker, strengthening the impression that these
ridges are points of muscle attachment. So far as is known,
however, these thick ridges are not extended as ventral processes.
The outer part of articulating and ring furrows is similarly
deepened (e.g. PI. 12, figs. 6, 18) but does not form so strong a
ridge on the inner surface.

Eye lohe, eye ridge, facial sutures: The external covering of
the eye of Ceratocephala has been described ("Whittington and
Evitt, 1954, p. 16) as having the outer surface smooth or faintly
divided into numerous closely spaced, slightly raised facets, the
inner surface showing shallow pits similarly arranged. The ex-
ternal covering of the eye of the silieified specimens of other
genera described here appears to be similar, the facets on the
outer surface sometimes clearly delineated (PL 2, figs. 18, 22;
PI. 3, fig. 21; PI. 16, fig. 23; PI. 17, fig. 21; PI. 23, figs. 9, 10).
This type of eye surface seems to be typical of odontopleurids.

The presence of the eye ridge, and of sutural ridges ("Whit-
tington and Evitt, 1954, pp. 13, 17-19), notably where the suture
approaches either border, is characteristic.

Fusion of the facial sutures, presumably a secondary phe-
nomenon, has been suggested as a criterion for distinguishing
genera of odontopleurids. It is true that in some specimens in
which the exoskeleton is preserved it is almost impossible to
detect the course of the suture, but whether or not this means
that secondary fusion has taken place is difficult to determine. In
any event this single character is not regarded as a reliable

170 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

criterion upon which to base a genus (Whittington and Evitt,
1954, p. 53).

Librigenal spine: Prantl and Pfibyl (1949, pp. 126, 131, etc.)
stated that, in Selenopeltis and miraspinids. the librigenal spine
arose from the "surface of the cheeks above the genal angle,"
rather than being a backward and outward extension of the
rolled borders, at the genal angle, as it is in odontopleurinids.
In Ceratocephala (AVhittington and Evitt, 1954, PI. 6, fig. 17;
PI. 25, figs. 12-14), the broad base of the librigenal spine merges
into the upper surface of the posterior and lateral borders and
the sutural ridge. In Ceratocephala {C cratocephalina n. subgen.)
(PI. 16, figs. 23, 24) it likewise arises from the upper surface of
the border, and a line of spines on the margin of the border is
continuous below it. The fact that the doublure extends beneath
the base of the librigenal spine (PI. 16, fig. 0) shows that the
spine arises from the border and not the surface of the cheek
(cf. Whittington and Evitt, 1954, p. 52). Thus I do not agree
with Prantl and Pfibyl (1949, p. 131, etc.) that the miraspinid
librigenal spine arises in a way that is fundamentally different
from, and originates from other cephalic segments than, that of
odontopleurinids. I l)elieve rather that the difference between
the librigenal spines in the two subfamilies is one of direction
and not place of origin. In apianurinids the librigenal spine also
arises from the cephalic border, but from a point farther for-
ward. The ontogeny of Diacanthaspis, of Apianurus n. gen., and
of Ceratocephala shows that the librigenal spine develops in
much the same way in each group, but does not show to which
cephalic segment it may belong.

Hypostome: The hypostomes of a species of Primaspis and
those of Apiayiurus n. gen. and Calipernurus n. gen. are described
here, and others in Whittington, 1956b. Thus the hypostomes of
more than half of the odontopleurid genera are known, and
seem to fall into four types, exemplified in Text-figure 2. The
"family resemblance" between them is shown particularly by
the small anterior wing (lacking the wing process), lateral notch
and pointed shoulder, middle body with large depression in
antero-lateral corner, from which the middle furrow runs inward
and backward, and tiny posterior wing. The types distinguished
form one of the main bases of the four subfamilies recognized.

WHITTINGTOX : SILICIFIED ODONTOPLEURIDAE

171

Prantl and Pfibyl (1949, p. 134, etc.) stressed the ditfereuce
between the odontopleurinid and the S elenopeltis-C eratocephala
types of hj'postome. The latter two present a considerable like-
ness, and in both the crescentic tip of the posterior lobe of the
middle body is inflated and extends outside the depression in the
antero-lateral corner of the middle body. In the odontopleurinid
(and apianurinid) type the tip of the posterior lobe of the mid-

Figure 2. HjTDOstomes charaeteristie of each of the four subfamilies of
Odontopleuridae. A, Odontopleurinae, Primaspis, based on that of P.
asffitus n.sp. B, Apianurinae, Apianuras, based on that of A. harbatus n.geu.,
n.sp. C, Miraspinae, C eratocephala, based on that of C. laciniata Whitting-
ton and Evitt. T), Selenopeltinae, Srlptiopeltis, based on that of S. hvchi
(Barrande).

die body does not extend as far forward. I hardly think this
difference is as great as Prantl and Pfibyl tend to make it by
saying that in odontopleurinids the middle furrow starts from
the lateral border furrow, whereas in Selenopeltis and Cerafo-
cephala it starts from the anterior border furrow.

The silicified specimens of hj^postomes have shown the exist-

172 BULLETIN : MUSEUM OF COMPABATn^E ZOOLOGY

ence of a cirealar bole through the doublure (PI. 3, figs. 9, 10;
PI. 10, figs. 21, 24; PI. 13, figs. 9, 11; PL 14, fig. 12; PI. 23,
fig. 15), situated ou or near the sharp flexure associated with the
shoulder. This hole is seen in species belonging to several genera,
but is not exhibited by all species of a single genus — for ex-
ample, Diacanthaspis. A similar-appearing opening in the ce-
phalic and thoracic doublure of other trilobites is the Pander-
ian opening, but whether or not this similarity implies that they
are analagous structures is uncertain.

The macula is not discernible in the odontopleurid hypostomes
here studied, except perhaps as the smooth area at the inner end
of the middle furrow of Pnmaspis ascitus, n. sp. (PI. 1, fig. 20).
The depression in the antero-lateral corner of the middle body
is conspicuous, however, in that it is smooth on the external
surface, and the quartz is often differently coloured, and on the
inner surface there may be some exfoliation. Thus it resembles
the similarly-situated area in Sphaerexochus (Whittington and
Evitt, 1954, p. 25) and may be an area of muscle attachment.

The problem of the manner in which the hypostome is attached
in odontopleurids has been discussed previously (Whittington,
1941, p. 516; Whittington and Evitt, 1954, pp. 20, 55-56). Iso-
lated rostra of the silicified species have not been found, but the
unique specimen of Acidaspis cincinnafievsis recently described
("Whittington, 1956b) has the rostrum in place. The posterior
edge of this rostrum is straight (except distally where it curves
to meet the inner edge of the cheek doublure), and obscurely
bevelled in this specimen. Measurement shows that the straight
edge fits exactly against the straight, bevelled, anterior edge of
the hypostome. The fit of the bevelled edges suggests that the
hypostome lies approximately in the horizontal plane (on the
orientation adopted here), or slopes slightly downward and
backward. The tip of the small anterior wing lies beneath the
intersection of the axial and preglabellar furrows, just in front
of the eye ridge. This intersection, and the border furrow just
outside it, is usually deepened in odontopleurids, though it does
not form a large projection on the inner surface. Thus the
attachment of the hypostome seems to be as previously suggested
— at the suture and vnth a muscular link between anterior
wing and/or pit at anterolateral corner of tbe middle bodv to

WHITTINGTON : SILICIFIED ODOXTOPLEUEIDAE 173

unspecified point or points on cranidiuni. One of the latter is
perhaps the preglabellar furroAv and border furrow outside the
inner end of the eye ridge (i.e. about where the anterior boss is
in other trilobites). The hypostome of Apianurus harhafus
n. gen., n. sp., is restored in approximately this position in Text-
figure 20. This attachment appears not to have been as rigid
as that in cheirurids and allied families (Whittington and
Evitt, 1954, pp. 19-21), and the evidence cited bj' Prantl and
Pfibyl (1949, p. 196) of many specimens of Miraspis mira in
which the hypostome is preserved Init rotated through 180° and
lying just in front of the cephalic margin, is interpreted by them
as indieating an ''easily movable connection." However, the
flat surfaces that are apposed along the hypostomal suture are
not suggestive of movement at the suture having been possible
during life.

Antennular notch: Figures 5, 7, of Plate 17 shoAv that there
is a -conspicuous notch in the anterior cephalic border immedi-
ately outside of where the connective suture crosses the border.
It has previously' been argued (Whittington and Evitt, 1954, p.
20) that the antenuule is attached to the surface of the anterior
boss and passes througli the lateral hypostomal notch. It may
also have passed through the notch (here called the antennular
notch) in the anterior cephalic border, W'hich would have per-
mitted the antennule to be forwardly extended, when the cephalic
margin w^as resting on the sea bottom, in the manner suggested
under "Mode of Life." The antennular notch is developed to a
greater or lesser extent in most odontopleurids (e.g. PI. 1. figs.
2, 5; PI. 12, fig. 2; PI. 22, figs. 3, 6).

Thorax

The eight thoracic segments of Leonaspis n. sp. (Whitting-
ton, 1956b ) are unusual, nine or ten being most commonly seen
in the family. The backwardly-couvex curve of the posterior
margin of the prominent axial ring (and the occipital ring) is
complemented by the curve of the articulating furrow. Medially,
the ring adjacent to this furrow is less elevated. The articulating
half ring- is long (sag.), particularly that of the first segment.
In some species and genera the antero-lateral part of the axial
ring is inflated, e.g. SeJenopeltis (Text-fig. 25), Miraspis (Text-

174 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

fig. 15), and Proceratocephala (Whittington, 1956b, Pi. 60, figs.
2, 10). A posterior band of the type seen on the occipital ring
in some genera has not been seen on axial rings.

The horizontal pleurae may be nnfurrowed as in Apianurus
n. gen. (PI. 18, figs. 6, 9) and Ceratocephala (Whittington and
Evitt, 1954, PI. 8, fig. 10), but are generally divided by the
straight, transverse pleural furrow into a narrow anterior and
wider (exs.) posterior band. The narrow, raised ridge of the
posterior pleural band of Selenopeltis, with its strong curve
convex forward, is distinctive. A posterior flange is present in
many species. At the fulcrum the pleural bands are continued
by spines: a stout, more or less horizontal, posterior spine, and
one or more slimmer, shorter, anterior spines, downwardly di-
rected. The doublure is curled under at the base of these spines,
and forms the fulcral articulating socket and process. In odonto-
pleurinid genera the broad-based librigenal spine is directed
horizontally outward and backward beside the thorax, and con-
sequently the first two or three thoracic pleurae are facetted
so that they fit beneath the posterior cephalic border and libri-
genal spine. Thus the anterior pleural spine is absent, and the
posterior pleural spine short (e.g. PI. 3, fig. 11; PI. 8, fig. 12).
The anterior pleural spines of the third or fourth, and successive,
segments are present and the posterior pleural spine is long and
backwardly directed. In miraspinids (Text-figs. 15, 18) and
apianurinids (Text-fig. 19), however, the librigenal spines are
respectively directed upwardly, and situated far forward on
the cephalon, and no facetting and tnnication of the thoracic
pleural terminations is necessary to permit fitting behind the
cephalon. Thus the posterior pleural spines are successiveh'
directed obliquely forward, outward, and obliquely backward.
This difference in development and direction of pleural spines
has been regarded by Prantl and Pfibyl (1949) as indicating a
distinction of family rank between odontopleurinids and mira-
spinids, but hardly seems to be of this magnitude (cf. Whit-
tington and Evitt, 1954, p. 52). Likewise the "stunting" of the
first two or three thoracic segments hardly seems an adequate
basis for the subgenus Leonaspis {Keftneraspis) Prantl and
Pfibyl (1949, p. 165; see Whittington, 1956b).

An inflation of the posterior pleural haiul at the fulcrum

WIIITTINGTON : SILICIFIED ODONTOPLET-RTDAE 175

characterises certain odontopleurinids, e.g. Priniaspis (PI. 1,
figs. 11-14), some species of Leonaspis, and Acidaspis (Whitting-
ton, 19561)) and may indicate relationship.

Pygidium

Major border spines, i.e. one notably larger pair among the
several arising from the pleural regions, are characteristic, but
may not be developed at all {Diacanthaspis secrefus n.sp., PI. 7,
fig. 10; Badiaspis radiata, see R. and E. Richter, 1917, text-fig.
10) or only slightly developed, as in some species of Cerafoce-
phoJa (Barrande, 1852, PL 38, figs. 5, 18). When present, the
major spine is connected at its base to the first axial ring by the
pleural ridge which runs across the pleural region. Commonly,
the major spine arises from the upper surface of the pygidial
border, but may take its origin from inside the border of the
pleural region (e.g. Diacanthaspis orandensis n.sp., PI. 10, figs.
17-19; Apianurus harbatns n.gen., n.sp., PL 18, fig. 12; Caliper-
nuriis insolitus n.gen., n.sp., PL 24, figs 27, 28). In this latter
case it seems quite independent of the border spines, though the
pleural ridge is present. When the major spine arises from the
border it is commonly the third or fourth spine, and in Miraspis is
unusually far forward on the border, though still the third spine
(cf. Prantl and Pfibyl, 1949, pp. 128-129). As remarked under
"Locus of segmental divisions" the major spine and pleural
ridge are homologous with the thoracic posterior pleural band
and spine.

The doublure of the pygidium is curled under, not wide, and
there is a small projection posteriorly, directed up toward the
axis.

External Surface

Short, thorn-like spines, tubercles, and granules are character-
istic; their form and arrangement are shown by the plates (cf.
Whittington and Evitt, 1954, Pis. 6-9, 25, 26). Rarely is the
external surface everywhere smooth, as it seems to be in
Primaspis kcyserlingi, but commonly the deeper parts of furrows,
and the doublure, are smooth. A symmetrical arrangement of the
prominent spines, making transverse rows across the exoskeleton,

176 BULLETIX : MUSEUM OF COMPARATIVE ZOOLOGY

is typical of early developmental stages (see below) and may
persist into the adult. In most adults this symmetry is less
obvious, and may be lost altogether (e.g. Primaspis ascitus, PI. 1,
fig. 1). Conspicuous paired spines on axial rings are common,
though rarely are they as large on the thoracic rings as in
Proceratocephala (Whittington, 1956b, PI. 60, figs. 2, 5, 10).

It was shown previously (Whittington and Evitt, 1954, pp.
56-59, text-fig. 1, etc.) that small openings occur on the distal
side of tubercles scattered along occipital and librigenal spines,
and at and near the tips of spines on the cephalic and pygidial
border. Such openings are characteristic of the specimens de-
scribed here (e.g. PL 7, figs. 12-11; PI. H, fig. 16; PI. 13, fig.
14; PL 23, fig. 6; PL 24, figs. 21, 24). It was further suggested
that in Ceratocephala laciniata there may have been a single
opening, possibly occupied by a hair, at the tip of the thorn-like
spines (Whittington and Evitt, 1954, p. 59; PL 8, fig. 11; PL 9,
figs. 2, 4; text-fig. 1) scattered over the external surface of the
cranidium. Among the few available cranidia of 3-. 'rarispina
n.sp., however, is one (PL 15, fig. 28) in which the thorn-like
spines, where not obviously broken, are closed at the truifcated
tip by a plate in which there are minute depressions (appearing
as dark spots in the photograph). The diameter of these de-
pressions is less than that of some of the quartz grains replacing
the exoskeleton, so that it is difficult to be sure of their nature.
It seems probable that they are the orifices of canals through the
exoskeleton. The thorn-like spines on the cranidium of Diacan-
thaspis are similar, sometimes showing an evenly eilt off tip
with a relatively large central opening (PL 7, fig. 15). A well-
preserved specimen of Z>. scitulus n.sp. (PL 13, figs. 16, 17), hoAv-
ever, has the truncated tip covered by a convex plate in which
there are tiny depressions. These may show a regular arrange-
ment of a central larger pit surrounded l)y smaller pits. A speci-
men of D. lepidus n.sp. (PL 4, fig. 22) shows a similar arrange-
ment. In Z>. cooperi and D. orandensis n.sp. (PL 9, fig. 6; PJ.
11, fig. 19) the tip is slightly expanded and subspherical, with a
large central depression or opening surrounded by smaller ones.
In B. aff". ulrichi n.sp. (PL 9, figs. 7, 9) some spines show only a
central opening, and others appear to show several openings.
One may conclude that in DiacantJiaspis, as in C. rarkpina n.sp..

WHITTIXGTON : SILJCIFIED ODOXTOPLEUEIDAE 177

the tip of the thorn-like spines was pierced by one or more minute
openings, but whether a hair issued from each opening is uncer-
tain. The tubercles of Primaspis ascitus n.sp. (PI. 2, fig. 21)
appear to be closed at the summit, and in Apionurus harhatus
n.gen., n.sp. (PI. 18, figs. 19. 22) the characteristically curved
tips of the thorn-like spines appear closed.

In Diacanthaspis lepidus n.sp. (PI. 5, fig. 1), D. secretus n.sp.
(PI. 6, fig. 22), and D. ulrichi n.sp. (PI. 8, fig. 27) the tip of the
librigenal spine is hooked. The hooked portion appears to be
smooth, without openings, but proximally to the tip the usual
type of opening is present. Such openings are numerous at the
tip of the libris'enal spines in species which do not have the hook
(PI. 13, fig. 14).

Tubercles on the external surface, as in Primaspis asciins n.sp.
or Calipernurus insolitus n.gen., n.sp.. seem never to have had
openings at the summit. The median occipital tubercle in ])oth
these species, however (PI. 2, figs. 21, 23; PI. 23- figs. 5, 7, 8), is
larger than any other, and has four pits, arranged to outline
a square (and rarely a faint, small, median pit) in its low,
domed surface. It appears early in ontogeny (PI. 2, figs. 6. 8;
PI. 24, fig. 1) and is retained in the largest holaspis known. The
pits, appearing as dark spots, do not seem to be the openings of
canals through the exoskeleton. The occipital doublure extends
close underneath the tubercle and well in front of it in P. ascitus
n.sp. (PI. 1, fig. 6). No similar tubercle is seen on the axial rings
of thorax, or pygidium. The same type of median occipital tu-
bercle has4jeen observed in other species of Primaspis, in Dia-
canthaspis hjoperi (PI. 11, fig. 18), D. ulrichi n.sp. (PI. 8, figs.
24, 30), D. aff. ulrichi n.sp. (PI. 9, figs. 8, 9), Z). orandcnsis n.sp.
(PL 11, fig. 20), and Leonaspis leonhardi Barrandc This
tubercle may be compared with the median glabellar tubercle of
Tretaspis seticornis (Stormer, 1930. pp. 85-87, figs. 36, 37),
which is siudlar but displays more clearly a fifth, centrally
situated pit, as does the four-celled sense-organ of Anaspides
(Hanstrom, 1934). Temple (1952, pp. 254, 258) has described
the median occipital tubercle of protaspid and early meraspid
stages of Dalmanifina olini as bearing five tiny tubercles, one
central and the others forming a square. This may be an anala-

178 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

<xo\\& organ to that of the odontopleurids, but represented by
tubercles rather than pits.

Hiipe (1953, pp. 80-81) sugge.sts that the median glabellar
tubercle is the specialized homologue of median tubercles of
succeeding segments, and this example in odontopleurids may
represent such a specialization (presumabh' for some sensory
function) of the median occipital tubercle.

Abnormal Specimens

Parts of exoskeletons showing marked deviation from the
normal structure are rare, as they also seem to be in other
collections of this type fcf. Ross, 1951, p. 134). The only ex-
amples among the present material are the single free cheek of
D. scitulus n.sp. (PI. 13, figs. 6, 7; the emargination apparently
resulting from an injury) and the three pygidia of Calipernurus
insolitus n.gen., n.sp. (PI. 24, figs. 11, 17, 18, 20). As compared
to the usual form (PL 24, fig. 27), these pygidia either lack one
of the three posterior border spines, or have a single spine which
bifurcates distally. This appears to be the result of abnormal
growth rather than injury, and another specimen (PI. 24, fig.
28) shows that even when three border spines are present, they
may be markedly unequal in size. Abnormal specimens of the
pygidium of C. insoUtus n.gen., n.sp. form some 15 per cent of
the total known, as compared to one injured free cheek of I).
scitulus n.sp. in 38 specimens.

Ontogeny

It has been shown elsewhere (Whittington, 1956a) that the
supposed odontopleurid protaspis described by Beecher is a
phacopid. Whittington and Evitt (1954, pp. 28-31) described
the meraspid and later ontogeny of two species of Ceratocephahj,
and made general ()l)servations on odontopleurid ontogeny. This
section supplements the earlier account. Here odontopleurid
protaspides are described for the first time, the three best-known
examples (PI. 3, figs. 1, 2; PI. 4, figs. 1-5: PL 6, figs. 1-5; Text-
figs. 9, 11) being different species of Diacanthaspis. Most re-
markable is the presence of fixigenal spines. The spines on the
dorsal surface are relatively large and arranged in a svmmetri-

WHITTIXGTOX : SILICIFIED ODOXTOPLEURIDAE 179

cal pattern (numbered and lettered here as in Text-fig. 1). there
being a median occipital, three pairs on the glabella (2-4). and
a pair on the anterior border just in front of 4 and slightly
farther apart. On the glabella the occipital furrow is shallow,
and in Z). coopcri and I), lepidus n.sp. the Ijasal glabellar lobe
is faintly delineated. The eye lobe is situated near the antero-
lateral border of the cheek, the tAvo bi-anches of the suture aligned
and isolating a narrow (tr.) free cheek, which bears a row of
border spines, the posterior the longer. The posterior branch of
the suture cuts the margin immediately outside the fixigenal
spine. A rostrum is present and the hypostome relatively large.
The succeeding Stage 0 of each of these three metaprotaspides
is known (Text-figs. 9C, IIB; PL 4, fig. 6) as is Stage 0 of
Apianurus harhatus n.gen., n.sp. (Text-fig. 22A). Fixigenal
spines persist, but long, stout librigenal spines are also present
in species of both genera. Some additional spines (including
axial 5 in A. harhatns) appear on the dorsal surface.

In the next largest specimens known of Diacanthaspis (Text-
figs. 9D, IIC) and Apianurus (Text-fig. 22B), specimens which
must represent the exoskeleton at Stage 1, the' stout fixigenal
spine has disappeared. There is no trace of where it has been.
unless it is represented in Apianurus, greatly reduced, by one
of the tiny spines between B and C on the extremity of the
posterior border (Text-fig. 22B). However, there are tiny
spines in this position in the smaller cranidia between B and the
fixigenal (Text-fig. 22A), and most probably these are the spines
seen at the extremity of the posterior border in the next stage.
One or two specimens (PL 19, figs. 4, 6) of Apianurus have the
fixigenal on one side and not on the other. It is hard to know
whether this is the result of accidental breakage (and the fragil-
ity of the specimen at the base of the spine makes this likely),
or whether it is showing the abrupt loss of the fixigenal on one
side before the other. The course of the posterior branch of the
suture is a curve across the ventral side of the base of the
fixigenal spine. After the fixigenal is lost, the curve is similar,
but runs across where the dorsal side of the base Avould have been.
Thus little modification of the course of the posterior branch of
the suture, and of free cheek outline, is caused by the loss.

In Diacanthaspis, spines A], B, C, D, and the fixigenal are

180 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

present on the postero-lateral part of the fixed cheek of the
protaspis and Stap:e 0 cranidia. In the next largest size the
fixigenal is gone and A, , B, C, and D remain. The free cheek
shows no sudden modification as the fixigenal is lost — the eye
lobe moves back and the course of the sutural branch is altered
slightly accordingly (Text-fig. 9D).

Thus the cephalon of Diacanthaspis, and of Apianuru!^, is, at
about Stage 1, "opisthoparian." Evidently the change takes
place rapidly, for there is little difference in size between cranidia
with and without the fixigenal spine (e.g. PI. 6, figs. 6, 8; PI. 19,
figs. 2, 4, 9). If the loss is by reduction, as it is in such genera
as Sphaerexochus (Whittington and Evitt, 1954, Pis. 17 and 32)
or Fle.ricalymene (Whittington, 1941, PI. 72), then this reduc-
tion is a far less gradual process in odontoj^leurids. No transi-
tional specimens showing such a gradual reduction have been
found, however, and the evidence points to the process being
one of abrupt loss taking place, perhaps between moults, in the
transition from Stage 0 to Stage 1.

The Stage 0 exoskeletons of Diacanthaspis and Apianvrus.
genera placed iii separate subfamilies, are so similar that we may
reasonably expect other odontopleurid protaspides to be like
those of Diacanthaspis — gently convex; fixigenal spine like the
two on border of protopygidium ; free cheek narrow (tr.), bear-
ing a librigenal spine ; eye lobe far forward on the antero-lateral
slope ; glabella parallel-sided, divided by the occipital furrow, in
front of which thei-e may be small, low, basal lobes, and bearing
a median occipital and, if any axial spines, numbers 2, ;?, 4, and
perhaps 5 ; fixed cheeks bearing some or all of the spines
lettered in Text-figure 1. The Stage 0 cranidium of D. ulrlclii
(Text-fig. 12A) shows how smooth the external surface may be,
compared to that of D. lepidus or D. sccrettts.

The meraspid development of tlie different genera also pro-
ceeds along parallel lines — as has already been observed in
species of Ceratocephala and Diacanthaspis (Whittington and
Evitt, 1954, pp. 28-31). In Diacanthaspis cooperi and D. lepidus
n.sp., small, low, basal lateral glabellar loI)es are present in the
metaprotaspis at the base of the gently inflated fronto-median
glabellar lobe, and the axial furrows are broad and shallow. The
second lateral lobes appear later, after Stage 0. In D. secretus

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 181

the basal glabellar lobes are not apparent until after Stage 0, bnt
this may be because they are obscured by the large spines on the
external surface. In Ceratocephala the fronto-median glabellar
lobe is strongly convex in the earliest known stages (Text-fig.
16). standing high above the broad axial furrows. At the next
known stage (PI. 14, fig. 2) basal glabellar lol)es appear in the
trough of the axial furrows. Thus, as pointed out previously
(Whittington and Evitt, 1954, p. 29), the strongly convex axial
region of the tiny Ceratocephala eranidiuni is homologous with
the fronto-median lobe only of later stages, and the true boun-
daries of the glabella lie somewhere just outside, in the axial
furrow. In the protaspis of Diaca7itliaspis the second segment
(axial 2) is the longest (sag.), and axial 5 unmarked. In early
meraspid stages of Ceratocephala (Whittington and Evitt, 1954,
fig. 16; 2a is labelled "2"), Apianurus (Text-fig. 22B), and
some species of Diacanthaspis (e.g. Text-fig. 9D), axial 2 is
conspicuously longer and axial 2a is present. Axial 3 and 4
are progressively shorter, and 5 (which appears in Apianunisi
at Stage 0 but later in some species of Diacanthaspis), is the
shortest. Subsequent growth of the glabella continues this trend,
the second segment expanding and lengthening, the third likewise,
but to a lesser extent, and the fourth and fifth remaining short.
Neither 4 nor 5 disappears completely, and it seems unlikely
that 4 disappears in Leonaspis, as implied by Hupe (1953, p.
102, fig. 62 (3)).

In the odontopleurid protaspis, as in those of, for example,
cheirurids, phacopids, calymenids, and pliomerids, the eye lobe
is situated inside the margin of the cheek, a short way up the
antero-lateral slope. The facial suture is completely developed
and is not marginal, and, in the late metaprotaspis (if not
before), rostral, connective, and hypostomal sutures are present.
The backward migration of the eye during ontogeny occurs in
all odontopleurids, and is most marked in odontopleurinids and
apianurinids.

A characteristic feature of odontopleurids is the way in which
the antero-lateral extremity of the occipital ring merges into
the inner corner of the cheek, and thus the posterior border is
widest (exs.) di.stally, narrowing proximally to disappear at the
axial furrow. At Stage 0 in Diacanthaspis cooperi (Text-fig.

182 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

9C), and in slightly later stages of otliei- species of Diacanthmpis
(Text-fig. lie, IID). in Geratocephala (Text-fig. 16), and in
Apianurus (Text-fig. 22C), there is a spine (A]) in the inner
corner of the free cheek where it passes into the occipital ring,
and the posterior border widens (exs.) rapidly outwards. In
later stages inflation of the inner corner of the fixed cheek ac-
companies the expansion of the posterior part of the glabella
and backward movement of the eye lobe.

The border spines of the protopygidium and early transitory
pygidium of Diacanthaspi'^, and of the early transitory pygidium
of Apianurus, become respectively posterior pleural, and pleural,
spines of the thorax. Anterior pleural spines are tiny and appear
low down on the edge of later transitory pygidia of Diacanthas-
pis (PI. 3, fig. 18), but never appear in Apianurus. The earliest
known transitory pygidia of Ceratocephala (Whittington and
Evitt, 1954, PI. 8, fig. 4; PI. 26, fig. 1) have the two types of
border spines, those corresponding to the posterior pleural being
relatively larger, as they are in the thorax.

Attention has been drawn to the likenesses between early
developmental stages of different odontopleurid genera, i.e. to
the generalized family characters appearing first in ontogeny.
Although the later developmental stages follow a somewhat
pai'allel course, it is also true that out of the general characters
the more special characters are developed, so that during onto-
geny each species diverges more and more from all others as it
takes on its distinctive holaspid form. No better example of von
Baer's "laws" could be wished for (cf. de Beer, 1951, pp. 2-3).
At Stage 0 we can clearly recognize not only family but generic
characters, and the ontogenies of the species of Diacanthaspis
suggest that specific characters are recognizable at this stage, and
l)robably still earlier, in the metaprotaspis. Thus, in the earliest
known ontogenetic stages, while general characters are most
obvious, special characters are expressed in the details of
morphology. Could we go farther back in ontogeny, into the
anaprotaspid stages and embryonic stages, we might find those
points at which only higher systematic categories — superfamily,
order, or even class — are recognizable.

Locus of Segmental Divisions
The occurrence of axial spine pairs, and of transverse, syni-

WHITTINQTOK : SILICIFIED ODONTOPLEURIDAE 183

metrical spine-rows in tlie protaspides and Stag'e 0 specimens of
odontopleurids is emphasized by the new evidence given here
(Text-figs. 9, 11, 22), and is manifestly a fundamental charac-
ter. On the cephalon (Text-fig. 1), spines Aj, B, and C form a
row with axial spines 1, Ao and D with axial 2, A^ with axial
3, Er and PI with axial 5. These rows seem without much doubt
to reflect segmentation, and the fixigenal spine, with B at its base,
belongs to the posterior row. Each of the two or three border
spines of the protopygidium has an upwardly-directed spine at
its base (PI. 3, figs. 1, 3; PI. 4, figs. 1, 4; PI. 6, figs. 1, 5, 10;
PI. 19, fig. 5), and the homology with the fixigenal and spine B
is clear. These border spines of the protopygidium will become
successive posterior pleural spines of the thorax. Thus there
seems no reason to doubt that the joint between cephalon and
protopygidium corresponds with a primary segmental division
in the animal, as do the joints between successive thoracic seg-
ments, and that the fixigenal and posterior pleural spines are
homologous.

The holaspid exoskeleton affords evidence in support of this
interpretation, and three examples may be offered : —

1) Primaspis ascitns (PI. 1, figs. 1, 11, 14). Posterior, most
convex part of occipital ring, curving forward distally to occi-
pital lobe, parallels structure and ornament of thoracic axial
ring; posterior border is narrow (exs.) and smooth proximally,
Avidened and inflated distally and tuberculate, as is the posterior
l)and of the thoracic pleura ; posterior border furrow is like
pleural furrow ; anterior band of pleura must correspond to most
posterior part of cheek.

2) Diacanthaspis ulrichi n.sp. (PI. 8, figs. 1, 12, 13, 21, 24)
affords corresponding evidence, the spines on occipital and
thoracic ring, and on posterior border and posterior band, being
similar in size and arrangement.

3) Calipernurus insolitus n.gen., n.sp. (PI. 23, tigs. 3, 5, 7;
PI. 24, fig. 16). The highest, median part of the posterior border
and lateral part of the occipital ring form a conspicuous smooth
band, which seems to pass behind the paired occipital spines,
and distally ends a short distance inside the suture. On the
thoracic segment the smooth band runs along the highest part of
the axial ring and pleural ridge.

184 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Many other examples could be used, and it is notable that, in
some, one could readih' assume that the librigenal and posterior
pleural spines Avere corresponding structures. This does not
appear likely, however, in the apianurinids, and early onto-
genetic stages of odontopleurinids seem to preclude this view
and to show that the librigenal spine belongs to some segment
in front of the occipital.

Study of the pygidium of P. ascitus n.sp. (PI. 1, fig. 9) and
of D. ulrichi n.sp. (PI. 8, fig. 6) shows how like the posterior
band and posterior pleural spine is the pleural ridge and major
border spine. The Stage 8 transitory pygidium of D. cooperi (PI.
3, figs. 17, 18) brings out strongly this likeness, and I have little
doubt but that pleural ridge and major border spine of pygidium
are homologous with posterior pleural band and spine. This
means that the anterior 2 or 3 pairs of border spines are homo-
logous with the anterior pleural spine, and their number is
added to during ontogeny. For example, in Apianuriis harhatus
n.gen., n.sp., a small holaspid pygidium (PI. 18, fig. 18) has one
border spine in front of the ridge bearing the major spine, a
larger holaspid has two (PI. 18. fig. 12).

The similarity in structure between occipital ring, posterior
border, and adjacent part of cheek and axial ring and pleura
of thorax in odontopleurids has been noted before (Warburg,
1933). These trilobites afford no support for the view of
Stormer (1942, p. 130) that "the transverse joint between the
pleurae of the thoracic segments, or between these and the
cephalon or pygidium, are secondary formations crossing the
primary segments." Stormer based his argument mainly on an
interpretation of Holmia and Faradoxides, and believed that in
these trilobites each thoracic pleural spine belonged to the axial
part of the segment in front of it (Stormer, 1942, text-figs. 14.
15a, 15c). In iSelenopeltis (his text-fig. 15b) he represented the
possible secondary segmentation as difterent in character, for
each thoracic pleural spine belonged to the segment it was
attached to, the "secondary segmentation" only affecting part of
the pleura. There seems no evidence in odontopleurids for such
a secondary joint, but in any genus other than SelenopeJfis.
which has the peculiar forward bend of the posterior pleural
ridge, it would be hard to detect. Hupe (1953, pp. 118-119) does
not support Stermer's suggestion.

WHITTINGTON: SILICIFIED ODONTOPLEURIDAE 185

Mode of Life

The spininess of the exoskeleton and the Avide and seemingly
rapid geographical distribution (many genera make their ap-
pearance almost simultaneously in two continents — Prantl and
Pfibyl, 1949, pp. 209-212) of many genera have led most students
to regard odontopleurids as floating in habit (e.g. Prantl and
Pfibyl, 1949, pp. 132, 209; Whittington and Evitt, 1954, pp.
32-33, and references). The present study provides additional
evidence for the view elsewhere expressed (Whittington, 1956b)
that the convex eephalon will rest on a level surface on the
anterior and lateral margins of the eephalon, the occipital ring
the highest region, the posterior margin approximately vertical.
In this position the thorax and pygidium, when stretched out
horizontally, lie a short distance above the level surface. The
outer surface of the hypostome lies roughly parallel to this sur-
face and must be close to it. Antennular notches are commonly
developed in the anterior cephalic margin, presumably s^ that
the antennae may protrude forward while the animal rests on
the cephalic margin. The eephalon is "propped" in this charac-
teristic attitude by the development of either of two structures,
or a combination of them. These are: (1) expansion of the
lateral areas of the cheek, as in Ceraiocephala (Whittington
and Evitt, 1954, fig. 14), Selenopeltis and Dicranurus (Text-fig.
18) ; (2) development of a fringe of spines on the lateral
cephalic border, increasing in length posteriorly, as in Acidaspis,
Dudleyaspis, Miraspis, and Diacanthaspis (Text-figs. 10, 13-15).
A combination of the two structures is seen in Whittingtonia
(Text-fig. 17), and in Apianurus (Text-fig. 19) the basal parts
of the librigenal spines act as ' ' props. ' ' The possibility of resting
in this position being universal among odontopleurids, and its
attainment by the development of a variety of structures, attests
that it was of fundamental importance, and that odontopleurids
commonly rested in this attitude. It may well have been a
feeding position on the sea bottom, movements of the appendages
causing food-carrying currents to move forward under the head
toward the mouth. This position — resting on the cephalic
margins and with the thorax and pygidium stretched out above
the bottom — is characteristic of diverse trilobites, such as
Harpidae (Whittington, 1950, pp. 25-26), Bathyuridae (Whit-

186 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

tington, 1953, p. 651, Text-fig. 1), and some illaenids and
cheirurids. It seems, therefore, that while the spinose exoskeleton
probably aided the animal in floating, and may have had a pro-
tective function, other features point to the odontopleurid as
having rested on the sea bottom for at least part of the time. We
need not necessarily regard odontopleurids as entirely pelagic,
then, but rather as drifting (and feebly swimming?) at some
depth in shallow seas, and resting at times on the bottom.
Odontopleurids occur in limestones ranging from light-coloured,
coarsely crystalline to dark, aphanitic, muddy and silty, and in
a variety of mudstones, shales, siltstones and fine sandstones.
They seem to have been preserved in a wide variety of shallow
water marine environments, and particular genera, and some-
times species, occur in a variety of rock types. This mode of
occurrence might be used as an argument for their being mainly
or wholly pelagic. The wide geographical distribution, however,
does not necessarily imply a pelagic mode of life, for the early
develtpmental stages were floating and would permit this wide
distribution. It is notable that, in the development of Diacan-
fhaspis cooperi (Text-fig. 9), the lateral cephalic spines develop
rapidly in the earliest meraspid stages, after the fixigenal spines
are lost. It may be at about this stage that the change from
entirely floating to a partially bottom-dwelling mode of life
occurs.

Origin and Evolution

Text-figure ;] summarizes my views on the taxonomy and evolu-
tion of the Odontopleuridae, and reasons for this arrangement
are given in the systematic section. The earliest representatives
of the family appear to be recorded from the Arenig (Canadian)
— Selenopeltis in Shropshire, England (Whittard, 1952, p. 158),
and pygidia of odontopleurinid type from bland, Sweden (Boh-
lin, 1949, pp. 539, 560, 566). A pygidium of the latter type was
described by Opik (1926) as Acidaspis solis, and came from
beds of early Llanvirn age in Estonia. The cranidium described
by Hintze (1953, PI. 19, fig. 16) is of about the same age —
late Canadian or immediatelj^ post-Canadian, and by this time
Selenopeltis is widespread in western Europe and North Africa.
By late Llandeilo time other genera have appeared — Primospis

WHITTIXGTON : SILICIFIED ODOKTOPLEUEIDAE 187

in Bohemia and Ceratocephala and Apianurus n.geu. in the
Lincolnshire limestone of Virginia. Thus in a relatively short
span of time odontopleurids appeared and diversified into the
types here regarded as constituting four subfamilies. These early
genera are at present known from widely separated geographical
areas, and correlation between the rocks containing them, and
thus relative times of appearance, can be only generalized. Thus
one cannot with any confidence single out the earliest stock or
a geographical area in which odontopleurids arose and from
whence they spread out. Prantl and Pfibyl (1949, p. 212)
pointed to the Upper Ordovician of Scotland as the time and
region where the first differentiation of odontopleurids took
place, but this evidently is not so. Comparative morphology
seems to throw no light on origins either, for other Canadian
trilobites seem not to be closely related. The peculiarities of
the odontopleurid protaspis serve to emphasize the lack of
obvious relationship between the family and its contemporaries.
In seeking the ancestors of the Odontopleuridae, ' ' The possibility
that groups hitherto soft-shelled were acquiring the power to
mineralize the exoskeleton cannot be overlooked" (Whittington,
1954, p. 198; cf. Rasetti, 1948, p. 5). Yet as Professor R. Koz-
lowski has pointed out to me (personal communication), the sud-
den appearance of new types is a frequent phenomenon in the
history of this and other animal groups, and alternative explana-
tions can be offered for these appearances.

The initial diversification of the odontopleurids established the
four subfamily types, illustrated in Text-figure 3 by Primaspis,
Ceratocephala, Apianurus n.gen., and Selenopelfis. Though these

Figure 3. Classification, relationships and range in time of the genera of
the Odontopleuridae. Characteristic members of each subfamily are
shown, reproduced from Figures 4, 5, 19A, and 25C. Ceratocephala from
Whittington and Evitt, 1954, figure 13. Range of each genus shown by
solid black bar, bar broken if range uncertain. Broken diagonal lines
suggest relationships between genera. Numbers of genera through time
proportional to width of liar at right. Lower, Middle, and Upper divisions
of the Ordovician correspond with the three Ordovician Series of Twenhofel
et al., 1954.

188

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Odontopleurinae

Apianurinae

o
>

Odontopleuro

I
I

52
a.

c
o
u
o

Q
D

Apianurus

ii'H

WHITTINGTON : SILICIFIED ODONTOPLBUBIDAE

189

Miraspinae

Sele n 0 pelt inae

Number of
Genera

)<U

i^:

C/5

Ce ratoc e p ha I a

I
I

O
O

0}
O
O

■

190 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

four groups have much in common, each has its distinctive
characters, one of the most important being the type of hypo-
stome (Text-fig. 2). There is no comparable diA^ersification after
this, for only two subfamilies survive the Ordovician and no new-
ones arise. In later Middle Ordovician times three (and possibly a
fourth) new genera and one new subgenus arise (making a total
of at least eight in the Middle Ordovician), and two new ones
appear in Upper Ordovician times. During the Silurian period
five new genera arise, and in the Devonian four. In the Middle
and earliest Upper Devonian the nine surviving genera become
extinct, this extinction being almost as rapid as the early diversi-
fication. Text-figure 3 shows the variation in time of number of
genera and reflects the abrupt appearance and extinction of the
group, though not the maximum morphological diversity of
Ordovician time. A striking feature of odontopleurid evolution
is the long range in time of certain genera — Ccratocepliala
endures for some 100 million j'ears, and each of four other
genera, Primaspis, Leonaspis, Acidaspis and 3Iiraspis, for some
50-70 million years (on present estimates of absolute time).
Other genera appear to have a relatively much shorter range.
It is suggested here (Text-fig. 3) that certain of these long-
ranging genera provided a root-stock from which the short-
ranging genera were derived. Ceratocephala affords the best
example, the other miraspinid genera differing from it in one
or more morphological characters (e.g. the big paired occipital
spines of Dicranurus, or the inflated median glabellar lobe of
Whittingtonia) but retaining the main structural plan. Not all
the seemingly derived genera are short-ranging, e.g. MirasjDis.
though this genus may prove to be rather a closely related but
distinct root-stock. Acidaspis and Primaspis-Leonaspis seem to
be the root-stocks of the Odontopleurinae. The other two sub-
families are relatively short-lived, not so diverse, and the Seleno-
peltinae especially seem an aberrant oft'-shoot.

If any general picture is afforded by the pattern of odonto-
pleurid evolution, it seems to be one of persistent main themes
and relatively brief appearances of variations on these themes.
Some of these variations, such as large, paired or single, occipital
spines, major pygidial spines, marked inflation of certain glabel-
lar lobes, appear more than once during the history of the family,
and in different combinations, suggesting a process of shuffling

WHITTTXGTOX : SILICIFIED ODONTOPLET'RIDAE 191

of inorphologrical characters.

The stratigraphical position of each of the species of Diacan-
fhaspis is shown in Text -figure 8. There appear to be three
main strands within the evolving plexus of species, viz. oran-
densis-cooperi, Icpidus-secretus-scitulus, and ulrichi-afl. idriclii.
The species are similar to each other, the distinctions being size
and length of median and/or paired occipital spines, outline of
glabella and degree of inflation of lobes, presence or absence of,
and position of origin of, major pygidial spine, size and distri-
bution of spines on external surface, etc. The illustrations show
that the.se characters are relatively minor distinctions within a
major framework, and that they appear and disappear, forming
(.litterent combinations. Thus the species of Diacanthaspis show
the same evolutionary pattern in a part of the family as in the
whole — the presence of several main themes, occurrence of
minor variations on these themes,* and shuffling and recombina-
tions of characters within the plexus. No other genus is repre-
sented in the collections by as many species as is Diacanthaspis.
In other genera represented by more than one species no particu-
lar trends of morphological change are evident.

Turning to consider the rate of morphological change among
the silieified trilobites, it is seen to be variable in different lines
during lower Edinburg to Oranda time — Apianunts harhafus
n.gen., n.sp., shows no change, and there is little in the Dia-
cantliaspis idricin — aff. nlrichi line, while the D. lepidus-secre-
tus-scitulns line shows considerably more. Cerafocephala is
present in the pre-Edinburg Lincolnshire limestone (C triacan-
theis Whittington and Evitt, 1954), a second species C. laciniata
is common in the lower Edinburg, and C. rarispiiia n.sp. is rare
in the Oranda. Thus there is change in this presumed line of
descent, though no particular trend is evinced. An apparent
oft'-shoot from this line in the lower Edinburg is C. {Ceratocepha-
lina) iridens n.subgen., n.sp. h\ Oranda to lower Martinsburg
time there is slight morphological change in the Diacanthaspis
orandensis-coopen line.

The ontogeny of certain fossil animals has been interpreted as
showing recapitulation of ancestral adult morphology, but no
examples of this process have been adduced from trilobites.
Following Stiibblefield's .suggestion (1936), Stflrmer (1942),

192 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

and most vigorously Hupe (1954), have urged that paedogenesis
or neotony is an important process in the evolution of trilobites.
In the case of Diacanthaspis we know the development in detail
of the oldest and youngest species (Text-figs. 9, 11; Whittington,
1941, text-figs. 2-6), and something of that of other species
(Text-fig. 12). However, not more than one complete ontoge-
netic series is known from any one of the three lines of descent
in DiacantJiasph. Thus we do not have sufficient evidence to
reveal whether or not recapitulation or paedogenesis are im-
portant processes in the evolution, but there are no indications
of either being operative. What is clear is that the earliest
developmental stages of the various species are remarkably
similar, though even at this stage specific differences may be
observed. As development proceeds through the meraspid stages
the peculiar specific characters of each (long, paired, occipital
spines, stout, median, occipital spine, shape of eye lobe, glabellar
lobation, etc.) become increasingly evident. Thus the develop-
ment of species of Diacanthaspis affords an excellent example of
von Baer's laws (De Beer, 1951, p. 8). Comparison of the ontog-
eny of Diacayiihaspis with that of Apianurus (Text-fig. 22) and
of CeraiocepJiaJa (Text-fig. 16; Whittington and Evitt, 1954,
text-fig. 16) shows that the early meraspid stages are remarkably
alike, though generic and subfamily differences are quite evi-
dent. As development proceeds the divergence between genera
increases. The protaspides of the Middle Ordovician odonto-
pleurids suggest that, if paedogenesis were an important evolu-
tionary process, we might expect, for example, holaspid Devonian
odontopleurids to bear fixigenal spines, or to have only basal
lateral glabellar lobes, or extremely short, paired, occipital spines.
These expectations are not fulfilled. A clearer understanding of
the evolutionary processes may be possible when ontogenetic
series of later Ordovician, Silurian, or Devonian odontopleurids
are known, but present evidence seems not to favour recapitula-
tion or paedogenesis, and seems to suggest that the importance
of this latter process in the evolution of trilobites in general may
have been overestimated by Ilupe (1954) and others.

WHITTINGTON : SILICIFIED ODONTOPLBITBIDAE 193

PART II: SYSTEMATIC DESCRIPTIONS

Family ODONTOPLEURIDAE BURMEISTER, 1843

Diagnosis: Cephalon convex, so that postero-mediau rpfjion
stands high above antero-lateral marjiin. Glabella with maximum
width generally at occipital ring, sub-parallel sided or tapering
forward, occipital ring may be elongated (sag.) and prominent,
lateral lobes may be present, median and/or paired spines or
tubercles characteristic; 2-3 pairs of lateral glabellar lobes.
Cheek convex, inner posterior corner merges with antero-lateral
corner of occipital ring; eye lobe prominent, situated centrally
on cheek or inside and behind this point, eye ridge present,
anterior branch suture runs forward and inward, posterior
branch outward and backward to cross posterior margin, sutural
ridges characteristic. Librigenal spine usually stout, broad base
merging with postero-lateral borders; commonly row of shorter
spines arising from oiTter edge of border of free cheek, y>vo-
gressively shorter anteriorly ; antennular notch in border of free
cheek adjacent to anterior branch of suture. Rostrum short
(sag. and exs.) and wide (tr.L llypostome of width greater
than, or equal to, length, postero-lateral margins rounded,
convex middle body, faint middle furrows running backward
and slightly inward from depression at antero-lateral corner
of middle body, lateral and posterior border of moderate and
similar width, small anterior Aving, no wing process, posterior
wing extremely small.

Thorax of 8-10 segments, convex axial ring and long articulat-
ing half -ring; pleura horizontal, gently or moderately convex,
undivided or divided into two bauds, long pleural spine on
posterior band, directed outward and progressively more strongly
backward posteriorly, short pleural spine or spines on anterior
band, directed outward and downward. Either or both pleural
spines maj* be missing from the tirst (and sometimes seccnid )
segment because of the size of the facet. Rygidiuin short, sub-
triangular in outline, 2 axial rings, may be faint third ring ; row
of horizontal border spines, one pair often larger than remainder,
may be upwardly directed, and with base connected by low
pleural ridge across pleural region to first axial ring; or major
spine may arise from pleural region.

External surface of exoskeleton rarely smooth, generally with

194 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

thorn-like spines or tubercles, granules between them, arrange-
ment of larger spines or tubercles may be symmetrical about mid-
line. Scattered openiiitis occur near tips of large spines (libri-
genal, posterior pleural, major pygidial), and sometimes also at
tips of thorn-like spines or tubercles. Latter may, however, be
closed at tip, or may exhibit several tiny openings. Four tiny
depressions, arranged at corners of square, maj^ occur at summit
of median occipital tubercle.

Doublure narrow, curled under. Appendifers not developed.

Geological Range: Late Lower Ordovician (late Canadian or
Arenig) to early Upper Devonian.

Discussion of Systematics of Odontopleuridae

Since Prantl and Pfibyl published (1949j their new arrange-
ment of the Odontopleuridae, it has been followed, with some
emendation, by Erben (1952a, 1952b), and adopted with reserva-
tions by Hupe (1953). Whittington and Evitt (1954, pp. 52-53)
voiced some criticisms, and these and others are explained in
detail here. Study of the silicified material, and of type speci-
mens of most genera (Whittington, 1956b), has provided a
wealth of new data, and my taxonomic scheme is summarized
in Text-fig. 3. This adds a new group, but retains the three main
groups of Prantl and Pfibyl, assigning to them subfamily rather
than familv rank. 1 consider the homogeneitv of the odonto-
pleurids, as well as the relatively small size of the group, sug-
gestive rather of family than superfamily rank (cf. Hupe, 1953,
p. 230). The similarity of early larval stages of different genera
may be interpreted as indicating a monophyletic origin for the
family and not "at least diphyletic" as maintained by Prantl
and Pfibyl (1949, p. 131 ). in subdividing odontopleurids those
authors laid emphasis on the form of the cephalon, and particu-
larly the hypostome, the '■different origin of the genal spines,"
and the direction of iIk^ i)osterior pleural thoracic spines. With
some of these 1 agree, while others seem unimportant. 1 have
based the .subfamilies here used principally on the following : —

a) Form of the hypo.stomes, which, while all are of a type
peculiar to the family, seem to fall into four sub-types corres-
ponding with the subfamilies (Text-fig. 2).

b) Form and inflation of the glabellar lobes and occipital ring.
The characters of this region are of prime impni-tance in sub-

WHITTIXQTO>r : SILICIFIED ODOXTOPLEt'RIDAE 195

family divisions, and their nature is summarized in the diagnoses.

c) Form of the cheek and position of eye lobe, which together
help to determine the course of the dorsal facial suture and the
angle between the two branches.

Such features as type of occipital spine or spines, form of
thoracic segments, direction of posterior pleural spines, sliapc
of pygidium and development of major border spines seem to
me to be of less importance. Whittington and Evitt (1954, p.
32) suggested that the librigenal spine of Ceratoccphnla did not
arise in a fundamentally different way from that of Diacan-
thaspis, and further evidence for this view is given here. Thus
1 do not accept the suggestion that a major systematic division
may be based on the supposedly different origin of this spine in
different genera (Prantl and Pi-ibyl, 19-19, p. 131).

A forward direction of the posterior pleural spines of the first
two or three thoracic segments is observed in some miraspinid
and in apianurinid genera. In these forms the librigenal spines
are well outside the distal parts of the anterior thoracic segments,
and so this direction of the pleural spines is possible. In
odontopleurinid genera the librigenal spine is sometimes broad-
based, and is directed back just outside the distal parts of the
anterior thoracic segments. Hence these segments are facetted,
and may lack pleural spines, or they are short. This difference
in direction of the pleural spines of the anterior thoracic seg-
ments was regarded as important taxonomically by Prantl and
Pfibyl, but it is really dependent on the form of the cephalon.
and in my view of little taxonomie value (cf. Whittington and
Evitt, 1954, p. 52).

Subfamily ODOXTOPLEURIXAE BURMEISTER, 1843
(=Odontopleuridae of Prantl and Pfibyl, 1949)

Diagnosis: Greatest width of glabella at occipital ring, taper-
ing forward slightly or moderately : occipital ring may be
elongated, median or paired tubercles or spines. Eye lobe far
back, variable distance out across cheek, angle between two
branches of suture near eye lobe 90-120". Librigenal spine slim
to stout, row of border spines on cheek Avell-developed. Hypo-
stome (when known) slightly wider than long, middle furrow
commences in front of mid-length and runs inward at a low
angle ; shoulder sharp. Lateral notch shallow. Pleurae of thorax

196 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

with broad (exs.), convex, posterior band, continuous with stout
posterior pleural spine. Pygidium usually with major pair of
border spines, 1-2 pairs small border spines between them.

Geological Range: Middle Ordovician to early Upper Devonian.

Discussion: Prantl and Pribyl (1949, pp. 135, 151, etc.) di-
vided this group into two parts, the first (their Odontopleurinae)
including Odontopleura, Acidaspis, rrimaspis, Radiaspis and
DiacantJiaspis. In this group the anterior branch of the facial
.suture is said to run just outside the eye ridge, so that no sub-
triangular area is enclosed between suture, eye ridge and the
anterior cephalic margin. In the second group (their Acanthalo-
minae) which included Leonaspis and Dudley aspis, a striking
subtriangular area was claimed to be developed. Text-figures 4,
5, 7, 10, 13, 14 suggest that such an area is variously developed
in all these genera, and that it scarcely aft'ords a basis for a sub-
family division. An additional distinction between the two
groups was the "configuration of the occipital ring and the
origin of the occipital spines" (Prantl and Pribyl, 1949, p.
135). 1 do not regard such differences as of this high a value.
An example of our different points of view is oui- treatment of
Dudleyaspis. Prantl and Pribyl (1949, pp. 171-172) regarded
this genus as allied to Leonaspis because of the presence of the
subtriangular area enclosed between the anterior branch of the
suture and the eye ridge. This area seems to me to be no larger
than that in Acidaspis, and the glabella lobation and cephalic
form of these two genera are extremely similar, and I therefore
regard them as allied, and Dudleyaspis as derived from Acid-
aspis.

Cienus Odontopleura Ennnrieh, 1839

Text-figure 4.

Type Species: Odontopleura ovata Emmrich, 1839 {=0. prc-
vosti Barrande, 1846). See Prantl and Pribyl, 1949, p. 135.

Discussion: Only the type and one very similar species, from
the \^'e^lockian of Europe, are known. Odontopleura is dis-
tinguished from Leonaspis (Text-fig. 7) by the larger, more
elevated, lateral glabellar lobes (particularly the basal pair), the
elevation and elongation of the median part of the occipital ring,
which bears paired spines, the faintly defined occipital lobes.

WHITTIXGTON' : SILICIFIED ODOXTOPLEURIDAE

197

the smaller eve lobes, situated farther outward and forward,
and consequent different direction of the anterior branch of the
suture, the slim librigenal spine and more numerous lateral
border spines, the long anterior pleural spine and slimmer, less
backwardly directed posterior pleural spine of the thorax, and
the relatively wider (tr.) pygidium.

In glabellar lobatiou. convexity of cheeks, position of eye
lobes, course of facial sutures, etc. Odontopleura approaches

Figure 4. Odontopleura ovata Emmrich, Middle Silurian, Motol Beds,
ea2, Bohemia. Eeeonstruetion, dorsal view, X IVs- Based on MCZ 4170,
Butovice; MCZ 4164, Lodenice.

Primaspis (Text-fig. 5), but is distinguished by the form of the
occipital ring and stouter occipital spines.

Prantl and Pfibyl (1949) reproduced some of Barrande's
drawings and also new photographs, but all these illustrations
are of specimens flattened in shale, as is that of Hupe (1953,
fig. 134). In my drawing I have endeavoured to portray the
convex cephalon in its true relation to the thorax, using uncom-
pressed specimens from Butovice, Bohemia.

198

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Genus PkimasPIS K. and E. Kichter, 1917
Text-figure 5.

Type Species : Odontopleura primordialis Barrande, 1846. See
l^rantl and Pfibyl, 1949, p. 144.

Diagnosis: Glabella with small third lateral lobes, greatest
width across occipital ring and basal lobes, occipital ring not

Figure 5. Frimaspis priinordialis (Barrande), Middle Ordovician, DraboT
quartzites, d5, Bohemia. A, B. C, cranidiuni, anterior, right lateral, and
dorsal views respectively, X 3. D, pygidium, dorsal view, X 3. Based on
MCZ 4139, Drabov.

greatly inflated or lengthened, bearing median tubercle or paired
spines, small occipital lobes. Eye lobe far back and at about
half width of cheek. Lateral border spines short, librigenal spine
broad at base. Thorax of ten segments, posterior pleural band
inflated at fulcrum and continued into stout posterior pleural
spine ; anterior pleural spine small.

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 199

Discussion: This genus is represented in Bohemia by species
ranging: in age from Upper Llandeilo to Ashgill, and perhaps
Silurian. Primaspis heyserlingi (Barrande, 1846) shows the
characteristic features of the genus including the anterior pleural
spines, which are not lacking (as stated by Prantl and Pfibyl.
1949, p. 14), but is unusual in that the lateral glabellar lobes
are fused, there are no occipital spines, and the external surface
is smooth. In North America, Primaspis is represented by P.
ascitus n.sp. (described below), /'. trentonensis (Hall, 1847)
(Whittington, 1941, p. 502, PI. 74, figs. 31-37; Prantl and
Pribyl, 1949, p. 149), and probably by P. crosotus (Locke, 1843)
from the Eden of the Cincinnati district. All these species have
a median occipital tubercle, but not paired spines. In the size
and convexity of the basal and median lateral glabellar lobes,
position of the eye lobe, course of the facial sutures, form of
the pygidium, etc., such species as P. crosotus approach typical
Leonaspis, and Prantl and Pribyl (1949, p. 146) have remarked
on the Primaspis features displayed by Leonaspis coronata and
L. deflexa. This seems to me to point to the derivation of
Leonaspis from Primaspis, as suggested in Text-figui'e 3. The
close relation of Odontopleura to these genera has been discussed.

Text-figure 3 also suggests that Acidaspis (Text-tig. 13; Whit-
tington, 1956b) is related to Primaspis. Such a species as P.
trentonensis, for example, has a strikingly similar thorax and
pygidium, with the posterior pleural bands inflated at the ful-
crum, and while the cephalon does not display the special fea-
tures of Acidaspis, it resemliles it in having the median lateral
glabellar lobe considerably smaller than the basal, in convexitj"
of cheeks, position of eye lobe, course of facial sutures, etc.

i'Ki]MA«pis ASCITUS Whittiugton, n.sp.
Plates 1 and 2 ; Text-figure 6

Uolotype: USNM 116515 (Plate 1, figures 1, 2, 5, 6) ; locality 10.

Other Material: Paratypes USNM 116516a-h; all figured speci-
mens in USNM.

Geological Horizo)i and Localitij: Lower Martinsburg .shale,
locality 10.
Description : Cephalon crescentic in outline, moderately convex

200 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

longitudinally, strongly convex transversely. Glabella widest
across basal lobes, slightly narrower at occipital ring, narrowing
rapidly- anteriorly, w^here it overhangs the border slightly.
Occipital ring longest (sag.) medially, where it slopes gently
forward to shallow occipital furrow ; laterally, behind basal
glabellar lobe, a low occipital lobe is developed, defined most
clearly on the inner side by a shallow depression. Low median
occipital tubercle. Occipital furroAv deeper laterally. Three
pairs of lateral glabellar lobes, basal (first) largest, sub-oval in
outline, inflated and standing above part of median lobe separat-
ing them; second lobes inflated, defined by basal and second
furrows directed inward and backward at about 45°, the basal
furrow deepest at the inner end ; inner ends of second and basal
furrows joined to each other and to occipital furrow by shallow
longitudinal furrow; anterior (third) lateral lobes with faint
independent convexit}', defined by a .short, shallow furrow run-
ning inward from the antero-lateral corner of glabella. Axial
furrows faint beside glabella, where fixed cheek merges with
antero-lateral corner of occipital ring, clearly defined beside
lateral lobes, dying out anteriorly where eye ridge merges with
glabella. Narrow anterior border separated from glabella by
shallow furrow. Cheek most strongly convex in inner corner
inside eye lobe, which is situated at about half width (tr.) and
close to posterior margin, opposite posterior portion of basal
glabellar lobe. Palpebral lobe short, low, defined on inner side by
deep palpebral furrow which curves around behind eye lobe and
dies out. Anteriorly palpebral lobe and furrow merge with broad
eye ridge and furroAV on inner side. Eye surface \\ith facets
faintly discernible on inner side (PI. 2, fig. 22). Anterior branch
of suture runs forward and then curves inward a short distance
outside eye ridge, crosses border furrow on low sutural ridge
and meets rostral suture at edge of border. Posterior branch
runs in "S" curve back from eye lobe, out across cheek, crosses
border furrow on sutural ridge, and runs over border to doublure
beneath base of librigenal spine. Anterolateral border of cheek
moderately convex, margin curves forward just outside anterior
branch of suture. Row of short border spines, increasing in
length posteriorly, directed down and slightly out from lower
.surface of border. Posterior border narrow (exs.) beside occip-

WHITTIXGTON : SILICIFIED ODONTOPLEURIDAE 201

ital ring, widening and swelling outward to merge with the
broad base of the librigenal spine. Doublure of cephalon of same
width as border. Low projections formed by inwardly projecting
outer part of occipital furrow and the three glabellar furrows,
particularly the deeper, inner part of the basal furrow. Small
depression, with sharply raised anterior edge, in doublure in
front of librigenal spine (PI. 1, fig. 4), is the socket for fulcral
process of anterior segment of thorax. Rostrum unknown, but
evidently a transverse plate on outer edge of anterior border,
perhaps curled under at hypostomal edge. Hypostome slightly
wider than long, gently convex middle body, middle furrow
running backward and inward a short distance from circular
depression at antero-lateral corner, macula faintly defined by
convexity, tips of crescentic posterior lobe slightly inflated. Nar-
row anterior border not defined by furrow ; lateral border with
small, triangular anterior wing directed dorsally, small lateral
notch and sharp, swollen shoulder projecting outward; postero-
lateral border widest, gently convex, small median projection in
faint, wide median notch. Doublure extending between shoul-
ders, of same width as borders, posterior wing a small, twisted
projection (PI. 1, fig. 18), inner edge of doublure curled in in
medial portion. External surface of exoskeleton tuberculate, ex-
cept in deeper parts of furrows, anterior border (which is gran-
ulate), inner part of posterior border, median area of posterior
lobe of middle body of hypostome and adjacent maculae, and
doublure (except median part of hypostomal doublure). Tuber-
cles vary in size, larger ones may display symmetrical arrange-
ment, including 5 or 6 pairs on fronto-median lobe, some on
lateral lobes, and fixed cheek inside eye lobe. Toward tip of genal
spine (PI. 2, fig. 10) tubercles become longer and distally di-
rected, and there appear to be tiny openings at the base on the
distal side of some of them. Other tubercles closed at top. Median
occipital tubercle with 4 tiny depressions (not the openings of
canals through the exoskeleton) arranged in a square, and some-
times a small central pit (PI. 2, figs. 21, 23).

Number of thoracic segments unknown. Convex axial ring
narrowest in mid-line, where it slopes gently forward to deep
articulating furrow. Latter with steep anterior slope, articulat-
ing half -ring long (sag.). Horizontal pleurae divided by slightly

202 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

diagonal pleural furrow into gently convex anterior band and
strongly convex posterior band. Anterior pleural spine blade-
like, shorter on anterior segments, facet at antero-lateral corner
of segment. Posterior pleural band becomes more swollen distally
at fulcrum, and is extended into spine, which is short on anterior
segments, progressively longer and more backwardly directed on
succeeding segments. Ring process and socket well developed,
but axial socket and process inconspicuous. Anterior surface of
pleura is flattened, with a projecting rim along the upper edge,
and this surface fits against the flattened surface of the posterior
flange, the projecting rim fitting against and above the curved
upper edge of the posterior flange (PI. 1, figs. 12, 13). Doublure
extends in to fulcrum, and has an anterior projection, the ful-
cral process, which fits into the notch (with raised anterior
edge to act as a "stop") in posterior part of doublure (PI. 1,
fig. 15). Axis of pygidium with strongly convex first ring, artic-
ulating furrow and half-ring as in thorax, second axial ring
lower and less convex, tip of axis merging into border. Pleural
regions flat, bounded by a low border anteriorly, strongest antero-
laterally, and a broader, more convex postero-lateral border;
crossed by a strong ridge connecting first axial ring to the base
of major border spine. Five smaller spines, longest posteriorly,
project horizontally from postero-lateral border and one major
spine (with a swollen base which merges with proximal part of
fourth horizontal spine) arises from upper, outer part of pos-
terior border and projects upward and backward. Doublure
curled under, as wide as border, at tip of axis most strongly
curled and with a slight inward projection, inner, anterior corner
forming fulcral articulating process. External surface of thorax
and pygidium tuberculate, tubercles densely packed on axial
rings, posterior pleural spines and pygidial border spines, less
closely packed in articulating furrow and on pleurae and pleural
regions, and absent from upper surface of posterior pleural
band, articulating half ring, and inner part of doublure. On
under side of proximal parts of spines tubercles smaller. Con-
spicuously larger tubercles as pair on posterior slope of axial
ring, including two rings of pygidium, and a third pair (or row
of 3) close together at the tip, suggesting a third segment. Large
tubercles also in row along anterior pleural band, and one at

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 203

about half ^Yidth (tr.) of posterior band, on posterior slope.
Towards tips of thoracic posterior pleural and pygidial border
spines tubercles elongated and distally directed, and some open-
ings occur at bases of these tubercles. Elsewhere tubercles
closed, and particularly the larger ones appear to expand slightly
distally to give a bulbous tip.

Discussion: Primasjyis asritus n.sp. is extremely similar to
P. trentonensis (Hall; see Whittington, 1941, pp. 501-502, PI.
74, figs. 31-37) from the Trenton, Sherman Fall formation, of
New York and Ontario. It differs from P. trentonensis in various
minor ways, of which the moi'e obvious are: 1) the greater con-
vexity of the posterior part of the cephalon, so that the posterior
surface of the basal glabellar lobes and fixed cheeks, particularly,
is longer and descends vertically; 2) the inflation at the fulcrum
of the posterior pleural band of the thorax is less; 3) the major
border spines of the pygidium of P. trentonensis are farther
apart, and the base does not merge with the distal part of the
minor border spine inside it, but is separate. Also in P. trento-
nensis there are two minor border spines in front of the major
spine, not three. Tubercles on the external surface are similarly
developed ^nd distributed in the two species, and the median
occipital tubercle of P. tre7itonensis exhibits the four tiny de-
pressions. Paired tubercles can scarcely be distinguished on
the axis of P. trentonensis, except on the posterior edge of the
two pygidial axial rings.

Development: Smallest cranidium of length (sag.) 0.6 mm.
Text-figure 6A (cf. PI. 2, figs. 1, 2) summarizes the characters:
glabella narrow (tr.), almost parallel sided, basal lobes outlined,
faint second lateral glabellar furrows, median occipital tubercle,
large median tubercle (with a suggestion of a subdivision indi-
cating that it will later divide and form axial spines 2a) opposite
basal glabellar lobes, 3 distinct pairs of tubercles in front (axial
spines 2-4) ; palpebral lobes about opposite axial spines 3, strong
eye ridge running forward to antero-lateral corner of glabella,
posterior border well-defined, widening (exs.) outward; tuber-
cles on fixed cheeks and border in sub-symmetrical arrangement,
including Ai, B, and C. Next largest cephalon known (Text-
fig. 6B; PI. 2. figs. 5, 6), length (sag.) 0.85 mm., shows consider-
able change. Both the cephalon as a whole, and individual parts,

204

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

are more convex. Fronto-median glabellar lobe iiov steep-sided,
standing highest, basal and median lateral lobes well defined and
inflated, width of glabella across basal lobes now 35 per cent width
of cranidium, rather than 22 per cent as at the smaller stage.
Eye lobe opposite anterior part of basal glabellar lobe, fixed

^O O W% n\
O).. o i ol '

Figure 6. Prima-^pis uitcitK.s n.sp. A, small i-ranidiuiii, e.xterior view,
ilrawn from orisiiial of Plate 2, fitruie 2. B, cranidium ami right free
ilieek, exterior view, drawn from original of Plate 2, figure 6, and free
cheek of appropriate size. X 38. Spines numbered and lettered as in Text-
figure 1.

cheek inflated at inner corner, which merges with occipital ring.
Median occipital tubercle large, low; on posterior margin of
occipital ring prominent pair of tubercles, other pairs of smaller
tubercles in front of these. Five pairs of tubercles on median
glabellar lobe — axial 2a, 2-5 — the most anterior not quite sym-

WHITTINGTON : SILICIPIED ODONTOPLBURIDAE 205

metrical, paired tubercles on lateral lobes, larger tubercles on
fixed cheeks and border sub-symmetrically situated. A larger
eranidium of length (sag.) 3.2 mm. (PI. 2, figs. 11, 12) shows
that these trends of change have continued — convexitv of lat-
I'ral glabellar lobes and inner part of fixed cheek haA'e increased,
and fixed cheek and basal glabellar lobes are level with median
glabellar lobe in transverse line. Width across basal lobes now
-14 per cent width of eranidium. Eye lobe farther back, paired
tubercles, including four on median glabellar lobe, may still be
distinguished. In this size of eranidium, third lateral glabellar
lobes and occipital lobes are distinct, as are 4 pits in median
occipital tubercle. Further increase in size shows slight change
continuing, so that width across basal glabellar lobes may become
almost 50 per cent of that of eranidium ; basal glabellar lobes
are inflated sufficiently to stand above part of median lobe be-
tween them (PI. 1, fig. 5) ; posterior slope of fixed cheek becomes
steeper. The conspicuous larger tubercles of earlier stages lose
their prominence and cannot be readily picked out in larger
specimens. The smallest hypostome (PL 2, figs. 8, 4) differs
little from larger ones.

Only transitory pygidium known (PI. 2, fig. 9) is probably
stage holaspid-1, since the portion behind the first segment is
like the true pygidium, lacking only the first two lateral border
spines. The axis is convex, each ring with a pair of prominent
tubercles on the posterior edge — recalling the pair on the
posterior margin of the occipital ring. Small true pygidia are
like larger ones, but may liave four border spines in front of
the major spines, the anterior extremely short. In larger speci-
mens this small anterior border spine is reduced, presumably
with enlargement of the facet, and disappears.

Genus LeoXASPIS R. and E. liiehter, 1917
Text-figure 7.

K. and E. Richter (1952) regard as undesirable the revival
of Conrad's (1840) name Acanthaloma by Prantl and Pribyl
(1949, p. 151). I am in agreement with them, and have sub-
mitted a proposal for the suppression of this name to the Inter-
national Commission on Zoological N'omenclature.

206

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Tj'pe Species: Odontopleiiraleonhardil^arvande, 1846.

Diagnosis: Glabella as AAide across large basal glabellar lobes
as across occipital ring, two pairs glabellar lobes, the anterior

Figure 7. Leonaspis new species, Lower Devonian, Haragan shale,
Arbuckle Mta., Oklahoma. A, complete exoskeleton, dorsal view. B, eephalon,
anterior view. C, hypostome, exterior view. X 2. Drawn from originals of
Whittington 1956b, Plate 57, figures 10-16.

WHITTIXGTOX : SILICIPIED ODOXTOPLEURIDAE 207

smaller than the basal, occipital ring not greatly lengthened or
inflated, median tubercle or short median spine. Eye lobe oppo-
site posterior part of basal glabellar lobe or occipital furrow,
large to medium size ; two branches of suture almost at right
angles adjacent to eye lobe, cross border furrows on sutural
ridges. Librigenal spines broad-based. Thorax of 8-10 segments.
pleurae with anterior pleural spines short, may be bifid distally,
posterior band continuous with stout posterior pleural spine;
anterior two or three segments may be facetted and pleural
spines reduced. Pygidium usually with prominent pair major
border spines, one or two pairs small border spines between them.

Geological Range: Lower Silurian to Middle Devonian.

Discussion: Elsewhere (Whittington, 1956b) I have described
two American species of Leonaspis, and one is illustrated here
(Text-fig. 7). Both are very like the type species. I have also
discussed the genus, and the unsatisfactory nature of the basis
for the two subgenera of Leonaspis recently proposed by Prantl
and Pfibyl.

Genus DiACAXTHASPIS Whittington, 1941

Type Species: Diacanthaspis cooperi Whittington, 1941.

Diagnosis: Glabella widest (tr.) across occipital ring and basal
lobes, length (sag.) greater than maximum width; two pairs of
lateral lobes, of which the anterior pair is the smaller, sometimes
faint, tiny third lateral lobes; occipital ring with paired and
median spines, of which one median and/or one pair may be
conspicuously longer and thicker than the remainder; small
lateral occipital lobes may be present. Eye lobe situated well
inwards opposite basal glabellar lobe, sutural and eye ridges
distinct. Row of spines on anterolateral cephalic border. Libri-
genal spine may be hooked at tip. Plypostome shield-shaped,
width (tr.) across anterior wings and shoulders about equal;
may be circular hole through doublure of shoulder. Number of
thoracic segments unknown; pleurae with narrow anterior and
broad posterior bands, latter not inflated at fulcrum. Pygidium

208

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

t/1

"5
re

iNJ 9unasNiiyviAi ubmoi

UIJ VQNVyO

WHITTINQTON : SILICIFIED ODONTOPLEURIDAE

209

Q,
to

z
<

-4-'

1— 1

«f-i

fcc

<»

-M

•t-t

g3
S

iNj ounaNias ubmoi

.1—

/^\

s,*-

.^H

r^^

.n;

■4-*

t|^

■^

c;i

■*>

r/1

«>

• rH

o
O

r-^

^ "3 -i^
.9 X

210 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

of three segments, two axial rings distinct. Border with 6 or 7
pairs of horizontally directed spines, of which the 4th may be
elongated; upwardly and backwardly directed major spine may
be present on upper surface of border postero-laterally ; major
spines linked b}' ridge to first axial ring.

External surface of exoskeleton with median and symmetri-
cally situated thorn-like spines on axial and pleural regions,
granulation between these spines.

Geological Range: Middle Ordovician.

Discussion: The diagnosis has been emended to embrace the
type and the new species shown in Text-figure 8. A recently
described Scottish species (Tripp, 1954, pp. 663-664, PI. 1, figs.
24-29) also lacks large paired occipital spines and major spines
on the pygidial border, but appears to belong here. It also has
faint third glabellar lobes, and such faint lobes are seen in
D. scitulus n.sp. On the other hand, three Esthonian and Swedish
species placed in this genus by Prantl and Pfibyl (1949, p.
150) are here placed in Apianurus n.gen.. and I consider it
unlikel}' that "Acidaspis" tremenda Barrande, 1852, belongs in
Diacanthaspis, as .suggested by Prantl and Pfibyl (1949, p. 150 ).

The similarities between DiacaniJiaspis and Primaspis have
previously been commented on (Whittington, 1941, p. 502;
Prantl and Pfibyl, 1949, p. 150), and the similarities and con-
trasts are well illustrated by the Virginia species described here.
The third glabellar lobes are clearly developed in Primaspis, and
species of this genus have stout major pygidial spines. The infla-
tion of the posterior pleural band at the fulcrum and of the
outer part of the posterior cephalic border and base of the
librigenal spine, are also chai-acteristic of Primaspis, but not of
Diacanihaspis.

The silicified species, though distinct from each other, have
much in common, and they also fall into groups related in time —
lepidus — secretus — scitulus, orandensis — cooperi, and ulrichi
— aff. ulrichi, as discussed in Part I. These groups give a glimpse
of the evolving plexus of Diacanthaspis, and might be regarded
as subgenera, though this step is not taken here. The basic unity
of all the species is shown by the similarity betAveen the develop-

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 211

mental stages of each, a similarity which is most marked in the
earliest stages. The basic unity is also shown by the presence of
many common characters which are differently developed in dif-
ferent species, e.g. a relatively large median occipital spine is
present in the tiny cranidia of D. secret us and Z>. ulrichi, but in
the former it loses its prominence in later stages, while in D.
ulrichi it is retained to the adult. The development of the paired
occipital spines, present in all the young stages, offers another
such example.

In all the species of Diacanthaspis there is a deep depression
separating the basal glabellar lobe from the outer part of the
occipital ring, the slope down to the depression from the occipital
ring being gentle, that from the glaljellar lobe steep. In D.
orandensis and D. cooperi the forward-sloping antero-lateral part
of the occipital ring develops a low, subcircular rise, bearing one
or two thorn-like spines. This is the occipital lobe, a faintly
developed feature peculiar to these two species.

The spines and granules on the external surface of the different
species are described and illustrated in detail, and have been
discussed in general terms in Part I. It should be emphasized
that only in the best preserved specimens is the tip of the thorn-
like spines seen to be truncated by a disk in which are tiny
pits or openings. Their arrangement is not as regular as that
of the four pits at the tip of the median occipital spine, and it
seems unlikelv that thev are different-sized versions of the
same structure.

DIACANTHASPIS COOPERI Whittiugtoii. 1941

Plate 3; Plate 9, fig. 6; Plate 11, figs. 16-18; Text-figures 8, 9.
Geological Horizon and Localities: Lower Martinsburg forma-
tion, localities 9-12.

Description: The new material enables the following additions
to be made to the original description (Whittington. 1941. pj).
O02-508, PI. 74, figs. 1-30, text-figs. 2-6).

The eye surface is preserved in some specimens (PL 3. fig.
21; PI. 11, fig. 17), the outer surface with the tiny facets faintly
convex, corresponding pits on inner surface deeper. Librigenal
spine ends in a sharp point, not a hook (PL 3, fig. 20).

212 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Slightly larger hypostomes (PI. 3, figs. 7, 9, 10, 13, 14) than
those described previou-sly have a length (sag.) little more than
two-thirds the maximum width across the anterior wings. There
is a distinct lateral notch, and the shoulders are sharply pointed,
the tip directed ventrally. The doublure is widest near the
shoulder, the inner edge extended in a small ventrally-directed
posterior wing. There is a circular hole through the doublure
at the shoulder. Posteriorly the doublure is narrowest. Orna-
ment of granules, but coarse tubercles on tips of crescent -shaped
posterior lobe of middle body, and along posterior margin.

Segments from all parts of thorax now known (PI. 3, figs. 11,
12, 15, 16) but not complete number. The anterior segment
(PI. 3, fig. 11), since it fits behind the cephalon, lacks the
anterior, and has an extremely short po.sterior, pleural spine.
The articulating half ring is, however, larger and more convex
than that of posterior segments. The photographs show the
progressive increase in size of both pleural spines in successive
segments, as well as the changes in direction.

Axis of the pygidium (PI. 3, fig. 19) with two rings and a
pair of spines at the tip, suggestive of a third segment. The
base of the large, upwardly directed spine on the margin of
the pleural region is situated between the bases of the 3rd and
4th pleural spines, and is connected by a low ridge to the first
axial ring.

Photographs at high magnifications (PI. 9, fig. 6j show the
dorsal external surface of the exoskeleton covered with sym-
metrically situated thorn-like spines with granules between.
The granulation is absent only from the posterior edge of the

figure 9. Diacanthnspis coopcri Whittingtoii. A, B, Protaspis, dorsal and
ventral views, drawn from the original of Plate 3, figures 1, 2. In A, the
right free cheek has been drawn in its correct position rather than displaced
as in B and the original. In B, the outline of the jjoorly preserA'ed hypostonie
is indicated by a broken line. C, Stage 0 exoskeleton with only right free
cheek restored in position. Drawn from original of Plate 3, figure 3, and
free cheek of appropriate size. D, small eranidiuni and left free cheek drawn
from originals of Plate 3, figure 8. Approximately X 62. Paired spines
numbered and lettered as in Text-figure 1; Lb is librigenal.

WHITTIXGTON : SILICIFIED ODOXTOPLEl'RIDAE

213

Figure 9

214 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGV

occipital ring, the upper part of the posterior cephalic border,
corresponding edge of the thoracic axial rings and the posterior
pleural bands, the border of the pygidium and the raised ridg(^
on the pleural regions, and the paired occipital, librigenal, and
posterior pleural spines. Distally directed openings may be seen
on the occipital and librigenal spines, and between and on the
granules at the tips of the lateral cephalic and pygiclial border
spines openings were present (PI. 11, fig. 16). The thorn-like
spines, when not obviously broken, appear bluntly rounded at
the tip, with a distinct, tiny central opening surrounded by an
irregular ring of tiny depressions or openings (PI. 9, fig. 6). The
median occipital tubercle is short and bluntly rounded at the
tip, and some specimens show four depressions, arranged in a
square, on the slopes of the tip (PI. 11, fig. 18). These appear
to be depressions, not openings through the exoskeleton, and
resemble those seen in, for example, D. aff. ulrkhi n.sp. and
D. orandensis n.sp. (see below).

Development: Protaspis (PI. 3, figs. 1, 2, 5, 6 ; Text-fig. 9 A,
B) of length (sag.) 0.42 mm., maximum width (tr.) at base of
fixigenal spines 0.51 mm. Shield subcircular in outline, gently
convex, most strongly so anteriorly ; divided by faint, curving
transverse furrow into larger cephalic and smaller pygiclial por-
tion ; doulilure narrowest posteriorly. Glabella outlined by
broad, shallow furrows, gently convex, a median occipital and
three pairs (2, 8 and 4 of Text-fig. 9A) of axial spines; occipital
furrow faint, in front of outer parts pair of extremely
faintly defined, subcircular, basal lateral lobes. Cheeks slope
quite steeply antero-laterally, connected by narrow anterior
border; low eye lobe far forward and outward, almost in line
with axial spines 4. Eye surface not preserved. Free cheek
narrow, the two branches of the suture forming a straight line
on the dorsal side, curving across the doublure as shown in
Text-figure 9B. Posterior border not clearly defined, extended
into long, broad-based fixigenal spine. Short spines or tubercles
arranged symmetrically on cheeks and anterior border include
Ai, A2, A3, B, C, D, Er, two pairs on the anterior border, and
additional small tubercles on the fixed cheek. Five spines on
vertical margin of free cheek, posterior (librigenal) the longest.
Protopygidium with strongly convex axis, especially posteriorly,
bearing two pairs of tubercles. Pleural regions depressed near

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 215

axis, convex toward rolled border ; two stout pairs of border
spines, outer pair with spine at base, inner pair directed back-
ward and inward. Rostrum and hypostome poorly preserved
(PI. 3, fig. 2), general outline as suggested in Text-figure 9B.

Exoskeleton at Stage 0 (PI. 3, figs. 3, 4; Text-fig. 9C) of
length (sag.) 0.57 mm., width (tr.) at base fixigenal spines
0.6 mm., length (sag.) of cephalon 0.34 mm. A dissociated free
cheek of appropriate size has been drawn in position in Text-
figure 9C. Glabella parallel-sided, more convex, defined by deep
axial furrows ; deeper occipital furrow and ring relatively wider,
with large median tubercle and slimmer but longer paired spines ;
low median tubercle in front of axial spines 4. Tiny basal lateral
glabellar lobe faint, almost concealed by increased inflation.
Cheek more convex, eye lobe larger and farther back, course of
cephalic suture similar. Posterior border defined distally, fixi-
genal spines as long as in protaspis. Tubercles (or short spines)
as on protaspis, on posterior border additional spine between
Ai and B. Free cheek with librigenal spine stouter and longer
than fixigenal, five spines on edge of border, decreasing in size
forward. A new row of spines has appeared on the border,
outside and below the other, directed outward and downward,
extending to base of librigenal spine and increasing in length
posteriorly. Transitory pygidium (width (tr.) at anterior mar-
gin 0.5 mm.) with well-defined axis, pleural regions fiat with
rolled border. Three pairs of spines on axis, and three on
borders, inner directed slightly inward. First interpleural fur-
row faint, two spines on pleura, outer at base of border spine,
third small spine on outer side border spine near base. Upright
spine at base second border spine.

Smallest cranidium previously described (Whittington, 1941,
PI. 74, fig. 1; text-fig. 2) slightly larger than those of Stage 0,
and without trace of fixigenal spines. A cephalon of length
(sag.) 0.51 mm. is here restored (Text-fig. 9D ; PL 3, fig. 8) and
is at same stage of development as original of my 1941, Plate
74, fig. 2 and text-figure 3. Basal part of glabella wider, basal
glabellar lobes much larger, axial spines 2a have appeared, as
have axial spines 5 (in place of the median tubercle of Stage 0).
Eye lobe larger and farther back, but most striking is absence
of fixigenal spine, so that librigenal spine is at genal angle.

216 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

There are no specimens showing any transition (e.g. a reduction
of the genal spines) between Stage 0 and this stage.

The smallest transitory pygidium previously described (Whit-
tington, 1941, p. 508; PI. 74, fig. 11) is probably that of Stage
1, while the larger one (op. cit., p. 508; PI. 74, fig. 12, 13)
may belong to a stage between 1 and 8. A new specimen (PL 3,
figs. 17, 18) appears to belong to Stage 8 (assuming that the
holaspid thorax is composed of 10 segments). Anterior segment
defined by interpleural groove, and right posterior pleural spine
curves backward and slightly inward. Proximal portions of the
posterior pleural spines of second segment may be seen, and
inside them there are four pairs of small marginal spines. On
the right side (the left is broken off), just inside the margin, is
a long upwardly and backwardly directed spine, its base in line
with the fourth pair of spines on the axis. The inclination of
this spine is different from that of the posterior pleural spines,
and it appears to be the upwardly directed major pleural spine
so characteristic of the true pygidium. On the posterior pleural
bands of the first two segments are two short spines, upwardly
directed, the outer at the base of the posterior pleural spine, and
forming a curved line between the axial ring and posterior
})leural spine. A spine corresponding to this outer spine is pres-
ent on the pleural portion of the third segment, just in front
of the upwardly directed major pleural spine. The curve be-
tween the axis and the base of the major spine corresponds with
that of the first two segments, and suggests that the major
pleural spine is the modified posterior pleural spine of the first
segment of the true pygidium.

DIACANTHASPIS LEPIDUS WhittiugtOn, ll.sp.

Plates 4, 5 ; Plate 7, figure 15 ; Text-figures 8, 10

Holotype: USN^I 116517 (PI. 5, figs. 1, 4, 7, 9), locality 2.

Other Material: Paratypes USNM 116518 a-c; all figured speci-
mens in USNM.

Geological Horizon and Localities: lower Edinburg limestone,
localities 2, 3, 4.
Description: At the localities given above, two species of

Diacanthaspis occur, D. lepidus n.sp., and D. secretus n.sp., the

WHITTINGTOX : SILICIFIED ODONTOPLELTIIDAE

217

latter described beloAv. Material from these localities, collected
at different times, has been prepared, and the relatii-e frequency
of occurrence of parts of the adult exoskeleton is as follows
(counts of complete or nearly complete parts only) :

Table 1

Xumbers of Exoskeletal Parts of Two Species of
Diacanthaspis at Localities 2-4.

Locality 2

Locality 3

Locality 4

Cooper and

Evitt and

Cooper

Evitt

Whittington

Collection

U. Ii pidu.s h.sp.

Cranidia

112

Free ohe; ks

•48

185

Pygidia

1122

D. secret Its n.sp.

Cranidia

Free cheeks

—

Pygidia

—

The rarity of D. secretns at locality 3, and the fit of the free
cheeks, has led to the associations made here (Text -fig. 8), and
support is afforded particularly by the nature of the ornament.
No specimens are intermediate in morphology between those
here called D. hijidus and D. secretns. and there is little variation
in even minor features of the ornament between individuals of
either species. It might be suggested that these two morphologi-
cal types, occurring together, are sexual dimorphs of one
species, rather than two distinct species. The fact that D.
secretns is rare at locality 3, but occurs in almost equal numbers
to those of D. lepielus at localities 2 and 4, seems to argue against
this suggestion.

Diacanthaspis lepielus is considerably older than the type spe-
cies, D. coo peri, and is distinguished from it at once in lacking the

218

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

WHITTINGTOK : SILICIFIED ODONTOPLEURIDAE 219

stout occipital spines. The cephalon is less convex transversely,
and both the width (tr.) between the palpebral lobes, and of the
occipital ring, are lesser fractions of the cranidial width. Occipi-
tal lobes are absent. Tip of librigenal spine hooked. The pygid-
inm of D. lepidus is relatively shorter than that of D. cooperi,
and has the elongated fourth marginal spines. That of D. cooperi
has a long upwardly-directed spine situated on the border inside
the base of the fourth marginal spine, which latter is not
elongated. One hypostome from locality 3 is referred to D.
lepidus, since this is the common species at this locality. It is of
the same general type as that of D. cooperi (compare PI. 3, figs.
7, 9, 10 with PI. 5, figs. 15-17), but different in detail — e.g.
the outline of the lateral and posterior margins, depth of lateral
notch, projection of shoulder directed laterally rather than
ventrally, greater convexity of middle body, and shorter middle
furrows. The doublure is incomplete postero-laterally, but at the
shoulders it shows no circular opening.

The long thorn-like spines on the external surface are hollow,
and the tips of even the longest and best preserved are not
pointed, but cut off (PI. 7, fig. 15). The end is usually open,
but in some is covered by a plate having several tiny holes or
depressions in it (PI. 4, fig. 22). Openings, directed distally,
occur on the librigenal spines, tips of the cephalic outer border
spines, on the posterior pleural spines of the thorax, and near
the tips of the pygidial border spines. These openings may have
been occupied by sensory hairs. On the cephalon, between the
thorn-like spines, the external surface is granulate, except on
the convex surface of the borders and occipital ring. Similarly,
granulation is absent from the convex surface of the axial rings,
the posterior pleural bands, and the border of the pygidium.

Development : A series of cranidia down to the length (sag.)
U.48 mm. is shown in Plate 4, figures 8-10, 12-14, 16-18. The
smallest (PI. 4, fig. 6) is distinguished from that of a smaller
size of D. secretus n.sp. (PI. 6, fig. 12; Text-fig. IIC) by the
presence of a median, as well as small paired, occipital spines,
faint basal glabellar lobes, a less marked indentation in the
anterior margin, a slightly different arrano-ement of the spines
on the fixed cheek on and inside the palpebral lobe (spines
Ao, A3, D, and Er), and a tiny additional spine at the extremity

220 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

of the posterior border near the posterior edge, between spines
B and C. In cranidia of 0.5-0.66 mm. in length (compare PI. 4.
fig. 7, with PI. 6, figs. 14, 15), some of these differences are more
prononnced — -the smaller indentation in the anterior margin
in D. lepidus, the different arrangement of the spines referred
to on the fixed cheek and the additional spine on the posterior
border. The first two of these differences persist into larger sizes,
and aid in differentiating the species. Another specific character
already evident at this size is the different shape of the palpebral
lobe. In larger cranidia the differences between D. Icpichis and
D. secretiis become more prononnced.

All the tiny cranidia with fixigenal spines (probably Stage 0)
known from localities 3 and 4 are of the type shown in Plate 6,
fignres 6. 7 (cf. Text -fig. IIB), i.e. all the spines are long, with
a thick base and blnntly terminated tip, there is a single occipital
and axial pairs 2, 3, and 4, and the anterior margin is indented.
This tjT)e is so like the cranidial portion of the protaspis from
locality 4 shown in Plate 6, fignres 1-5 that the conclusion that
they form part of a series is inevitable, and this series is traced
into the holaspid D. secrcfus. However, a second pi'otaspis known
from locality 4 (PL 4, figs. 1-5) is clearly odontopleurid, and
differs from that of D. secretus only in the following characters :
(a) the spines on the dorsal surface are low, appearing rather
as high tubercles ; ( b ) the occipital ring bears a median spine
and a suggestion of small, paired spines just behind the median
spine; (c) small basal glabellar lo])es are present; (d) axial
spines 5 are represented by tiny tubercles. The characters (a),
(b), and (c) are those which distinguish the smallest known
cranidium of D. lepidus, the spines on the external surface being
less massive than those of D. secretus. Axial paired spines 5 are
not known in this size of D. lepidus cranidium, but there is a
median axial 5. This second protaspis from locality 4 is thus
tentatively regarded as that of D. lepidus, for comparison be-
tween it and smallest known cranidia of other species of Dia-
canthaspis, and of Ceratocephala, Apianurus n.gen., and Ccdi-
pernurus n.gen., reveals far greater differences.

DiACANTHASPis SECRETUS Whittington, n.sp.
Plate 6 ; Plate 7, figures 1-14 ; Text-figures 8, 11.

Holotype: USNM 116519 (PI. 7, figs. 1, 3, 5, 11), locality 2.

WHITTINQTOX : SILICIFIED ODONTOPLEt'RTDAE 221

Other Material: Paratypes USNM 116520 a-d; all figured speci-
mens in USNM.
Geological Horizon and Localities: lower Edinburg limestone,
localities 2, 3, 4.

Description : The occurrence of this species is discussed under
Z>. Icpidus n.sp. The seemingly minor but persistent differences
that distinguish D. secretus from D. lepidus, summarized in Text-
figure 8, are: 1) the longer, lower palpebral lobe, more strongly
curved course of the anterior branch of the suture, and better-
defined eye-ridge (compare PI. 5, figs. 1, 4, 7, with PL 7, figs.
1, 3, 5) ; 2) the different direction and arrangement of the four
long spines on the fixed cheek between the palpebral and lateral
glabellar lobes (compare PI. 5, fig. 1 with PL 7, fig. 1) ; 3) the
l)ackward curve of the tips of the outer spines on the border of
tlie free cheek, and the lesser number of spines on the upper
surface of this border (compare PL 5, fig. 9 with PL 7, fig. 7) ;
4) the relatively longer pygidium, in which the fourth marginal
spines are not elongated (compare PL 5, fig. 12 with PL 7, fig.
10). A single small ornamental spine, upwardly directed, is
present at the base of the 4th, 5th, and 6th marginal spines;
in D. lepidus two such spines are present on the 3rd, 4th, 5th.
and 6th marginal spines. Other ways in which the exoskeletal
parts of the two species can be discriminated may be seen from a
study of Plates 5 and 7. It will be seen that the minor orna-
mental spines of D. lepidus tend to be longer and sharper, as
well as more numerous in certain regions (compare PL 5, figs.
1, 7, 9 with PL 7, figs. 1, 3). The thorax of neither species is
eompletety known, but isolated segments of D. secretns have
shorter, blunter ornamental spines, and the lateral barbs on the
anterior pleural spine arc fewer (^compare PL 5, figs. 2, 3, lO, 11
with PL 7, figs. 4, 6).

The librigenal, cephalic, thoracic pleural and pygidial boi'der
spines of the exoskeleton show openings (PI. 7, figs. 12-14) like
those described in D. lepidus. The thorn-like spines have the
tips cut off and closed, with tiny openings or pits in the end.
The granulation is distributed as in Z>. lepidus. Intermediate
between the thorn-like spines and granules are high tubercles,
and these apparently are closed at the tip.

Development: Metaprotaspis (PL 6, figs. 1-5; Text-fig. IIA)

222

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

>=*• /

'^o'
^ %

\ '■?

^1 N so o\

Figure 1 I

WHITTINGTON : SILICIFIED ODONTOPLEl^RIDAE 223

subcircular in outline, gently convex ; free cheeks, rostrum and
h3^postome unknown. Cranidial portion divided by shallow,
broad, poorly-defined axial furrows into glabella, with very
gentle convexity, and gently convex fixed cheeks. Glabella not
divided into rings, but with stout median occipital spine and
three pairs in front, a fourth pair on the anterior border, these
four pairs forming tAvo lines diverging slightly forward. Pos-
terior area of cheek with prominent spines on dorsal surface
symmetrically arranged, lettered A,, Ao, A3, B, C, D, in Text-
figure 11 A. At margin, at end of curving row formed by median
occipital and spines A| and B, is posteriorly directed fixigenal
spine. On anterior area of cheek, spines included are those
labelled PI, A3, and Ei- respectively in Text-figure 11 A, and
one on anterior border. x\nterior margin faintly indented at
midline. Only a gentle slope divides cranidium from proto-
P3'gidium. Axis prominent, two pairs of spines, tAvo horizontal
pairs directed back from margin of pleural regions, and one pair
on dorsal surface at base anterior marginal spine.

Stage 0 specimen reconstructed (Text-fig. IIB) from originals
of Plate 6, figures 6, 10, and similar specimens. It is larger in
size (length (sag.) about 0.5 mm., length cranidium 0.32 mm.)
but otherwise exceedingly similar, except for the presence of a
third pair of border spines on the pygidium. Free cheek ex-
tremely narrow (tr.), small eye lobe far forAvard, roAv of five
spines on upper surface of convex border, roAv of seven on outer
surface of border, long librigenal spine Avith hooked tip. Rostrum
and hypostome unknoAvn. Text-figures IIC-F (cf. PL 6, figs. 8, 12,
l-Jr, 15) shoAv larger cranidia and associated free cheeks. Not
until the cephalon is about 0.5 mm. in length (sag.) (Text-fig.
HE, PI. 6, fig. 14) do small paired occipital spines appear, faint

Figure 11. Diacanthaspis secretus n.sp. A, Pi'otaspis, dorsal view, drawn
from the original of Plate (5, figures 1-4. Approximately X .52. B, Stage 0,
dorsal vIbav, left free cheek separated from cranidium, draAvn from originals
of Plate 6, figures 6, 10. Approximately X 32. C, D, E, F, cranidium and
right free cheek of individuals of increasingly larger size, exterior A'iews.
Drawn from originals of Plate 6, figures 8, 9, 12, 14, 15 respectively. Ap
proximately X 32. Paired spines numbered and lettered as in Text-figure 1;
Lb is librigenal.

224 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

basal glabellar lobes, and a median axial spine in front of axial
spines 3. The eye has moved back at this stage, and considerably
farther back at the next stage shown, when the second lateral
glabellar lobes and axial spines 5 appear (Text-fig. IIF; PI. 6,
fig. 15). The indentation in the anterior margin is marked in
small cranidia but becomes much less obvious, though not quite
obliterated, as size increases. Additional paired spines appear
on cranidia of length (sag.) about 1 mm. (PL 6, fig. 16) upwards.
A slightly larger transitory pygidium than Stage 0 (PI. 6, fig.
13) has three pairs of spines on the axis, the border with three
pairs of spines directed horizontally (posterior pleural spines),
progressively shorter backward. At the antero-lateral corner is
the broken base of a tiny anterior pleural spine. On the pleural
region a short spine is situated at the base of the first two
posterior pleural spines, remainder of region granulated, and
showing the first pleural furrow and interpleural groove. A
considerably larger transitory pygidium (PI. 6, fig. 20) shows
4 pairs of axial spines and 4 posterior pleural spines. Short,
blade-like anterior pleural spines are present on the first three
segments (mostly concealed in dorsal view) ; there are two spines
on the posterior pleural band of the first segment, the outer one
longest.

Discussion of Ontogeny of Diacanthaspis

If these ontogenetic series of D. lepidus and D. secret us be
accepted, it is remarkable that the protaspis of the latter (PI. 6,
figs. 1-5) has far the more massive and prominent spines on
the external surface, while the reverse is the case in the holaspis,
where the spines, though massive, are fewer and shorter and thus
less prominent than the long, thin, thorn-like spines of D.
lepidus (compare PI. 7, figs. 3, 15). Comparison of the metapro-
taspis and Stage 0 specimens of D. lepidus and D. secretus with
corresponding specimens of D. cooperi (PI. 3, figs. 1-6; Text-figs.
9A-C) reveals the close similarity — in outline and form, shape
of glabella, presence of fixigenal spines, num])er and arrange-
ment of spines on the external surface, and axial and border
spines of the protopygidium. Distinctive of D. secretus is the
size of the spines on the external surface, while in both the
others these spines are short and rounded. In D. lepidus and

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 225

D. cooperi basal glabellar lobes are present in the metaprotaspis.
Comparison between later developmental stages (PI. 4, figs. 6-10,
12-14, 16-18; Text-figs. 9D, IIC-F; Whittington, 1941, text-figs.
2-6) reveals both the parallelisms and specific divergence. Fixi-
genal spines are lost abruptly in all. In D. secretus lateral glabel-
lar lobes and axial spines 2a and 5 appear at a later stage than
in either of the other species. While paired spines are present
on the occipital ring of the older species, they never become
bigger than the median occipital tubercle, whereas in D. cooperi
their growth is rapid.

DiACANTHASPis ULRiCHi Whittington, n.sp.

Plate 8 ; Text-figures 8, 12.

Holotijpe: USNM 116521 (PI. 8, figs. 1-5), locality 7.
Other Material: Paratypes USNM 116522a-c : all figured speci-
mens in USNM.
Geological Horizon and Localities: lower Edinburg limestone,
localities 3, 4, 7. This species has not been found at locality 2,
and only 2 pygidia are known from locality 3. From locality
4, 8 cranidia and 14 pygidia, together with 26 immature
cranidia, have been recovered. At locality 7 D. ulrichi is not
accompanied by D. lepidus and D. secret us, whereas at locality
4 these latter two species are rather more abundant (see Table
1, p. 217) than is D. ulrichi.

Description: D. ulrichi is readily distinguished from con-
temporary and later species (except D. aff. ulrichi) by: (1) the
convex fronto-median glabellar lobe and narrow (tr.), low lateral
lobes, wbich latter are separated from the fixed cheeks by ex-
tremely faint axial furrows; (2) the straight course of each
branch of the suture adjacent to the eye lobe, which gives a
distinctive outline to both cranidium and free cheek; (3) the
short, thick, bluntly terminated median and paired main occipi-
tal spines; (4) the convex axis and outwardly -sloping pleural
regions of the pygidium; (5) the short spines on the outer edge
of the cheek border and border of the pygidium.

The external surface of the cephalon (PI. 8, figs. 24, 30)
bears, besides the main occipital and outer border spines referred
to, smaller, thick, blunt-tipped spines. Inside the cephalic

226 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

borders the surface between the spines is covered with fine gran-
ules. The blunt spines are symmetrically arranged on the lateral
glabellar lobes and cheeks, but on the fronto-median lobe, while
the number and general disposition is similar in different speci-
mens, it is not exactly the same, nor is the arrangement sym-
metrical (PI. 8, figs. 1, 30). On the posterior edge of the occipital
ring (PI. 8, fig. 24) there is a median and a pair of these spines,
and in some specimens a second pair is situated in front of the
main pair.

Hypostome unknown.

Pleurae of thoracic segments (PI. 8, figs. 12, 13, 17, 21) with
narrow (exs.) anterior and wide (exs.) posterior band, former
bearing short anterior pleural spine, latter bearing larger spine,
progressively longer and more backwardly curved on successive
segments. Axis with 2 pairs spines, 3 on posterior band, area
between smooth ; articulating groove, anterior band, and pos-
terior flange tuberculate. Axis of pygidium with two rings, each
with pair of spines. Six pairs of spines on outer edge of border,
the first very small, the fourth elongated and connected by a low
ridge to the first axial ring. Prominent spine on upper surface
of border at base of elongated fourth spine. One specimen (PI.
8, fig. 9) has only one, rather than two, pairs of spines between
the major border pair.

The distal portions of the main median and paired occipital
spines (PI. 8, figs. 24, 30) display an irregular, hummocky sur-
face at high magnifications. There may have been openings at
the base of the tubercles on the paired spines, and the tip of the
median spine shows the four tiny depressions arranged in a
square. The distal tips of the posterior outer border spines of the
free cheek, and the longer border spines of the pygidium, may
have been like those of the paired occipital spines. Other, shorter,
spines are rounded at the tip and may be closed, and the distal
part of the librigenal spine does not show any openings.

Development: Protaspis unknown. Development of cephalon
from Stage 0 shown in Text-figure 12 A-D (compare PL 8, figs.
10. 14-16). The smallest cranidium has the characteristic
trapezoidal outline, low, parallel-sided glabella divided only
by the occipital furrow, low palpebral lobe far forward, faint
eye ridge, and fixigenal spine. The presence of the latter suggests

WHITTIXGTON : SILICIFIED ODONTOPLEURIDAE

227

that it belongs to Stage 0. The median occipital spine is repre-
sented by a large, low tubercle, and the paired occipital spines,
present in some specimens of this size, are tiny. Remainder of
glabella without paired spines. On fixed cheeks paired spines
Aj, Ao, A3, B, and C have been recognized. Surface of cranidium
between spines covered with fine granules. In the next size of
cranidium found, fixigenal spines are absent, the palpebral lobe
is farther back, and paired spines 2. 3, 4 appear on the glabella

Figure 12. Diacanihaspis ulrichi n.sp. A, B, C, D, crauidia of increasing
size, exterior views, right free c-heek shown in D. Approximately ^ 38.
Drawn from the originals of Plate 8, figures 10, 14, 15, 16, 28 respectively.
Paired spines numbered and lettered as in Text-figure 1.

in front of the occipital furrow. At this stage the cranidium is
quite like that of D. cooperi (AYhittington, 1941, text-fig. 2)
and D. secret us (PI. 6, fig. 12; Text-fig. IID^, but the large
median occipital tubercle and fewer and shorter paired spines
on fixed cheeks and borders are distinctive. The later develop-
ment parallels that of other species, except that beyond a length
(sag.) of about 0.6 mm. the sj'uimetrical arrangement of spines
on the glabplla in front of the occipital ring is lost.

228 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

On Plate 8, figures 25, 26 show a transitory pygidium of an
unknown stage. Axis with three pairs of spines ; two long border
spines are posterior pleural spines of two segments that will be
released into thorax; anterior pleural spine of first is short,
curved. Behind axis, two pairs of tiny border spines. A small
holaspid pygidium is shown in Plate 8, figures 22, 23.

DiACANTHASPTs afi:. ULRiCHi Whittiugton, n.sp.
Plate 9, figures 1-5, 7-9 ; Text-figure 8.

Material: all figured specimens in USNM.

Geological Horizon and Localities: Oranda formation, locality 8.
Description: a few fragments only of this extremely rare
species are known, and deemed insufficient as the basis for a
new specific name. It is like D. ulrichi n.sp., but ditfers (compare
PI. 9, figs. 1, 5, with PI. 8, figs. 1, 6), for example, in the more
convex fronto-median glabellar lobe, the longer median occipital
spine, the deeper axial furrows of the cranidium, particularly
outside the lateral glabellar lobes, and the more triangular
outline of the p^^gidium. The spines and granules on the external
surface of the two species are similar both in appearance and
distribution (compare PL 9. figs. 7-9 with PI. 8, figs. 24, 30). The
smooth tip of the median occipital spine shows clearly the four
depressions arranged in a square. The smooth, blunt tips of the
spines on the glabella and free cheeks show a tiny central
opening.

DiACANTHASPis OBANDENsis Whittiugton, n.sp.
Plate 10 ; PI. 11, figures 1-15, 19, 20 ; Text-figure 8.

Holotype: USNM 116523 (PI. 10, figs. 1, 3, 5-7), locality 8.
Other Material: Paratypes USNM 116524a-f ; all figured speci-
mens in USNM.

Geological Horizon and Locality: Oranda formation, locality 8.
Three species occur at this locality, and approximate numbers
of complete or fairly complete parts of pxoskeletons obtained
are as foUoAvs :

WHITTINGTON" : SILICIFIED ODONTOPLEURIDAE

229

Table 2

Numbers of Exoskeletal Parts of Three Species of
Diacanthaspis at Locality 8.

D. orandcnsin n.sp.

I), scitiilas u.sp.

D. aff. idrichi n.sp.

Cranidia

120

Free cheeks

150

Pygidia

100

The most common hypostome, 130 specimens, is regarded as
belonging to D. orandensls, the less common, 16 specimens, to
D. scitulus.

Description: Diacanthaspis orandensis is exceedingly similar
to D. cooperi, differing only in minor but persistent characters,
of which the more obvious are : (1) relatively longer outer border
spines on free cheek and pygidium (compare PI. 10, figs. 2. 17.
18, with PL 3, fig. 19, and ^Yhittington, 1941, PL 74, fig. 23) ;
(2) more prominent axis of pygidium; (3) thorn-like spines on
external surface tend to be longer and sharper; (4) particularly
axial, but also longitudinal, furrows bounding first and second
lateral glabellar lobes tend to be deeper and lobes to be more
prominent (compare PL 10, figs. 1, 3, 5; PL 11, fig. 19 with PL
9, fig. 6, and Whittington, 1941, PL 74, figs. 24. 25, 29). As might
be expected, the early developmental stages of the two species
are indistinguishable (compare PL 11, fig. 1 with PL 3, fig. 8) ;
only later do the specific differences become clear.

No complete specimens of the librigeual spines are known,
i.e. the extreme tip is always broken, but almost certainly it is
not hooked. The dorsal external surface of D. cooperi has been
described in detail, and most of the remarks apply to D. oran-
densis. Specimens of the latter, however, are better preserved,
and there is no doubt of the presence of distally-directed open-
ings, with raised rims, on the paired occipital, librigenal,
posterior pleural, and pygidial border spines. On the rounded
tip of the median occipital spine of D. cooperi tiny depressions,
arranged in a square, were observed. Similar depressions are

230 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

present on the occipital spine of D. orandensis (PI. 11, fig. 19),
and further, the tips of the thorn-like spines are likewise bluntly
rounded -with a group of tiny depressions, of which the central
may be larger (PI. 11, fig. 19). The depressions are extremely
small, and whether or not they are the openings of canals
through the exoskeleton is uncertain.

DiACANTHASPis sciTULUs "Whittington, n. sp.

Plates 12, 13 ; Text-figure 8.

Holotype: USNM 116525 (PL 12, figs. 1-3), locality 8.
Other Material: Paratypes USNM 116526 a-c ; all figured speci-
mens in USNM.
Geological Horizon and Localities: Oranda formation, locality 8.
Description: As Table 2 (p. 229) shows, this species is much less
common than D. orandensis at locality 8. The general plan of
the exoskeleton, and particularly the outline of the glabella,
absence of prominent paired occipital spines, and arrangement
of thorn-like spines on the external surface of the exoskeleton
(and particularly of the fixed cheek), ally it with D. lepidiis
n.sp. and D. secretus n.sp. rather than with D. orandensis and
D. cooperi. The lack of large median or paired occipital spines,
and the form of pygidium, distinguish it from D. ulrichi. D.
scitulus differs from D. lepidus and D. secretus in that: (1)
the cephalon is more convex transversely between the eye lobes,
the eye ridges and sutural ridges are stronger and more clearly
defined, and the row of border spines on the free cheek extends
back onto the base of the librigenal spine (compare PI. 12, figs.
1-3 with PL 5, figs. 1, 7, 9 and PL 7. figs. 1. 3) ; (2) there are
seven pairs of horizontally directed border spines on the pygid-
ium, approximately equal in length, and a stouter, longer spine
arises from the upper surface of the border and is directed back-
ward and slightly upward (compare PL 12, fig. 14 with PL 5,
fig. 12 and PL 7, fig. 10). In the type of major pygidial spine,
convexity of the cranidium, strength of sutural ridges, and
extension of lateral border spines on to the base of the librigenal
spines, D. scitulus approaches the D. orandensis-cooperi group
(compare PL 12, figs. 1-3, 14 with PL 10, figs. 1, 3, 5, 27). Thus
the species of Diacanthaspis, while they fall into groups, never-

WHTTTINGTON : SILICIFIED ODONTOPLBURIDAE 231

theless betray their close relationship. Large specimens of B.
scitidus show, particnlarly on the inner snrface (PI. 12, fig. 7),
a faintly impressed furrow directed transversely on the anterior
glabellar lobe Avhere it slopes down to the second glabellar fur-
row. This furrow is not well enough developed to be termed a
third glabellar furrow, but may be the incipient stage of such
a furroAv. The hypostome is of Diacanthaspis type, and dis-
tinguished from that of D. orandensis (with which it occurs, but
in lesser numbers) by the more curved outline of the anterior
margin, the less angulate postero-lateral outline, and the nar-
rower (sag.) posterior border (compare PI. 13, fig. 8 with PL
10, fig. 22). The doublure of the shoulder shows the large circu-
lar opening. The one known specimen of the hypostome of D.
lepidus is of similar type, but does not show the opening through
the doublure.

The dorsal external surface of the exoskeleton is shown in
Plate 13, figures 12, 14-17. The symmetrically arranged thorn-
like spines are truncated and closed at the tip, and there are
tiny depressions (which may be openings) in the tip. Granula-
tion extends between the thorn-like spines, but is absent from
the cephalic borders, central part of occipital ring and inner
corner of fixed cheek around large spine, upper surface of lateral
glabellar lobes and eye ridge, axial rings, posterior pleural l:)ands.
borders of pygidium and pleural ridge, and border and librigenal
spines. Tiny tubercles are present particularly near the tips of
the librigenal and border spines. Distally directed openings
with a raised rim also occur on the distal parts of these spines,
including the tip of the librigenal spine (PI. 13, fig. 14). which
is not hooked as in D. lepidus, I), secretus and D. idrichi.

Development : The smallest known cranidium of D. sciiidus
is 0.93 mm. in length (sag.) (PL 13, figs. 3-5), and is very like
cranidia of about the same size of D. lepidus (PL 4, figs. 12-14 j
and D. secretus (PL 6, figs. 16-18), even in the size and position
of the main thorn-like spines. On the fronto-median glabellar
lobe axial spine pairs 2, 3, 4 are present, a median spine only in
the position of pairs 2a and 5. On the fixed cheeks spines A^, Ao,
A3, B, C, and PI may be seen. The shape of the palpebral lobe,
and the strength of the eye ridge and sutural ridges, are espe-
cially distinctive of D. scitulus, while examples of smaller dif-

232 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

ferences are the greater number of spines on the cranidium of
D. lepidus, and the fewer and differently situated spines of
D. secretus. The small free cheeks of D. scitulus (PI. 13, fig. 13)
have the distinctive border spines, as also do the small pygidia
(PI. 12, fig. 16).

Genus AciDASPIS Murchison, 1839

Text-figure 13.

Synonym (objective) : Pseudomonaspis R. and E. Richter, 1917.
Type Species : by monotj^py, Acidaspis brightii Murchison, 1839

(Wliittington, 1956b).

Diagnosis: Two pairs of lateral glabellar lobes well developed,
separated from median lobe by deep longitudinal furrows, in
type species frontal glabellar lobe rounded, projecting well in
front of anterior lateral glabellar lobes. In Ordovician species
small third lateral glabellar lobes are present. Median part of
occipital ring inflated and prolonged backward as thick median
spine, separated from rest of glabella by shallow median part of
occipital furrow; lateral part of occipital ring with low, gently
convex occipital lobe in inner corner, behind deeper outer part
of occipital furrow. Ej^e lobe elevated, situated opposite most
posterior part of basal glabellar lobe ; anterior branch of the
facial suture runs straight forward and inward, diverging from
the course of the eye ridge, crosses border furrow on sutural
ridge and curves over the anterior border. Posterior branch
curves downward and outward across the fixed cheek in front
of the posterior border furrow, and on to the inner side of
sutural ridge, against which this furrow ends. It then curves
over posterior border, inside base of librigenal spine. Convex
anterolateral border with row of stout spines directed down-
ward, longest posteriorly and diminishing forward ; antennal
notch present, anterior border between sutures projects slightly.
Librigenal spines long, curved. Thorax of 10 segments, posterior
pleural band convex and inflated at fulcrum. Pygidium with
7 pairs border spines, 5th the major. External surface tuber-
eulate.

Geological Range: Middle Ordovician to Middle Devonian.

WHITTINGTON : SILICrFIED ODONTOPLEURIDAE

233

Figure 13. Acidaspis brightii Murchison, Wenlock limestone, Middle Silu-
rian, England. A, B, C, cephalon, dorsal, anterior, and right lateral views
respectively, approximately X 3. (After Whittington 1956b. text-figure 1.)

234

BULLETIN : MUSEUM OF COMPARATI\^ ZOOLOGY

Genus DUDLEYASPIS Prantl and Pfibyl 1949
Text-figure 14.

Type Species: Acidaspis quinquespinosa Lake, 1896
(Whittington, 1956b).

Diagnosis: Occipital ring with short median spine and two
further pairs spines on posterior margin; not prolonged back-

Figure 14. Dudleyaspis quinquespinosa (Lake), Wenlock limestone, Mid-
dle Silurian, England. A, B, cephalon, dorsal and right lateral views re-
spectively, approximately X 6. After Whittington 1956b, text-figure 2.

ward and lacking lateral lobes. In front of large basal lateral
lobes glabella narrows rapidly; small third lobes, third lateral
furrows short, transversely directed, frontal glabellar lobe pro-
jecting in front of third lobes. Eye lobe situated opposite mid-

WHITTINGTON : SILICIFIED ODON'TOPLEURIDAE 235

part of basal lateral glabellar lobes, anterior branch suture runs
on sutural ridge straight for^Yard and slightly inward to anterior
border; suture then runs more directly inward and over outer
edge of border to meet rostral suture about where anterior border
projects forward. Sutural ridge also connects eye lobe and
swollen base librigenal spine, and posterior branch suture runs
along inner side of this ridge and over border inside base
librigenal spine. Convex anterolateral cephalic border with row
vertical spines as in Acidaspis; librigenal spines slim, curved;
two pairs spines on posterior border backwardly directed. Thorax
of 10 segments, pygidium with 2 pairs spines between major
pair.

Geological Range: Silurian (mainly Middle).

GTenus RadIASPIS R. and E. Richter, 1917

Type Species: A7^ges radiatus Goldfuss, 1843
(Prantl and Pi-ibyl, 1949, p. 142).

Discussion: The type species has been described by R. and E.
Richter (1917, pp. 468-472, text-figs 9, 10; 1926, pp. 109-110;
1930, text-fig. 2). The form of the cephalon is like that of
Acidaspis — in convexity, lobation of glabella, elongation of
occipital ring, presence of lateral border spines, etc., but is dis-
tinguished by the paired occipital spines. The thorax is of 9
segments, and the pygidium is notable for the absence of major
border spines (there being 8 pairs of equal length) and the
bilobed form of the posterior part of the axis. I regard Badiaspis
as most closely related to Acidaspis, not Odontopleura, contrary
to Prantl and Pfibyl's opinion. The genus is knoAvn from Lower
to earliest Upper Devonian of Germany and Bohemia.

Subfamily MIRASPINAE R. and E. Richter, 1917
(=Ceratocephalidae of Prantl and Pfibyl, 1949 ; of Erben, 1952)
Diagnosis: Glabella wide; occipital ring long (sag. and exs.),
convex, with prominent paired spines arising from swollen base ;
well-defined, sub-parallel sided median lobe, gently to strongly
convex, 2 pairs lateral glabellar lobes, small third pair usually
present. Convex cheek of characteristic subrectangular outline,

236 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

antero-lateral portion projecting, librigenal spine arising from
upper surface of border, directed upward and outward, may or
may not be spines on cheek border ; eye lobe about centrally situ-
ated, may be pedunculate ; two branches of suture usually inclined
to each other at an obtuse angle. Hypostome subrectangular in
outline with median posterior notch; middle furrow in form of
triangular depression in anterolateral corner of middle body.
Thorax of 9-10 segments; pleurae lacking pleural furrow or di-
vided by it into narrow convex anterior and broader convex
posterior bands, anterior pleural spine characteristically blade-
like, with lateral barbs, and downwardly directed; posterior
pleural spine much larger, 'horizontal, first 2 or 3 directed out-
ward and in some eases slightly forward, successive spines di-
rected more strongly backward. Pygidium with border spines,
may or may not be major pair, may or may not be unpaired
median posterior border spine.

Geological Rayige: Middle Ordovician to Middle Devonian.

Discussion: Since this group includes Miraspis, the oldest
available name Miraspinae is used for it. Prantl and Pribyl and
Erben include here the three best-known genera — Miraspis,
Ceratocephala and Dicranurus. As a result of the present study
I include Proceratocephala, Whitti7igtonia, and Ceratocephala
(Ceratocephalina) n. sub. gen. I have no new information on
Ceratonurus, Koneprusia, Orphanaspis, or Selenopeltoides, but
presume that they may also belong here. It appears to me to be
an overestimate of the value of morphological differences to
divide this group into three subfamilies, and some of the criteria
used by Prantl and Pribyl — e.g. supposed fusion of facial
sutures in Ceratocephala, lack of anterior pleural spines of
thorax in Dicranurus — are either of doubtful value or errone-
ous.

Genus MiRASPIS R. and B. Richter, 1917

Text-figure 15.

Type Species: Odontopleura mira Barrande, 1846.
(Whittington, 1956b.)
Diagnosis: Occipital ring with wide (sag.) posterior band.
Small third lateral glabellar lobes. Eye lobe situated opposite

WHITTINGTON : SILICIFIED ODONTOPLELTJIDAE

237

mid-part of basal glabellar lobe, pedunculate. Row long, slim
spines, diminishing in length forwards, on antero-lateral cephalic
border. Nine thoracic segments, well-marked pleural furrow ;

Figure 15. Miraspifi mira (Bariande). Motol Beds, ea2, upper middle
Silurian, Bohemia. A, B, cephalon, anterior and dorsal views, respectively ;
C, thoracic segment, dorsal view; D, pygidium, dorsal view. Approximately
X IV2. After Whittington 1956b, text-figure 3.

slim additional pleural spine between anterior and posterior.
Pygidium with pleural ridge running out at first transversely
from first axial ring, then turning abruptly to join base of curved

238 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

major spine. Four to eight pairs small border spines between
major pair, 2 in front.

Geological Range: Middle ( ?) Ordo^ician to LoAver Devonian.

Discussion: In addition to the material described elsewhere is
the single silicified pygidium discussed below, and the incomplete
thorax and Miraspis-type pygidium from the Middle Ordovician
of Wales (Whittington and Williams, 1955, p. 425, PL 40, fig.
119). Complete exoskeletons of Ordovician odontopleurids with
Miraspis-type pygidium are as yet undescribed, so the range of
the genus is uncertain.

?MiRASPis sp.ind.
Plate 14, figures 1, 7.

Material: one incomplete pygidium and one free cheek, both from

locality 3, lower Edinburg limestone.

Discussion: The pygidium is of characteristic Miraspis form,
the short, convex axis with a prominent first ring, bearing a pair
of spines. Pleural region crossed by a ridge connected to the
axial ring, this ridge directed at first outward and then turning
sharply to merge with the base of the major border spine. In
front of ridge is first pleural furrow, separating it from low
convex band bearing row of three spines. Major border spine
long, directed backward and slightly upward, two small spines
outside the major spine, 4 pairs inside, upper surface of border
with small spines, most prominent pair behind axis. Small spines
scattered on pleural regions.

The free cheek has the librigenal spine arising from the upper
surface of the border, latter bearing a row of short spines, and
the eye lobe is pedunculate. These features recall Miraspis, and
suggest that possibly this cheek belongs to the same species as
the pygidium.

Genus CeeATOCEPHALA Warder, 1838

Synonyms (subjective) : Onchaspis (Onychaspis) Raymond,
1925; see Whittington and Evitt, 1954, p. 53. Trapelocera
Corda, 1847 ; see Prantl and Pribyl, 1949, pp. 180-181, for sum-
nmvy of argument.

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE

239

Ceratocephala laciniata Whittington and Evitt, 1954

Plate 14, figures 2-6, 8-15 ; Text-figure 16.

Discussion: Two cranidia from locality 3 (PI. 14, figs. 2, 3)
are smaller (length (sag.) 0.46 and 0.57 mm.; maximum width
0.75 and 0.85 mm. respectively) than those previously described,
though not as small as that of C. triacantheis (Whittington and
Evitt, 1954, p. 60). They do not have fixigenal spines, but are
larger than the cranidia with fixigenal spines of Apianurus n.gen.
and Diacanthaspis described here, and thus seem to represent
the immediately succeeding stage when these spines are lost.
Notable is the convexity of the glabella of the smallest, on which
there are axial pairs of spines 2. 3, and 4 (Text-fig. 16 ; equivalent

Figure 16. Ceratocephala laciniata Whittington and Evitt. Cranidium,
exterior view, X 38. Drawn from the original of Plate 14, figure 3. Paired
spines numbered and lettered as in Text-figure 1.

to 3, 4, and 5 of Whittington and Evitt, 1954, p. 61, fig. 16), the
glabella being highest where spines 3 are situated. Lateral
glabellar lobes cannot be distinguished. The palpebral lolje is
far forward, bearing a long, curved spine (PI in Text-fig. 16)
and there are three spines on the eye ridge. On the convex
fixed cheeks spines Aj, A2, A3, B, C, and D may be recognized,
and between the spines on both glabella and cheeks there are tiny
close-spaced spines. The larger of these two cranidia (PI. 14, fig.
2) shows characters intermediate between the .smaller and the
originals of Whittington and Evitt (1954, PI. 7, figs. 1-8). Pairs
of spines 2a, 2, 3, 4, and 5 (equivalent to 2-6 of Whittington and
Evitt, 1954, text-fig. 16) are now present on the fronto-median
glabellar lobe, as well as a pair on the flanks between 2 and 3.

240 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Basal lateral lobes are faintly developed, and bear a stout median
spine, and with this inflation the fixed cheek no longer slopes
inward to the axial furrow.

Additional and better-preserved hypostomes (PI. 14, figs. 11,
12) have shown that the inner edge of the posterolateral doub-
lure is flexed upward, and just behind the lateral notch it is also
thickened. This slight projection may be regarded as the
posterior wing. The doublure is flexed (the flexure running
almost transversely) at the shoulder, and on the anterior side of
the flexure, near the outer edge, a hole pierces the doublure. This
hole is present in small hypostomes, is variable in size and usually
elongated transversely.

From localities 2, 3, and rarely 4, come the objects shown in
Plate 14, figures 5, 6, 10, 13-15. They show a range in size and
are both right and left-handed. The two divergent blade-like
spines, with a row of thorn-like barbs, curved distally, along
each side, are like the anterior thoracic pleural spines and the
pygidial border spines of C. laciniaia. As in these latter spines,
there are openings at the tip of the barbs (PI. 14, fig. 14). The
fused base of the two spines is attached to a flat strip which
resembles the outer part of a thoracic pleura of C. laciniata; it is
without a "pleural" furrow, the transverse edge is flattened and
bears a row of tiny tubercles along the "upper" edge. The anter-
ior edge of the strip is also very like the anterior edge of the pyg-
idial pleural region, notably in the way the flattened edge slopes
forward at the "fulcral articulating process" (PI. 14, figs. 8, 10).
On the opposite ("ventral") side the end of the "pleura" is
rolled under and there is an antero-lateral projection ("fulcral
articulating process") and a "fulcral" socket beneath the fused
base of the spines. AVhen placed as in Plate 14, figures 4, 8, 9,
between the outer part of a posterior thoracic segment and a pyg-
idium, the resemblance of this object to the outer part of a pleura
is extremely close. Further, the angle made by the lateral margin,
and parallel inner margin, of the "doublure," with the sagittal
line is a large one, larger than that of an obviously posterior
(with backwardly directed posterior pleural spine) thoracic
segment. This angle is such as to make the line of the inner edge
of the "doublure" fit between that of the p.ygidium and of a
posterior segment, as shown in Plate 14, figure 4. Thus this

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 241

object, if it is part of a thoracic pleura, belongs to the posterior
thoracic segment. It is not crossed by a pleural farrow, but such
furrows, though present in some posterior segments (AVhitting-
ton and Evitt, 1954, PI. 6, fig. 6), are extremely shallow and not
seen in posterior segments in which the posterior pleural spine
points inward and backward. When this object is arranged as
in Plate 14, figures 4, 8, 9, it is thus difficult to think of it as
anything but the outer part of the posterior thoracic segment.
There are difficulties to be surmounted before accepting this con-
clusion, viz. :

(1) Most serious is the fact that the ''pleura" of the object
is of constant length and the inner termination is a V-shaped
edge, the anterior limb of the "V" the longer, the edge bevelled
(PI. 14, figs. 13, 15). If this object is the distal portion of a
segment, why is it terminated in this way. when all other distal
parts of segments found are obviously broken (PI. 14, fig. 4; see
also Whittington and Evitt, 1954, PI. 6, fig. 6) ? The flattened
edge is like that of the anterior and posterior edges, which would
be the sutures between segments. Is the inner edge of this object
also a suture? If so these sutures would run symmetrically on
either side of an axial ring and innermost pleural parts, a
peculiar situation seemingly unique among trilobites. Xo such
median part of this supposed segment has been found.

(2) Why are the two "pleural spines" of the object similar
to each other, and not of the distinct anterior and posterior types
of all other known segments ? The border spines of the pygidium
are also of the thoracic anterior pleural spine type, and so, if
the object is the outer part of the last segment, it also represents
this posterior simplification of spine type.

I have examined two entire specimens of C verneuili in the
Museum of Comparative Zoology collections, but they show no
peculiar last thoracic .segment, and no such segment is illustrated
by Barrande (1852, PI. 38, figvs. 5, 6). The objects discussed
above occur in material in which C. laciniata is fairly abundant,
and they also occur in the lower Lincolnshire limestone with
C. triacantheis Whittington and Evitt, 1954. C. rarispina n.sp.
is not abundant in the Oranda formation, and these objects have
not been found at locality 8. Tentatively I regard these objects
as part of a thoracic segment of Ceraiocephaln. but the various

242

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

problems mentioned above remain at present unsolved. If they
are truly distal parts of the last thoracic segment, the reconstruc-
tion given by Whittington and Evitt (1954, fig. 13) requires
modification on the lines suggested by Plate 14, figure 8.

Ceratocephala rarispina Whittington, n.sp.

Plate 15, figures 1-25, 28, 29.

Holotype: USNM 116527 (PI. 15, figs. 1, 4, 5, 7), locality 8.
Other Material: Paratypes USNM 124698 a-d ; all figured speci-
mens in USNM.
Geological Horizon and Locality: Oranda formation, locality 8.
This species is rare at locality 8, as shown by the following
table : —

Table 3

Numbers of Odontopleurid Cranidia at Locality 8.

Diacanthaspis orandensis n.sp 120

Diacanthaspis scitulus n.sp. . . .
Diacanthaspis aff. ulrichi n.sp. .
Apianurus harhaius n.gen., n.sp.
Ceratocephala rarispina n.sp. . .

Description: This species differs from both Ceratocephala
triacantheis and C. laciniata (Whittington and Evitt, 1954, pp.
54-60, Pis. 6-9, 25, 26, figs. 1-17), bearing perhaps more re-
semblance to the former, older, species than the latter. Points
of discrimination are :—

1) Occipital ring lacks posterior band, and paired spines are
more strongly curved outward.

2) Third lateral glabellar lobes not developed.

3) First and second lateral glabellar lobes more inflated and
separated from both median lobe and cheeks by well marked
furrows.

4) Eye lobe situated farther back than in C. laciniata.

5) Border of free cheek subdivided by groove into two convex
bands of about same width, both of which merge into base of
librigenal spine.

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 243

6) Few, larger spines on external surface of cephalon. Six
pairs on median glabellar lobe include 2a, 2, 3, 4, 5, the latter
being far apart, and an additional pair far apart between 2 and
3. All have counterparts on other two species (e.g. compare PI.
15, fig. 1, with Whittington and Evitt, 1954, PI. 8, fig. 1 ; PL 25,
fig. 10), as has swollen base of pair 3. Two spines on eye ridge
rather than three. Spines along margin of free cheek longer.
Conspicuous on C. rarispma are the tiny spines that line the
edges of occipital furrow and furrows on cheeks (PL 15, figs. 1,
29).

7) Hypostome with shallower lateral notch and narrower
postero-lateral border.

8) Pygidium with much shorter median border spine.

In Ceratocephala laciniaia some of the thorn-like spines scat-
tered over the glabella and cheeks appear to have a single open-
ing at the tip (Whittington and Evitt, 1954, p. 59). In at least
one specimen of C. rarispina (PL 15, fig. 28), however, some
of these spines are covered over at the truncated tip, there
being several minute depressions or openings in the plate cover-
ing the tip (appearing as darker spots in the photograph). This
structure is like that in Diacanthaspis described above, and
whether or not hairs emerged from the tips of these spines is
uncertain.

Subgenus CerATOCEPHALINA Whittington, n.subgen.

Type Species: Ceratocephala {Ceratocephalina) tridens
Whittington, n.subgen., n.sp.

Discussion: C. {Ceratocephalina) tridens displays many of the
exoskeletal characters of Ceratocephala — the subtrapezoidal
outline of the convex cephalon, broad, long (sag.) occipital ring,
large basal, smaller second, and extremely small third lateral
glabellar lobes, prominent eye ridge and eye lobe, and librigenal
spine originating on the postero-lateral border of the free cheek ;
the thoracic pleurae are not furrowed and bear two pleural
spines, the anterior shorter, with lateral barbs; the pygidium
is short, triangular; paired axial spines are prominent on each
segment, and the ornament is of large, well-spaced spines. Yet
it is distinguished from any of the three species of Ceratocephala

244 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

in the Middle Ordovieian of Virginia (and other species of this
genus) by a well-marked group of characters: (1) the lesser
inflation of the lobes (fronto-median and lateral) of the glabella :
(2) the position of the large eye lobe, far back opposite the
midpoint of the basal glabellar lobe, and the consequent align-
ment of the branches of the dorsal suture; (3) the prominent
median occipital spine, as large as the paired spines; (4) the
relatively longer pygidium, axis well-marked, and lacking the
median border spine. The differences between C. laciniata, C.
iriacantheis ("SVhittington and Evitt, 1954) and C. rarispina
n.sp., seem to be less than those between any one of them and
C. {€.) tridens, and I have recognized these differences as of
subgeneric rank. If the hypostome associated with C. (C.)
tridens is correctly placed, the subsquare outline and well-defined
subtriangular anterior lobe of the middle body afford further
distinguishing characters of the subgenus.

Only the cephalon is known of Whittingtonia hispinosa, from
the Upper Ordovieian of Eire and Sweden (Text-fig. 17). It
is distinguished from Ceratocephala (Ceratocephalina) by the
far more prominent median glabellar lobe, the relatively narrow
basal glabellar lobe, the more anterior position of the eye lobe,
lack of median occipital spine, and less prominent ornament.
Proceratoeephala (Whittington, in56b), also an Upper Ordo-
vieian genus, has more inflated median and lateral glabellar
lobes, lacks the median occipital spine and has the paired spines
dift'erently situated, has well-marked pleural farrows in the
thoracic segments, and a pygidium with more prominent border
spines.

Ceratocephala (Ceratocephalina) tridens Whittington,

n.subgen., n.sp.
Plate 16.

Holotype: USXM 124699 (PI. 16, tigs. 1-3, 18), locality 4.

Other Material: Paratypes USXM 124700 a-c, all figured speci-
mens in USXM.

Geological Horizon and Locality: lower Edinburg limestone,
localities 3 and 4.
Description: Outline and form of cephalon shown in Plate 16,

figures 1-7. Occipital ring set off' from posterior border by

WHITTIXGTOX : SILICIFIED ODOXTOPLEURIDAE 245

change in slope and backward curve of posterior margin, longest
(sag.) medially; occipital furrow deep only behind basal lateral
glabellar lobes. Lobes of glabella gently convex, separated from
each other by broad, shallow furrows, only the inner parts of the
first and second glabellar furrows relatively deeper, and the
third furrow represented by a subcircular pit behind the eye
ridge (PL 16, figs. 3, 6) ; third glabellar lobe extremely small.
Change of slope only separates glabellar from anterior border,
axial furrows shallow beside first glabellar lobes, elsewhere
glabella merges into cheeks. Eye lobe large, subspherical, situ-
ated on highest jDoint of cheek, and opposite a point on midline
just in front of occipital furrow. Palpebral lobe vertical,
anteriorlj- merging with eye ridge, which curves forward to
join with most anterior part of glabella. Eye surface (PI. 16.
figs. 23, 24), with each facet convex externally, concave in-
ternally. Anterior branch of suture curves forward, diverging
slightly from eye ridge, crosses anterior border furrow on low
sutural ridge, and turns abruptly to run inward and downward
across anterior slope of border to meet rostral suture at obtuse
angle. Rostral suture runs along outer edge of narrow (sag. and
exs.) anterior border. Posterior branch of suture (PI. 16, fig.
18) runs down the vertical cheek, in a curve convex outwards,
and across posterior border furrow and l^order. Two branches of
suture are approximately aligned and thus do not make an
angle at eye lobe. Free cheek (PI. 16, figs. 23, 24) with broad
border occupying almost half the width at the librigenal spine,
shallow border furrow, midpart of border, between rolled margin
and inner part, concave upward. Librigenal spine arising inside
edge of border, base merging into border, the low swelling ex-
tending forward especially conspicuous. Interior view (Pi. 16.
figs. 6, 9) shows doublure widest laterally, and ridges formed by
deepest parts of occipital and glabellar furrows. Rostrum must
be short (sag. and exs.) but broad (tr.). Hypostome tentatively
placed here (PI. 16, figs. 19-22) sub-square in outline, middle
body convex, divided by shallow furrows into sub-triangular
anterior lobe and crescentic posterior lobe, well-marked depres-
sion in antero-lateral corner. Lateral notch extremely shallow,
shoulder small, pointed, posterior border wide (exs.) with
median notch. Major cephalic spines are librigenal and paired

246 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

and median occipital spines. The next size, thorn-like, include
those on the major spines, those projecting from edge of lateral
border, paired spines of glabellar lobes, palpebral lobe, eye
ridge, free cheek, etc. Fine granular ornament extends over
dorsal surface between these spines.

Number of thoracic segments unknown. Axis broad, pleura
relatively narrow (tr.). Construction of axial rings and pleurae
like Ceratocephala (cf. Whittingtoii and Evitt, 1954, PL 8, figs.
8, 10), but appendiferal pits shallower and anterior pleural
spines curved, directed outward and downward, not steeply
downward as in C. laciniata and C. triacantheis. The posterior
pleural spines are directed slightly upward and progressively
more strongly backward, so that on a posterior thoracic segment
(PI. 16, fig. 12) they point almost directly back. The pair of
spines on the axial ring are long and curve outwards. Thorn-
like spines occur on them and on the pleural spines. Pygidium
(PI. 16, figs. 15-17) with axis undivided, but the two pairs of
axial spines suggest it may be composed of at least two segments.
Pleural regions without border, one major pair of spines arising
behind axis on upper surface, directed backward and upward ;
doublure narrow with broader (sag.) "tongue" projecting
toward axis. Many short spines on edge of border, smaller spines
on pleural lobes.

Discussion: From locality 4, where most of the specimens
of C. (C.) tridens have been found, comes a single small cranid-
ium of the same general type, but displaying certain differences
(PI. 15, figs. 26, 27, 30). Most striking are the long curving
pair of occipital spines, the bases closer together, and the small
median occipital spine. The external surface is covered with
tiny tubercles, and the spines in the row on the fixed cheek are
longer than those of C. (C.) tridens. The different appearance
of the external surface of this cranidium may be the result of
preservation, but the difference in the occipital spines is sugges-
tive of a specific difference. Perhaps this cranidium represents
a second species of Ceratocephala (Ceratocephalina) .

WHITTINGTON : SILICIFLED ODONTOPLEURIDAE

247

Gemis PkOCERATOCEPHALA Prantl and Pfibyl, 1949

Synonym : Drummuckaspis Prantl and Pfibyl, 1949

Type Species : Acidaspis terribilis Reed, 1914.

Discussion: The type species has been described in detail

elsewhere (Whittington, 1956b), and is like Ceratocephala, but

differs in the presence of long paired spines and lateral lobes

on axial and occipital rings, possessing 9 thoracic segments with

deep pleural furrows, and a pygitlium with a long major and

short median border spine. The cephalon displays some features

recalling Miraspis, but is too poorly known to allow detailed

distinctions to be made. P. terrihilis is known only from the

Upper Ordo^dcian of Scotland.

Figure 17. Whittingtonia bispinosa (M'Coy), Upper Ordovician, Chair
of Kildare, Eire. A, B, C, incomplete cephalon, exterior, anterior, and left
lateral views, approximately X 3. Dra'VNii from originals of Whittington
1956b, Plate 59, figures 4, 5, 7, 8.

248 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Genus WhITTIXGTONIA Prantl and Pfibyl, 1949

Text-figure 17.

Type Species : Acidaspis bispmosus M'Coy, 1846.

Discussion : Only the cephalon is known, and a well-preserved
specimen has recently been described (Whittington, 1956b). It
is distinguished from Ceratocephala by the three narrow (tr.)
pairs of lateral lobes and wide (tr.), strongly convex fronto-
median lobe which overhangs the anterior border and has a
convex band across the base, and the short, thick spines on the
lateral cephalic border. Specimens are known from the Upper
Ordovician of Eire and Sweden.

Genus DiCRANUEUS Conrad, 1841

Text-figure 18.

Type Species : Acidaspis hamata Hall, 1859 (Whittington,

1956b).

Diagnosis: Occipital ring without posterior band, paired
spines thick and long, recurved in a hook extending over the
thorax. Small third lateral glabellar lobes. Eye lobe situated
opposite basal glabellar lobe on highest part of convex cheek.
No spines on lateral cephalic border. Nine thoracic segments;
anterior pleural spine blade-like, with lateral barbs, curved
downward and backward ; convex posterior pleural band con-
tinued into stout posterior pleural spine, anterior directed out-
ward, remainder outward and backward. Pygidium with pair
major border spines only, connected by strong pleural ridge to
first axial ring.

■Geological Range: Lower to Middle Devonian.

Notes on Other Miraspinid Genera

Selenopeltoides Prantl and Pfibyl, 1949, type species by orig-
inal designation Acidaspis hawlei Barrande, 1852, from the

Figure 18. Dicranurus monstrosus (Barrande), Prokop limestones, goo,
Middle Devonian, Lochkov, Bohemia. A, anterior view of cephalon, B, C,
dorsal and right lateral views of entire exoskeleton, approximately X 1%.

After Whittington 1956b, text-figure 4.

WHITTINQTON : SILICIFIED ODONTOPLEURIDAE

249

Figure 18

250 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

upper Middle Silurian of Bohemia. No material other than the
original incomplete holotype (and an isolated pygidium, now
missing) is known, so that the basis for a genus is most unsatis-
factory. Erben (1952b) suggested that the type species was
most closely related to Dicranurus, which appears probable.

Orphanaspis Prantl and Pribyl, 1949, type species by original
designation Trilohites orphanus Barrande, 1852, Middle Silurian.
Beside the type species, two others have been recognized in the
Lower and early Middle Devonian of Germany (Erben, 1952a.
pp. 306-308, 314-316, text-figs. 52, 54, PI. 20, figs. 7, 13; 1952b),
but all are based on pygidia only, of miraspinid type.

Koneprusia Prantl and Pfibyl, 1949, type species by original
designation Acidaspis fuscina Novak, 1883, lower Middle De-
vonian. Three species originally described by Barrande were
placed in this genus by Prantl and Pribyl (1949, pp. 199-202),
but of neither the type species nor these other species was new
material described or new figures given. The pygidium of the
type species is distinctive, the border spines being only one
stout pair and a median posterior, but the cephalon is poorly
known. Tentatively this material may be recognized as repre-
senting a separate genus, presumably of miraspinid type, but
until it is better known its position will remain uncertain.

Ceratonurus Prantl and Pfibyl, 1949, type species by original
designation Acidaspis krejcii Novak, 1883. I have no new obser-
vations to add to those of recent authors (Prantl and Pfibyl,
1949, pp. 189-192; Erben, 1952a, pp. 308-313 ; Erben 1952b), and
all are agreed that this Lower and Middle Devonian genus is of
Ceratocephala type. It may be derived from this latter genus
or from Miraspis.

?Subfamily MIRASPINAE R. and E. Richter

Odontopleurid Protaspis

Plate 21, figures 23, 24.

Material: USNM, one incomplete specimen from locality 4.

Discussion: This sjDecimen is smaller than either of the pro-
taspides of Diacanthaspis lepidus n.sp. (PI. 4, figs. 1-5) and D.
secretus n.sp. (PI. 6, figs. 1-5) from the same locality, but is of
extremely similar form. It difi:'ers principally in that the fixi-

WHITTINGTON: SILICIFIED ODONTOPLEURIDAE 251

genal spiue and those on the external surface are slimmer and
less swollen at the base. On the parallel-sided glabella only the
thick median occipital and axial spine pair 3 are present. On
the cheek, which is most convex in the inner part, the fixigenal
and spines Ai, A2, A3, B, C (directed almost straight outward
and somewhat upward), D (about midway between C and palpe-
bral lobe), PI and Er may be seen, and there is an additional
spine about midway between Ao and B. The protopygidium has
a markedly triangular outline, convex axis bearing two pairs of
spines, two pairs of spines on borders, at base of anterior of
which is short upwardly-directed spine.

The presence of only spine pair 3 on the glabella and the
outline of the protopygidium, as well as the slimmer spines
on the external surface, make it unlikely that this protaspis is
that of B. lepidus or D. secretus. The Stage 0 cranidium of
D. ulrichi n.sp. (PI. 8, fig. 10), has stout fixigenal and median
occipital spines, but others are faint, and the outline also is
unlike that of the cranidium of this protaspis. The Stage 0
cranidium of Apianurus harhatus n.gen., n.sp. (PI. 19, fig. 2),
has stouter spines, including paired occipital, pairs 2 and 4, and
those of the anterior border, all of which distinguish it. Thus

1 am driven to suggest that this protaspis might be that of
Ceratocephala laciniafa, which is fairly abundant at this locality,
but cannot exclude the possibility that it is that of the rare
C. (CeratocephaUna) tridens, n.subgen., n.sp. No evidence can be
offered for or against this latter possibility, since tiny cranidia
of C. (C.) tridens are not known. The smallest known cranidia
of C. laciniata (PI. 14, figs. 2, 3) have an outline like that of the
protaspis, notably the curve of the lateral sutural margin,
anterior border without spines, eye lobe far forward, fixed cheeks
lacking fixigenal spine but having all lettered spines and some
additional. The glabella of these small C. laciniafa cranidia,
however, is stronglj^ convex, spine pair 3 most prominent, but
long, curved paired occipital spines are present as well as pairs

2 and 4, and the median occipital spine is small. It is chiefly
this difference in the glabellae that makes me hesitate to identify
this protaspis as that of C. laciniata, though there are also dif-
ferences in the disposition of the lettered spines on the fixed
cheek. Only the discovery of cranidia intermediate between this

252 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

protaspis and the smallest of C. laciniata would show how. if
they belong to the same species, the transition takes place.

Subfamily APIANURIXAE Whittington, n.subfam.

Diagnosis: Glabella narrows forward, occipital ring long
(sag.), convex, long paired spines and median tubercle; well
defined, parallel-sided median lobe, two pairs of lateral lobes
fused. Eye lobe situated far back and about midway across
cheek, two branches of suture forming a straight line inclined
inward and forward to the sagittal line, free cheek narrow,
librigenal spine arising about midway along lateral border and
curving back. Hypostome shield-shaped, middle furrow arises
at antero-lateral corner of middle body and runs inward and
backward; small, pointed shoulders and shallow lateral notch.
Thorax of unknown number of segments, pleurae convex (exs.),
single large pleural spine. Pygidium with paired border spines,
unpaired median border spine may be present, long major spine
upwardly directed.

Geological Range: Middle to Upper Ordovician.

Genus ApiAXURUS Whittington, n.gen.

Type Species: Apianurus harhaius Whittington, n.gen., n.sp.

Diagnosis: Long occipital spines diverge at 60-80°; fused
lateral lobes kidney-shaped. Large eye lobe opposite basal glabel-
lar lobe. Hypostome widest anteriorly, convex middle body
divided into triangular anterior and crescentic posterior lobe ;
small anterior, tiny posterior, wings. Pygidium with six or seven
pairs border spines, flat pleural region bearing centrally-situated
upright major spine. Long spines (except occipital) with thorn-
like lateral spines, remainder of exoskeleton tuberculate or
spinose.

Geological Range: Middle and Upper Ordovician. An incom-
plete cranidium and fragmentary free cheeks and thoracic seg-
ments testify to the presence of a species of this genus in the
Lincolnshire limestone, below the Edinburg limestone.

WHITTINGTON : SILICIFIED ODOXTOPLEURIDAE

253

254 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Apianurus barbatus Whittington, n.gen., n.sp.

Plates 17-19 ; 20, figures 1-17, 19 ; Text-figures 19-22

Holotype: USNM 124701 (PI. 17, figs. 3, 4, 6), locality 2.
Other Material: Paratypes, USNM 124702 a-e; all figured speci-
mens in USNM.
Geological Horizon and Localitiei^: Edinburg limestone, localities

2, 3, 4, 6, 8.

Description: Cephalon moderately convex, outline in dorsal
and anterior aspect elliptical. Glabella moderately convex trans-
versely, gently convex longitudinallj', maximum width across
occipital ring, narrowing forward to half this width at anterior
margin ; length about two-thirds maximum width. Occipital ring
of length (sag.) in dorsal aspect half that of rest of glabella,
becoming narrower (exs.) laterally, outline of posterior margin
a curve strongly convex posteriorly, outline of anterior margin
a curve more gently convex anteriorly. Stout occipital spines
diverge at about 60° and curve upward and backward, extend-
ing to a length three times the sagittal length of the cephalon.
Small median occipital spine just behind occipital furrow. Lat-
ter shallow medially, deep behind basal lateral lobes. First and
second glabellar lobes fused to give a kidney shape, the second
the smaller, the first glabellar furrow represented by a subcircu-
lar pit adjacent to the median lobe. A pit anterior to the second
glabellar lobe represents the second glabellar furrow, and is
bounded anteriorly by the eye ridge as it fuses with the frontal
glabellar lobe. Outline of anterior margin of latter curve con-
vex forward, and separated by change of slope from narrow
anterior border. Fronto-median lobe gently convex, change of
slope separating it from lateral lobes, faint additional swelling
running transversely across it between first glabellar lobes.
Axial furrow not deep, but a narrow, unornamented band, mark-
ing reversal of slope between cheek and glabella. Cheek semi-
circular in outline, maximum width opposite second glabellar
lobe, rising steeply to large eye lobe. Transverse line through
mid-point of eye lobe passes through mid-point of first glabel-
lar lobe and runs just in front of occipital furrow. Anterior
branch of facial suture runs straight forward and inward
at about 45° to the midline, curving inward a little more as it

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 255

crosses the border on a sutural ridge. Posterior branch runs out
and back in line with anterior branch for a short way, then
curves in over posterior border. Latter longest (exs.) at suture,
convex, and becomes shorter inward. Thus the occipital ring
merges with inner corner of cheek rather than with posterior
border. Fixed cheek .slopes vertically behind eye lobe. Palpebral
lobe with rim becoming well-defined anteriorly, and passing into
broad, convex eye-ridge. Latter runs with a slight curve inward
and forward to merge with frontal glabellar lobe. Eye surface
almost hemispherical, external surface (PI. 17, fig. 21) showing
faintly the tiny facets. Lateral cephalic border rolled, shallow
border furrow interrupted at about mid-length by swelling at
the base of librigenal spine. Latter directed outward and slightly
forward at first, then curving and slimming to point backward,
and reach to a length about three times that of the cephalon
(sag.). Just outside anterior branch of suture is antennular
notch (PI. 17, fig. 5), with a sharp projection at outer margin.
Anterior border between sutures narrow (sag. and exs.), rostral
suture running along outer edge. Rostrum unknown, but pre-
sumably wide (tr.) and short, sloping downward and inward.
Doublure narrow on free cheek, absent on posterior border of
cranidium, articulating half-ring long (sag.). On inner surface
of cranidium (PI. 17, fig. 12) outer part of occipital ring, first
and second glabellar furrows make rounded projections but are
not extended as appendifers. Width of hypostome (PI. 18, figs.
1-5) at anterior margin greater than length, rounded postero-
lateral outline, faint median posterior notch. Anterior wing
small, tip rounded, no wing process. Lateral border narrow,
gently convex, deep border furrow, shallow lateral notch and
sharply-pointed shoulder. Po.sterior border wider and in median
portion separated by shallow depression from median body.
Outermost part of posterior border bent up sharply near mid-
line. From the broad depression at antero-lateral corner the
middle furrow runs straight inward and backward to define tri-
angular, convex anterior lobe of the middle body. Posterior lobe
crescentic in outline, inflated. Doublure of hypostome widest
laterally, disappearing at anterior wing and pressed close to
median part of posterior border so that edge is sharp. Ridge
crosses doublure from point of shoulder to inner edge, immedi-
ately behind ridge doublure extended inward to form small tri-

2r>6 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

r abf

Figure 20. Apianurus barbatm n.gen., n.sp. Outline reeonstruetioiis of
cephalon, approximately X 10. A, ventral view, right half of rostrum and
left half of hypostome shown, anterior portion of left free cheek incomplete.
B, anterior view, rostrum not shown and anterior portion of left free cheek
incomplete. C, left lateral view, rostrum and left free cheek not shown, a,
antennular notch; abf, ridge on inner surface corresponding to anterior
liorder furrow; d, doublure of right free cheek; 2gf, ridge on inner surface
corresponding to second glabellar furrow; h, hypostome; m, boss corre-
sponding to depression at outer end of middle furrow; r, rostrum.

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 257

angular posterior wing. Perforation through doublure posterior
to ridge (PI. 17, fig. 19).

By mounting a cranidium and free cheek of appropriate size
iu juxtaposition, as shown in Plate 17, figures 5, 7, 9-11, the
approximate position and attitude of the hypostome has been
determined (Text-fig. 20). The gently-curved hypostomal suture
(PI. 18, figs. 1, 5) fits against the rostrum, and the anterior wing
is directed upward and outward, the tip lying beneath the deep-
^i-ned outer part of the anterior border furrow where it meets the
furrow bounding the outer side of the eye ridge. There may have
been muscles linking the wing and inner side of the furrow,
but there is no wing process or socket. The depression at the
outer end of the middle furrow forms a considerable projection
on the inner surface (PI. 18, fig. 2), and may have been a
point of muscle attachment. Perhaps it was linked by muscles to
the outer end of the anterior border furrow. The shape of the
rostrum is suggested, and is consistent with the form of the few
odontopleurid rostra knoAvn (Whittington, 1956b, PI. 58, fig. 7;
PI. 59, fig. 3). If the cephalon rests on a flat surface on the
antero-lateral cephalic border and librigenal spines, as portrayed
in Text-figure 20 B, C, the middle body of the horizontal hypo-
stome would lie on this surface.

Ornament of cephalon of close-spaced spines which curve
characteristically toward the closed tip (PI. 17, fig. 21; PI. 18.
fig. 22). The spines A'ary in size, the exoskeleton between their
bases smooth, as are also the axial, occipital, and glabellar fur-
rows. Arrangement of spines not perfectly regular, but sym-
metrical pairs of larger spines ma}^ be distinguished even on
largest cranidia, on the occipital ring, median and lateral glabel-
lar lobes, eye ridge, etc. (PI. 17, fig. 2). Single row of s.ymmetri-
eal spines on anterior border of cranidium. Large, curved spines
are also regularly arranged on the border of the free cheek and
the proximal portion of the librigenal spine (PI. 17, figs. 1, 7, 14,
15) — e.g. four long and one short spine on the border behind
the librigenal spine ; stout spine at edge of autennular notch, one
of same size midway to base of librigenal spine ; 6 spines along
outer side of proximal part of librigenal spine, etc. The character-
istic curved spines are not present on the distal part of the
librigenal spine, nor on the occipital spines. Much shorter, blunt,

258 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

spines are present, directed outwards on the proximal portion of
the spine, but distally inclining more and more in the direction
of the main spine, so that at the tip they lie at a low angle to
the surface (PI. 17, fig. 3). Toward the tip of the main spine, on
the distal side of the base of the small spine, a minute opening
is sometimes seen — apparently the locus of a sensory hair.
Other cephalic spines do not seem to show such openings.
Proximally, in the region where there are the large lateral spines,
the under .side of the librigenal spine is smooth, as is the doublure
of the free cheek. The characteristic curved spines occur on the
lateral and posterolateral borders of the hypostome (PL 18, fig.
1), but are reduced to granules on the posterior border. Similar
spines are scattered on the lateral areas of the middle body, but
the central portion is smooth.

Number of thoracic segments unknown. A sufficient number
of the delicate segments are preserved to permit the reconstruc-
tion (Text-fig. 19 A, B), which is based on the assumption of
a total of ten. Axis broad, two-thirds total width (excluding
pleural spines) at anterior segment (PI. 18, fig. 6), narrowing
back to half its width at posterior segment (PI. 18, fig. 9). Axial
ring moderately convex, posterior margin forms a curve, convex
posteriorly. Articulating furrow more strongly curved in same
sense, outermost part deepened. Articulating half-ring longer
(sag.) than axial ring. Pleura narrowest (tr.) in anterior seg-
ments, horizontally extended, without pleural furrow. Narrow
(exs.) flanges developed, outer surface of each flat and inward
and downward sloping. Ring and axial articulating sockets and
processes scarcely at all developed, but lateral margin of pleura
rolled under and fulcral socket and process prominent. Base of
pleural spine swells out from upper, outer surface of pleura.
Length, curvature and direction of pleural spines indicated in
reconstruction. The characteristic curved ornamental spines are
close together on the axial ring, varying in size, some larger
ones paired. Similar spines more widely scattered on pleurae.
Pleural spines ornamented like librigenal, with row of lateral,
curving hooks on each side of proximal part, and the tiny in-
clined spines distally (PI. 18, fig. 9). Similar openings to those
on the tips of the occipital and librigenal spines seem to be
present.

Pygidium (PI. 18, figs. 11-13) semi-oval in outline, length

WHITTINGTOX : SILICIFIED ODOXTOPLEURIDAE

259

(sag.) a little more than one-third width at anterior margin.
Axis moderately convex, a little wider than long, rounded pos-
teriorly, rising from the flat pleural regions and not defined by
deep furrows. Articulating furrow and half-ring like those of
segments. First ring furrow broad and deep, curving forward
to midline, extremities widened and deepened. Second ring

Figure 21. Triangular graph comparing relative proportions of dimensions
of cranidia of Apianttrus barhatus n.gen., n.sp. Those from the lower Edin-
luirg formation, localities :!, 3, sliown in dots enclosed by solid line.
Cranidia from the Oranda formation, locality 8, shown by open circles
enclosed by broken line. Two tiny cranidia from locality 3 shown by X
(originals of Piate 19. figures 2, 9). L, sagittal length of cranidium. "W,
width between spines B on posterior border. E. length (exs.) from anterior
margin to midpoint of palpebral lobe. Measurements were made in ex-
terior view, i.e. with isolated cranidium resting on a horizontal surface.

260 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

furrow represented only by pair of oval, unornamented areas on
terminal part of axis, corresponding in position to wider outer
part of first furrow. In many specimens these two pairs of oval
areas are translucent, the quartz much thinner over them, or
broken, indicating- that they may be areas of muscle attachment
(cf. Whittington and Evitt, 1954, pp. 24-25). Long spine rises
vertically from flat pleural region, opposite first ring furrow,
and distall}^ curves a little backward. Border of pleural regions
with 6 pairs of spines directed outward and slightly upward,
first pair short, second longer, third longest, and remaining
pairs shorter inwards. Doublure of pygidium narrow, rolled
under at antero-lateral corner forming articulating process.
Ornament of curved spines on pygidium (PI. 18, figs. 12, 13, 19)
disposed as on thorax — dense, varying in size, and with some
conspicuous pairs on the axis, more widely spaced on pleural
regions. Upright and border spines with lateral barbs, and
towards tips the tiny inclined spines are present, and the open-
ings may be present.

Discussion : In addition to the material from the lower Edin-
burg formation (localities 2, 3, 4; Pis. 17-19), that from the
Oranda formation (locality 8; Pi. 20, figs. 1-17, 19) also repre-
sents Apianurus harbatus. Comparison of parts of the exoskele-
tons from all localities reveals no consistent differences between
them. Measurements of length and width of eranidia and of posi-
tion of palpebral lobe in specimens from both horizons fall within
almost the same area of a triangular graph (Text-fig. 21). Local-
ity 8 in the Oranda formation is some 450 ±: feet higher in the
section than localities 2-4 in the lower Edinburg, so that A.
harbatus evidently has a long range in time. A few specimens
(see Table 4) are knoAvn from locality 6, at an intermediate point
in the section.

Figure 22. Ainannrus harhatus, u.geu., ii. sp. A, Stage 0 exoskeleton,
dorsal view, drawn from originals of Plate 19, figures 1, 2, 5. Approximately
X 32. B, t-ranidium, exterior view, drawn from original of Plate 19, figure
9. Approximately X 32. C, cranidium and right free cheek, exterior view,
drawn from originals of Plate 19, figures 7, 8. Approximately X 32. D,
cranidium and right free cheek, exterior view, free cheek from original of
Plate 19, figure 11. Approximately X 32. E, cranidium, exterior view. Ap-
liroxiniately X 32. F, cranidium, exterior view. Approximately X 16.

WniTTIXGTOX : SILICIFIED ODOXTOPLEURIDAE

261

262 BULLETIN : MUSEUIM OF COMPARATIVE ZOOLOGY

Development. Stage 0: Protaspis unknown, but the smallest
known cranidium, free cheek and pygidium from localities 3 and
4 (PI. 19, figs. 1, 2, 5) have been associated (by analogy with
Diaca7ithaspis) to show the probable nature of the Stage 0 exo-
skeleton (Text-fig. 22A). Outline of cranidium (length (sag.)
0.41 mm.) trapezoidal, gently convex longitudinally, more gently
transversely. Glabella outlined by broad, straight, axial furrows,
short (sag. and exs.) occipital ring with prominent median and
short curved pair spines, occipital furrow defined only distally,
remainder of glabella without furrows, four prominent pairs
short, curved, axial spines (2-5 in Text-fig. 22A). Narrow an-
terior border, fixed cheeks gently convex, palpebral lobes at two-
fifths cranidial length (exs.) from anterior margin. Posterior
margin of fixed cheek runs directly outward from the axiai
furrow, and just beyond half the width bends abruptly to run
outward and forward to the base of the long fixigenal spine.
On the outer, postero-lateral area of the fixed cheek there are
4 paired spines (A, B, C, D in Text-fig. 22A), and tiny spines
along the inner side of the fixigenal spine and adjacent outer
part of the fixed cheek. Certain specimens (PI. 19, figs. 4, 6) have
the fixigenal spine retained on one side only. Their significance
is discussed under "Ontogeny" in Part I. Free cheek of ap-
propriate size to fit this smallest cranidium (PI. 19, figs. 1, 3;
Text-fig. 22 A) narrow, large eye surface situated at half length,
big librigenal spine curving outward and backward. Anterior
branch of the suture straight, posterior runs in curve convex out-
ward, doublure projecting inward and fitting against doublure at
base of fixigenal spine. Notch in anterior border outside suture
very small. Long spines arranged in a constant pattern along
outer edge of border and proximal edges of librigenal spine —
conspicuous is variation in size of those along antero-lateral edge,
curved spine at anterior basal edge of librigenal spine, numbers,
curvature and/or inclination of remainder. Transitory pygidium
of width at anterior margin 0.51 mm., length (sag.) 0.30 mm.
Prominent axis bearing three pairs of curved spines, first two
rings distinct, tip indistinct where posterior part of pygidium is
bent sharply down. Pleural regions flat, borders bent down, 8
border spines arise on dorsal surface inside border, curving out-
ward and backward. Doublure rolled under. Small barbs along
sides of these border spines.

WHITTINGTON* : SILICIFIED ODONTOPLEURIDAE 263

Further development of cephalon: Next largest cranidia (PI.
19, figs. 9, 10; Text-fig. 22B) from localities 3 and 4 only. Par-
allel-sided glabella more convex, defined by deep axial furrows,
tiny fifth pair of axial spines (2a in Text-fig. 22B) present just
in front of occipital furrow Small lateral pairs of spines also
present outside pairs 2, 3, and 4. Extremity of occipital furrow
a shallow pit, in front of which is small swelling representing
the beginning of basal glabellar lobe. Large palpebral lobe
situated at about the same position. Most striking is the absence
of fixigenal spine and rounded outline of genal angle of fixed
cheek. Paired spines A, B, C and D (Text-fig. 22B) are de-
veloped as in the smaller cranidium, and there are one or two
additional spines at the genal angle between B and C, as well as
a small one on the fixed cheek between A and B. Free cheek
associated with this size of cranidium very like that of Stage 0,
but has eye surface situated behind mid-length, posterior branch
of suture running in curve convex outward and seemingly little
or not at all modified despite absence of fixigenal spine. Notch
in anterior border is wider, spines along margins and proximal
part of librigenal spine similar in size and arrangement to those
on the smaller cheek.

The next largest cephala (PI. 19, figs. 7, 8 ; Text-fig. 22C) show
a marked increase in convexity of glabella and cheeks and in
height of eye lobes. Occipital ring projects back farther behind
cheeks, and paired spines now reach back to a length equal to that
of cranidium. Midpoint of eye lobes is at about half the length,
cheeks slope steeply behind them to border furrow, which runs
out to margin between spines A and B. Eye ridge faintly de-
fined. Median occipital and five paired spines of glabella
prominent, additional lateral pairs developed, including 2 pairs
(one central and one just in front) on convex basal glabellar
lobe. On cheek and borders main spines developed as liefore
but additional pairs also present. Next largest cephalon (PI. 19,
figs. 11, 14-16; Text-fig. 22D) similar, eye lobes at about .same
position, eye ridge more clearly defined. Occipital ring with
longer paired spines, and relatively reduced median spine. In
front of pits at extremity of occipital furrow are convex, ovate,
basal glabellar lobes. In front of these, in line with anterior part
of eye lobe and at base of side of glabella, a prominent pair of

264 BULLETIN' : MUSEUM OF COMPARATIVE ZOOLOGY

spines. In slightly larger cranidia (PL 19, figs. 18-20, 22, 23;
Text-fig. 22E ) the area around the base of these spines is slightly
swollen, and represents the second glabellar lobes. A deep pit
represents the first glabellar furrow, and a faint swelling outside
it connects the two lobes. Thus the bean-shaped fused lateral
lobes are present at this stage, and become more prominent as
the cranidium increases in size, swelling up to partly fill the de-
pression between eye lobes and glabella. Other changes as size
increases include appearance of swelling on median glabellar
lobe just in front of occipital furrow, backward movement of eye
lobes with consequent steeper slope behind them, and increased
prominence of eye ridges. Many more ornamental spines appear,
mostly paired, and a median row on the median glabellar lobe.
The relative size of these spines diminishes, but even in the larg-
est cranidia (PI. 17, figs. 2, 3) the numbered and lettered pairs
may be recognized by their slightly larger size, and most spines
retain the characteristic curve so prominent in tiny specimens.
The chief change in the free cheek from the size shown in Text-
figure 22A is the backward movement of the eye lobe, so that the
postero-lateral slope behind it is steep or overhanging. The
arrangement of the spines on the border and proximal part of
the librigenal spine remains basically the same, though the curved
spine at the anterior basal edge of the librigenal spine is reduced,
and an additional large spine (making six in all) appears along
the anterior proximal margin of the librigenal spine. The margin
of the antennular notch remains smooth, but the outer edge bears
a large spine. Because the broad, swollen base of the librigenal
spine merges into the cheek a border becomes defined by a shallow
border furrow only near the branches of the suture.

Smallest known hypostome (PI. 19. figs. 12, 13) 0.4 mm. in
length (sag.), 0.55 mm. in width across anterior wings, propor-
tions of various parts similar to adult. There is little change with
increasing size.

Thorax and pygidium. Next largest transitory pygidium (PI.
19, fig. 24) to that assigned to Stage 0 (PI. 19. fig. 5) has three
pairs of spines on the axis, 3 pairs long border spines, with
lateral barbs, similarly directed to those of the earlier stage.
An upright spine at the base of the anterior two border spines.
Larger transitory pygidia (PI. 19, figs. 25, 27-29) have four

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 265

pairs axial spines and four pairs border spines, the fourth border
pair short, blunt, situated at tip. Latter less steeply bent down.
First three border spines with lateral barbs and upright spine
at base, direction and curvature shows slight variation, perhaps
associated with stage. Latest transitory pygidium (PL 19, fig.
26) of length (sag.) 0.36 mm., width 0.7 mm. with 4 pairs of
axial spines, posterior two close together at termination. Long,
curved spines arising from the antero-lateral portions of pleural
regions are pleural spines of what is to become last thoracic seg-
ment. Bases of broken-oif pair of upright spines on pleural
regions are opposite 3rd and 4th axial spines. Four pairs of
border spines, third pair larger than others. Posterior tip of this
pygidium not bent down. Small true pygidium (PI. 18, figs. 23,
24) has first axial ring defined, with its pair of axial spines, and
posterior part of axis bears two pairs of axial spines (and pre-
sumably therefore includes at least 2 segments). Swollen base
of upright spine on the pleural region is connected with first
axial ring. Margin with tiny anterior pair of spines and 4
larger pairs, third of these distinctly longer. As size increases
(PI. 18, figs. 16-18, 20, 21) the adult appearance is attained,
except that there are still only 5 pairs of marginal spines, the
anterior short, the others becoming equal in size. There are
three pairs of axial spines, and a median spine on the tip of the
axis. Large pygidia (PI. 18, figs. 11-13) have six pairs of mar-
ginal spines, the first much shorter and slimmer than the rest.
Behind the ring furrow the axis includes two pairs of axial
spines, and a median spine at the tip.

Specimens from Localities 2 and 3
Differing from Apianurus barhatus n.gen., n.sp.

Under this heading may be grouped the following :

(a) Cephala and pygidia from both localities which are so
strikingly different as to be regarded as representing a species
of a distinct genus, described below as Calipernuriis insolitus
n. gen., n. sp.

(b) Cranidia from locality 2 which are like A. harhatus but
are smoother — i.e. the spines are short, reduced to tubercles, or
absent. These are described below as A. glaber n.sp.

266 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

(c) Pygidia occurring at both localities, distinguished from
A. harhatus b}' being smoother and having shorter, inwardly-
curved upright spines. These are here regarded as belonging
with cranidia (b).

(d) A second type of pygidiuni occurring at locality 2, hav-
ing the typical curved spine ornament of A. larhatus, but the
upright spines are very short, and there are 7 pairs of spines
on the border. This pygidium is described as Apianurus sp.ind.

(e) Three types of hypostome from locality 2, each different
from that of A. harhatus. The one differing most has been
assigned to Calipernurus insolitus n.gen., n.sp., the smooth one
to A. glaher n.sp., the third described under '' Apianurus sp.
ind."

Each of the groups of specimens included under (a) to (e) is
distinct, and there is no gradation between groups or between
any one of them and A. harhatus. Within each group except (e)
there are specimens of different sizes, showing that the dis-
tinctive characters of the group are not those of one particular
size (i.e. groAvth stage). Table 4 shows the relative abundance of
the different species at the different localities. Only A. harhatus
is found in strata above the lower Edinburg. If the specimens
grouped under (b), (c), and (d) were expressions of the
morphological range of variation of A. harhatus one might expect
them to be represented in the sample from locality 8. The fact
that they are not suggests that they represent distinct species.
This same pattern of distribution is displayed by species of
other genera from localities 2 and 3, e.g. among the hundreds
of specimens of Dimeropyge virginiensis Whittington and Evitt
(1954, pp. 37-42, PI. 2, PI. 3, figs. 1-30) are a small number that
are distinctly different and do not grade morphologically into
the common form, and which I regard as a separate species. One
possible explanation of the ''common" and "rare" species of
the same genus at the same locality is that they represent sexually
dimorphic forms of the same species, but this explanation can
hardly be removed from the realm of speculation. In this con-
nection Hintze (1953, p. 150) has discussed the occurrence of
two pliomerid species at the same horizons of the Pogonip group,
but in this case the two species are apparently equally abundant.

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE

267

id
<u

O
OS

u.
Ph

F— I

=4-1

o
CO

sarao^sodjCjj

■^

BTpiS-fj

sauio;so(LvH

eipiS^jj

O
C3

BTpuiBJ3

CO
I— 1

sanio^socKg;

5, B

BipiS.Vcj

1-H

BTpmB-tQ

r— 1
CI

1— (

P.
d

sarao^sodAjj

t— 1

1—!

BipiSij

to
in

BrpTUBI3

r— '
r-i

268 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Apianurus glaber Whittington, n.sp.

Plate 21, figures 1-15.

Holotype: USNM 124703 (PL 21, figs. 1-3), locality 2.

Other Material: Paratypes, USNM 124704a, b; other figured

material in USNIM.
Geological Horizon and Locality: lower Edinburg limestone,

localities 2, 3.

Description: Cranidium (PI. 21, figs. 1-7, 11, 12) differs from
that of Apianurus harhatus n.gen., n.sp. (PI. 17, figs. 1-12) in
that :

(1) Occipital ring with slightly depressed posterior band,
widest (exs.) laterally, extremely narrow medially. In front
of it a low ridge connects inner, posterior bases of occipital
spines. Median occipital spine prominent, area between it and
bases of occipital spines flattened, not evenly inflated. Paired
occipital spines more divergent proximally, but curved so that
distally they are directed slightly inward, not outward. Ex-
tremity of occipital furrow deeper, forming a more prominent
projection on inner surface.

(2) Palpebral lobe less steeply inclined, eye ridge running
forward and inward slightly farther out from lateral glabellar
lobes, then curving more strongly in to meet frontal lobe of
glabella. Anterior part of eye ridge less convex, and shallower
depression outside it. Posterior border juts out slightly farther,
and outer, sutural edge more acutely rounded.

(3) Larger cranidia with low tubercles on external surface,
not curved spines. Tubercles are fewer in number than spines
of A. hariatus, and on the glabella do not show any obvious
paired arrangement. Lateral and postero-medial parts of occipi-
tal ring smooth. Smallest cranidia bear short spines on the
glabella, few on occipital ring, but the fixed cheek bears spines
with the characteristic A. harhatus curve, symmetrically ar-
ranged. These latter are reduced to rounded tubercles in the
large cranidia, but retain the symmetrical arrangement.

A hypostome (PI. 21, figs. 13, 14) from locality 2 differs
slightly from that of Apianurus harhatus (PL 18, figs. 1-5), and
is assigned to A. glaher principally because it is almost without
ornament, there being only a few tubercles on the lateral border

WHITTINGTON : SILICIFIED ODONTOPLBURIDAE 269

adjacent to the shoulder. The outline is more rectangular, the
median part of posterior margin being straight, the postero-
lateral border broader. Middle furrow of middle body is distinct
to midline, and does not fade out there, as in A. harlatus, and
the same is true of the posterior border furrow.

Free cheek and thorax unknown. Pygidium attributed to this
species (PI. 21, figs. 8-10, 15) of the same form as that of A.
harbatus (PI. 18, figs. 11-13), with the same number of border
spines but directed slightly more upward. Distinctive are the
upright spines on pleural regions, short and inwardly curved,
rather than long and curving back. Low tubercles on external
surface, not short spines, a few on axial rings and scattered
over pleural regions. Border and upright spines with lateral
barbs. The reduction of the ornament on the axis reveals clearly
the supposed areas of muscle attachment, over which the quartz
is thinner. Three segments are distinct, the ring of the third
outlined by an inverted V-shaped line of tubercles which enclose
the tip of the axis. There is a median tubercle at the extreme
tip, and between this and the ring of the third segment the
quartz is also thin. This triangular area presumably represents
the muscle attachments of a 4th segment.

Apianurus Whittington sp.iud.
Plate 21, figures 16-22, 25, 26.

Material: Figured specimens USNM 124705a-e, lower Edinburg

limestone, locality 2.

Description: Pygidium known only from locality 2 and differs
from that of Apianurus harbatus n.gen., n.sp. (PI. 18, figs. Il-
ly) in that the pleural regions are relatively wider, the upright
spine small and short, and there are one short and six long pairs
of border spines, longer and slimmer and slightly more upwardly
directed. The ornament is of typical curved spines like those
of A. harbatus, and there are lateral barbs on upright and border
spines. The smallest specimen (PI. 21, fig. 21) is 1.6 mm. in
width at the anterior margin, and exhibits the same distinguish-
ing characters, except that there are 1 short and 5 longer pairs
of border spines. However, pygidia of A. harbatus of this size
also have one less pair of border spines.

270 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

A few specimens of a hypostome (PI. 21, figs. 20, 22, 25, 26),
occurring at locality 2, differ from that of Apianurus harhatus
(PI. 18, figs. 1-5) in the following respects: (1) smooth even
curve of outline of anterior margin without the projection in
front of middle body; (2) smooth curve, rather than V-shape,
made by confluent middle furrows; (3) posterior lobe of middle
body evenly inflated, not divided by a flattening near the mid-
line; (4) lateral notch less well defined, anterior edge merging
with adjacent border.

Both these exoskeletal parts are distinct from corresponding
parts of Apianurus harhatus and A. glaher, and seem to repre-
sent at least one additional species.

Apianurus aff. furcata (Linnarsson, 1869)

Plate 20, figures 18, 20-25.

Material: PMO 66691a, incomplete cranidium; 66691b, small

fragmentary cranidium; 66690, cranidium with incomplete

occipital spines, all from Chasmops limestone, Guttorms-

berget ; 3673, cranidium from Upper Chasmops limestone, Fer-

neholmen, Asker; 5647, pygidium from Bygdoy, near Oslo,

Norway.

Geological Horizon: Chasmops limestone, middle Caradoc (Stor-

mer, 1953, p. 130), or approximately late Black River or early

Trenton (Twenhofel et al., 1954). The block from Bygdoy

contains, besides the pygidium, cranidia and pygidia of "Bron-

teopsis" gregaria type (Cooper, 1953, PI. 9) and Bemopleuri-

des, genera characteristic of the lower Edinburg of Virginia.

Description: Cranidium typical of the genus, differing from

that of Apianurus harhatus n.gen., n.sp. (PI. 17, figs. 2-12)

principally in that : ( 1 ) occipital ring relatiA'ely a little longer

and occipital spines more divergent; (2) fronto-median and

fused lateral glabellar lobes less inflated; (3) palpebral lobe

larger and less steeply sloping, and eye ridge just in front of

lobe appears broader and more prominent. The tubercles of A.

aff. furcata may be the bases of spines like those of A. harhatus

n.gen., n.sp., and the number and distribution is similar.

The small pygidium (PI. 20, fig. 20) is characteristic, having
the large upright spine on the pleural region (broken off at the

WHITTINQTON : SILICIFIED ODONTOPLEURIDAE 271

base), and 6 pairs of border spines, the anterior pair small.

Discussion: The following species have been described from
Sweden and the East Baltic :
Apianurus furcaia (Linnarsson, 1869, p. 65, PI. 1, tig. 18)

cranidium from the Chasmops (=Beyrichia) limestone of

Vastergotland, Sweden.
Apianurus kuckersiana (Schmidt, 1885, pp. 4-5, PL 1, figs. 2,

3; bpik, 1987, p. 47, PL 24, figs. 3, 4) from the Kukruse

(C2) stage of Estonia. Opilv figured the characteristic pygidium

as well as the cranidium.
Apianurus kuckersiana var. mickwitzi (Schmidt, 1907, pp. 23-

24, PL 1, fig. 19), cranidium from the Keila (Do) stage of

Estonia.
Apianurus askJundi (Thorslund, 1940, pp. 154-155, PL 6, fig.

14), cranidium from the lower Chasmops limestone, Jemtland,

Sweden.

All are from the Chasmops limestone or its equivalents, and
presumably closely related. I have not had an opportunity to
examine the original material, and so cannot venture an opinion
as to how many species are represented. I haA'e used the oldest
specific name for the Norwegian material, rather than create
another name. It is notable that many of the genera of trilobites
of the lower Chasmops limestone listed by Thorslund (1940, pp.
184-185) occur also in the Edinburg limestone — e.g. Trinodus,
Remopleurides, "Bronteopsis" (^Stygina'! of Thorslund, 1940,
p. 137), illaenids, Dimeropyge, Ampyx, Lonchodomas, Sphaer-
exochus (see Cooper, 1953; Whittington and Evitt, 1954).

Apianurus clevei (Warburg, 1925, pp. 243-245, PL 6, fig. 1).
a cranidium from the Upper Ordovician Boda limestone of the
Siljan district, central Sweden, appears to represent the young-
est known species of Apianurus. Warburg also described a hypo-
stome (1925, pp. 253-254, PL 6, fig. 7) and an incomplete pygid-
ium (pp. 241-242, PL 6, fig. 9) from other localities in the Boda
limestone which may well represent this same species.

Genus CaLIPERNURUS Whittington, n.gen.
Type Species : Calipernurus insolitus Whittington n.gen., n.sp.
Diagnosis: Differs from Apianurus in that: (1) Cephalon is

272

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

relatively wider, six-sided rather than ovate in outline, and less
deep; (2) occipital ring with lateral regions more sharply set
off from swollen median portion, occipital spines diA'erge at 90°
or more and directed close above thorax ; ( 3 ) eye lobe farther f or-

-Figure 23. Triangular graph comparing relative dimensions of cranidia
of Apianuriis harhatus n.gen. n.sp. (shown by dots enclosed by solid line,
the dot in the lower right-hand corner has been aceidentaJly omitted —
compare Text-fig. 21.) and Calipepnurii.i insolitus n.gen., n.sp. (shown by
crosses enclosed by broken line). L, sagittal length of cranidium. W, width
between spines B on posterior Ijorder. E, length (exs.) from anterior mar-
gin to midpoint of palpebral lobe. Measurements were made in exterior
view, i.e. with isolated cranidium resting on a liorizontal surface.

ward, opposite first glabellar furrows; (4) pygidium with 4
lateral and one median border spine, posterior three of equal
size, major spines on pleural regions directed backward and

WHITTINGTON : SILICIFIED ODONTOPLEITRIDAE 273

slightly upward, curving inward in caliper-shape ; (5) ornament
of large tubercles rather than slim, thorn-like spines or small
tubercles.

Geological Range: jNIiddle Ordovician.

Discussion: Text-figure 23 illustrates the generally greater
width of cranidia of Calipernurus compared to those of Apia-
nurus, and the greater distance from the anterior margin back
to the palpebral lobe in Apianurus.

At least four species of Apianurua are here recognized, from
the Appalachian. Scandinavian and Baltic areas. All have the
less divergent occipital spines, and the associated pygidium has
pairs of border spines and the upright major spines. The Caliper-
nurus type of craniclium and p3^gidium seems to be known only
from Virginia, and is here regarded as representing a different
but evidently closely allied group, of generic rank.

CaXiIpernurus insolitus Whittington, u.gen., n.sp.

Plates 22-24 ; Text-figures 23, 24.
Holotype: USNM 124711 (PI. 22, figs. 1-3, 6; PL 23, figs. 1, 3).

Locality 3.
Other Material: Paratypes, USNM 124712 a-d; all figured ma-
terial in USNM.
Geological Horizon and Localities: Lower Edinburg limestone,
localities 2, 3.

Description: Cephalon wider than long, outline (ignoring
major spines) roughly six-sided; cranidium trapezoidal in out-
line, anterior margin less than half width of posterior. Glabella
widest at occipital ring, occipital furrow shallow medially, outer
part deep, diagonally directed outward and forward. Median
part of occipital ring considerably higher than lateral, and bear-
ing a short, stout median tubercle and the occipital spines ; latter
diverge at 90° or more (more in most larger specimens), proxi-
mal part straight and directed low over thorax, distal part
curved inward and tapering. Pronto-median glabellar lobe sub-
parallel-sided, gently convex posteriorly, moderately convex
anteriorly, sloping steeply down to shallow pregiabellar furrow.
Portion between basal lateral lobes merges with these lobes, and
is set off by faint transverse depression almost in line with pits
representing basal glabellar furrows; greatest convexity of

274

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

fronto-median lobe in front of this depression. First and second
glabellar lobes fused, kidney-shaped in outline and gently con-
vex, separated by change in slope from fixed cheeks, posterior
lobe slopes vertically to occipital furrow, second separated by
sharp change in slope from median. First and second glabellar

Figure 24. Calipernurus insoUtus u.gen., n.sp. Reconstruction, dorsal
view, number of thoracic segments unknown. Approximately X 10.

furrows represented by circular depressions, the first at the
inner margin of the fused lateral lobes, the second at the inner,
anterior corner of these lobes. Highest point of large eye lobe
on transverse line passing just behind first glabellar furrows
and level with crest of glabella in this line. Palpebral lobe slopes

WHITTINGTON: SILICIFIED ODONTOPLEURIDAE 275

steeply, broad rim passes into broad, convex eye ridge which runs
straight inward at 45°, passes close to antero-lateral margin of
fused lateral lobes and merges into frontal glabellar lobe. Eye
surface hemispherical (PI. 23, figs. 9, 10), outer surface almost
smooth, facets faintly visible on inner surface. Anterior branch
of suture runs on low sutural ridge straight forward and inward
to anterior margin, where it meets rostral and connective suture.
Posterior branch of suture runs back in line with anterior branch
across border, then curves inward across doublure. Anterior
border narrow (sag.) medially, becoming wider laterally and
merging with sutural ridges. Deep depression parallels eye ridge
on outer side, separating it from border and sutural ridge. Fixed
cheek inside eye lobe slopes gently inward to margin of fused
lateral glabellar lobes, inside and behind eye lobe it slopes
steeply, almost vertically, dow^n to occipital ring and posterior
border furrow. Latter commences at lateral margin of occipital
ring and runs forward and outward, so that convex border is
widest at suture. Free cheek narrow, librigenal spine arises in
front of eye lobe and curves outward and backward, the distal
part directed straight back. Lateral border gently convex,
defined by shallow border furrow, and runs from posterior
branch suture to merge with swollen base of librigenal spine.
No antero-lateral border is defined, though there is a depression
between the sutural ridge and sw^ollen base of the librigenal
spine. Shallow antennal notch in vertical border of free cheek
adjacent to anterior suture (PI. 23, fig. 1). Doublure of free
cheek widest behind notch, narrowing back and ending at
posterior suture. Posterior border without doublure, articulating
socket just inside suture. On inner surface of cranidium (PI.
23, fig. 4) doublure of occipital ring and ridges, made by outer
parts occipital furrow, are seen. First, and especially second
glabellar furrows, and depression outside eye ridge, also project
ventrally. Rostrum unknown, but evidently (PI. 23, fig. 1)
broad (tr.) and short (sag. and exs.), lateral margins converg-
ing backward, fitting so that outer surface faces almost directly
downward. Two specimens only are known of the hypostome
(PI. 23, figs. 11-15). Anterior margin projects forward in front
of convex middle body. Latter divided by shallow middle fur-
rows which run inward and backward to about three-quarters

276 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

of the length and do not reach the midline. Posterior lobe
crescentic, convex, especially toward the tips. Lateral borders
horizontal, shallow border furrows, shallow lateral notch and
sharply-pointed shoulder. Posterior border little wider than
lateral, bent to slope steeply dorsally, border furrow shallow.
Doublure widest at shoulder, where it is crossed by a sharp ridge,
narrowest posteriorly. Tiny perforation just behind shoulder
ridge (PI. 23, fig. 15). The smaller specimen (PI. 23, figs. 13, 14)
has short, triangular anterior wings, but these are absent in the
larger specimen (PL 23, figs. 11, 12, 15) and both antero-lateral
corners are cut off. This truncation seems not to be accidental
breakage, but is symmetrical. This hypostome differs from that
of Apianurus (PI. 18, figs. 1-5) in the much greater inflation of
the middle body posteriorly, less distinct middle furrows, inclina-
tion of posterior border, smaller lateral notch, smaller opening
in doublure, and ornament. It is accordingly regarded as prob-
ably belonging to Calipernurus.

External surface of cephalon with scattered tubercles, area
between them, and all furrows, smooth. Largest is the median
occipital tubercle, in the rounded top of which are four tiny
depressions set at the corners of a square (PI. 23, figs. 7, 8).
Next largest in size are various symmetrically placed tubercles
— pairs on the fronto-median and lateral glabellar lobes, pair at
antero-lateral corners of cranidium, pair on outer edge of eye-
ridge at midlength, three just inside eye on fixed cheek, and one
about at midpoint, one at extremity, of posterior border. Smaller
tubercles are scattered, some symmetrically, on the cranidium
and free cheeks, including base of librigenal spines. On the
borders of the free cheek and proximal part of the librigenal
spine are thorn-like spines, constant in number, position and
direction. On the librigenal and occipital spines are short spines
directed distally at a low angle to the axis of the spine, and
becoming longer distally. There are openings at the base of these
spines on the distal side (PI. 23, fig. 6), but tubercles and spines
elsewhere on the cranidium seem to be imperforate, and the
bases of the tiny depressions in the occipital tubercle seem to
be closed. Hypostome with tubercles on lateral borders only.

FcAV thoracic segments known, and number in thorax un-
known: reconstruction (Text-fig. 24) assumes number was 10.

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 277

One seofinent from anterior part of thorax (PI. 2-4, figs. 15, 16.
19) shows relatively wide, convex axis and horizontal, iinfnr-
rowed pleurae. Extremity of articulating furrow deep, articulat-
ing half ring as long (sag.) as axial. Pleural spines curving out
and down, directed slightly forward. Doublure rolled under at
base spine, and articulating process on anterior edge, socket on
posterior edge. Smooth liand runs across highest part of axial
ring and pleurae into base pleural spine. Tubercles outside this
band, some larger ones on posterior edge axial ring and on slope
of articulating furrow paired. Lateral margins of pleural spine
with row of close-spaced, curved spines. Two incomplete seg-
ments from the posterior part of the thorax, and an incomplete
segment, possibly the posterior (PI. 24, figs. 22, 23, 26), have
the pleural spine curved like that of the major pygidial pleural
spine. Smooth median band, and larger tubercles on posterior
edge of axial ring and flanges of pleura are rounded like those of
the cephalon, some on the axis paired. These segments are
placed in Calipernurus because smooth band resembles that on
occipital ring and posterior border (PL 23, fig. 5), deep extremi-
ties of articulating furrow resemble those of occipital furrow,
and because pleural spines and bands on them are like those
of major pleural spine of the pygidium.

Pygidium (PI. 24, figs. 25, 27, 28) more than twice as wide
as long, axis of same width as pleural region at anterior margin.
First axial ring prominent, in front of it shallow articulating
furrow and long (sag.) articulating half -ring. Posterior edge of
axial ring descends vertically to gently convex posterior part of
axis, which merges into pleural regions except antero-laterallj',
where shallow depressions occur. First axial ring connected
across pleural lobes hy low jileural ridge to base of major pleural
spine, which arises inside margin and curves upward and in-
ward. Rest of pleural regions flat, margin rolled, no border
furrow. From vertical margin of border, four pairs and one
median posterior spine arise — a pair of similar size flanking
the median spine, a short pair just inside and beneath the major
spine, and two pairs in front, the anterior tiny. Two specimens
(PI. 24, figs. 11, 18) lack the median border spine, and one
(PL 24, fig. 17) has the median bifurcate at the tip, but lacks
the pair flanking it. Such variation is unusual in this material.

278 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Doublure sharply bent up, widest posteriorly, appendifers not
developed, but areas of muscle attachment (shown by thinness
of c^uartz) include extremities of articulating furrow, and cor-
responding areas behind first ring. In some species the entire
area of the axis behind the ring is of thinner quartz, suggesting
that it is all an area of muscle attachment. Ornament of rounded
tubercles, paired and median on axial ring, pair at center of
rest of axis, row of two or three at tip. Other tubercles on
anterior region of pleural regions and base of major pleural
spines. Latter with row of curved, thorn-like spines along lateral
margins, also tubercles and distally-directed sharp tubercles,
near tip. Latter have openings at base on distal side (PI. 24,
figs. 21, 24). Other border spines with short spines scattered on
them, not in regular rows, directed distally at tips and with
openings.

Development. Cranidium. Smallest known cranidium (PI.
24, fig. 1) of length (sag.) 0.78 mm. Occipital spines diverging
at about 90°. Glabella convex, parallel-sided, occipital ring prom-
inent, with median tubercle and paired spines ; f ronto-median
lobe with 5 pairs of axial spines — 2a, 2, 3, 4, and 5. Extra spine
on left side between 3 and 4, median spine between 4 and 5.
Median lobe between spines 2a and 2 with slight extra inflation.
Basal glabellar lobe small, gently inflated ovate area low on side
of f ronto-median lobe, one median spine ; second glabellar lobe
smaller and fainter. Palpebral lobe rsituated in front of half
length of cranidium, strongly raised, low eye ridge runs forward
and inward. Narrow anterior border; posterior border widens
(exs.) outward. Most prominent spines on fixed cheek include
Ai, Ao, A3, B, C, PI, Er, and one of similar size at extremity of
anterior border. Larger cranidium 1.08 mm. in length (sag.)
(PI. 24, figs. 5-7) has similar glabella, but the lateral lobes are
larger, more inflated, separated by the subcircular pit of the
first lateral furrow. Palpebral lobe larger, higher, slightly
farther back, eye ridge more prominent, and steeper slope of
cheek behind palpebral lobe. Free cheek is like larger ones,
except eye surface is farther forward.

With further increase in size (PI. 24, figs. 2-4, 9) the main
changes are widening and elevation of the lateral glabellar lobes
so that they become fused outside the first lateral furrow, the

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE 279

axial furrows stand higher, and thus the fixed cheeks slope less
steeply inward. The eye lobe moves back, and the slope behind
it becomes steep. Additional spines appear, and the larger ones
are reduced to high tubercles. In cranidia of length (sag.)
1 to 2 mm. some 40 per cent of the specimens have the occipital
spines diverging at less than 90°, whereas the remainder and
all larger specimens have them diverging at 90° or more.

The close similarity between the development of the cranidium
of Calipernurus and that of Apianurus is evident.

Pxjgidium: A transitory pygidium (PI. 24, fig. 12) of Stage
holaspis-1 has the first segment marked off by interpleural
grooves, and the long pleural spines curve inward and back-
ward. The pleural spines of the second segment are only slightlj^
curved, and more inwardly directed. The posterior border bears
three border spines. Axis of 3 rings, anterior with 2 pairs and a
median spine, second two with one pair each. This specimen
shows that the major pleural spines of the true pygidium corre-
spond to the pleural spines of the thorax. Smaller transitory
pygidia have not been recognized, but they would evidently be
extremely similar to some of those here regarded as representing
Apianurus harhatus, n.gen., n.sp. (PI. 19, figs. 24, 25, 27-29).
Small true pygidia (PL 24, figs. 8, 13, 14) have the additional
pairs of border spines, though the anterior is extremely small.

Subfamily SELENOPELTINAE Corda, 1847

(^Selenopeltides Corda, 1847, p. 33, and Selenopeltidae

Prantl and Pribyl, 1949, p. 172)

Genus SelEXOPELTIS Corda, 1847

Text-figure 25.

Synonym : Polyeres Kouault, 1847, type species by monotypy
P. dufrenoyi Rouault, 1847 ; see Clarke, 1892, p. 96 ; Prantl
and Pinbyl, 1949, pp. 173-175.

Diagnosis: Cephalon transverse^ subrectangular in outline.
Glabella gently convex, tapering slightly forward, occipital ring
short (sag. and exs.), low median tubercle; wide fronto-median
lobe, convex band across base ; apparently three lateral glabellar
lobes, defined by faint axial furrow, shallow first lateral and

280 BULLETIN' : MUSEUM OF COMPARATIVE ZOOLOGY

deeper second lateral furrow, and by longitudinal furrow joininsr
first lateral to occipital furrow, anterior part of this longitudinal
furrow deep ; basal lobe divided subequally by faint transverse
furrow and posterior portion subdivided by longitudinal fur-
row. Inner corner of clieek inflated, merging into antero-lateral
part of occipital ring. Crescentic eye lobe at about half length
(exs.) of cheek and in inner part. Librigenal spine stout, long,
directed upward and outward, no spines on anterior and lateral
borders of cheek. Hypostome subrectangular in outline, wider
than long, shallow median posterior notch, wide postero-lateral
border ; middle furrow broad, shallow, running in from antero-
lateral corner of middle body. Thorax of 9 segments, axial rings
with prominent lateral lobes, horizontal pleurae with ridge run-
ning in curve convex forward which distally runs out into long
posterior pleural spine ; anterior pleural spine downwardly and
outwardly directed, curved. Pygidium with short axis and one
pair border spines only, connected to first ring by prominent
pleural ridge. External surface tuberculate or granulate.
Geological Range: Lower to Upper Ordovician.

Discussion: The peculiarities of this genus have long been
recognized, and include the shortness of the occipital ring, the
partial fusion of the basal and median glabellar lobes and sub-
division of the former, the conspicuous lateral lobes of the thoracic
axial rings, the forward curve of the main ridge of the pleura,
and the lack of spines along the borders of the free cheek and
pygidium (excepting the major pair). The anterior pleural
spine was figured by Barrande (1852, PI. 36, fig. 6; PI. 37, fig.
25), and is curved, projecting downward and outward below the
posterior pleural spine of the preceding segment.

Selenopeltis is here regarded as belonging within a separate
subfamily (rather than family). 1 agree with Prantl and Pi'ibyl,
(1949, p. 173) than Selenopeltis has affinities with the Mira-

Figure 25. Selenopeltis hucJii (Barrande), Middle Ordovician, Bohemia.
Approximately X %. A, cephalon, anterior view, based on MCZ 4317,
Chlustina Beds, de2b, "Brdatka" near Beraun. B, hypostome, e.xterior view,
based on MCZ 4316, Drabov quartzites, d5, Drabov. C, exoskeleton in dorsal
view, based on MCZ 4319, Sarka Shales, d7i, Osek, MCZ 4316, MCZ 4317,
etc.

WHITTINGTON : SILICIFIED ODONTOPLEURIDAE

281

Figure 25

282 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

spinae rather than with other subfamilies, particularly in the
general form of the cephalon, and notably the hypostome. But
Selenopeltis shares many odontopleurid characters with other
genera, and some of its peculiar features appear in these other
genera. For example, Dicranurus also lacks spines on the lateral
cephalic border; the curve of the pleural ridge in Selenopeltis
is an accentuation of the same curve seen in such genera as
Miraspis and Dicranurus; fusion of median and basal glabellar
lobes occurs in Apianurus, n.gen.

Selenopeltis is well known from the Llanvirn to Ashgill of
Bohemia, has been recorded in France, and more recently in
Morocco (Termier and Termier. 1950, PI. 194, figs. 1-4), and
Shropshire (Whittard, 1952, p. 158), the latter in rocks of
Arenig age.

Other Genera, Subgenera and Species Sometimes Referred

to Odontopleuridae

Acidaspis ulrichi Bassler, 1919 (pp. 355-356, PI. 37, figs. 6-8)
is part of an Upper Cambrian trilobite (Wilson, 1952, p. 317),
presumably the free cheek, and of unknown affinities.

Acidiphorns Raymond, 1925. Not an odontopleurid, probably
a bathyurid (Whittington, 1953, p. 669).

Anc]jropyge Clarke, 1892. Based on a pygidium, recently re-
described (Stumm, 1953, p. 126, PI. 6, figs. 1, 2), and may be an
odontopleurid.

Bounyongia Etheridge and Mitchell, 1917, type species by
monotypy B. hoivningensis Etheridge and Mitchell, 1917. Based
on two poorly preserved specimens of the cephalon, one with a
few thoracic segments attached, recently said by Gill (1948,
p. 18) to be a subgenus which "must now lapse," since the
character upon which it was founded — a pair of cephalic spines
arising from the glabella — is a misinterpretation, the spines
being occipital. Prantl and Pfibyl (1949, p. 181) regarded
Bounyongia as a synonym of Ceratocephala.

Glaphurus Raymond, 1905, and Glaphurina Ulrich, 1930. In
1913 Raymond (p. 723) placed the former genus in the Odonto-
pleuridae, but later (1916, p. 138) thought it should be excluded.
Hupe (1953, p. 229) comments on the apparent odontopleurid

WHITTIXGTON : SILICIFIED ODOXTOPLEURIDAE 283

characters. I consider the resemblance superficial, and agree with
Hupe that these two genera may he placed in a separate family,
allied to Telephidae (Hupe, 1953. pp. 228-230).

Glohidaspis Reed, 1931, type species Acidaspis (Glohulaspis)
prommens Reed, 1931 (pp. 100-101, PI. 5, figs. 5, 5a, 5b), from
the Lower Silurian of southern Scotland. The holotype, the
internal mould of an incomplete cranidium, appears to be unique.
I do not think it shows much resemblance to Whittingtonia, as do
Prantl and Pi^-ibyl (1949, pp. 133-134). and am not sure that
it is an odontopleurid.

Pharostoma Corda, 1847, and Ptychometopus Schmidt, 1894.
have sometimes been excluded with question from the Calymeni-
dae, and comments have been made regarding their odontopleu-
rid-like appearance (cf. Shirley, 1936, pp. 385-386; Opik, 1937.
p. 24). I consider this resemblance probably superficial, and that
these genera belong with the calymenids (ef. Hupe, 1953, p. 232 i.

EEFEEEXGES

Additional references on Odontopleuridae are contained in Prantl and
Pribyl (1949), and Whittington, 1956b.

Barrande, J.

1846. Xoticc pn'iliminaire sur le Systeme Silurien et les Trilobites de

Boheme. Leipzig.
1852. Systeme Silurien du centre de la Bohgme, Vol. I, Prague and

Paris.
1872. Systeme Silurien du centre de la Bohenic. Supplement to vol. I.
Prague and Paris.
Bassler, E. S.

1919. Cambrian and Ordovician deposits of Maryland. Md. Geol.
Survey, Baltimore, pp. 1-424, pis. 1-58, 27 figs.
DE Beee, G.E.

1951. Embryos and ancestors. Eev. ed., Oxford.
BOHLIN, B.

1949. The Asaph)(.<i limestone in northernmost Oland. Bull. Geol. Inst.
Upsala, vol. 33, pp. 529-570, pis. 1, 2, 10 figs.
Clarke, J. M.

1892. Notes on the genus Acidaspis. X. Y. State *Mus., 44th Ann. Eep.,
pp. 91-101. pis. 1-3.
Conrad, T. A.

1840. Third Annual Eeport, Palaeontological Department of Geologi-
cal Survey of Xew York, Albany. Assembly Xo. 50, pp. 199-207.

284 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Cooper, B. N.

1953. Trilobites from the Lower Champlainian formations of the
Appalachian Valley. Geol. Soc. Am., Mem. 55, pp. 1-69, pis.
1-19, 2 figs., table.

Cooper, G. A., and B. N. Cooper

1946. Lower Middle Ordovician stratigraphy of the Shenandoah Val
ley, Virginia. Gcol. Soc. Am., Bull., vol. 57, pp. 35-114, pis. 1-3,
9 figs.

CoRDA, A. J. C.

1847. In I. Hawle and A. J. C. Corda: Prodrom einer Monographie
der Bohmischen Trilobiten. Prague, pp. 1-176, pis. 1-7.

ElIMRICH, F.

1839. De Trilobitis, Dissertatio Petrefactologica. Berolini.
Erben, H. K.

1952a. Zur Gliederuug der Ceratocephalidae Eud. und E. Richter, 1925,

emend. Prantl und Pribyl, 1949 (Tril.). Neues Jb. Geol.

Palaeont., Mh., vol. 7, pp. 304 317, 3 figs.
1952b. Trilobiten aus dem Alteren Hereyn (Unterdevon) des Unter-

liarzes. Neues ih. Geol. Palaeont., vol. 94, pp. 150-362, pis.

17-24, 64 figs.

Etheridge, E., and J. Mitchell

1917. The Silurian trilobites of New South Wales, with references
to those of other parts of Australia. Proc. T-inn. Soc. New
South Wales, vol. 42, pp. 480-510, pis. 24-27.

EviTT, W. E.

1951. Some Middle Ordovician trilobites uf the families Cheiruridao,
Harpidae and Lichidae. Jour. Paieont., vol. 25, pp. 587-616, pis.
85-88, 1 fig.
1953. Observations on the trilotiite Ciraunis. Jour. Paieont., vol. 27,
pp. 33-48, pis. 5-8, 1 fig.

AND H. B. Whittington

1953. The exoskeleton of Flcxical.ijiiiciK (Tiiloliitn ). Jmir. I'aleoiit.,
vol. 27, pp. 49-55, pis. 9, 10, 1 fig.

Gill, E. D.

1948. A new trilobite from the Yeringian (Lower Devonian) rocks of
Kinglake, Victoria. Proc. Royal Soc. Victoria, vol. 59, (N. S.),
part 1. pp. 8-19. pi. 3.

Hanstrom, B.

1934. Uber das Vorkommen eiues Naekenschildes und eines vierzelli-
gen Sinnesorganes bei der Trilobiten. Lunds Univ. Arsskrift,
N.F., Avd. 2, vol. 30, no. 7, pp. 1-12.

WHITTINQTON : SILICIFIED ODOXTOPLBURIDAE 285

HiNTZE, L. F.

1953. Lower Ordovician trilobites from western Utah and eastern
Nevada. Utah Geol. and Mineral. Riirvey. Bull. 48, pp. 1 240.
pis. 1-28, figs., tables.
HUPE, P.

1953. Classe des TrDobites in: Traite de Paleontologie, vol. 3, pp.

44-246, Ed. J. Piveteau. Paris.
1954-5. Classification des Trilobites. Ann. Paleontol., vol. 39, pp. 1 110,
figs. 1-92; vol. 41, pp. 111-345, figs. 93-247.
Lake, P.

1896. The British Silurian species of Acidaspis. Geol. Soc. London,
Quart. Jour., vol. 52, pp. 235-245, pis. 7, 8.

LiNNABSSON, J. G. O.

1869. Om Yestergotlands Cambriska oeh Siluriska aflagringar. K.
Svenska Vet.-Akad. Handl., N.F., vol. 8, no. 2, pp. 1-89, pis.
1, 2.
Locke, J.

1843. Notice of a new trilobite, Ceraurvs crosotus. Am. .Jour. Sci., vol.
44, p. 346.
MURCHISON, E. I.

1839. The Silurian system. Parts I and II. London.
Opik, a.

102fi. T'ber die Knlksandsteinfacies des Vaginatenkalkes auf der
Halbinsel Baltischport und iiber ein Acidaspis-Fy gidium au?
denselben Schichten. Pub. Geol. Inst. Univ. Tartu, no. 4, pp.
1-7, 2 figs.
1937. Trilobiten aus Estland. Pub. Geol. Inst. Univ. Tartu, no. 52,
pp. 1-163, pis. 1-27.
Prantl, F., and a. Pribyl

1949. A study of the superfamily Odontopleuraeea nov. superfani.
(trilobites). Eozpravy, sv. 12, Stiit. Geol. Ust. Ceskoslovenske
Eep., Pralia, pp. 1-221 (Czech and English texts, Eussian sum-
mary), pis. 1-11, 1 fig.
Easetti, F.

1948. Lower Cambrian trilol)ites from the conglomerates of Quebec.
Jour. Paleont., vol. 22, no. 1, pp. 1-24, pis. 16.
Eaw, F.

1025. The development of Lt ptopJastus xulteri and other trilobites.
Geol. Soc. Londcm, Quart. Jour., vol. 81, pt. 2, pp. 223-324,
pis. 15-18.
Eaymond, P. E.

1913. Trilobita in Text-book of Paleontology. Eastman-Zittel, 2nd ed.,
vol. 1, pp. 692-729. London.

286 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

1916. The genera of the Odontopleuridae. Ottawa Natur., vol. 29, pp.
135-139.

1925. Some trilobites of the Lower Middle Ordovician of east-ern
North America. Miis. Comp. Zool.. Biil!., vol. fi7. no. 1, pp.
1-180, pis. 1-10.
Reed, F. R. C.

1925. The classification of the Acidaspidae. Geol. Mag., vol. 62. no.
735, pp. 416-430.

1931. Additional new trilobites from Girvan. Ann. Mag. Nat. Hist.,
ser. 10, vol. 7, pp. 97-104, pis. 4 and 5.

RiCHTEB, R.

1932. Crustacea in: Handworterbuch der Xaturwissenschaften, 2nd
ed. Jena, vol. 2, pp. 840-864, figs. A, 1-65.

and E. Richter

1917. Uber die Einteilung der Familie Acidaspidae und iiber einige
ihrer devonischcn Vertreter. Centralbl. f. Min., Jahrg. 1917,
pp. 462-472. 10 figs.

1926. Die Trilobiten des Oberdevous. Abh. Preuss. Geol. Landesanst.,
X.F., H. 99, pp. 1-314, pis. 1-12, 18 figs.

1930. Xeue Horner-Trilobiten in unserer Schausammlung. Xatur

und Mus., vol. 60, pp. 31-45, 10 figs.
1952. Die Typen von Proetus tenuimargo und Leonaspis aries. Sencken-

bergiana, vol. 33, pp. 109-113, pi. 4. figs. 25-27.
Ross, R. J.

1951. Stratigraphy of the Garden City formation in northeastern

Utah, and its trilobite faunas. Peabody Mus. Xatur. Hist., Yale

Univ., Bull. 6, pp. 1-161, pis. 1-36.
Schmidt, F.

1885. Revision der Ostbaltischen Silurischen Trilobiten. Abt. 2, Mem.

Acad. Imp. Sci. St.-Petersbourg, Ser. 7, vol. 33, no. 1, pp.

1-124, pis. 1-6.
1907. Revision der Ostbaltischen Silurischen Trilobiten. Abt. 6. Mem.

Acad. Imp. Sci. St.-Petersbourg, Ser. 8, vol. 20, no. 8, pp. 1-104,

pis. 1-3.
Shirley, J.

1936. Some British trilobites of the family Cah-menidae. Geol. Soc.

London, Quart. Jour., vol. 92, pp. 384-422, pis. 29-31, 4 figs.

ST£fRiIER, L.

1930. Scandinavian Trinucleidae with special references to Xorwegiau
species and varieties. Vid.-Akad. Oslo, Skr. T, Mat.-Xaturv.
Klasse. no. 4. pp. 1-111, pis. 1-14, 47 figs.

WHITTINQTOX : SILICIFIED ODOXTOPLEURIDAE 287

1942. Studies on trilobite morphology, Pt. II. The larval develop-
ment, the segmentation and the sutures, and their ])earing on
trilobite elassitication. Norsk. Geol. Tide., vol. 21, pp. 49-164,
pJs. 1, 2, 19 figs.
1953. The Middle Ordovician of the Oslo region, Norway. Norsk Geol.
Tids.. vol. 31, pp. 37-141. pis. 1-6, 16 figs.
Stubblefield, C, J.

1936. Cephalic sutures and their bearing on current classifications of
trilobites. Biol. Eev., vol. 11, pp. 407-440, 9 figs.
Stumm, E.

1953. Trilobites of the Devonian Traverse group of Michigan. Contrib.
Mus. Paleont., Univ. :Mich., vol. 10, no. 6. pp. 101-157, pis. 1-12.

Temple, J. T.

1952. The ontogeny of the trilobite iJalmanitina oliiii. Geol. Mag.,
vol. 89, no. 4, pp. 231-262, pis. 9 and 10, 1 fig.
Termier, H., and G. Termier

1950. Paleontologie Marocaine, vol. 2. Actual. Sci. Industr. no. 1095.
Paris.
Thorsluxd, p.

1940. On the Chasmops series of Jemtland and Sodermanland
(Tvaren). Sveriges Geol. Under., ser. C, no. 436, pp. 1-191,
pis. 1-12, 56 figs., map.
Tripp, E. P.

1954. Caradocian trilobites from mudstones at Craighead Quarry,
near Girvan, Ayrshire. Eoyal Soc. Edinburgh, Trans., vol. 62,
part 3 (no. 16), pp. 655-693, pis. 1-4.

Ttvexhofel, W. H., et al.

1954. Correlation of the Ordovician formations of North America.
Geol. Soc. Am., Bull., vol. 65, pp. 247-298. 1 pi., 2 figs.
Warburg, E.

1925. The trilobites of the Leptaeua limestone in Dalarne. Geol.

Instit. Upsala, Bull., vol. 17, pp. 1-446, pis. 1-11, 23 figs.
1933. On the structure of the occipital ring of the Odontopleuridae.
Ark. f. Zool., vol. 25A, no. 9, pp. 1-19, 4 figs.
Warder, J. A.

1838. New trilobites. Am. Jour. Sci., vol. 34, pp. 377-380, 3 figs.
Weller, S.

1907. The paleontology of the Niagaran limestone in the Chicago area.
The Trilobita. Chicago Acad. Sci., Bull. 4, part 2, pp. 166-281,
pis. 16-25.
Whittard, W. r.

1952. A geology of South Shropshire. Geol. Assn., Proc, vol. 63, part
2, pp. 143-197, tables 1-6.

288 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Whittington, H. B.

1941. Silicified Trenton trilobitcs. Jour. Paleout., vol. 15, pp. 492-522,

pis. 72-75, 13 figs.
1950. British trilobites of the family Harpidae. Paleontogr. See,
London, pp. 1-55, pis. 1-7, figs. 1-16.

1953. North American Bathyuridae and Leiostegiidae (Trilobita).
Jour. Paleont., vol. 27, no. 5, pp. 647-678, pis. 65-69, 5 figs.

1954. Status of invertebrate paleontology. Part VI. Arthropoda:
Trilobita. Mus. Comp. ZooL, Bull., vol. 112, no. 3, pp. 193-200,
2 figs.

1956a. Beecher 's supposed odontopleurid protaspis is a phacopid. Jour.

Paleont., vol. 30, pp. 104-109, pi. 24, 2 figs.
1956b. Type and other species of Odontopleuridae (Trilobita). Jour.
Paleont., vol. 30, in press.

and W. E. Evitt

1954. Silicified Middle Ordovician trilobites. Geol. Soc. Am., Mem.
59, pp. 1-137, pis. 1-33, 27 figs.
■ and A. Williams

1955. The fauna of the Derfel limestone of the Arenig District, North
Wales. Eoyal Soc. London, Phil. Trans., ser. B, no. 658. vol.
238, pp. 397-430, pis. 38-40, 6 figs.
Wilson, J. L.

1952. Upper Cambrian stratigraphy in the central Appalachians.
Geol. Soc. Am., Bull., vol. 63, pp. 275-322. pis. 1-4, 4 figs.

EXPLANATION OF PLATES 1-24.

To make the photographs, the larger specimens were mounted
on insect pins with gum arabic, the smaller specimens laid on a
black surface. A light coating of ammonium chloride was used.
Stereoscopic pairs of views were made by tilting the specimen,
and only the left-hand view of a pair is numbered. The heads
and shafts of the pins have been blacked out ; otherwise the photo-
graphs have not been retouched. It is arbitrarily decided that
the plane passing through the margin of the occipital or axial
ring shall run in the dorso-ventral direction, and views are de-
scribed accordingly. Exterior or interior views have been taken
in a direction lying in the sagittal plane to give the fullest possi-
ble view of the exoskeletal surface. Oblique exterior or interior
views have been taken in directions at an angle to the sagittal
plane. The initials USNM refer to the U. S. National Museum,
Washington, D. C, and the initials PMO to the Museum of
the Paleontological Institute, Oslo, Norway. These initials are
followed by the catalogue numbers.

Plate 1

Primaspis ascitus n.sp.
Lower Martinsburg shale, locality 10

Figure

1,2,5. Holotype cranidium, paratype free cheek: dorsal aud antero-

lateral stereographs, anterior view, X 6. USNM 116515,
116516a.

3,7,9,10. Paratype pygidiuni: posterior, left lateral views; dorsal,
ventral stereographs, X 6. USNM 116516e.

4. Paratype free cheek: oblique interior view, X 9. USNM

116516a.

6. Holotype cranidium: interior stereograph, X 6. USNM

116515.

8,11,12,13. Paratj-pe segment: right lateral view; dorsal stereograph;
anterior, posterior views, X 4. USNM 116516c.

14. Paratype thoracic segment: dorsal stereograph, X 4. USNM
116516d.

15. 17. Incomplete posterior thoracic segment; interior, exterior views,

X9.

16. Small pygidium: dorsal view, X 15.

18, 19, 20, 21. Paratype hypostome: posterior, anterior views; exterior, in-
terior stereographs, X 6. USNM 116516b.

PLATE 1

PLATE 2

Figure

Plate 2

Primaspis ascitus n.sp.
Lower Martins);uig shale, locality 10

1, 2. Small ci-anidium : interior, exterior views, X 220. Original of

Text-figure 6A.

3, 4. Small hypostome : exterior, interior views, respectively X 30,

X 20.

5, 6. Small cranidium : anterior, exterior views, X 20. Original of

Text-figure 6B.

7, 8. Small cranidium : anterior, exterior views, X 20.

9. Transitory pygidium : dorsal view, X 30.

10. Tip of posterior pleural spine showing openings between
tubercles, X 30.

11-12, 23. Small cranidium : anterior view, X 15 ; dorsal views, respec-
tively X 15, X 30.

13-14. Small cranidium: anterior, dorsal views, X 10.

15. Small pygidium : dorsal view, X 15.

16-17. Small cranidium: anterior, dorsal views, X 10.

18,22. Paratype free cheek : ol)lique exterior, oblique interior views

of eye surface, X 30. USNM 116516a.

19, JO. Small free cheek: oblique interior, oblique exterior views, X 20.

Original of text-figure 6B.

21. Holotype cranidium: basal portion of glabella showing external

surface, X 20.

Figtire

Plate 3

Biacanihaspis cooperi Whittington
Lower Martinsburg shale

1,2,5,6. Protaspis: exterior, interior stereographs; postero-dorsal, an-

tero-dorsal views, X 50. Locality 12. Original of Text-figure
9A, B.

3-4. Stage 0 exoskeleton: exterior, interior, stereographs, X 50.

Locality 12. Original of Text-figure 9C.

7,9-10. Ilypostome: exterior stereograph; interior, oblique interior

views, X 10. Locality 10.

8. Small eranidium and free cheek: exterior stereograph, X 20.

Locality 10. Original of Text-figure 9D.

11. Anterior thoracic segment: dorsal view, X 6. Locality 10.

12. Median thoracic segment: dorsal view, X 6. Locality 10.

13. Small hypostome: exterior view, X 30. Locality 10.

14. Small hypostome: ol>lique interior view showing hole in
doublure, X 30. Locality 10.

15-16. Incomplete posterior thoracic segment: dorsal, left lateral

views, X 6. Locality 10.

17-18. Stage 8 transitory pygidium : exterior, interior views, X 30.

Locality 12.

19. Ilolaspid pygidium: dorsal stereograph, X 4%. Locality 12.

20. Tip of genal spine : oblique exterior view, X 30. Locality 9.

21. Eye surface: oblique exterior view, X 30. Locality 9.

PLATE 3

PLATE 4

•''iin^

»»*'•

Plate 4

Diacnnthaspis hpidu.s n.sp.
Lower Edinburg limestone

Figtire

1,2,3,4,5. Incoiuplete protaspis : exterior stereograph, anterior view, left
lateral view, postero-dorsal stereograph, antero-dorsal stereo-
graph, X 33.6 Photograph by W. E. Evitt. Locality 4.

6. Small eranidimu: exterior stereograph, X 30. Loeality 3.

7. Small cranidium: exterior stereograph, X 30. Locality 3.

8. 9, 10. Small cranidium : dorsal, anterior, right lateral views, X 15.

Locality 3.

Small pygidium : dorsal \'iew, X 20. Locality 2.

Small cranidium: dorsal, anterior, left lateral views, X 15.
Locality 2.

Transitory pygidium: duisal vicAV, X 30. Locality 4.

Small cranidium: dorsal, anterior, left lateral views, X 15.
Locality 2.

Cranidium: exterior stereograph, anterior view, X G. Locality 2.

Cranidium: exterior stereograph, X 6. Locality 2.

Posterior part of cranidium: exterior view, X 30. Locality 2.
Arrow points to spine witli truncated tip and tiny depressions
or openings.

11.

12,

13,

14.

15.

10.
19,

17,
21.

18.

20.

Figure

Plate 5

Diacanthaspis lepidus n.sp.
Lower Edinburg limestone

1,7,9. Holotype cranidium and paratype free cheek: dorsal stereo-

graph; anterior, oblique exterior views, X 10. USNM 116517,
116518a. Locality 2.

2.3,0,6. Incomplete thoracic segment: exterior, anterior, posterior, in-

terior views, X 12.5. Locality 4.

i. Holotype cranidium: interior view, X 15. USXM 116517.

Locality 2.

8,13,1-4. Pygidium: interior view, X 6; posterior and right lateral

views, X 4. Locality 2.

10,11. Part of thorax and pygidium, paratype: dorsal and ventral

stereographs, X 6. USNM 116518c. Locality 3.

12. Pygidiiun : dorsal stereograph, X 9. Locality 2.

15, 16, 17, 18. Paratype hj^ostome: interior stereograph, X 20; exterior
stereograph posterior and anterior views, X 15. USNM
116518b. Locality 3.

PLATE 5

PLATE 6

Plate 6

BiacantlMspis secretus n.sp.
Lower Edinburg limestone

Figure

1,2,3,4,5. Incomplete protaspis: exterior stereograph; anterior, left
lateral, antero-dorsal, posteero-dorsal views, X 33.6. Photo-
graphs by W. E. Evitt. Locality 4. Original of Text-figure
llA.

6. Stage 0 eranidium and free cheek: exterior stereograph, X 30.
Locality 4. Original of Text figure IIB.

7, Same eranidium as figure 6.: interior view, X 30. Locality 4.

8,11. Small eranidium: exterior stereograph; interior view, X 30.

Locality 4. Original of Text-figure IIC.

9. Small free cheek: oblique exterior view, X 30. Locality 4.
Original of Text-figure IIC.

10. Stage 0 pygidium : dorsal ^•iew, X 30. Locality 4. Original of
Text-figure IIB.

12. Small eranidium and free cheek: exterior stereograph, X 30.
Locality 3. Original of Text-figure IID.

13. Transitory pygidium: dorsal view, X 30. Locality 4.

14. Small eranidium and free cheek: exterior stereograph, X 25.
Locality 2. Original of Text-figure HE.

15. Small eranidium and free cheek: exterior stereograph, X 15.
Locality 2. Original of Text-figure IIF.

16,17,18. Small eranidium: dorsal, right lateral, anterior views, X 9.
Locality 2.

19. Same eranidium as figure 14: interior view. X 20. Locality 2.

20. Transitory pygidium : dorsal view, X 30. Locality 4.

21,22. ParatjT)e free cheek: oblique interior views, X 15. L^SX^^M

116520a.. Locality 2.

Figure

PlxiTE 7

Diacanthaspis secretus n.sp.
Lower Edinburg limestone

1,3,5,11. Holotype cranidium : dorsal and anterior views, X 9; interior
view, X 6; left lateral view, X 9. USNM 116519. Locality 2.

2, 7. Paratype free eheek : dorsal and oblique exterior views, X 10.

USNM 116520a. Locality 2.

4. Paratype thoracic segment: doi'sal view, X 6. USNM 116520b.

Locality 2.

6. Paratype, incomplete thoracic segment : dorsal view, X 6.

USNM 116520b. Locality 2.

8. Pygidium : interior view, X 9. Loe^ity 2.

9. Pygidium: dorsal view, X 9. Locality 4.

10. Paratype pygidium: dorsal stereograph, X 9. USNM 116520c.
Locality 2.

12. Pygidium, exterior view, tips of border spines shoAving open

ings, X 30. Locality 2.

13, 14. Free cheek: oblique exterior views, border spines and librigenal

spine showing openings, X 30. Locality 2.

Diacanthaspis lepidus n.sp.
Lower Edinburg limestone

15. Cranidium : anteiior view to show spines X 30. Locality 2.

PLATE 7

PLATE 8

Plate 8

Diacatnthatipis iilriclii n.sp.
Lower Edinburg limestone

Figure

1,2,3,4,5. Holotype cranidium: dorsal stereograph, X 15; right lateral,
interior, anterior, posterior views, X 10. USNM 116521.
Locality 7.

(5, 7, 8. Paratype pygidium : dorsal, posterior, right lateral views, X

15. USNM 116522c. Locality 7.

Pygidium: dorsal view, X 30. Locality 7.
Stage 0 cranidium: exterior stereograph, X 30. Locality 4.
Original of Text-figure 12A.
Pygidium: ventral view, X 20. Locality 7.
Two paratype segments from anterior part of thorax: exterior
view, X 15.USXM 116522b. Locality 7.

Paratype thoracic segment: dorsal view, X 15. USNM
116522b. Locality 7.

Small cranidium: extei ior view, X 30. Locality 4. Original of
Text-figure 12B.

Small cranidium: exterior view, X 30. Locality 4. Original
of Text-figure 12C.

Small cranidium: exterior view, X 30. Locality 4. Original
of Text-figure 12D.

Paratype thoracic segment: dorsal view, X 15. USNM
116522e. Locality 7.

Small cranidium: anterior, dorsal, right lateral views, X 15.
Locality 4.

Paratype thoracic segment: dorsal view, X 15. USNM 116522b.
Locality 7.

Small pygidium: exterior and interior views, X 20. Locality
7.

Cranidium: two exterior views to show spines and depressions
at tip of occipital spine, X 30. Locality 7.
Transitory pygidium: exterior and interior views, X 20. Lo-
cality 7.

27. Free cheek: oblique exterior view, X 9. Locality 7.

28. Small free cheek: oblique exterior view, X 30. Locality 7.
Original of Text-figure 12D.

29. Free cheek: oblique exterior view, X 30. Locality 7.

10.

11.

12.

13.

14.

15.

16.

17.

18,

19, 20.

21.

23.

24,

30.

25,

26.

Figure

Plate 9

Diacanthaspis aff. ulrichi n.sp.
Oranda formation, locality 8

1-3, 7, 9. Cranidium : 1, 2, 3, exterior stereograph, dorsal and left lateral

views, X 9 ; 7, oblique exterior view showing spiiies and granu-
lation on external surface, X 30; 9, oblique exterior view show-
ing pits in smooth tip of median occipital spine, spines and
granulation on external surface, X 30.

4, 8. Incomplete cranidium : 4, exterior stereograph, X 9 ; 8, ex-

terior view showing pits in the smooth tip of median occipital
spine, X 30.

Pygidium : dorsal stereograph, X 15.

IJiacantliaspis cooperi Whittingt
Lower Martinsburg shale, localit

itv 10

6. Incomplete cranidium: exterior view showing openings in tips

of spines and granulation, X 30.

PLATE 9

PLATE 10

Figure

Plate 10

Diacanthaspis orandensis n.sp.
Oranda formation, localitv 8

1,3,0-7. Holotype cranidium: exterior stereograph, dorsal, anterior,

left lateral, posterior views, X 9. USNM 116523.

2,4,20. Paratype free cheek: oblique exterior, dorsal, and oblique

interior views, X 9. USNM 116524a.

8-12. Paratype segment from anterior part of thorax: dorsal, ven-

tral, posterior, anterior, right lateral views, X 10. TJSX^M
116524c.

13, 14. Cranidium : interior and dorsal views, X 9.

15. Paratype segment from posterior part of thorax: dorsal view,
X 15. USXM 116524d.

16. Paratype, right pleura of segment from posterior part of
thorax, dorsal view, X 15. USNM 116524e.

17. Incomplete pygidium, dorsal view, X 9.

18. Incomplete pygidium, dorsal view, X 9.

19. Incomplete pygidium, dorsal view, X 9.

21-24. Paratype hypostome: interior, exterior, posterior, right lateral

views, X 15. USNM 116524b.

25. Incomplete pygidium, ventral view, X 6.

26,27. Paratype pygidium: 26, posterior view, X 6; 27, dorsal stereo-

graph, X 9. USNM 1165241

Plate 11

Diacanthaspis orandensi^ n.sp.
Oranda formation, locality 8

Figure

I. Smallest cranidiuui ; exterior view, X 30.

2,3. Small hypostome: interior, exterior views, X 20.

4. Paratype free cheek (original of Plate 10, figures 2, 4, and
20) : oblique exterior view, X 9. USNM 116.524a.

5. Small cranidium: exterior view, X 30.

C. Small free cheek: oblique exterior view, X 10.

7. Small cranidium : exterior view, X 30.

8. Pygidium (original of Plate 10, figure 18) : jjosterior view,
X 6.

9. Small pygidium : exterior view, X 20.

10. Small cranidium: exterior view, X 15.

II. Small cranidimu : interior vieAV, X 15.

12-15. Small cranidium: exterior stereograph; riglit lateral, dorsal,

anterior views, X 15.

19,20. Incomplete cranidium: 19, exterior view of part of external

surface ; on right fixed cheek rounded tips of spines show
minute dark spots, the locii of depressions or openings, X 30;
20, exterior view of part of ()cci])itai ring, showing pits in tip
of median spine, X 30.

Diacanthaspis cooperi Whittington
Lower Martinsburg shale

1(5,17. Free cheek: 16, oblique exterior view of base of librigenal

spine and adjacent border showing openings in tips of border
spines, X 30; 17, oblique interior view of eye surface showing
facets, X 30. Locality 9.

18. Postero-median part of incomplete cranidium showing pits in

tip of median occipital spine, X 30. Locality 10.

PLATE 11

PLATE 12

Figure

Plate 12

Diacnnthaspis scitulus n.sp.
Oranda formation, locality 8

1-3. Holotype cranidium and left free cheek: dorsal stereograph,

anterior and anterolateral views, X 9. USXM 116525.

4-6. Paratype segment from anterior part of thorax: posterior,

dorsal, interior views, X 10. USXM 116526b.

7. Cranidium: interior view showing short third glabellar fur-

lows, X 20.

8-10. Paratype segment from posterior i^art of thorax : anterior,

dorsal, right lateral views, X 10. USX^M 116526c.

11. Small pygidiuni : dorsal view, X 15.

12-1-t. Paratj-pe pygidiiun: posterior, right lateral views, dorsal

stereograph, X 9. USNM 116526e.

15. Pygidium: interior view, X 9.

16. Small pygidiuni : dorsal view, X 30.

17-19. Paratype, last two segments of thorax and pygidium: 17, left

lateral view, X 15; 18, ventral view, X 10; 19, dorsal view,
X 15. USXM 116526d.

20. Free elieek: oblique exterior view showing broad flattened

edge along suture, X 15.

Plate 13

Biacanthaspis scitulus n.sp.
Oranda formation, locality 8

Figure

1, 2. Small eranidium: dorsal, anterior views, X 9.

3-5. Smallest knovm eranidium: dorsal, anterior, right lateral

views, X 15.

6,7. Free cheek showing injury: oltlique interior and oblique ex-

terior vieAvs, X 9.

8,10,11. Paratype hypostome: exterior stereograph, posterior and left

lateral views, X 15. USXM 116526a.

9. Hypostome : interior stereograph, X 15.

112.1"). Paratype pygidium (original of Plate 12, figures 12-14):

enlarged views of external surface to show tips of spines, X 30.
USXM 116526e.

13. Small free cheek : oblique exterior view, X 30.

14. Free cheek, tip of librigenal spine showing openings in spines
on external surface, X 30.

16,17. Cranidium: 16, part of glabella, right fixed cheek and an-

terior border showing spines and granules, X 30; 17, part of
frouto-niedian glabellar lobe showing minute openings in tips
of paired axial spines, X 50.

PLATE 13

PLATE 14

Plate 14

■'Miraspis sp.ind.
Lower Edinburg formation, locality 3

Figtire

1. Pygidium: dorsal stereograph, X 9.

7. Free cheek, oblique exterior view, X 9.

Ceratoceplmla lacinmta Whittington and Evitt

2. Small eranidium : exterior stereogi'aph, X 30.

3. Small eranidium: exterior stereograph, X 30. Original of Text
figure 16.

4. Outer parts of pleurae of two posterior thoracic segments,
pygidium, and objects of uncertain affinities (pin supporting
this object not blacked out) arranged in approximate relative
positions: ventral view, X 6.

5. (i. Objects of uncertain affinities, smallest size known: exterior

and interior views, X 10.

8, 9. Left- and right-hand objects of uncertain affinities, arranged

in juxtaposition with a iiygidinm : dorsal and posterior views,
X 6.

10. Objects of imcertain affinities: "dorsal" view, X 9.

11. Small liypostome: interior view showing tiny opening through
doublure at shoulder, X 10.

12. Paratype hypostome (original of Whittington and Evitt, 1954,
PI. 6, figs. 10-13) : enlarged view of part of interior, showing
openings through doublure at shoulder, X 20. USXM 116503e.

13. Object of uncertain affinities: "ventral" view showing pos-
terior edge of "pleura" X 10.

14,15. Objects of uncertain affinities: 14, tips of "pleural" spines

showing openings at tips of barbs, X 30; 15, oblique interior
view showing anterior and inner edges of "pleura" X 15.

Plate 15

Ceratocephala rarispina u.sp.
Oranda formation, locality 8

Figure

1, 4, 5, 7. Holotype cranidium : exterior stereograph, anterior, interior

and dorsal views, X 9. USNM 116527.

2,12. Paratype free cheek: exterior and right lateral views, X 9.

USNM 124698a.

3, 6. 8. Fragment of left side of cranidium with iacomplete free cheek

attached, showing course of suture and poorly preserved eye
lobe: oblique exterior, exterior, and anterior views, X 9.

9. Pleural portion of segment from anterior part of thorax:

dorsal view, X 9

10,13. Paratype segment from anterior part of thorax: posterior

and dorsal views, X 9. USNM 124698c.

11,14-16. Cranidium: left lateral, anterior, posterior, exterior views,
X 10.

17. Incomplete segment from posterior part of thorax: dorsal

view, X 9.

18,20,21. Paratype pygidium: dorsal, posterior and left lateral views,
X 15. USNM 124698d

19. Pygidium: dorsal view, X 30

22, 23. Paratype hypostome: interior and exterior views, X 30. USNM

12469Sh.

24. Small oranidiuni and left free cheek: exterior stereograph, X

20.

25, 28. Oranidiuni, incomplete but showing entire anterior border

(compare with Figure 1) : 25, exterior view, X 10; 28, part of
axial region showmg depressions (appearing as dark spots)
in tips of spines, X 30.

29. Free cheek: oblique exterior view, X 20.

Ceratoceplmla (Ceratocephalina n.subgen.) sp.iml.
Lower Edinburg formation, locality 4.

26,27,30. Cranidium: anterior, right lateral, dorsal views. X 15.

PLATE 15

PLATE 16

Plate 16

Ccratocephahi (Ceratoceplialina) tridens u.subgeu., n.sp.
Lower Edinburg limestone

Figure

1,3. Holotype eranidiuni and yaratype free cheek: exterior and

anterior stereographs, X 9. Locality 4. USNM 124699,
124700a.

2,18. Holotype cranidium: right lateral and dorsal views, X 9.

Locality 4. USNM 124699. Ostracod carapace adhering to
base of glabella on right side (figure 18) was removed before
photographs forming figures 1-3 were made.

4, o, 7. Cranidium : right lateral view, anterior stereograph, dorsal

view, X 9. Locality 4.

6. Incomplete cranidium : interior view, X 10. Locality 4.

8. Small cranidium and free cheek: exterior stereograph, X 20.
Locality 4.

9. Paratype free cheek: oblique interior view, X 9. Locality 4.
USNM 124700a.

10,11,13. Paratype, 2 thoracic segments: dorsal, ventral and anterior
views, X 9. Locality 4, USNM 124700b.

12. Paratype thoracic segment : dorsal view, X 9. Locality 4.

USNM 124700b.

14. Small free cheek : oblique exterior view, X 30. Locality 4.

15-17. Paratype pygidiuui: dorsal stereograph, right lateral, pos-

terior views, X 15. Locality 4. USNM 124700c.

19,20. Hypostome: interior and exterior views, X 30. Locality 4.

21,22. Hypostome: interior and exterior views, X 30. Locality 4.

23. Free cheek: oblique exterior view showing eye surface, X 20.
Locality 3.

24. Free cheek: oblique exterior view, X 20. Locality 4.

Figure

Plate 17

Apianurus barbatus n.gen., n.sp.
Lower Edinburg limestone

1,16. Paratype free cheek: exterior and left lateral views, X 4.

Locality 2, USNM 124702a.

2. Cranidium : exterior stereograph, X 4. Locality 3.

3, 4, 6. Holotype cranidium : exterior stereograph, dorsal and left

lateral views, X 4. Locality 2, USNM 124701.

5,7,9-11. Cranidium from Locality 2, free cheek from locality 3: an-
terior view, anterolateral stereograph, right lateral view, left
antero-lateral view, dorsal stereograph, X 4.

8,212. Cranidium: posterior and interior views, X 4. Locality 3.

13. Eight free cheek: lateral view, X 9. Locality 2.

14. Left free cheek: exterior view, X 9. Locality 2.

15. Left free cheek: ventral view, X 4. Locality 3.

17,19. Hypostome: exterior view, oblique interior view showing

opening through doublure at shoulder and posterior wing, X
9. Locality 2.

18. Free cheek: interior view, X 6. Locality 3.

20. Free cheek with complete librigcual spine: dorsal view, X 6.
Locality 2.

21. Right free cheek: exterior view showing eye surface, X 30.
Locality 2.

PLATE 17

PLATE 18

Figure

Plate 18

Apianurus barbatus n.gen., n.sp.
Lower Edinburg limestone

1-5. Paratype hypostome: exterior stereograph, interior, left lat-

eral, posterior, anterior views, X 4. Locality 2. USNM
124702b.

6. Paratype anterior thoracic segment: dorsal view, X 4. Local-
ity 2. USXM 124702d.

7, 8. Anterior thoracic segment : dorsal and posterior views, X 10.

Locality 2.

9. Paratype thoracic segment : dorsal stereograph, X 4. Locality

2. USNM 124702c.

10,14,15. Paratype incomplete thoracic segment: left lateral, dorsal,
anterior views, X 4. Locality 2. USXM 124702d.

11-13. Paratype pygidium: left lateral view, dorsal stereograph, pos-

terior view, X 4. Locality 2. USNM 124702e.

16-18. Small pygidium: posterior, left lateral, dorsal views, X 10.

Locality 2.

19. Pygidium: enlarged view of anterior part of axis showing

spines on external surface, X 30. Locality 2.

20,21. Small pygidium: dorsal and posterior views, X 10. Locality 2.

22. Cranidium : enlarged view of median part of occipital ring

between occipital spines showing external surface, X 30. Lo-
cality 2.

23,24. Small holaspid pygidium: dorsal and ventral views, X 30.

Locality 4.

Figure

12,

13.

14 IG.

17,

.21.

2U.

Plate 19

Apianurus harbatus n.gen., n.sp.
Lower Edinburg limestone

1,3. Stage 0 free c-hcek: exterior and interior views, X 30. Local

ity 4. Original of Text-figure 22A.

2. Stage 0 cranidium : exterior stereograph, X 30. Locality 3.

Original of Text-figure 22A.

4,6. Stage 0 (?) cranidium lacking right fixigenal spine, exterior

and interior views, X 30. Locality 4.

5. Stage 0 transitory pygidiuni: exterior stereograph, X 30.

Locality 4. Original of Text-figure 22A.

7. Small cranidium: exterior view, X 30. Locality 2. Original
of Text figure 22C.

8. Small free cheek: exterior view. X 30. Locality 4. Original
of Text-figure 22C.

9,10. Small cranidium: exterior and interior views, X 30. Locality

3. Original of Text-figure 22B.

11. Small free cheek: exterior view, X 15. Locality 4. Original

of Text-figure 22D.

Small hypostome: exterior and interior views, X 30. Locality 3.

Small cranidium: dorsal, anterior, right lateral views, X 15.
Locality 2.

Small hypostome: exterior and interior views, X 30. Locality 3.

Small cranidium: dorsal, anterior and left lateral views, X
10. Locality 2.

22, 23. Small cranidium: exterior and left lateral views, X 15. Local-

ity 2.

24. Transitory pygidium : dorsal view, X 30. Locality 4.

25. Transitory pygidium: dorsal view, X 30. Locality 4.

26. Transitory pygidium, Stage holaspis-1 ; dorsal view, X 30.
Locality 2.

27. 28. Transitory pygidium: dorsal and ventral views, X 30. Locality

29. Transitory pygidium: dorsal view, X 30. Locality 4.

PLATE 19

PLATE 20

Figure

Plate 20

Apicunurus harbatus n.gen., ii.sp.
Oranda formation, locality 8

1-3,11. Cranidium: exterior stereograph, X 4; anterior, right lateral,

and interior views, X 6.

4. Cranidium: exterior stereograph, X 6.

5,7. Free cheek: dorsal and right lateral views, X 4.

6, 8, 10. Hypostome : right lateral view, exterior stereograph, posterior

view, X i.

9. Small hypostome: exterior view, X 30.

12.16. Segment from anterior part of thorax: dorsal and left lateral
views, X 4.

13,14,19. Pygidium: dorsal stereograph; ventral and posterior views,

vieys, X 4.

15.17. Segment from posterior part of thorax; right lateral and
dorsal views, X 4.

Apianurus aff. furcata (Linnarsson)
Middle Ordovician, Chasmops limestone, southern Norway

18,21. Incomplete cranidium: anterior view, exterior stereograph,

X 2. Guttormsberget. PMO 66690.

20. Pygidium: dorsal view, X 10. Bygd£?y. PMO 5647.

22,24,25. Incomplete cranidium: dorsal stereograph, anterior and ex-
terior views, X 4. Guttormsberget. PMO 66691a.

23. Cranidium: exterior view, X 2.2 Fernholmen. PMO 3673.

Figure

Plate 21

Apianurus glaber ii.geii., u.sp.
Lower Edinburg limestone, locality 2

1,3. Holotype eranidiiim: exterior stereograph, right lateral, dorsal

views, X 4. USXM 124703.

4,5. Incomplete cranidlum: dorsal and exterior views, X 4.

6,7. Incomplete cranidium: interior and posterolateral views, X 6.

8, 10. Paratype pygidium: posterior, right lateral and interior views,

X 4. USXM 124704a.

11,12. Incomplete cranidium: exterior and left lateral A'iews, X 4.

13,14. Paratype hypostome: interior view, exterior stereograph, X 6.

ITSXM 124704b.

13. Pygidium: dorsal stereograph, X 4.

Apianurus sp.ind.
Lower Edinburg limestone, locality 2

16,17,19. Pygidium; dorsal stereograph, X 6; posterior and left lateral
views, X 4. USNM 124705a.

18. Pygidium: dorsal stereograph, X 6. TSXM 124705b.

20, 25. Hypostome : exterior stereograph, interior view, X 9. USXM

124705d.

21. Pygidium: dorsal view, X 15. USXM 124705c.

22,26. Hypostome: interior and exterior views, X 9. USNM 124705e.

Odoutopleurid protaspis, ? subfamily Miraspiuae
Lower Edinburg limestone, locality 4

23, 24. - Exterior stereograph, right lateral view, X 33.6. Photograph

by W. R. Evitt.

PLATE 21

PLATE 22

Figure

Plate 22

Calipernurus bisoUtus n.gen., u.sp.
Lower Edinburg limestone

1-3, H. Holotyye, incomplete ceplialon: exterior stereograph, left

lateral, anterior, posterior views, X 9. Locality 3. USNM
124711.

4,5,7. Cranidiiun (Locality 3) and free cheek (Locality 2): dorsal

and antero-lateral stereographs, exterior view, X 6.

Fi(/iire

Plate 23

Calipernurus insolitus ii.gen., ii.sp.
Lower Edinljurg limestone

1,3. Holotype, incomplete cepbalon: ventral, dorsal views, X 9. Lo-

cality 3. USNM 124711.

2. 4. Crauidium (original of Plate 22, figures -4, 5, 7) : posterior,

interior views, X 4. Locality 3.

5.7. C'ranidiiim: dorsal and exterior views, X 9. Locality 3.

(i. Tip of libiigenal .si)ine slio\vin<;' openings at base of barbs, X

Md. Locality 3.

8. Cianidiuin ; exterior view showing pits in median occipital
tul)ercle, X 15. Locality 3.

9, lU. Free cheek (original of Plate 22, figures 4, 5 and 7) ; oblique

interior and oldique exterior views of eye surface, X 30. Local-
ity 2.

11, 12,15. I'aratype hypostome: exterior stereograph, X G; interior view,
X 6; oblique interior view, X 15 (arrow points to tiny open-
ing in doublure). Locality 2. USNM 124712a.

13,14. Hypostome: exterior stereograph, interior view, X 9. Locality

PLATE 23

PLATE 24

Plate 24

Calipernurus insolitus n.gen., n.sp.
Lower Edinburg limestone

Figure

I. Small cranidiuiu : exterior stereograph, X 30. Locality 3.

2-4. Small craiiidiuin: right lateral, anterior, dorsal views, X 10.

Locality 2.

5-7. Small erauidiimi (Locality 2) and tree cheek (Locality 3) : left

lateral view, exterior stereograph, anterior view, X 10.

8.13,14. Small pygidium : left lateral, posterior, dorsal views, X 1,.

Locality 3.

9. Cranidium : exterior stereograph, X 10. Locality 2.

10. Small segment from posterior part of thorax: dorsal view,
X 10. Locality 3.

II. Small pygidinm lacking median border spine: dorsal view,
X 12. Locality 3.

12. Transitory pygidium : dorsal view, X 15. Locality 2.

15,16,19. Paratype segment: anterior, dorsal, posterior views, X 4.5.
Locality 3. USNM 124712b.

17,20. Pygidimn with median border spine distally bifurcated:

dorsal and ventral views, X 10. Locality 3.

18. Pygidium lacking median border spine : dorsal view, X 9.

21,25,27. Paratype pygidium: 21, oblique posterior view showing tips
of major border spines and openings at base of barbs, X 30;
25, posterior view, X 6 ; 27, dorsal stereograph. X 6. Locality
2, USNM 124712d.

22,23. Incomplete segment from posterior part of thorax: dorsal and

left lateral views, X 6. Locality 2.

24. Tip of major border spine of pygidium showing openings at

base of barbs, X 30. Locality 2.

26. Paratype, 2 incomplete segments from posterior part of thorax:

dorsal view, X 9. Locality 3. USXM 124712c.

28. Pygidium: dorsal stereograph, X 6. Locality 2.

Bulletin of the Museum of Comparative Zoology

AT H A R V A E T) COLLEGE
Vol. 114, No. 6

A CAT SKELETON WITH AN ANOMALOUS
THIRD HIND LEG AND ABNORMAL VERTEBRAE

Thomas S. Parsons

and
John M. Stein

WiTH Four Plates

CAMBRIDGE, MASS., U.S.A.
PRINTED FOR THE :\r U S E U M

i\l.\Y, 1956

Publications Issued by or in Connection

WITH THE

MUSEUM OF COMPARATIVE ZOOLOGY
AT HARVARD COLLEGE

Bulletin (octavo) 1863 — The current volume is Vol. 114.

Breviora (octavo) 1952 — No. 53 is cui-rent.

Memoirs (quarto) 1864-1938 — Publication was terminated with Vol. 55.

Johnsonia (quarto) 1941 — A publication of the Department of Mollusks.
Vol. 3, no. 35 is current.

Occasional Papers of the Department of ^Iollusks (octavo) 1945 —
Vol. 1, no. 18 is current.

Proceedings of the New England Zoological Club (octavo "i 1899-
1948 — Published in connection with the Museum. Publication terminated
with Vol. 24.

Bulletin of the Museum of Comparative Zoology

AT HARVARD COLLEGE
Vol. 114, No. 6

A CAT SKELETON WITH AN ANOMALOUS
THIRD HIND LEG AND ABNORMAL VERTEBRAE

Thomas S. Parsons

and

John M. Stein

With Four Plates

CAMBRIDGE, MASS., U.S.A.

printed for the imuseum
May, 1956

No. 6 — A Cat Skeleton with an Anomalous Third Hind Leg

and Abnormal Vertebrae

By Thomas S. Parsons and John M. Stein

Harvard Biological Laboratories
Cambridge 38, Massachusetts

INTRODUCTION

The cat skeleton here described (see Plate 1) displays on the
right side a highly abnormal innominate bone articulating with
two legs, the more anterior of which has two tibiae and is evi-
dently the product of a fusion of two legs. The posterior verte-
brae show various fusions and abnormalities. The skeleton was
kindly lent to us by its owner, Mr. Elling 0. Eide, of Sarasota,
Florida. The cat was a family pet which lived over six and a
half years and could run and climb trees despite its extreme
pelvic abnormalities. Its appearance in life is shown in Plate 3A.
It was buried shortly after its death and the skeleton dug up
fifteen months later. Although a few of the smaller bones were
not recovered, the completeness of the recovery is shown by the
finding of twenty-eight of the sesamoids.

The mother of the cat was apparently normal and produced
many litters of kittens, probably many by the same father whose
only abnormality was Polydactyly. The one described here had
two normal littermates. The mother was slightly over seven and
a half years old when this litter was born. She was very sick for
several weeks after their birth, but recovered and had normal
litters afterwards. Most of her kittens were normal except for
numerous cases of Polydactyly, but one born approximately three
years before the one here described had two symmetrically fused
heads. It was found dead, and whether or not it was alive at
birth is not known.

Due to the burial, only the skeleton of the animal was re-
covered, and no attempt has been made to reconstruct any of the
soft parts. The skeleton has, however, been compared with five
normal ones, one articulated and four disarticulated, as well as
with the extensive descriptions and illustrations in Jayne (1898).

The only definite evidence on the articulation of this skeleton
in life comes from the nature of the articulatory surfaces and
the excellent x-ray shown in Plate 2. The skeleton was mounted
as shown in Plate 1 ; no attempt was made to include most of

294 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

the tail, the ribs, sternum, phalanges, or hyoid apparatus in this
reconstruction.

Although the work of making the original descriptions was
divided between the authors, both of us have gone over all the
parts and are equally responsible for this account. In order to
give morphological descriptions, it is necessary for us to give
names to several of the anomalous elements before the reasons for
our interpretations of their nature can be stated. However, we
hope that the reader can separate our theories from the more
objective accounts of the various bones.

We would like to thank several people for their kind assistance
in the writing of this paper : first, Mr. Elling 0. Eide, who not
only loaned us the skeleton, but also gave us its history, provided
the photograph of the live cat, and made the drawings ; Dr. C. K.
Newton, D.V.M., of Bradenton, Florida, who provided the x-ray
of the hind legs; and especially Dr. Ernest E. Williams of Har-
vard University, who offered much helpful advice throughout the
course of the study.

DESCRIPTION

General

Much of the skeleton, such as the pectoral girdles and limbs,
is normal and needs no detailed description. The skull and jaws
are also normal, but have a very stocky appearance. Both are
short and very wide ; the areas of attachment of the jaw muscles
are noticeably larger than average. The canine teeth are very
large. The third right lower premolar, the third right upper
incisor, and the third left upper premolar are missing and the
sockets filled with l^one, but these probably were lost after ma-
turity was reached. The two second upper premolars are also
missing, but they seem never to have been present, a relatively
common occurrence, according to Jayne (1898). The first and
second left upper incisors were broken oft', apparently before
death. Only the tAvo stylohyals of the hyoid apparatus were re-
covered. They appear normal.

The ribs seem normal except that the twelfth right one is lack-
ing due to the malformation of the vertebra. The recovery of the
sternum was incomplete, but the manubrium and five sternebrae
(two of them fused) are present. All appear slightly thicker

PARSONS AND STEIN : ANOMALOUS CAT SKELETON 295

than usual for their length, especially the fused pair. Their ends
appear rather asymmetrical, but in the absence of the cartilages
between them this indicates little. The fused ones each send out
a short prong of bone, on either side, at the fused ends which must
liave articulated with the costal cartilages.

Vertebral column

The anterior lialf of the vertebral column appears normal; the
posterior part shows abnormalities. There are seven eervicals,
thirteen thoracics, .seA'en lumbars, four sacrals (one more than
usual), and nineteen caudals. The last number, although low, is
within the range given by Jayue (1898), and some of the terminal
ones ma}' not have been recovered. Only those vertebrae which
show marked peculiarities Avill be described here.

The first such is the tenth thoracic. The neural spine and post-
zygapophyses are twisted approximately 15 degrees to the left.
The posterior articular surfaces are more nearly horizontal and
the anapophyses shorter than usual, but this vertebra is still
relatively normal in general appearance.

The last three thoracic vertebrae are fused into one mass (Fig.
1). The eleventh is joined to the tAvelfth by the dorsal part of

A B

Fig. 1. Eleventh, twelfth, and thirteenth thoracic vertebrae. Anterior end
to right. A, dorsal view. B, ventral view. xl.

the centrum and a small flange of bone just ventral to the left
anapophysis. In the former the two sides of the neural arch
meet dorsally in the midline, but are not fused, and there is no
trace of a neural spine. All of the articular processes appear quite
normal in shape, although the postzj-gapophyses are deflected
slightly to the left and the right metapophysis is more anterior

296 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

than lateral to the prezygapophyseal facet. Well developed tuber-
cular facets are present and the capitula of the ribs articulate
intervertebrally as is usual in the more anterior thoracic verte-
brae, but not common in the eleventh (see Jayne, 1898, p. 84;
however, his fig. 55, p. 82, shows a situation similar to that in
this specimen).

The twelfth thoracic vertebra is represented ventrally only
by the wedge-shaped left half of the centrum (Fig. IB). On the
right side the eleventh and thirteenth thoracics come together
without any gap between them. Dorsally (Fig. lA), both sides
of the neural arch of the twelfth are present and tightly fused
to the last thoracic, but the right side is far smaller than the left.
Only a left rib articulates with this vertebra, and that to a large
convex capitular facet. The prezygapophyses are quite normal,
but the right metapophysis is much reduced. The postzygapo-
physes are fused with the prezygapophyses of the following verte-
bra forming dorsal mounds. An anapophysis is present only on
the left. The small neural spine points straight dorsally. This
vertebra is thus anticlinal, using Jayne 's (1898) definition; he
gives the eleventh as the normal anticlinal vertebra, while Gott-
lieb (1914) states that it can be either the tenth or the eleventh.
The intervertebral notch on the right appears double, so pre-
sumably the normal number of spinal nerves was present.

The thirteenth thoracic vertebra (Fig. 1) is the most normal
looking of the three fused ones. The prezygapophyses lack dis-
tinct metapophyses due to their fusion with the postzygapophyses
of the preceding vertebra, but there are raised roughened areas
which probably correspond to them. The neural spine is quite
small and directed anteriorly. The postzygapophyses and ana-
pophyses are displaced to the left with the right ones dorsal to
the left. The latter are normal, but the right postzygapophysis
is elongate, while the right anapophysis is higher than long, ar-
ticulating with the entire height of the following prezygapo-
physis.

The first lumbar vertebra (Fig. 2A) is markedly asymmetrical
with its centrum about 9 mm. long on the right side and 12 mm.
on the left. The neural arch appears slanted to the left so that
the left pedicle is abnormally thick while the right, although in
contact with the centrum, is not fused to it. The neural canal is
thus displaced to the right. The right prezygapophysis is vertical

PARSONS AND STEIN : ANOMALOUS CAT SKELETON

297

and massive while the left is thinner and projects dorsolaterally.
Only the latter bears a normal metapophysis. The postzygapo-
physes are both displaced to the left as are the anapophyses. The

Fig. 2. Lumbar vertebrae. Anterior end to right in B, D, and E and to
left in C. A, first in posterior view. B, second in dorsal view.
C, third in posterolateral view. D, fourth and fifth in ventral view.
E, sixth in dorsal view. xl.

latter are quite short and round in section with the right one
well dorsal to the left. The rather small neural spine arises from
the normal position on the neural arch and slopes towards the
right to the midline. The pleurapophyses are rudimentary prongs
about 2 mm. long on the right, and 4 mm. on the left. The former
is somewhat dorsal to its usual position while the latter is curved
slightl}' posteriorly.

The second lumbar vertebra (Fig. 2B) is also highly asym-
metrical. The prezj'gapoph^'ses, massive structures with nearly
vertical articular facets, are displaced to the left with only the
right bearing a normal metapophysis. The postzygapophyses
and anapophyses appear quite normal, the only irregularity being
a slight elongation of the right postzygapophysis. As in the first
lumbar, the neural spine leans to the right, almost meeting the

298 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

right prezygapophysis. The pleurapophyses are very small and
do not extend ventral to the centrum. The right one points lat-
erally and the left anteriorly.

The third lumbar vertebra (Fig. 2C) displays spina bifida with
both sides of the neural arch forming a separate neural spine.
The middorsal cleft is exceedingly narrow, but the two sides do
not actually touch at any point. The left neural spine is directed
anteriorly and appears much like the anterior half of a normal
neural spine. The more posterior right one is directed straight
dorsally so that a V-shaped notch is seen between them in lateral
view. All the zygapophyses and metapophyses are normal with
the left prezygapophysis and right postzygapophysis slightly
larger than the others. The anapophyses, especially the left, are
reduced in size and are spinous rather than blade-like. The
pleurapophyses while normal in shape are also reduced, being
intermediate in size between those of the first and second lumbar
vertebrae of the four normal cats used for comparison.

The fourth and fifth lumbars (Fig. 2D) are fused by a large
rounded mass of bone ventral to the two centra. The normal parts
of the two, which are held in the articulated position, show no
fusion. The ventral mass is fused to the sides of the centrum of
the fourth lumbar leaving a small midventral canal between it
and the centrum ; no such canal is present in the following
vertebra. The anterior face of the fourth lumbar centrum faces
anteroventrally producing a distinct angle in the articulated
spine at this joint. The pleurapophyses are small and more
ventrallj' directed than usual. The right one points slightly more
laterally than tlie left. The metapophyses are not distinctly de-
veloped and the left anapophysis is rudimentary; otherwise the
processes of this vertebra are normal.

The fifth lumbar vertebra also has quite normal dorsal proc-
esses. The right anapophysis is almost completely absent, a con-
dition which was also seen in the normal skeletons used for com-
parison. The left postzygapophysis is slightly elongated, and the
posterior end of the neural spine is deflected to the right. The
other processes show no modifications. The pleurapophyses are
somewhat reduced (Fig. 2D), and the right one especially does
not extend laterally as far as usual. The complex of the fourth
and fifth lumbars shows a slight curvature with the length of
the two centra slightly greater on the left than on the right. It

PARSONS AXD STEIN : ANOMALOUS CAT SKELETON 299

is impossible to say exactly how this curvature is brought about
because of the mass of bone ventral to the two centra obscuring
their articulation.

The sixth lumbar vertebra (Fig. 2E) also has greatly reduced
pleurapophyses with the right one considerably smaller than the
left. Both extend anteroventrally but not laterally, as would be
the normal condition. The posterior face of the centrum is some-
what dorsally directed, thus making another sharp angle in the
articulated spine. The prezygapophyses and metapophyses are
normal with those on the left slightly larger. The neural spine
leans towards the left. On the right, the postzygapophysis is
much reduced and the anapophysis completely absent, while on
the left the former is greatly enlarged and the latter present
although very small.

The seventh lumbar vertebra is tightly fused to the sacrum
dorsally, although ventrally the centra remain separate (Fig. 3).
It does not enter into the sacroiliac joint. The left pleurapophy-
sis, while short, is fairly normal, but the right one is absent
except for a broad truncated base extending just anterior to the
side of the centrum. The left prezygapophysis is a simple dorsal
projection lacking a metapophysis completely (as in one of the
four disarticulated normal cats used for comparison). On the
right the entire process is a low mound reaching posteriorly as
far as the right neural spine. The left postzygapophysis is com-
pletely fused to the prezygapophysis of the first sacral vertebra,
the two being represented by a small mound. There is no trace
of any corresponding articulation on the right. The neural arch
apparently did not fuse middorsally until very late in develop-
ment since there is a distinct suture curving to the left poste-
riorly. Two neural spines are present, their arrangement being
similar to that in the third lumbar except that in the seventh the
right is anterior. The latter expands anteroventrally where it
has a concave anterior face into which the prezygapophysis ex-
tends. The neural canal is triangular rather than quadrilateral
in end view.

The sacrum is shown in Figure 3. It is, as a unit, somewhat
twisted. The posterior part is slightly curved to the left, mainly
by unequal growth of the sides of the second sacral vertebra, and
accompanying this there is a counterclockwise rotation about the
long axis of the body when seen in posterior view. The left side

300

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

of the sacrum is quite normal despite the presence of a fourth
sacral vertebra. There are three small mounds representing:
fused zygapophyses and three sacral foramina both dorsally and
ventrally. The sacroiliac joint on this side is apparently normal

Fig. 3. Seventh lumbar vertebra and sacrum. Anterior end to right,
dorsal view. B, ventral view. C, lateral view. x%.

although the auricular facet is slightly larger than figured by
Jayne (1898, fig. 75, p. 103) or in the four normal cats exam-
ined. The neural spines are fused into a low longitudinal ridge
with a definite spine for the first, and small gentle rises for the
middle two, but no distinct spine on the fourth sacral. The left
pleurapophysis of the fourth sacral is slightly smaller and more
laterally directed than the right.

The right side of the sacrum is highly abnormal. There are no
signs of the three pairs of fused zygapophyses and only the pos-
terior two pairs of sacral foramina are normal. The surface
articulating with the ilium is split into two separate parts. The
larger and more posterior is an irregular, posterolaterally di-
rected, oval facet on the expanded transverse process of the sec-
ond sacral. The other is a small circular facet at the posterior
end of the first sacral about midwav dorsal. It is laterallv di-
rected. The first sacral lacks a large expanded transverse process
on the right, liaving only this articular facet on a small process
which does not extend as far laterally as the second sacral (from
which it is separated by a narrow groove), and a posteriorly
directed ventral process somewhat resembling a caudal rib. This

PARSONS AND STEIN : ANOMALOUS CAT SKELETON 301

ends just ventral to, but not fused with, the transverse process
of the second sacral. There is thus neither a dorsal nor a ventral
sacral foramen l)etween these two vertebrae, but a large lateral
intervertebral foramen.

Of the nineteen caudals, only the first six show any abnormali-
ties. The first three show slight asymmetries in the zygapophyses,
but nothing else unusual. Numbers four through six are fused
together by their neural arches, with their centra remaining quite
distinct. As in the last three thoracics, the middle one, the fifth
caudal, is represented only by one side, in this case the right.
Apparently the left side is completely absent even dorsally, the
postzygapophysis of the fourth articulating directly with the
prezygapophysis of the sixth on that side. On the sixth there is
a small prong of bone on the centrum representing part of a
haemal arch ; a similar condition was observed in one of the nor-
mal cats used for comparison.

Pelvic girdles and appendages

The left innominate bone seems quite normal except that the
auricular articulation is unusually large and the spine of the
pubis for the attachment of the rectus abdominis is absent. One
of the normal cats u.sed for comparison also lacked the latter
structure. The left femur, tibia, and fibula all appear quite
normal.

The right innominate bone (Fig. 4) is a very deformed struc-
ture, seemingly divided into two parts : one part consists of the
right half of the more normal or left pelvis; the other is a very
disfigured complete second pelvis. This second or teratological
pelvis and the more normal right innominate are solidly fused
in two places : one between the spine and horizontal ramus of the
more normal pubis and the medial portions of the teratological
right ischium, the other between the anteromedial portion of the
more normal ilium and the anterolateral portion of the tera-
tological right ilium. A distinct opening occurs between the two
l)oints of fusion (Fig. 4A, oi). Thus the teratological pelvis is
suspended by the more normal right innominate in a parasagittal
plane of the animal. The part of the teratological pelvis normally
ventral faces laterally to the right.

The ilium of thp more normal right innominate is about 1 cm.

302

BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

shorter than the left. The suture (Figs. 4A and B, sut) between
the two ilia is solid yet clearly defined. Approximately 7 mm.
posterior to the suture on the medial side is an oval articulation
with the sacrum (4x7 mm. in size; Fig. 4A, al).

.Starting about 10 mm. from the anterior end of the ilium and

extending for 40 mm. on the lateral side and even overlapping for

al 0' so sp

Fig. 4. More normal right innominate and entire teratological pelvis. A,
medial view. B, lateral view. C, ventromedial view. Abbrevia-
tions: al, regular auricular facet; a2, right auricular facet of tera-
tological pelvis ; act, acetabulum of teratological pelvis ; ft, ischial
tuberosities of teratological pelvis; Iti, left teratological ilium;
oi, gap l)etween regiilMi- right innominate and teratological pelvis;
ot, obturator f enestrae of teratological pelvis ; sp, spongy bone
mass; sitt, suture between regular right innominate and teratological
pelvis ; nyin, regular pubic symphysis ; xs, bony ridge ventral to
pubic sjTnphysis. x^

cy2.

12 mm. on the medial surface lies a spongy mass of bone (Figs.
4A and B, up). Part of this spongy mass forms the articulation
for the more normal right femur. The acetabulum is irregular
in shape and has a very rough surface as does the head of the
articulating fenuir. This spongy bone seems to have replaced
arthritic cartilage and to have been applied to the more normal
bone underlying it. Some rough bone also appears on the medial
side of the ischium, dorsal to the obturator fenestra.

PARSONS AND STEIN : ANOMALOUS CAT SKELETON 303

The iscliiiim of the more normal innominate is bent ventrally
and laterally to the right more strongly than in any of the four
cats used for comparison. The bone on the ventral side of the
ischium is rough and the tuberosity is unusually large. At the
symphysis, the fit between the right and left halves of the left
pelvis is good, but a small ridge of bone extends ventrally along
the entire right part of the symphysis (Fig. 4C, xs) . The descend-
ing ramus of the pubis of its right innominate is much thicker
dorsoventrally than that of the four normal specimens. The hori-
zontal ramus is fused with the teratological right ischium for
about 17 mm. This suture is very solid, but the line of fusion can
be seen quite clearly by the grain of the bone.

The teratological pelvis looks like a fat, symmetrical "Y" from
a lateral view. There is a small fenestra near the base of the Y,
this presumably being the fused obturator fenestrae from both
sides (Fig. 4A, of). The right ilium (the posterodorsal arm of
the Y) is shorter than the left and has an oval auricular facet
(Fig. 4A, a3) about 3x5 mm., located 9 mm. from the right upper
end of the Y, and 4 mm. from the line of fusion with the more
normal right ilium. The teratological left ilium is located on
the right side of the cat 's central axis. The left teratological ilium
looks slightly more normal in shape than the right teratological
one does. The right pleurapophj'sis of the seventh lumbar verte-
bra is fused to the left teratological ilium. It will be remembered
that the rest of the seventh lumbar is fused to the sacrum.

From ventrolateral aspect, the teratological innominate has one
fairly smooth acetabular articulation (Fig. 4B, act). This articu-
lation has two distinct sides, right and left. Posterior to the fused
obturator fenestrae are the fused tuberosities of the isehia (Fig.
4A, ft). The ilia on both sides are of smooth solid bone.

The articulation of the femur with the more normal right in-
nominate is a knotty, gnarled mass of bone (Fig. 5 A, ar). The
articular surface itself seems worn enough to have allowed some
slight movements. The abnormal area seems to have been formed
on top of smoother bone and is probably ossified cartilage. Only
the posterolateral part of the head is present and that is formed
in rough bone. An irregular pit occurs in place of the rest of
the head (Fig. 5A, pt). The lesser trochanter (Fig. 5B, It), also
covered with rough bone, seems also to have articulated with
the innominate slightly posterior and medial to the head articu-

304

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Figure 5

PARSONS AND STEIN : ANOMALOUS CAT SKELETON 305

latiou. This places the femur in a position approximately per-
pendicular to the central axis of the cat. The shaft is almost
C3'lindrical, i.e., not graded in thickness as were those of the cats
used for comparison. Although slightly rough in appearance,
the greater trochanter has a normal shape. Distally, the condyles
are bent about 15 degrees medially, considering the bone in the
normal position (Fig. 5B). The articulatory surfaces of the
condyles are smaller than those of the four normal cats ; they
are also slanted inward towards the axis of the bone and do not
extend outward as they should. The medial condyle has a long
curved articulation (Fig. 5B, ic) ; the lateral condyle has a short
oval articular surface (Fig. 5B, oc).

Proximally the more normal right tibia has unevenly developed
articular surfaces. The lateral articular surface (Fig. 5C, oa) is
less extended than in the normal eats, whereas the medial surface
(Fig. 5C, ia) extends in an arc with the lateral articulation as
a center. Considering the shape of this articulation and that of
the distal end of the femur, it seems as though the knee joint had
a swivel action as well as a hinge in order to compensate for the
very slight motility of the acetabular articulation. The rest of the
bone appears normal, as does the fibula of this leg.

The fusion of two femora, one right and one left, apparently
formed the double femur which articulates at act, Figure 4B.
with the fused teratological pelvis. The articulation is smooth
and allowed hip movement. The distal portions of the femora are
fused at an angle of about 60 degrees (Fig. 5D). Proximally
the two component bones have turned so that their heads and tlu-
greater and lesser trochanters all have fused with their counter-
parts from the other side, the double bone having one head, one
greater trochanter, and one lesser trochanter all located in the

Fig. 5. A and B, more normal right femur in anterior and posteroventral
views. C, more normal right tibia in posterodorsal view. D, E, and
F, fused teratological femur in actual ventromedial (morphological
posterior), actual anterior (morphological left lateral), and actual
lateral (morphological anterior) views, respectively. Abbreviations:
ar, irregularly formed head; c, canal; gt, greater trochanter; h,
head; ia, medial tibial condyle; ic, medial femoral condyle; It.
lesser trochanter; oa, lateral tibial condyle; oc, lateral femoral
condyle; pt, rough pit; sr, sharp ridge. B, C, and D foreshortened
to show articular surfaces. x%.

306 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

medial plane of the bone {Fig. 5F). On one side at the proximal
end in the medial plane there is a sharp thin ridge (Fig. 5F, sr)
running for about 25 mm. In this region the bone is thin latero-
medially and thick anteroposteriorly. Moving distally, the bone
gradually widens lateromedially and at the distal end is as thick
that way as it was anteroposteriorly before (Figs. 5E and F).
A canal (Fig. 5F, c) perforates the bone starting anteriorly
(here and for the rest of the description of the long bones of this
leg, anterior, posterior, medial, and lateral refer to the normal
position, not that in this specimen) about 25 mm. from the
distal extremity and ending 10 mm. from that end on the pos-
terior side. This canal is directly in the midline of the bone. A
large articulation for a sesamoid is present on the lateral surface
of the right component. All four condyles are well formed (Fig.
5D). The x-ray shown in Plate 2 gives the relationships at this
joint clearly.

The two separate tibiae which articulate with the double femur
are almost exact mirror images of each other. They lie so that
their anterior surfaces are pointed outward from their mutual
articulations. These articulations between the two tibiae lie one
at the posterior side of the head (Plate 2) and the other at the
distal extremity. There are two proximal articular surfaces on
each of the bones ; they lie on the posterolateral and anteromedial
parts of the head. It appears that the joint between the double
femur and the two tibiae must have been almost completely im-
mobile. The shafts are unusually thin and the bones are dispro-
portionately long in comparison to the other bones in the cat.
The tubercle is prominent and the crest is short. Distally all
the parts are recognizable, although modified.

The overall organization of this double leg is rather baffling
to the authors. The anterior halves of the teratological pelvis
and of the double femur appear to represent the morphological
left side, an interpretation which fits not only the shape of the
elements, but also the most probable pattern of fusion. Nonethe-
less, certainly in the foot and very probably in the lower leg, the
morphological left is posterior. It is as though the knee were
mistakenly articulated backwards in the mounting of the skele-
ton ; however, the nature of the articular surfaces and the photo-
graphs of the living cat seem to preclude this latter possibility.

The double fibula is a symmetrical structure formed by the

PARSONS AND STEIN : ANOMALOUS CAT SKELETON 307

complete fusion of two fibulae, one right and one left, along their
medial sides. At the proximal end, the fusion has caused the
loss of the medial parts of both bones, leaving only the lateral
parts recognizable. The shaft is thin and bent concavely from the
posterior aspect. Distally the median parts of the bones are not
lost. There is a fossa with a central ridge in the line of fusion
and indications of the lateral maleoli are present on both sides.
The astragalar articulations form one single articular surface
which is quite rough.

The feet of both the regular hind legs appear completely
normal, but the tarsus of the abnormal limb is much modified by
being in the flexed rather than the normal extended position so
as to form an angle of about 120 degrees with the tibiae. This
tarsus is a double structure bearing metatarsals 1, 2, 3, 4, 3, 2, and
1, hence with two sides corresponding to the normal median and
without any normal lateral. In the description, dorsal, ventral,
lateral, and median will be used for the condition in this foot ;
all are the reverse of the normal morphological pattern. Although
only the disarticulated bones are available, the articular surfaces
can be matched very precisely in all cases so that the restoration
is almost certainly correct although two elements are missing
(Plates 3B, C, and 4).

The calcanea, astragali, and scaphoids are fused into one sym-
metrical mass, but in most places the sutures can be distinguished.
The first form a single midventral bar which expands posteriorly
in a rough and immovable articulation with the single fibula and
both tibiae. Each astragalus is an L-shaped element with the
two arms of approximately equal length. It lies along the side of
the calcaneum and has one arm extending dorsally from its
proximal end to form the main articulation with the tibiae. The
dorsal surfaces are fused into one triangular facet with the apex
anterior. The scaphoids are fused together and cover the distal
ends of the astragali, but not the more ventral calcanea. The
tuberosities are thus dorsal and lateral.

The cuboids are represented by a single midventral element
which articulates proximally with the calcanea and dorsally with
the scaphoids and ectocuneiforms. It is more or less cylindrical
with the length about twice the diameter. The distal end is con-
cave to receive the proximal end of the single fourth metatarsal.

On the left, the ectocuneiform and mesocuneiform are fused

308 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

into a single element which covers almost all of the distal surface
of the left scaphoid. The part representing the mesocuneiform
is far larger than usual being almost equal to the ectocuneiform
in size. However, the former is, as usual, shorter in the proximal-
distal direction. The articular surfaces have the same general
pattern as in normal specimens, but those of the mesocuneiform
are much expanded, especially in height. The left entocuneiform
and probably also the vestigial first metatarsal are represented
by a single fairly normal-appearing element which lies in the
expected position dorsal and lateral to the mesocuneiform and
proximal end of the second metatarsal. It also articulates with
the scaphoid.

On the right, only the ectocuneiform is present. It is, like most
of the elements of this tarsus, smaller than usual, compared to
normal cats of the same overall size. It lies distal to the ventro-
median half of the right scaphoid. The hook on the plantar (here
dorsal) surface is much reduced and the ventrolateral corner
is drawn out into a small prong, but the element is immediately
recognizable as an ectocuneiform. The mesocuneiform was lost
but, judging fi'om the articular surfaces of the scaphoid and sec-
ond metatarsal, must have been quite large and similar to the
one on the left. The right entocuneiform and first metatarsal
are also lost and were probably similar to the left, although
there is no smooth articular surface for them on the right
scaphoid. It is possible that these elements were never present.
The overall symmetry of the foot and the notch on the dorsolateral
surface of the proximal end of the second metatarsal strongly
suggest their presence however.

The five metatarsals require little description except for the
middle one (number 4, probably representing a fusion of that
element from both sides). The others, numbers 2 and 3 of each
side, bear normal proximal articular surfaces although those
of the third metatarsals are in contact with each other dorsally
and both have a small facet for this articulation. The distal ends
of these third metatarsals are normal while the seconds both end
in rather amorphous rounded swellings. These four are all curved
with the concave surface dorsal ; that is, they are not modified for
the inverted position of the foot. The outer pair are also curved
laterally at their distal ends. The right second metatarsal is
somewhat shortened, but the other three are all normal length

PARSONS AND STEIN : ANOMALOUS CAT SKELETON 309

although considerably thinner than is usual.

The central, fourth, metatarsal is a symmetrical element with
both ends much modified. Proximally, it bears a small convex
head articulating with the cuboid, while distally it has a larger
convex swelling with a small dorsal projection. The curvature is
the reverse of the other metatarsals with the concave surface
ventral. It lies ventral as well as median to the third metatarsals
and is in contact with them proximally. This gives it a triangular
shape in section although it is rounded distally. Its length is
equal to that of the third metatarsals.

There are also three phalanges presumably associated with
this foot. However, their position cannot be told from the dis-
articulated bones and it is possible that some were lost. Two
of the three have fairly normal proximal concave faces, and
distally small swellings showing none of the usual features. The
third is a short thin cylindrical bone without any distinctive
features.

DISCUSSION

General

The main significance of highly abnormal specimens is the in-
formation they may give concerning developmental processes or
inheritance; however, detailed consideration of these points is
not possible in this case although we will mention possible genetic
mechanisms. It is hoped that the description is detailed and
accurate enough to give future workers another example which
will be useful in reaching am' general conclusions.

The literature of teratology and the genetics of abnormal de-
velopment is very extensive and no attempt at compiling a com-
plete bibliography of the described abnormalities was made. The
papers cited include only those available ones which seem to
bear most directly on the problems discussed. More extensive
bibliographies of teratology can be found in such general works
as Bateson (1894), Schwalbe (1906-1913), Schlegel (1921), and
Stroer (1937). Griineberg (1947) includes many references to
the genetical literature, especially concerning rodents.

310 bulletin : museum of comparative zoology

Vertebral column

The anomalies shown by the vertebrae of this cat are all well
known and have been much studied, especially in humans. How-
ever, single specimens with so many deformities are rare and
seldom reach maturity. While some of the peculiarities may be
a direct result of the pelvic malformation, others, though quite
possibly caused by the same factors, are not. Possibly some, such
as the fusion of the fourth and fifth lumbars, were caused by
disease after maturity was reached and are thus outside the prov-
ince of this discussion.

Spina bifida (or rachischisis) is a well known defect, but in
most studies on teratology only extreme cases are considered (as
in Kermauner, 1909). Examples such as the eleventh thoracic
and third lumbar of the present specimen should have little or
no adverse effect upon the animal and hence are of small interest
medically. Probably such cases are of more frequent occurrence
than the literature would indicate since Jayne (1898) apparently
found at least one in the cats he used (see his fig. 83, p. 125).
Unfortunately he does not saj' how many were examined, but he
had at least part of fifty or more spinal columns. In mice there
are many mutations which can cause spina bifida (Griineberg,
1952).

The two half-vertebrae, the twelfth thoracic and fifth caudal,
also represent a well known anomaly of great theoretical interest.
Baur (1891), on the basis of several reptilian skeletons, put forth
a theory of increase in the number of segments bj* intercalation
with such hemivertebrae representing incomplete cases. Brash
(1915) studied a human spine with this defect in a cervical, and
reviewed the literature on all cases reported in man. He con-
cluded that the hemivertebrae could arise from either the addi-
tion or loss of half of a normal vertebra and that in his specimen
the condition was secondary, with the absence of the spinal nerve
on one side as the primary cause. Similar cases have been re-
ported in many forms including the cat (Jaj'ue, 1898, fig. 84.
p. 125). The twelfth thoracic in our specimen is presumably an
example of the failure of half of the centrum to form, since
counting it there are the usual thirteen thoracies. Judging from
the foramina, the normal number of spinal nerves was present,
so here the condition is primary. Concerning the fifth caudal, no

PARSONS AND STEIN : ANOMALOUS CAT SKELETON 311

definite statements can be made.

Even commoner and better known are variations in the num-
ber of sacral vertebrae. Bateson (1894) gives many cases of this
and Dwight (1901) discusses its occurrence in man. Jayne (1898)
mentions its occurrence in the cat. The incorporation of the
first caudal into the sacrum is not very common according to
the last author, but no figures are given. On the other hand,
partial fusion of the seventh lumbar and sacrum, often with a
tendency for the lumbar to become involved in the sacroiliac
joint, is exceedingly common. Klatt figures such a condition in a
dog and a cat (1939, figs. 7 and 8, pp. 230 and 231). In the
present case there is no actual participation of the last lumbar
in the joint, but the right side shows modifications of the pleura-
pophysis. This seems to approach sacral articulation with the
extra pelvis.

More recent workers such as Dauforth (1930b) and Backman
(1934) consider the boundaries between the various regions of
the vertebrae to be determined by gradients in differentiation.
Thus the homology of individual vertebrae cannot be determined
by simple counting, whether from the anterior end or any other
selected point. This opinion seems w^ell confirmed by the work
of Sawin and his co-workers, especially Green, on the lumbar
region in rabbits and mice (see Griineberg, 1947, for references).
The twenty-seventh vertebra in the cat here described is, accord-
ing to this conception, partly sacral as w' ell as lumbar ; the name
seventh lumbar is used merely as a convention. This viewpoint
also casts grave doubts on the earlier ideas concerning hemi-
vertjbrae mentioned above.

The spinae bifidae and hemivertebrae could be independent of
tHe leg anomalies with the sacral malformations directly de-
pendent on the latter, although similar cases of deformed sacra
cire known where there are no pelvic abnormalities. However
ir is also possible that all the spinal abnormalities are a result
of the pelvic duplication since they are all consecutive, running
from the tenth thoracic through the sixth caudal. The general
asymmetry of the posterior part of the vertebral column may
also be partly or wholly dependent on the other anomalies since
certainl}" the muscles must have been very abnormal in this area.
However such asymmetry is well known especially in the tail
where it is very common in some forms such as certain dogs which

312 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

are otherwise quite normal (see Klatt, 1939, and Stoekard et al.,
1941, for a discussion of this). Berg (1912) describes a series
of cats in which the tail is greatly deformed. However, in almost
all these cases the tail is markedh' shortened and the malforma-
tion usually is greatest distally, neither of which is true of the
cat under considei-ation. The genetics of such deformed tails
has been studied by a large series of workers, but is not well
understood and need not be considered here. Bamber (1927)
reviews the genetieal literature on eats.

In mice and other rodents there are many mutants listed by
Griineberg (1947 and 1952) which show similar deformation of
the caudal vertebrae. The morphological expression and genetic
action of these various mutations are very diverse. While most
of them tend to shorten the tail, some such as the mutation ' flexed '
(Hunt, Mixter, and Perniar, 1933) have no such effect. In rats,
'stubs' (Ratcliffe and King, 1941) affects not only the tail but
also causes abnormalities in the hind limbs. Both of these are
recessive genes. There are also several dominants such as
'Brachyury' or 'T' (Chesley, 1935) and 'Sd' (Gluecksohn-
Schoenheimer, 1943) which are lethal when homozygous. It is of
interest that in a mouse of genotype T/+ Sd/+ there is a cumu-
lative effect producing a more highly aberrant vertebral column
than either gene does alone, although both normally produce a
similar effect ( Gluecksohu-Schoenheimer, 1943). Most of these
mutations cause not only asymmetry of the vertebrae, but various
other defects, such as spina bifida, are common in mice bearing
these mutant genes.

It therefore seems possible, though in no wise proven, that the
vertebral column of this cat developed abnormally under the
control of a mutation similar to some of those known for mice,
and that this abnormality was accentuated by the occurrence of
another developmental anomaly, genetically controlled or not,
resulting in the formation of a second and much modified pelvis
and pair of limbs. Another possibility is that all the anomalies
are the result of a single event, either a mutated gene or some
environmental factor operating during the very early develop-
ment. If it is a single mutation, then it would probably be dom-
inant although there is the possibility of a recessive or the ex-
treme expression of a homozygote in which the heterozygote is
relatively normal as in Wright's (1935) polydactylous guinea

PARSONS AND STEIN : ANOMALOUS CAT SKELETON 313

pigs. Any sucli theory postulating homozygosity for a mutation
in this cat Avould ahnost certainly require that the father was
not the one suspected but a half-brother of the cat being studied,
since the chance of the same mutation appearing in two unrelated
cats is very small. Another problem which arises in considering
possible genetic mechanisms is that any genes involved may show
a very low penetrance, as is the case in Danforth's (1925 and
1930a) strain of mice to be discussed below. Thus, for example,
the mother and ai^proximately half her offspring could be carry-
ing a gene responsible for the abnormalities described in this
specimen without their showing any phenotypic effects.

Pelvic girdles and appendages

Like most described polymelous specimens, this cat exhibits
posterior duplication — however, in a much modified form. The
commonest pattern of such abnormalities appears to be one in
which two approximately equal pelvic girdles are present, pro-
ducing a Y-shaped main axis of the animal. However, the spinal
column is usually single and much deformed posteriorly. In such
cases there is a marked tendency for the reduction and fusion
of the inner pelvic elements of the two sides. Cases in point have
been described by Conrow (1917) and Mainland (1929) where
six legs were clearly present, and by Lesbre (1900) and Horsley
(1920) where fusion has reduced the number to five.

Reports of cases of asymmetry like the present one are some-
what rarer and far less complete. Murie (1934) reports a five-
legged rabbit, but very little is told of the nature of the extra
leg. Three other papers describe animals in which the doubling
is confined to the appendicular skeleton on one side with the
other side essentially normal: Bateson's (1894) macaque (his
no. 504) shows this doubling extending into the pelvic girdle
while in Weimer's (1942) sheep only the limb, and in Dwight's
(1892) human, only the forearm and hand are affected.

The cat under consideration conforms to neither of these
patterns. There is posterior duplication resulting in four pelvic
limbs and two girdles. Reduction and fusion characterize the
right girdle (the teratological pelvis of the descriptive part) and
limbs, while the left side remains essentially normal except for
the right innominate, thus forming the more normal pelvic girdle
of the cat. It is impossible to tell whether any duplication oc-

314 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

curred in the soft parts. However, the viscera probably did not
show much duplication since the right pelvic canal is completely
closed by extensive fusion.

As mentioned previously, fusion in the limb is a common fea-
ture of polymelous specimens, there being great variation in de-
gree. Here the femora, fibulae, and feet are very closely fused
but the two tibiae are completely separate. The macaque of
Bateson (1894) is very similar in this respect except that there
the positions are reversed and the fibulae are separate and the
tibiae joined together.

There is very little known about possible genetic mechanisms
causing such posterior duplication. Danforth (1925 and 1930a)
has described a mutant strain of mice with this character, but
the exact mode of inheritance is not known. By selecting for
it, he was able to develop only a line with an incidence of 12 per
cent, so it is considered probably recessive. The expression was
very variable with extensive fusion normally occurring. The
spinal column was also abnormal with the tail generally kinky,
and occasional cases of spina bifida. In the case of the cat de-
scribed in this paper, there is no evidence that the cause of the
duplication is genetic as opposed to environmental.

The functional problems arising from the extra leg are of
some interest, but little can be definitely said about them. It
seems most unlikely that the much deformed double foot could
have borne any weight, so that leg was presumably of little use
despite the very smooth hip articulation. The more normal right
leg, on the other hand, has much irregular and apparently
arthritic ossification at the hip resulting in what must have been
an almost immovable joint. Despite the presumably compensa-
tory freedom at the knee evidenced by the articular surfaces
there, this leg too would seem greatly inferior to the normal
condition. Thus it seems probable that the cat depended on the
three normal limbs, making little use of the two right hind legs.

SUMMARY

The cat skeleton described in this paper shows an unusual type
of posterior duplication in which the left pelvic girdle and limbs
are relatively normal and the right ones much modified. The
right girdle is a single Y-shaped element fused to the antero-
ventral border of the right innominate of the left pelvis, and bears

PARSONS AND STEIN : ANOMALOUS CAT SKELETON 315

one limb. This consists of a single element representing two
femora, two tibiae, one fibula, and a highly abnormal foot in
the flexed position bearing metatarsals 1, 2, 3, 4, 3, 2, and 1. The
vertebrae from the tenth thoracic through the sixth caudal are
all abnormal showing various degrees of asymmetry and fusion
and cases of spinae bifidae and hemivertebrae. Possible genetic
causes of such a condition are discussed, but no definite conclu-
sions can be reached concerning this case.

BIBLIOGRAPHY

Backman, G.

1934. Die Abliangigkeit morphologischer Variationeu von Differenzie-
rungs- und Waehstunisgradienten. Anat. Anz., vol. 79, pp. 78-87.

Bamber, E. C.

1927. Genetics of domestic cats. Bihliog. Genetica, vol. 3, pp. 1-86.

Bateson, W.

1894. Materials for the study of variation. London.

Baur, G.

1891. On intercalation of witchrae. Jour. Morph., \'ol. 4, pp. 331 33G.

Berg, W.

1912. Ueber stummelsehwanzige Katzen und Hunde. £in Beitrag zu
der Lehre von der Eeduktion des Schwanzes. Zeitschr. Morph.
Anthrop., Sonderheft 2, pp. 227-267.

Brash, J. C.

1915. Vertebral column witli six and a half cervical and thirteen true
thoracic vertebrae, with associated abnormalities of the cervical
spinal cord and nerves. Jour. Anat. Physiol., vol. 49, pp. 243-273.

Cheslet, p.

1935. Development of the short-tailed mutant in the house mouse. Jour.
Exp. Zool., vol. 70, pp. 429-459.

CONROW, S. B.

1917. A six-legged rat. Anat. Rec, vol. 12, pp. 365-370.

Danforth, C. U.

1925. Hereditary doubling suggesting anomalous chromatin distribu
tion in the mouse. Proc. Soc. Exp. Biol. Med., vol. 23, pp. 145
147.
lySOa. Developmental anomalies in a special strain of mice. Anier. Jour.
Anat., vol. 45, pp. 275-287.

yl6 BULLETIN : MUSEHJM OF COMPARATIVE ZOOLOGY

1930b. Numerical variation and homologies in vertebrae. Amer. Jour.
Phys. Anthrop., vol. 14, pp. 463-481.

DWIGHT, T.

1892. Fusion of hands. Mem. Boston Soc. Nat. Hist., vol. 4, pp. 473-
486.

1901. Description of the human spines showing numerical variation in
the Warren Museum of the Harvard Medical School. Mem. Bos-
ton Soc. Nat. Hist., vol. 5, pp. 237-312.

Gluecksohn-Schoenheimer, S.

1943. The morphological manifestations of a dominant mutation in
mice affecting tail and urogenital system. Genetics, vol. 28,
pp. 341-348.

Gottlieb, H.

1914. Die Antikliuie der Wirbelsiiule der Saugetiere. Morph. Jahrb.,
vol. 49, pp. 179-220.

Gruneberg, H.

1947. Animal genetics and medicine. London.

1952. The genetics of the mouse. 2nd ed. The Hague.

HORSLEY, J. S., Jr.

1920. A description of a six-legged dog. Anat. Rec, vol. 19, pp. 1-27.

lIi'NT, H. R., R. MixTER, and D. Permar

1933. Flexed tail in the mouse, Mus musculii^s. Genetics, vol. 18, pp.
335-366.

Jayne, H.

1898. Mammalian anatomy. Part I. The skeleton of the cat. Phila-
delphia.

Kermauner, F.

1909. Missbildungen des Rumpfes (Fortsetzung). In Schwalbe, E.,
Die Morphologie der Missbildungen des Menschen und der Tiere.
Jena. vol. 3, Lief. 1, pp. 86-112.

Klatt, B.

1939. Erbliche Missbildungen der Wirbelsiiule beim Hund. Zool. Anz.,
vol. 128, pp. 225-235.

Lesbre, F. X.

1900. Observations de deux vaches et d'une cane pygomeles, avec con-
siderations generales sur la pygomelie. Jour. yied. Vet., ser. 5,
vol. 4, pp. 625-652.

PARSONS AND STEIN! ANOMALOUS CAT SKELETON '517

Mainland, D.

1929. Posterior duplicity in a dog, with reference to mammalian tera-
tology in general. Jour. Anat., vol. 63, pp. 473-495.

MURIE, A.

1934. A five legged rabbit. Jour. Mammal., vol. 15, p. 162.

Ratcliffe, H. L., and H. D. King

1941. Developmental abnormalities and spontaneous diseases found in
rats of the mutant strain, stubs. Anat. Rec, vol. 81, pp. 283-305.

SCHLEGEL, M.

1921. Die Missbildungen der Tiere. Ergebn. AUgem. Path. u. Path.
Anat., vol. 19, pp. 650-732.

Schwalbe, E.

1906-1913. Die Morphologie der Missbildungen des Menschen und der
Tiere. 3 vols. Jena.

Stockakd, C. E., et al.

1941. The genetic and endoerinic basis for differences in form and
behavior as elucidated by studies of contrasted pure-line dog
breeds and their hybrids. Amer. Anat. Mem., no. 19.

Stroer, \V. F. H.

1937. Die Extremitatenmissbildungen und ihre Beziehungen zum
Bauplan der Extremitiit. Zeitschr. Anat. Entw.-Gesch., vol. 103.
pp. 136-160.

Weimer, B. E.

1942. A case of hind-limb duplication in the sheep. Proc. West Virginia
Acad. Sci., vol. 15, pp. 82-84.

Wright, S.

1935. A mutation of the guinea pig, tending to restore the peutadactyl
foot when heterozygous, producing a monstrosity when homo
zygous. Genetics, vol. 20, pp. 84-107.

PLATE 1

Plate 1. Mounted skeleton of the eat. The greater part of the tail, the ribs,
sternum, phalanges, clavicles, patellae, and hyoid apparatus are
not included.

PLATE ■:?

I'LATK 1^. X-i-,iy of the (losteiidr p;nt of tlu' cat sliow'.iig tlic regular right
anil teratological h'lul linilis in lato'al view. x%.

PLATE 2

PLATE 3

Plate ?>. A. snapslicit of the cat in life. B and C, teratological foot in
ventral (normal dorsal) and dorsal (normal i)lantar) views. 7-5
••nid (" x4/.T.

*•, , I-

PLATE 3

PLATE 4

Plate -i. Teiatological foot in laternl view compared to ii iKirmal left foot
in dorsal view. x6/5.

PLATE 4

Bulletin of the Museum of Comparative Zoology

AT HARVARD C O L L t: G E

Vol. 114, No. 7

POST-TRIASSIC NAUTILOID GENERA

By Berniiard Kummel

With Twenty -eight Plates

CAMBRIDGE, MASS., U.S.A.
PRINTED FOR THE MUSEUM

August, 1956

Publications Issued by or in Connection

WITH THE

MUSEUM OF COMPARATIVE ZOOLOGY
AT HARVARD COLLEGE

Bulletin (octavo) 1863 — The current volume is Vol. 114.

Breviora (octavo) 1952 — No. 57 is current.

Memoirs (quarto) 1864-1938 — Publication was terminated with Vol. 55.

JoHNSONiA (quarto) 1941 — A publication of the Department of Mollusks.
Vol. 3, no. 35 is current.

Occasional Papers of the Department of Mollusks (octavo) 1945 —
Vol. 1, no. 18 is current.

Proceedings of the New England Zoological Club (octavo) 1899-
1948 — Published in connection with the Museum. Publication terminated
with Vol. 24.

Bulletin of the Museiun of Comparative Zoology

AT HARVARD COLLEGE

Vol. 114, No. 7

POST-TRIASSIC NAUTILOID GENERA

By Bernhabd Kummel

With Twenty-eight Plates

CAMBRIDGE, MASS., U.S.A.
PRINTED FOR THE MUSEUM

August, 1956

No. 7 — Post-Triassic Nautiloid Genera

By Bernhard Kummel
TABLE OF CONTENTS

Page

Introdiu'tion 324

Classification of Post-Triassic Nautiloids 325

Evolution 344

Systematic Paleontology 360

Family Xautilidae d 'Orbigny, 1840 3f50

Subfamily Nautilinae d 'Orbigny, 1840 361

Genus Cenoceras Hyatt, 1883 361

Genus Eutrephoceras Hyatt, 1894 377

Genus Pseudocenoccras Spath, 1927 384

Genus Carinonautilus Spengler, 1910 385

Genus Ohinautilus Kobayashi, 1954 386

Genus Nautilus Linne, 1758 387

Subfamily Pseudaganidinae nov 389

Genus Pseudaganides Spath, 1927 390

Genus Pseudonautilus Meek, 187fj 397

Subfamily Paracenoceratinae Spath, 1927 399

Genus Paraccnoceras Spath, 1927 402

Genus Aulaconautilus Spath, 1927 409

Genus Tithonoceras Eetowski, 1894 410

Genus SomalinautiJns Spath, 1927 411

Subfamily Cymatoceratinae Spath, 1927 412

Genus Cymatoceras Hyatt, 1884 417

Genus Paracymatoceras Spath, 1927 427

Genus Procymatoceras Spath, 1927 428

Genus Cymatonautilus Spath, 1927 430

Genus Anplonautilus Spath, 1927 430

Genus Eucymatoceras Spath, 1927 431

Genus Syrionautilus Spath, 1927 433

Genus Eeminautilus Spath, 1927 434

Genus Deltocymatocera.s n.gcn 438

Genus Epicymatoceras n.gen 439

Subfamily Hercoglossiuae Spath, 1927 440

Genus Cimomia Conrad, 1866 445

Genus AnguUthes Montfort, 1808 452

Genus Hercoglossa Conrad, 1866 457

Page

Genus Aturoidea Vredenburg, 1925 462

Subfamily Aturinae Hyatt, 1894 464

Genus Aturia Bronn, 1838 464

Addendum 468

Bibliography 469

ILLUSTRATIONS

PLATES
Plate

1. Cenoceras orbignyi

2. Cenoceras trechmanni

3. Cenoceras striatus

4. Cenoceras simillimus

5. Cenoceras inornatus

6. Cenoceras hurtonenais

7. Cenoceras and Syrionautilus

8. Paracenoceras and Cenoceras

9. Ohinautilus and Cenoceras

10. Heminautilus and Pseudocenoceras

11. Carinonautilus ariyalurensis

12. Pseudonaiitilus geinitzi

13. Paracenoceras hexagonnm

14. Aulaconautilus and Somalinautilus

15. Tithonoceras zitteli

16. Cymatoceras pseudoelegans

17. Procymaioceras suhtrwicatus

18. Cymatonautilits julii

19. Paracymatoceras asper

20. Anglonautilus undulatus

21. Eucymatoceras plicatus

22. Deltocymatoceras leiotropis

23. Epicymatoceras vaelsensis

24. Cimomia burtini

25. Angulithes and " Bisiphytes' '

26. Angulithes triangularis

27. Hercoglossa orhiculata

28. Aturoidea and Pseudaganides

FIGURES ^

Figure Page

1. Bar chart of nautiloid genera per period 345

2. Phylogenetie relationships of late Triassic and early Juras-
sic nautiloids • • • • 347

3. Phylogeny of post- Triassic nautiloids 348

4. Bar chart of number of genera of post-Triassic nautiloids

per series 353

5. Bar chart of number of species of post-Triassic nautiloids

per series 357

6. Distribution map of Cenoceras 358

7. Distribution map of Eutrephoceras 359

8. Cenoceras 363

9. Cenoceras 365

10. Cenoceras 367

11. Cenoceras excavatus 369

12. Cenoceras burto?ievsis 371

13. EiitrepJioceras 378

14. Pseudaganides 391

15. Ps.eudaganides and Pseitdonautilu-s 393

16. Pertnoceras and Pseudonautilus 399

17. Distribution map of Paracenoceratinae 401

18. Paracenoceras hexagonum • 403

19. Paracenoceras and Aulaconautilus 404

20. Paracenoceras and Reminautilus 406

21. Distribution map of Lower Cretaceous Cymatoceratinae . . . 414

22. Distribution map of Upper Cretaceous Cymatoceratinae . . 415

23. Cymatoceras, Paracymatoceras, Deltocymatoceras, Procyma-
toceras, Epicymatoceras 419

24. Cymatoceras and Paracymatoceras 422

25. Procymatoccras suhtruncatus 429

26. Anglonautilus undidatus 431

27. Eucymatoceras pUcatus 432

28. Reminautilus saxbii 436

29. Distribution map of Cretaceous Hercoglossinae 442

30. Distribution map of Tertiary Hercoglossinae 443

31. Hercoglossa, Cimomia, and Atiimidcs 446

32. Cimomia and Angidithes 448

33. Angidithes 454

34. Hercoglossa and Aturoidea 458

35. Distribution map of Aturinae 465

INTRODUCTION

Studies on tetrabranchiate cephalopods have added much data
to our understanding of the patterns and modes of evolution as
displayed by fossil animals. Few invertebrate phyla show such
marked periods of expansion in evolutionary intensity con-
trasted with periods of extreme contraction to the point of near
extinction. The generalized evolutionary history of the ammon-
oids with periods of near extinction at the end of the Paleozoic
and again at the end of the Triassic, and their final extinction
at the end of the Cretaceous is well known. Among the nautiloid
cephalopods a similar evolutionary history exists but with some
important differences.

The nautiloid cephalopods had their greatest radiation soon
after their first appearance in the Upper Cambrian, that is in
the Ordovician. In this early phase of their history numerous
distinct phyletic trends are well established. However, by
Silurian time there is already a marked decline in the number
of genetic groups. This decline in the viability of the nautiloids
continued, until today we have only a single genus (Nautilus)
living as a relic form in the southwest Pacific. In this long,
gradually diminishing diversity the nautiloids nearly became
completely extinct in the late Triassic, a period of extinction
coinciding with a similar phase in the evolution of the ammon-
oids. A single stock derived from the Triassic Syringonautilidae
survived into the Jurassic. From this surviving stock all post-
Triassic nautiloids are directly or indirectly derived.

The purpose of this study is to analyze the evolutionary his-
tory of the family Nautilidae which includes the surviving stock
from the Triassic and all post-Triassic nautiloid genera. This
phyletic group is of particular interest since it includes an evo-
lutionar}^ complex which has survived a period of near extinc-
tion of the class, went through a minor eruptive phase in the
development of new adaptive types, and is now represented by
a single genus. No explanation is given as to the cause of the

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 325

decreasingr viability of the nautiloids, but the patterns anrl modes
of evolution of this declining group are instructive.

Post-Triassic nautiloids have received but scant attention,
most workers having merely referred all species to the genus
Nautilus. In 1927, L. F. Spath presented a very comprehensive,
but short, review of post-Triassic nautiloids and proposed a
taxonomic scheme including many new genera. In 1951, I spent
eight months at the British Museum of Natural History, and at
that time had the opportunity of studying the very large and
excellent collections of that institution including the types of
many of the genera proposed by Spath. At the time Spath
(1927a) proposed his taxonomic scheme for post-Triassic nau-
tiloids he did not give any diagnosis of his genera or illustrations
of his type species, many based on Sowerby species not previously
illustrated except in woodcuts. In the present report all genera
of post-Triassic nautiloids are diagnosed and illustrated. An
intensive search has been made of the literature to bring together,
as far as possible, all species names that have been proposed over
the past hundred odd years. In this way data was obtained on
the variability, geologic range and geographic distribution of
each generic group.

This study is an outgrowth of the preparation of a chapter
on Mesozoic nautiloids for the Treatise on Invertebrate Paleon-
tology, edited by R. C. Moore. A similar study on Carboniferous
to Triassic nautiloids has already been published (Kummel,
1953e). I wish to express my very sincere appreciation to Dr.
L. F. Spath for many long hours of stimulating conversations
and for facilitating in every way my work at the British Mu-
seum. Fruitful discussions with Dr. Curt Teichert and Dr.
Ernest Williams on taxonomic and evolutionary problems were
most helpful.

CLASSIFICATION OF POST-TRIASSIC NAUTILOIDS

The only previous comprehensive discussion of post-Triassic
nautiloids is that of L. F. Spath (1927a). In recent years several
large works on Tertiary nautiloids have been produced by Miller
which have included descriptions of many new Tertiary faunas.
Spath 's discussion of post-Triassic nautiloids was only a small
and incidental part of his brilliant work on the famous Jurassic

326 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

fauna of Kachh (Cutch). Even so, he laid the framework for
future work on these animals.

Nautiloids of Mesozoic age and, until recently, Tertiary age
have received only brief incidental treatment in paleontologieal
publications. For one thing post-Triassic nautiloids are not
common fossils compared to the contemporaneous ammonoid,
pelecypod, or gastropod faunas. They likewise are seemingly
not nearly so diverse. The evolutionary decline of the nautiloids,
already very marked in the Silurian, progresses steadily towards
the Recent. Thus in studying Mesozoic and Tertiary nautiloids
we are closely examining the decline and near extinction of a
large animal group.

Until the contribution by Spath (1927a) there had been few
genera proposed for post-Triassic nautiloids and most of these
were not generally accepted or used. The most common practice
was to place all post-Triassic species in the genus Nautilus.
Those genera that had been proposed before Spath 's 1927 re-
vision include Nautilus Linne, 1758, Bisiphytes Montfort, 1808,
Cenoceras Hyatt, 1883, Eutrephoceras Hyatt, 1894, Digonioceras
Hyatt, 1894, Angulithes Montfort, 1808, Hercoglossa Conrad,
1866, Pseudonautilus Meek, 1876, Cimomia Conrad, 1866, Cyma-
toceras Hyatt, 1883, Tithonoceras Retowski, 1893, Carinonautilus
Spengler, 1910, and Aturia Bronn, 1838. As mentioned above,
most of these genera were never used after their original pro-
posal ; however, such generic names as Eutrephoceras, Herco-
glossa, Cymatoceras, and Aturia are fairly common in the early
literature. Most of these genera were considered as members of
the Nautilidae.

In his revision of post-Triassic nautiloids Spath (1927a) pro-
posed 16 new genera and adopted 5 families, namely the Nautili-
dae d'Orbigny, 1840, the Hercoglossidae Spath, 1927, the Cyma-
toceratidae Spath, 1927, the Paracenoceratidae Spath, 1927, and
the Aturidae Hyatt, 1894. On a straight morphological basis,
the Nautilidae include those genera directly or indirectly in the
main stock giving rise to Nautilus. The Hercoglossidae include
those genera with '"goniatitic'* or very sinuous sutures. The
Cymatoceratidae include those stocks characterized by ribbing
— the only stock of po.st-Triassic nautiloids with anj' ornamen-
tation. The Paracenoceratidae are characterized by differentia-
tion of the periphery, and the Aturidae are characterized by

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 327

the unique and peculiar dorsal siphuncle.

There have been no significant changes made to this taxonomic
arrangement since its proposal in 1927. Restudy of the large
nautiloid collections in the British Museum and the Museum of
Comparative Zoology plus a comprehensive review of the litera-
ture have brought forth further data on the range of variation
within the various genera, and also new data on the geographic
and stratigraphic range of most of the species described.

The nautiloids reflect no significant change in their evolution-
ary development from the late Paleozoic into the Triassic. In
fact, the evolutionary pattern of Triassic nautiloids is merely
a culmination of trends begun back in the Carboniferous ( Kum-
mel, 1953c). Towards the end of the Triassic most of the long-
lived stocks became extinct. A single persisting stock repre-
sented by Cenoceras survived the Triassic, and it represents
the root form from which, directly or indirectly, all post-Triassic
nautiloids are derived. On examination of the available data on
post-Triassic nautiloids, one is impressed by the essential homo-
geneity of the group. Distinctive radiations are recognizable,
each representing an elaboration of particular morphological
characters. Homeomorphous developments within the group and
with pre-Jurassic genera are common. The time-space relation-
ships of the post-Triassic nautiloid "species" and their mor-
phological modifications, interpreted in terms of adaptive radia-
tion, make possible a constructive phylogenetic interpretation of
the group. The interpretation presented in this report gives a
much clearer understanding of the various generic groups and the
kinds and ranges of variation within the groups. The classifica-
tion adopted reflects the phylogeny of post-Triassic to Recent
nautiloids. The classification used in this report is as follows:

Family Nautilidae d'Orbigny, 1840
Subfamily Nautilinae d 'Orbigny
Genus Nmitilus Linne, 1758

Type species: XctidiUis ponipiHns Limn'
Genus Cenoceras Hyatt, 1883

Type species: Nautilus orbignyi Prinz
Genus Eutrephoceras Hyatt, 1894

Type species: Nautilus delayi Morton
Genus Pseudocenoceras Spath, 1927

Type species: Nautilus largilliertianus d'Orbigny

328 , BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Genus Carinonautilus Spengler, 1910

Type species: Carinonautilus ariyalurensis Spengler
Genus Obinautilus Kobayashi, 1954

Type species: Obinautilus pulclira Kobayashi
Subfamily Pseudaganidinae nov.
Genus Pseudaganides Spath, 1927

Type species : Nautilus Tcutchensis Waagen
Genus Pseudonautilus Meek, 1876

Type species: Nautilus geinitzi Oppel
Subfamily Paracenoceratinae Spath, 1927
Genus Paracenoceras Spath, 1927

Type species: Nautilus hexagonus J, de C. Sowerby
Genus Aulaconautilus Spath, 1927

Type species: Tautilus sexcarinatus Pictet
Genus Somalinautilus Spath, 1927

Type species: Nautilus antiquum Daeque
Genus Tithonoceras Retowski, 1894

Type species : Tithonoceras zitteli Retowski
Subfamily Cymatoceratinae Spath, 1927
Genus Cymatoceras Ilyatt, 1883

Type species: Nautilus pseudoelegans d'Orbigny
Genus Procymatoceras Spath, 1927

Type species: Nautilus subtruncatus Morris and Lycett
Genus Cymatonautilus Spath, 1927

Type species: Nautilus julii d'Orbigny
Genus Paracymatoceras Spath, 1927

Type species : Nautilus asper Oppel
Genus Syrionautilus Spath, 1927

Type species: Nautilus libanoticus Foord and Crick
Genus Anglonautilus Spath, 1927

Type species : Nautilus undulatus J. Sowerby
Genus Eucymatoceras Spath, 1927

Type species: Nautilus plicatus Fitton
Genus Heminautilus Spath, 1927

Type species : Nautilus saxbii Morris
Genus D.eltocymatoceras n.gen.

Type species: Nautilus leiotropis Schliiter
Genus Epicymatoceras n.gen.

Type species: Nautilus vaelsensis Binckhorst
Subfamily Hercoglossinae Spath, 1927
Genus Hercoglossa Conrad, 1866

Type species : Nautilus orbiculatus Tuomey
Genus Cimomia Conrad, 1866

Type species: Nautilus burtini Galeotti

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 329

Genus Angulithes Montfort, 1808

Type species: Nautilites triangularis Montfort
Genus Aturoidea Tredenburg, 1925

Type species: Nautilus parl-in,'<oni Edwards
Subfamily Aturinae Hyatt, 1894
Genus Aturia Bronn, 1838

Type species: Nautilus aturi Basterot

Because of the essential homogeneity of post-Triassic nauti-
loids their grouping into a single family — Nautilidae — seems
most appropriate. Thus each of the distinctive radiations from
the main evolving stock is reflected at the subfamily level and
the radiations within the subfamilies are expressed at the generic
level.

Several genera that have been proposed are considered to be
synonyms of previously described forms. Bisiphyfes Montfort,
1808, is suppressed because of the great ambiguity connected
with the type species. Sphaeronautilus Spath, 1927, Digonio-
ceras Hyatt, 1894, Ophionautilus Spath, 1927, and Nautilites
Prinz, 1906, are considered to be synonyms of Cenoceras. Her-
coglossoceras Spath. 1927, is considered to be a synonym of
Pseudagayiidcs Spath. The single species assigned to Xeocymaio-
ceras Kobayashi, 1954, falls within the range of variation of
Cymatoceras and the genus is suppressed. Vorticoceras Scott.
1940, was established before its author knew of Heminautilus
Spath, 1927, and was suppressed by Scott at a later date (Scott,
1943). PlafynaKtilus Yabe and Ozaki, 1953, is a perfect synonym
of Heminautilus. Ddtoidonautdus Spath, 1927, is a synonym
oi Angulithes Montfort, 1808 (Kummel, 1953a). Enclimatoceras
Hyatt, 1883, is a synonym of Hercoglossa Conrad, 1866 (Spath,
1927a). Woodringia Stenzel, 1940, also is here considered a syn-
onym of Hercoglossa. Paraturia Spath (1927) was proposed
prior to his ( Spath 's) knowledge of Aturoidea Vredenburg.
1925, and was shortly thereafter properly suppressed for Aturoi-
dea which has priority.

The basic framework of the above classification is the work
of Spath (1927a) but differs in some important aspects involving
the general interpretation and phylogenetic relationships of
post-Triassic nautiloids. Spath (1927a) originally proposed his
major units as families, but later (1935b) he did use a subfamily
rank while describing a species of Paracenoceras (Paraceno-

330 niTLLETIN : MUSEUM OF COMPARATIVE ZOOLOGV:

ceratinae). Within the Nautilinae the main difference in the
above classification from that of Spath is placing Angulithes in
the Hercoglossinae, and including Carinonautilus in the Nau-
tilinae, which Spath had placed in the Paracenoceratinae. Also
several genera of 8path's Nautilidae {Bisiphytes, Sphaeronau-
tiliis, Digonioceras, and Ophionautilits) are suppressed or placed
in synonymy of other genera.

The Pseudaganidinae is a new subfamily proposed to dif-
ferentiate those nautiloids with sinuous sutures that arose as a
separate phyletic stock in the early Jurassic, and are distinct
from the Cretaceous-Tertiary radiation including Hercoglossa
etc. Spath (1927a) had included the two genera of the Pseuda-
ganidinae {Fseudaganides and Paendonautilus) in the Herco-
glossidae. From Spath 's Paracenoceratidae are removed Cari-
nonautilus to the Nautilinae and B cm inaufilus to the Cymato-
ceratinae. The subfamily Aturinae includes only the genus
Aturia. The Cymatoceratinae remains essentially as conceived
by Spath except for the inclusion of Heminautilus and two new
genera described in this report.

The relative rarity of post-Triassic nautiloids has led to a
situation where most species have been deseril)ed on the basis of
one or very few specimens. To my knowledge, no large collection
representing a population has ever been assembled from a single
horizon and locality nor are any known Avhere the fossils are
sufficiently Avell preserved to allow a thorough study of intra-
specific variation. The great majority of species are defined in
terms of being slightly thinner, fatter or more involute or
evolute than other already established species ; or there are
slight variations in the whorl shape or character of the suture.
However, in none of these species is the range of variations in
any of their morphological features known. Because of these
factors, the so-called species of Mesozoic and Tertiary nautiloids
have little reality in terms of our concepts of neontological
species. Thus, in attempting to decipher the evolutionary pat-
terns and modes of these nautiloids, it is impossible to use the
species as a working unit. One's thinking and correlation of data
need to be at the generic level. The species (representing the
named units accumulated in the literature) when brought to-
gether in generic groups give very useful data in regard to range
and direction of variation, and relative abundance of the various

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 331

adaptive types. Since so many of the so-called species are
merely descriptions of specimens, they are here treated as ex-
amples of morphologic types brought together into genera —
which are the only feasible taxonomic units to handle with this
sort of material. Another difficulty is the complete lack of
understanding of the adaptive values of the various morphologi-
cal features. Most features of the conch are thought to be
adaptive, but the exact nature of this adaptation is not known.

To properly appraise the evolutionary history of post-Triassic
nautiloids it was essential that all of the named units be brought
together and placed in generic groups. Thus, through the named
units (species), the geologic history and range of variation,
origin, and evolution of each generic group could be interpreted
in terms of the whole nautiloid fauna. In assembling a list of
described species, there are numerous problems of synonymy and
other taxonomic irregularities that cannot possibly l)e determined
just from the literature, and it is doubtful at this stage if such
efforts Avould be worthwhile. Thus in compiling a list of the
species of post-Triassic nautiloids most names that have been
introducetl are listed, except for several very obvious and long
recognized cases of synonymy. The following list of 531 species
of ])Ost-Triassic nautiloids gives the units upon which much of
this study is based. Because of the great difficulty in tracking
down many of these forms the author, date, original generic
assignment, and the i)resent generic assignment are given. The
list is reasonably complete but even after several years of ac-
cumulating these data it is certain that there are still species that
have escaped my attention. This list of forms does represent the
great majority and from it certain interesting observations on
relative abundance of the various adaptive types can be made.
This problem will be discussed later.

SPECIES OF POST-TRIASSIC NALTILOIDS

Xautiluti (Faiacenoceras) acklini Jcannet, 1951 — Panircnoceras

y . adneticus Pia, 1914 — Cenoceras

y. (Hercoglossa) aegyptiacus Foord, 1891 — Angulithes

A", affinlis Chapuis and Dewalque, 1853 — Cenoceras

A", aganiticus Schlotheim, 1820 — Pscn(laga>nides

.V. ahltcnensis Schliiter, 1876 — Eutr.ephoceras

y. alabamensis Morton, 1834 — Aturia

332 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

Aturia alaskensis Schenck, 1931 — Aturia

N. alhensi.s d 'Orbigny, 1850 — • Cijmatoceras

Eutrephoccras nJcescnse Eeeside, 1927 — Eutrephoceras

N. allani Fleming, 1945 — Eutrephoceras

N. allioni Miehelotti, 1840 — Eutrephoceras

K. altavensis Pomel, 1889 — Cymatoceras

N. altifrovti Chapman, 1915 — - Eutrephoceras

N. altisiphites Prinz, 1906 — Cenoc.eras

N. amasianus Gugenberger, 1928 — Cenoceras

N. ammoni Loesch, 1912 — Pseudaganides

N. amorettii Parona, 1897 — Cenoceras

Cymatoceras andranofotsyense CoUignon, 1951 — Cymatoceras

N. anguliferous Schliiter, 1876 — Cymatoceras

N. angustata Conrad, 1849 {in Dana) — Aturia

N. angustus Blanford, 1861 — Cimomia

X. anomplialus Pia, 1914 — Cenoceras

X. (iiifiqifus Daeque, 1910 — Somalinautilus

X. (ipplanal i(s Wanner, 1902 — Fseudocenoceras

X . arariforin'm Pia, 1914 — Cenoceras

X . araris Dunioitier, 1869 — Cenoceras

X. aratus Quenstedt, 1846 — Cenoceras

X. <ircliiacia7iHs d'Orliigny, 1840 — Eutrephoceras

X. arcuatus Deshayes {in Leymerie, 1842) — Angidithes

X. ardiiennensis Loeseli, 1914 — Faraccnoceras

X. argoviensis Loesch, 1912 — Pseudaganides

Carinoiiaitlihis ariyalnrensis Spengler, 1910 — Carinonautilus

X. (irlhahrri Gugenberger, 1928 — Cenoceras

X. (irtictdatus Pultency, 1813 {in Cox, 1940) — Cymatoceras

X. aspcr Oppel, 1865 — Paracymatoceras

X . a.stacnides Young & Bird, 1828 — - Cenoceras

X. (ithis Whiteaves, 1876 — Cymatoceras

X. atari Basterot, 1825 — Aturia

X . aturioides Pictet, 1867 — Pseudonautilus

Atnvia australis McCoy, 1867 — Aturia

X . austriacus Hauer, 1856 — Cenoceras

X. averilli Anderson, 1938 — Cymatoceras ?

.V. haheri Morris & Lycett, 1850 — Proeymatoceras

X. baconicus Vadasz, 1911 — Cenoceras

Dcltoidcnautihi.s hakeri Teichert, 1947 — Angulithes

X. balcombensis Chapman, 1915 — Eutrephoceras

X. balsamoorivellii Faroiia, 1897 — Cenoceras

Aturia basteroti Benoist, 1888 — Aturia

X. bniifieldi Foord and Crick, 1890 — Cymatoceras

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 333

A', begudensis Kilian and Eeboul, 1915 — Anglonautilus

N. bellerophon Lundgren, 1867 — Eutrcphoceras

N. hcrriasensis Pictet, 1867 — Pseiulocenoceras

Eiitrephoceras hcrryi Miller, 1947 — EutrcpJioceras

Aturia (Afuria) herryi Stenzel, 1940 — Aturia

N. ( Aidaconmitilus) hicarinatus Jeannct, 1951 — Anlacovautilu.i

X. hifurcatus Ooster, 1858 — Cymatoceras

Deltoidnnautilus biyogoren.sis Haas and Miller, 1952 — Angitlithc.'i.

N. blakei Avnimeleeh, 1947 — Cimomia

N. blanfordi Douville, 1929 — Eutrephoceras

N. bodeni Loesch, 1914 — Pseudaganides

N. boissieri Pictet, 1867 — Eutrephoceras

N. boucJmrdianus d'Orbigny, 1840 — Eutrephoceras

N. bradfordensis Crick, 1898 — Cenoceras

X. brancoi Genmiellaro, 1884 — Cenoceras

Aturia {Brazaturm) brasocnsis Stenzel, 1935 — Aturia

X. breislaclci Parona, 1897 — Cenoceras

X. broitzemensis Miiller and Wollemann, 1906 — Cymatoceras

Aturm (Sphenaturia) briiggeni Ihering, 1921 — Aturia

X. brunliuberi Loesch, 1914 — Pseudaganides

Aturia brunlechneri Frauscher, 1895 — Aturoidea

X. bruntrutanae Kuhn, 1936 — Paracenoceras

X. bryani Gabb, 1877 — Eutrephoceras

Cimomia buccinaeformis Haas and Miller, 1952 — Cimomia

X. burTcarti Castillo and Aguilera, 1895 — Eutrephoceras

X. burtini Galeotti, 1837 — Cimomia

X. burtoncnsis Foord and Crick. 1890 — Cenoceras

X. butoncnsis Martin, 1933 — liutreplioceras

Deltoidonautilus caheni Miller, 1951 — AnguUthes

X. cnlloviensis Oppel, 1858 — Paracenoceras

X. culvimontensis Bede, 1948 — AnguUthes

X. campbelli Meek, 1861 — Cymatoceras ?

A*, cumpichei Karakaseh, 1907 — Pseudocenoceras

X. cantabrigiensis Foord, 1891 — Cimomia

X. (Cymatoceras) carlottensis Whiteaves, 1900 — Cymatoceras

X. caroliameghinoi Ihering, 1902 — Aturia

Eutrephoceras caroUnense Kelluni, 1926 — Eutrephoceras

X. (Eercoglossa) cassinianus Foord and Crick, 1890 — AnguUthes

X. catonis Gemmellaro, 1886 — Cenoceras

X. cenomanensis Schliiter, 1876 — Cymatoceras

X. centraUs J. Sowerby, 1812 — Eutrephoceras

Aturia charlesworthi Foord, 1891 — Aturia

X. charpentieri Leymerie, 1851 — Eutrephoceras

334 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

A", chilennis Huppe (in Gay) 1854 — Cenoceras

N. chudeaui Douville, 1920 — Angulithes

Aturia clarkei Teichert, 1944 — Aturia

N. olausus d'Orbigny, 1842 — P.seudafjanides

A', clementimtfi d'Orbigny, 1840 — Eutrephoceras

Cymatoceras colombiana Durham, 1946 — Cymntoceras

A', colvmhinus Fritsch and Schlonbach, 1872 — Cimomia

N. compressxLs Tavani, 1942 — Cymatoceras

N. cookana Whitfield, 1892 — Eiitrephoceras

N. oossmanni Vredenburg, 1928 — Cimomia

Paracenoceras costatum Scott, 1943 — Procymatooeras

Aturia ? coxi Miller, 1947 — Aturia

A^. crassiconcha Vogl, 1908 — Cimomia

N. crassisinuatus Crick, 1898 — Pseudaganides

N. crassus Schafhautl, 1863 — Eutreplioceras

X. crebricostatiis Blanford, 1861 — Cymatocerax

X. cuhaensis Lea, 1841 — Aturia

Aturia curvilineata Miller and Thompson, 1937 — Aturia

X. cyclotus Oppel, 1865 — Eiitrephoceras

X. danicus Schlotheini, 1820 — Eercoglossa

Eutreplioceras darteveUei Miller, 1951 — Eutrephoc^ras

X. darupensis Schliiter, 1876 — Eiitrephoceras

X . decipiens Michelotti, 1861 — Eutrephoeeras

X. delcayi Morton, 1834 — Eutrephoeeras

X. deluci d'Archiac 1854 — Angulithes

X. demonensis M. Gemmellaro, 1911 — Cenoceras

X. depressus Binckhorst, 1861 — Eutrephoeeras

X. desertorum Quaas, 1902 — Eutrephoeeras

X. deslongcha mpsianns d 'Orbigny, 1840 ■ — • Cymatoceras

Aturia dickersoni Sehenck, 1931 — Aturia

Eercoglossa diderrichi Vincent, 1913 — Eercoglossa

X. dietricM Zwierzycki, 1914 — Eutrephoeeras

X. (Paracenoceras) dilatatus Jeannet, 1951 — Paracenoceras

X. dispansiis Morris and Lycett, 1850 — Cenoceras ?

Aturoidea distans Teichert, 1943 — Aturoidea

X. distefanoi Gemmellaro, 1884 — Cenoceras

X. divesianus Kuhn, 1936 — Paracenoceras

X. domeykus d'Orbigny, 1842 — Cenoceras

X. d'orbignyanus Forbes, 1846 — Cimomia

X. dorsatus Roemer, 1836 — Paracenoceras

X. dorsoexcavatum Parona and Bonarelli, 1897 — Paracenoceras

Eutrephoeeras douvillei Spath, 1927 — Eutrephoeeras

X. drepanensis Tagliarini, 1901 — Pseudaganides

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 335

A', dubaleni Peyrot, 1932 — ■ Eutrephoeeras

N. dubiu.s Zieten, 1830 — Ccnoeeras

X. (luilii Gemmellaro, 1886 — Psendaganides

A', dumasi Pictet, 1867 — Pseudac/anidrs

X. egregius Pia, 1914 — CenoceraH

X. eichwaJdi Karakasch, 1907 — Cymatoceras

X. elcgans J. Sowerby, 1816 — Cymatoceras

X. elcganioides d'Orbigny, 1840 — Cymatoceras

Deltoidonautilus eUiotti Stenzel, 1940 — Angvlithes

X. elliptioiis Scliafhaiitl, 1852^ — Cimomia

X. ejvnianum Dacque, 1905 — Paraeenoceras

X. ercyeinus Tagliarini, 1901 — Cenoceras

Heminautilus etheringtoni Durham, 1946 — Eeminaiitilus

X. euthymi Pictet, 1867 — Eutrephoeeras

X. excavatus, J. de C. Sowerby, 1826 — Cenoceras

X. exiguus Crick, 1898 — Cenoceras

X. expansus J. de C. Sowerby, 1824 — Eutrephoeeras

X. expletus Zwierzycki, 1914 — Cimomia

X. exterebratiis Crick, 1898 — Cenoceras

Eutrephoeeras fax(\ense Hyatt, 1894 — Eutrephoeeras

X. felix Chapman, 1915 — Eutrephoeeras

Aturia (Sphenaturia) felschi Ihering, 1921 — Aturia

X. fischeranus Foord and Crick, 1890 — Cenoceras

X . fittoni Sharpe, 1853 — Pseudocenoceras

X. flammeus Eonehetti, 1947 — Eutrephoeeras

X. fleuriausian us d'Orbigny, 1840 — Angidithes

X. forbesi d'Arehiac and Haime, 1854 — Cimoynia

X. forhesianus Blanford, 1861 — Hercoglossa ?

Aturia formae Parona, 1899 — Aturia

X. formosus Blanford, 1861 — Cymatoceras

X. fourneti Dumortier, 1874 — Cenoceras

X . franeo)uonta<nus Kuhn, 1939 — Eutrephoeeras

X. jranconicus Oppel, 1865 — Pseudaganides

X . (Pseudaganides) fricTcensis Jeannet, 1951 — Pseudaganides

X . fuscus Crick, 1898 — Somalinautilus

X. gabhi Anderson, 1902 — Cymatoceras

X . galea Fritsch and Sehlonbach, 1872 — Angidithes

X . galicianus Alth, 1850 — Pseudocenoceras

Hercoglossa gardnerae Stenzel, 1940 — Hercoglossa

Atitria (Brazaturia) garretti Stenzel, 1940- — Aturia

X . geelongensis Foord, 1891 — Eutrephoeeras

X. geinitzi Oppel, 1865 — Pseudonautilus

X . geyeri Prinz, 1906 — Cenoceras

336 BULLETIN : MUSEUM 01' COMPARATIVE ZOOLOGY

iV. giganfeun d'Orbigny, 1825 — Paracennceras

N. girardoti de Loriol, 1903 — • Psendar/anides

K. glaber Foord and Criek, 1890 — Pseudaganidi\<:

N. gosavicus Eedtenbacher, 1873 — Eutrcphocerns

Aturia grandior Schenck, 1931 — Atiiria

Afuria (Aturia) grangei Fleming, 1945 — Aturia

N. granidosum d'Orlngny, 1843 — ■ Paraccrwcrras

N. gravesianus d'Orbigny, 1843 — Pseudaganidcs

N. guilielmi telli Oostcr, 1858- — Cymatoceras

N. hallidayi Waring, 1914 — Eutrephnrcras ?

N. hallstattensis Spengler, 1919 — - CenoccraH

N. 'haltomi Aldricli, 1931 — Cimomia

Euireplioceras hannai Vokes, 1937 — Eutrephoceras

Ilercoglossa hnrrisi Miller and Thompson, 1937 — Hercoglosaa

N. haughti Olsson, 1928 — Cimomia

A^ hazaraensis Das-Gupta, 1916 — Angulithes

N. lieberti Binekhorst, 1861 — Cimomia

N. helvcticus Loesch, 1914 — Pseudnfianidcs

N. (Cymntnceras? ) Jiendersoni Etheridge (1901) — Eutrcpltoceras

N. (Paracenoceras) hersnachensis Jeaunet, 1951 — Paracenoceras

Cimomia hesperia Miller and Downs, 1950 — Cimomia

Paracenoceras hexagonoides Spath, 1927 — Paracenoceran

N. hexagonus J. de C. Sowerby, 1826 — Paracennceras

N. hilli Shattuck, 1903 — Cymatoceras

N. hiinstamtonensis Foord and Crick, 1890 — Cyviatoceraft

Cimomia Imnti Haas and Miller 1952 — Cimomia

N. huxleyanus Blanford, 1861 — Cymatoceras

N. imbricatus Crick, 1907 — Cymatoceras

Bisiphytes (Cenoceras) imlayi Kummel, 1954 — Ccnoceras

N. impendens Crick, 1898 — Cenoceras

N. imperialis J. Sowerby, 1812 — Cimomia

N. indicum Spengler, 1910 — Eutrephoceras

Hercoglossa innoviinanda Fleming, 1945 — Ilercoglossa

N. inoimatus d 'Orbigny, 1842 — Cenoceras

N. intermedius J. Sowerby, 1816 — ■ Cenoceras

N. interstriatvs Strombeck, 1863 — Cymatoceras

N. intumescens Waagen, 1873 — • Procymatoceras ?

A'. (Javanoceras) intuscatenatus Martin, 1932 — Cimomia

N. izumoensis Yokoyama, 1913 — Evtrcplwceras

N. japonicus Shimizu, 1926 — Eutrephoceras

N. javanus Martin, 1879 — Eutrephoceras

Paracenoceras jeanneti Sanchez Eoig, 1951 — Paracenoceras

Eutrephoceras johnsoni Miller, 1947 — Eutrephoceras

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 331

E litre phoceras joiw.'^i Miller and Thompson, 1933 — Eutrephoceras
N. jordani Wanner, 1902 — Cimomia
N. jourdani Dumortier, 1874 — Cenoceras
N. juliamis Fucini, 1895 — Cenoceras
X. jiilii d 'Orbigny, 1850 — Cymatonautilus
X. jumarensis Waagen, 1873 — Paracenoccraa ?
X. jur.en.sis Qucnstedt, 1858 — Cenoreras
X. Justus Blanford, 1861 — ■ Eutrephoceras
Cimomia l-ar1:arensis Haas and Miller 1952 — Cimomia
X. Icarpinsl-yi Karakasch, 1907 — Faracymatoceras
X. kayeanus Blanford, 1861 — Cymatoceras
Aturia herniana Anderson and Hanna, 1925 — Aturia
X. Mehelshergi Loesch, 1914 — Pseudagnnides
Aganides kochi Prinz, 1906 — Fscudaganides
Aturia koenei Cagel, 1928 — Aturia

X. (Cymatoceras) A-oss7n.afi SpeiigltT, 1910 — Cymatoceras
X. krenkeli Jeannet, 1951 — Pseudaganides
Cimomia kugleri Miller, 1947 — Cimomia
X. kwmagunense Waagen, 1873 — Paracenoceras
X. kutchensis Waagen, 1873 — Pseudaganides
X. labechei d 'Arehiac and Haime, 1854 — Eutrephoceras
X. lallierianus d 'Orbigny, 1840 — ■ Heminautilus
X. lamarckii Deshayes, 1824 — Angulithes
Uercoglossa lamegoi Oliveira, 1953 — Eercoglossa
X. lavdanensis Vincent, 1913 — Cimomia
X. largillicrtianus d 'Orbigny, 1840 — Pseudocenoceras
Aturia (Brazaturia) laticlavia Stenzel, 1935^ — Aturia
X. laiifrons Zwierzycki, 1914 — Paracenoceras
Eutrephoceras Javerdei Durham, 1946 — Eutrephoceras
X. ledonicus de Loriol, 1903 — Pscudagatnides
X. Jehardyi Binckhorst, 1861 — Pseudocenoceras ?
.V. leiotropi-s Schliiter, 1876 — Deltocymatoceras
iJeltoidonautiJus lemoinei Miller, 1951 — Angulithes
X. lentiformis Stoliczka, 1866 — Eutrephoceras
X. Iconei Negri, 1934 — Eutrephoceras
X . leonicensis de Zigno, 1881 — • Cimomia
X. Ubanotieus Foord and Crick, 1890 — Syrionautilus
X. Uneatus J. Sowerby, 1813 — Cenoceras
X. lineolatus Foord and Crick, 1890 — Cenoceras
_Aturia linicentensis Vincent, 1907 — Aturia ?
Cymatoceras loehlichi Miller and Harris, 1945 — Cymatoceras
X. loricatus Schliiter, 1876 — Cymatoceras
Aturia lotzi Bohm, 1913 — Aturia

338 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

Aturm luculoensi'S Miller, 1938 — Aturia

Bisiphytes (Cenoceras) hipheri Kummel, 1954 — Cenoceras

N. lutatii Gemmellaro, 1886 — Cenoceras

Cimoniia macfadyeni Haas and Miller, 1952 — Cimomm

Aturi-a (Brazaturin) vinckayi Fleming, 1945 — Aturia

N. macrocephalus Schafhautl, 1863 — ■ Cimomia

N. (Cymatoceras) madagascarensis Yabe and Shimizu, 1924 — Cymatoceras

Herooglossa madgascnriensis CoUignon, 1951 — Hercoglossa

N. malbosi Pictet, 1867 — Pseudonautilus

N. malherbii Terquem, 1855 — Cenoceras

iV. manissadjiani Gugenberger, 1928 — Cenoceras

N. manuanensis Crick, 1907 — Cymatoceras

Hercoglossa maracaihoensis Miller and ColUnson, 1951 — Hercoglossa

N. mariani M. Gemmellaro, 1911 — Cenoceras

N. marii Gemmellaro, 1886 — Cenoceras

EutrepJioceras marJcsi Miller, 1947 — Eutrephoceras

Parace')ioceras marocense Miller and Collinson, 1952 — Faracenoceras

N. marucoensis Giovine, 1950 — Eutrephoceras

Cimomia marylandensi-s Miller and Thompson, 1933 — Cimomia

Aturia matheivsonii Gabb, 1864 — Aturoidea

N. mazsarensis Tagliarini, 1901 — Cenoceras

Hercoglossa mcglameryne Miller and Thompson, 1933 — Hercoglossa

Hercoglossa merriami Dickerson, 1914 — Hercoglossa

N. metafleuriausi Douville, 1929 — EutrepJioceras

N. meyrati Ooster, 1858 ■ — Cenoceras

N. mikado Krenkel, 1910 — Cymatoceras

Paraoymatoceras milleri Humphrey, 1949 — Paracymatoceras ?

N. trijatschkowanus Loeseh, 1914 — Paracenoceras

X. mojsisoviosi Neumayr, 1870 — Cymatonauiilus

X. mokattamensis Foord, 1891 — Cimomia

X. molli Douville, 1920 — Angidithes

EutrepJioceras montanensis Kunmiel, 1954 — EutrepJioceras

X. montmollini Pictet and Canipic-he, 1859 — EutrepJioceras

X. moreaui d'Orbigny, 1842 — Paracenoceras

Aturia morrissi Michelotti, 1847 — Aturia

Paracenoceras midlerriedi Sanchez Roig, 1951 — Paracenoceras

X. midtiseptatus Foord and Crick, 1890 — Cenoceras

X. munieri Choffat, 1886 — AngulitJies

Aturia myrlae Hanna, 1927 — Aturia

Aturia narica Vredenburg, 1925 — Aturia

X. nebrascense Meek and Haydcn, 1862 — Cymatoceras

X. neckerianus Pietet, 1847 — Cymatoceras

X . negama Blanford, 1861 — Cymatoceras

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 339

.V. negritensi.s Olsson, 1928 — Cimomia

X. neocomiensis d'Orbigiiy, 1840 — Cymntoceras

A". neohispanieuiH Burekhardt, 1925 — Ci/mntoceras

X. neiiherfjicus Rodtenlindior. 1873 — EutrcpJmcern.'i

N. obesus J. Soweiby, 1816 — Cenoceras

N. obstriictus Deslongelininp.s, 1878 — Cenoceras

Xautilus (CymatoccruH.' ) dcclusus Cric-k, 1907 — Cynialoceras

Aturoidea oIh.soiiI Miller, 1947 — Atiiroidca

X. oppell Zittel, 18(58 — Pscu(ki;!(uiid( a

X. orbiciihil us Tuoniey, 1854 — llervnglossa

N. orbignyi Prinz, 1906 — Ccnoceias

Eiitrephoceras orrr/oniiifK .Miller, 1947 — Ealrephocerai^

X. ornatus Foord and Crick, 1890 — Cenoceras

X. urnattis vai'. atdiiati iisis Pin, 1914 — Cenoceras

Eiitrephoceras ovoideimi Cricdv, 1907 — T^itlreplioeeras

Aiuria pamamensis Miller, 1947 — Aturia

X. parahoUcus Sehafhiiiitl, 1863 — Angulithcs

X. parallelus Sehafhiiiitl, 1863 — Ciinouiui

X. paretoi M. Geiiinielhuo, 1911 — Cenoceras

X . parisiensis Deshayes, 1866 — Eiitrephoceras

X. parl:inso7ii Edwards, 1849 — Aturoidea

.lluria paronai Eovereto, 1900 — Aturia

X. patens Kner, 1850 — Cymatoceras

X. paucifcx Cope, 1866 — ^Itiiroidea

X. pavlowi Arkhanguelsky. 1904 — Hercoglossa

X. pellerensis Vialli, 1937 — Pseudaganides

X. perinflaius Foord and Crick, 1890 — Cenoceras

X . perlatiis Mortou, 1834 — Eiitrephoceras

X. pernambucensis Maury, 1930 — Cimomia

X. perornatiis Crick, 1894 — Cenoceras

X. perstriatiis Steiier, 1921 — Cymatoceras

X. pertextus Dumortier, 1867 — Cenoceras

Aturia peruviana Olsson, 1928 — Aturia

Hercoglossa peruviana Berry, 1923 — Hercoglossa

X. (Hercoglossa) pJiosphaticus Bede, 1933 — Cimomia

X. pictcti Karakasch, 1907 — Pseiidocenoceras

X. picteti Oppel 1865 — Aulaconautihis

X . piersantii Sergio, 1933 — Eutrephoccras

Aturoidea pilsbryi Miller and Thompson, 1935 — Aturoidea

X. pisanus Fucini, 1895 — Cenoceras

Eutrephoccras planovcnter Stephenson, 1941 — Eiitrephoceras

X. plicatus Fitton, 1835 — Eucymatoceras

X. polygonalis J. de C. Soweiby, 1826 — Cenoceras

Hercoglossa popenoei Miller and Downs, 1950 — Hercoglossa

340 BULLETIN: MUSEUIVI OF COMPARATIVE ZOOLOGY

i\'. portlandicus Foord and Crick, 1890 — Psetidaganides

N. postriatus Prinz, 1906 — Cenocenis

Aturia praezicsac Oppenheim, 1903 — Atnria

AtiiiHa preaturi (Cuvillier) 1935 — Aturia

N. profundiniphites Prinz, 1906 — Cenoeeraa

Paracenoceras prohexagonum Spath, 1933 — Paracenoceras

N. (Cymatoccras) pseudoatlas YalK' and Sliiniizu, 1924 — Ciimdtoceia.s

N. pseiidohouchardianus Spengler, 1910 — Cimomia

N. pseudoelegans d 'Orbigny, 1840 — Cymatoceras

N. pseudolineatus Foord and Crick, 1890 — Cerioceras

N. (Cymatoccras) psendoncgama Spengler, 1910 — Cymatoceras

Cytnatoccras pseudoneolcoymense Shimizu, 1931 — Cymatoceras

N. pseudorugosus Pia, 1914 — Cenoceras

N. pseudotrimcatus Crick, 1921 — Ccnoceras

N. (Pscudaganides) pidcJtellus Jeannet, 1951 — Pseudagavides

Ohinautilus pidchra Kobayashi, 1954 — Obinautihts

Cimomia piisilla Haas and Miller, 1952 — Cimomia

N. quadrungularis Pia, 1914 — Ccnoceras

X. quadriUncatns Favre, 1869 — Entrrphocerns

Aturia radiata Bellardi, 1872 — Aturia

N. radiatus J. Sowerby, 1822 — Cymatoceras

X. rangei Hoppe, 1922 — Heminautilus

Eutrephoceras reesidel Stenzel, 1940 — Eutrephoccras

X. rcgalis J. de C. Sowerby, 1843 — Eittrcplioccra-^

X. regvlaris Schafhautl, 1863 — Eutrephoceras

X. restrictu.s Griepenkerl, 1889 — Eutrephoceras

X. resupinatiis Eedtenbacher, 1873 — Eutrephoceras

X. rhodami Boiix, 1848 — Paracenoceras

Aturia ricJiardsi Miller, 1947 — Aturia

N. ricordeanus d 'Orbigny, 1847 {in Kilian and Reboul 1915) — Hercoglnssal

N. rohnstus Foord and Crick, 1890 — Ccnoceras

X. roemeri I^oescli, 1914 — Pscudaganides

Dcltoidonautilus rogeri Miller, 1951 — Angidith< s

X. rogeri Loesch, 1914 — Pseudaganides

X. rollandi Leymerie, 1846 — Angutiihes

X . rolUeri Loesch, 1914 — Paracenoceras

X. romeroi Ihering, 1903 — Cimouiia

X. rota Blanford, 1861 — Paracymatocerus

Bigonioceras rotundum Hyatt, 1894 — Cenoceras

N. rotundus Crick, 1898 — Cenoccras

Aturia rovasendiana Parona 1899 — Aturia

X. royeri de Loriol, 1872 — Pseudaganides

X. rugatus Fritscli and Schlonbacli, 1872 — Dcltocymatoceras

X. rugosus Buvignier, 1852 — Cenoceras

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 341

.Y. .sahariensis Keller, 1932 — Cimomia

Ctjmatoceras sakaJavum Colliguon, 1949 — Cymatoceraa

X. sanfiUppoi Sorreutiuo, 1932 — Eutrephoceras

\ . -^attleri Krenkel, 1910 — Paracenoceras

X. siiKfisurcanii.'i Pietet, 18-17 — C if mat oc eras

X. .saxbii Morris, 1848 — Heminautilus

X. scecquren-^ia Tavaui, 1942 — Cymatoceras

X. achattenhcryi Kului, 193G — Paraci noceras

N. scJilosseri Loescli, 1914 — P.seudaganidex

X. .schlotheiini Lo:'Sch, 1914 — Psetulaganides

X. sehlumbergeri Terqucni, 1855 — Cenoceras

X whmidti Giebcl, 1852 — Coioreran

N. schneidi Loeseli, 1914 — PseudaganUles

X. schusteri Loeseli, 1912 — Paracenoceras

X. scJiuahiii Priiiz. 1906 — Cenoceras

X. sclnreinfiirtJii Quaas, 1902 — Atiiroidca

X. schwertschhigeri Loesch, 1912 — Pseiidaganides

X . Heceiviendus Pia, 1914 — Cenoceras

X . scelandi Peuecke, 1884 — Cimomia

.V. (Cymaloccrus) semilobatus JSpeugler, 1910 — Cymatoceras

X. semiornatus Crick, 1898 — Cenoceras

X . semistriatus d'Orbigny, 1843 — Cenoceras

.V. srmiimdatns Foord, 1891 — Cymatoceras

X. scitiseyi Prinz, 1904 — Cenoceras '

X. scuegalonsis Douville, 1920 — Angulilhes

Cimomia septeincastrensis Haas and Miller, 1952 — Cimomia

X. sf rpoifinits Blaiiford. 1861 — Atiiroidca

X . SI xcarinatits Pietet, 1867 — Aulaconautilus

X. sliarpci Schliiter, 1876 — Cymatoceras

X. slcihis Gemmellaro, 186S — Paracenoceras

llercoglos.sa simiensls Yokes, 1937 — JJercoglossa

Eiitreplioct rax simile Spatli, 1953 — Eutrephoceras

X. simlUimits Foord and Cricdi, 1890 — Cenoceras

X. sindieiisls Yredenburg, 1928— Cimomia

Diltoidonautihts singularis Haas and ^liller, 1952 — Angtdithcs

X . filnuatoplicutas (jeinitz, 1843 — Cymatoceras 1

X. sinuatus Sowerby, 1818 — Pseudaganides

X. sinuosiis Roemer, 1836 — Paracenoceras

Eutrephoceras sloani Eeeside, 1924 — Eutrephoceras

X. smitlii Foord and Crick, 1890 — Cenoceras

X. somaliensis Newton, 1925 — Angulithes

Aturia somaliensis Haas and Miller, 1952 — Aturia

X. sou-erhyaniis d'Orbigny, 1840 — Angulithes

X. sowerhyi Wotlierell, 1836 — Angulithes

342 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

Deltoidonautilus spathi Haas aud Miller, 1952 — Angulithes

N. (Paraturia) spathi Vredenburg, 1928 — Aturoidea

X . spliaericus Forbes, 1846 — Eutrephoceras

IVoodringia splendens Stenzel, 1940 — Uercoglossa

N. spreaficoi Parona, 1897 — Cenoceras

N. staadti Cossniann, 1902 — Eutrephoceras

N. staff elbergensis Kulm, 1936 — Paracenoceras

Vorticoceras stantoni Scott, 1940 — Heminautilus

A\ steinmanni Moricke, 1894 — Cenoceras

N. stephensoni Dickersoii, 1914 — EutrepJioceras

N. steveni Karakaseh, 1907 — Eucymatoceras

N. stoppanii Parona, 1897 — Cenoceras

N. strambergensis Oppel, 1865 — Pseudaganides

A^. straiicostatus Crick, 1907 — Cymatoceras

N. striatus J. Sowerby, 1817 — Cenoceras

N. stricteumhiUcatus Stchepinsky, 1943 — Eutrephoceras

N. stromeri Loesch, 1914 — Pseudaganides

N. stschurousTcii Milasehevitch, 1877 — Eucymatoceras

N . siuri Haiicr, 1856 — Cenoceras

N. suhalbensis Sinzow, 1913 — Anglo-nautilus

X. suhbiangulatus d'Orbigny, 1850 — Pseudaganides

N. subfleuriaiisianus d'Archiac, 1850 — Angulithes

N. (Paracenoceras) subhexagonus Jeaunet, 1951 — Paracenoceras

Eutrephoceras subinflatus d'Orbigny, 1850 — Eutrephoceras

N. sublaevigatus d'Orbigny, 1840 — Eutrephoceras

N. subplicatus Philippi {in Sleinniann, 1895) — Eutrephoceras

Cimomia subrecta Miller and Thompson, 1933 — Cimnmia

N. subrotundus Crick, 1898 — Cenoceras

N. s'ubsinuatus d 'Orbigny, 1850 — Pseudaganides

N. subtruncatus Morris and Lycett, 1850 — Prncyntatnceras

N. subtruncatus Prinz, 1906 — Cenoceras

N. suciense Whiteaves, 1879 — Cymaiocera.s

Cimomia sudanensis Miller, 1951 — Cimoiiiia

X. szontaghi Vogl, 1910 — Eutrephoceras

X. tamulicus Kossmat, 1897 — Angulithes

X. tenuicostatus Sehliiter, 1876 — Cymatoceras

X. terebratus Dumortier, 1874 — Cenoceras

Cimomia tessieri Miller, 1951 — Cimomia

X. texanum Sliumard, 1860 — Pnracymatoceras

Eutrephoceras thomi Reeside, 1927 — Eutrephoceras

X. thyrrenus Tagliarini, 1901 — Cenoceras

X. toarccnsis d'Orbigny, 1849 — Cenoceras

Deltoidonautilus togoensis Miller, 1951 — Angulithes

Aturia tolcunagai Shiuiizu, 1926 — Aturia

KUMMEL : POST-TRIASSIC XAUTILOID GENERA 343

y. tourtiae Schliiter, 1876 — Cymatoceras

Aturia (Brazaturia) triangulata Stenzel, 1935 — Aturia

X. triangularis Montf ort, 1802 — Angulithe.'i

N. tricurinatus Yadasz, 1911 — Cenoeeras

N. trichinopolitcnsis Blaufoid, 1861 — ■ Paracymatoceras

X. truncatus J. Sowerby, 1816 — Cenoeeras

K . tsTcaltsithelensis Eouchadze, 1931 — Cymatoceras

N eocymatoc.eras tsuTcusliiense Kobayashi, 1954 — Cymatoceras

N. tuhingensis Loesch, 1914 — Pseudaganides

N. tumescens Frauscher, 1895 — Eutreplioceras

Uercoglossa tuomeyi Clark and Martin, 1901 — Hercoglossa

N. turcicus Krumbeck, 1905 — Cimomia

N. turJceryi Gemmellaro, 1886 — Pseudaganides

Aturia (Brazaturia) turneri Stenzel, 1940 — Aturia

Platynautilus tyosiensis Yabe and Ozaki, 1953 — Heminautilus

Eutreplwceras vitenhagense Spatli, 1930 — Eutreplioceras

EncUmaioceras ulriclii White, 1882 — Hercoglossa

X. umhilicaris Deshayes, 1835 — Eutreplwceras

Cymatoceras undulatiformis Spath, 1927 — Cymatoceras

N. undulatus J. Sowerbj-, 1813 — Anglonautihis

X. urhanus J. de C. Sowerby, 1843 — Eutreplioceras

X. vaelsensis Binckliorst, 1861 — Epicymatoceras

Xautilopsis vaniixemi Conrad, 1847 — Aturia

X. vastus Kner, 1850 — Eutreplwceras

X. vaughani Gardner, 1923 — Cimomia

X. v.entroplicatus Foord, 1891 — Cymatoceras

Cimomia vestali Miller and Thompson, 1933 — Cimomia

X. vicentinus Oppenheim, 1901 — Eutreplwceras

X. victorianus Teiehert, 1943 — Eutreplwceras

Aturoidea vieirai Miller, 1951 — Aturoidea

X. vinassai Venzo, 1937 — Eutreplioceras

X . (Cym.atoceras) virgatus Spengler, 1910 — Cymatoceras

V. volgeMsi-s Xikitin, 1888 — Paracenoceras

V. u-aageni Gemmellaro, 1886 — Cenoeeras

Hercoglossa walteri Miller, 1947 — Hercoglossa

X. v:andaense Waagen, 1873 — Paracenoceras

Hercoglossa waringi Miller, 1947 — Hercoglossa

X. westphalicu^ Schliiter, 1872 — Angidithes

X. (Paracenoceras) wilmae Jeannet 1951 — Paracenoceras

X. tooodsi V. Hoepen, 1921 — Cymatoceras

X. wyllei Ne-wtou, 1925 — Cimomia

Aturia yohoyamai Nagao, 1926 — Aturia

X. ziczac J. Sowerby, 1812 — Aturia

X. zignoi Gemmellaro, 1886 — Cenoeeras

344 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

.v. zitteli Gemmellaro, 1886 — Cenoceras
Tithonoceras zitteli Eetowski, 1894 — Tithonoceras

The following are species for which the available data is insuf-
ficient to place them in a oeneric group with any degree of
certainty.

N. astierianua d'Orbigny, 1850
A', delpliinus Forbes, 1846
N. fricator Beck, 1835 (nomen nudum)
A', liartmanni Loesch, 1914 (nomen nudum)
A', herbertinus d'Orbigny, 1850
N. inaequalis J. Sowerby, 1813
A'. krenJceli Loesch, 1914 (nomen nudum)
A', marcoui d 'Orbigny, 1850
A'. mathero7iianus d'Orbigny, 1841
N. navfragus Craigin, 1905

A', nohilis Aliinster (see Foord, 1891, p. 326) (nomen nudum)
A', normannicas Loesch, 1914 (nomen nudum)
A', paeudoganiticus Loesch, 1914 (nomen nudmn)
A', reussii Fritsch and Sclilonl^ach, 1872
A', sfaffelhfirgcnsis Loesch, 1914 (nnmen nudum)
A', sitinatraiius Zwierzj'cki, 1915
.V. tenuiplanatus Dana, 1849
A', valenciennii Hupe (in Gay, 1854)
X. varusensi^ d'Orlngny, 1850
A', wcpferi Loesch, 1914 (nomen nudum)

EVOLUTION

Evolutionary patterns, characterized l.y almost complete ex-
tinction followed b}- adai)tiv(' phases witli both eruptive and
stable periods, are well displayed in the general history of the
Cephalopoda. Pew other invertebrate groups display' such
marked expansion and contra.-rion in their evolution. Within
the ammonoids, periods of aljrupt euiit faction, characterized by
mass extinction of most evolving lines, occurred in the late
Permian and in the late Triassic. (^'omplete extinction of the
group came at the end of the Cretaceous. Witliiu the ammonoids
each of the two earlier periods of near extinction was followed
l)y an even greater evolutionary radiation. This is Avell illus-
trated by the number of genera involved in each of these phases;
there are 172 genera of ammonoids of Devonian to Permian age,
:]70 in the Triassic, and 1228 in the Jurassic and Cretaceous.

KUMMEL : POST-TRIASSIC XAUTILOID GEXERA

845

There is some parallelism to the ammonoid history in tIk
evolution of the nautiloids; however, in the latter sroiip insteai

275

W%A/Vv/VVV

150
140
130
120
110
100
90
80
70
60
50
40
30
20
10

NAUTILOID
GENERA

qM.lpilR

u.-e.

ORD.

SIL,

DEV.

MISS.

PENN PERM TRIAS.

JURA.

CRET.

Cenozoic

Fig. 1. Bar chart showing imniher of genera of naiitiloid r-pphalopoils
present in each periorl.

346 BULLETIX : MUSEUM OF COMPARATIVE ZOOLOGY

of an ever expanding evolutionary complex, the nantiloids ex-
perienced their widest radiation in the Ordovician relatively
soon after their appearance in the Upper Cambrian. Soon after
this period of maximum expansion, there set in a steady, gradual
contraction in numbers and diversity of taxonomic units (Fig.
1). There is no wave of extinction toward the end of the Paleo-
zoic; in fact the whole evolutionary pattern of Triassic nautiloids
is really a culmination of trends begun back in the Carljoniferous
(Kummel, 1953c). Thus the evolutionary phases at the transi-
tion from tlie Paleozoic to the Mesozoic are quite ditferent in
the ammonoids and the nautiloids. However, toward the end of
the Triassic most of the long-persisting Carboniferous to Triassic
stocks became extinct. This jjhase of the evolutionary history
of the nautiloids has recently been discussed by Kummel (1953c).
There are, in fact, no Rhactic nautiloids known.

The earliest marine deposits of the Jurassic contain a verj-
homogeneous nautiloid iauna of nearh' world-wide (distribution.
Insofar as nautiloids are concerned, there must have existed
at the transition period from the Triassic to the Jurassic an
ecologic vacuum. A single stock survived this transition period
from the Triassic (Kummel, 1953b), and in the early Jurassic
(Lias) an intensified and new evolutionarj^ radiation took place
(Fig. 2). The early Jurassic stocks are quite clearly of a single
genetic complex, but at the same time show a wide adaptive
range in terms of conch shape, size, suture, and shell characters.
All these features reflect and express the wide adaptive radiation
that took place at this time. Whereas there is very little factual
data available as to the adaptive significance of most morphologi-
cal features it seems only logical to conclude that the various
conch shapes, etc., represent adaptations to a specific niche in
the marine environment. The repopulation of the early Jtirassic
seas by nautiloids was uninhibited as far as other nautiloid
groups are concerned, an<l the diversity of conch types is cogent
testimony to the adaptation to and occupation of many environ-
mental niches.

Eruptive phases expressed in terms of great plasticity of a
group are quite common among the ammonoids. Spath (1934, p.
19) has drawn attention to the "sudden burst of Gephurocera-
tids in the Upper Devonian of the Domanik, Southern Timan.
In that localitj^ there appeared in large numbers depressed and

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

oil

compressed, smooth and ornamented, evolute and invohite,
rounded, square, or acute-ventered forms, and even highly

to
o

-o

1 0

j.-o

1 0

■^—

c
o

M ' '

1 1 "i V"\

CO
CO

c
.0

'c
0

—

c
0

"c
<

^4
0

m
0

0
■0

<3J
D

C
0

0
0
a.

c
0
cr

>•

Cenoceras
trechmanni

0
0

^%^

^^^r^'^'.

*^-.

•:_

f-„'-

Fig. '2. Phylogenetic diagram illustrating the relationships of the Liassio
Xautilidae with the Triassie nautiloid families. The columns representing
the Triassie families have been terminated at an arbitrary even line in the
Xorian. The exact relative extinction dates of these families in the Xorian
are not known.

specialized oxycones" (See Holzapfel, 1899). A similar situa-
tion is found with the early Triassie ophiceratids of the Hima-
layas (Diener, 1897) and Greenland (Spath, 1930a, 1935a). More
examples can he found within the Jurassic and Cretaceous am-
monoids.

348

BULLETIX : ML7SEUM OF COMPARATIVE ZOOLOGY

Recent

A'^^

££.

>-
DC
<

\-
(T.

"-»«

"^z,;.

•O

^b-^"-

O
UJ
O

o
E
O

"^ ^
^ ^

^«„

>>'

3
O

c
o

0)
Q.

o.
3

<
q:

-3

?S*^

c^^

^*^

Cenoceras

ipv

.e^

,i'^<*

A^'

Fig. 3. Diagram showing inferred phylogeny and geologic distribution
of post-Triassic nautiloids.

KUMMEL : POST-TRIASSIC NAUTILOLD GENERA 349

Most early Jurassic nautiloids are here considered species
of a single genus, Cenoceras. The earliest species of Cenoceras.
C. trechmanni. is from Carnian strata of Nbav Zealand and is
derived from the Syringonautilidae (Kiimmel, 1953b). Xo
species of any other Triassic stock survived into the Jurassic.
In the Lias there is thus a Avorld-wide fauna of involute to
evolute, strigate to smooth forms with a wide range in whorl
shapes — a genetic com])lex of great plasticity. From this homo-
geneous but plastic stock there arose from diverse parts of the
complex several distinct evolutionary lines built on specializa-
tion of one or more morphological features. From this Cenoceras
complex arose the persisting stable stock (Eutrephoceras) which
gave rise to other members of the Nautilinae. Likewise the
Pseudaganidinae, Paracenoceratinae and most probably the
Cymatoceratinae arose directly out of the Cenoceras complex
(Fig. 3).

The Pseudaganidinae are characterized bj' their sinuous su-
tures. In this respect they are adaptive types like the Triassic
Clydonautilidae, Gonionautilidae, and Siberionautilidae. They
are likewise similar to these Triassic groups in that they show
stability in the general form and patterns of the suture and
great varialulity in conch shape. The Pseudaganidinae range
from the Liassic into the Lower Cretaceous. There are only two
genera involved in this radiation, of which Pseudaganides is the
main line and PsciidonauHliis is a specialized offshoot of Upper
Jurassic and Lower Cretaceous age. Tn the evolutionary history
of the uautiloids from the Carboniferous to the Recent, adaptive
trends expressed in terms of sinuous — ''goniatitic" — sutures
occurred at four separate times. The first is known through
the genus Permoccras Miller and Collinson based on Aganides
hitauniensc Haniel from Middle Permian strata of Timor. In
luy opinion. Pcrmoceras is related to the Grypoceras-Doma-
toceras evolving .stock and is an aberrant development of this
•Stock. Grgpoceras iPlummeroceras) Kummel (1953c) likewise
had a highly sinuous suture with a deep ventral lobe and a deep
lateral lobe but it has a very evolute conch. The second radiation
repeating this adaptive trend is tliat in the Upper Triassic in-
cluding the Clydonautilidae, Gonionautilidae and Siberionautili-
dae. All of the genera of these families include species with very
involute conchs but show great variability in conch shape. The

350 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

only other Triassic species with a "'gouiatitic'' suture is Cly-
menonautUus ehrlichi Mojsisovics which in all features but the
suture is allied to the Syringonautilidae. The suture however has
a deep, tongue-shaped, narrow-, lateral lobe. The Upper Triassic
families listed above which include those involute species with
highly sinuous sutures, evolved from involute, smooth forms
with nearly straight sutures (Paranautilidae).

The third radiation of this series is that of the Pseudaganidi-
nae of Lower Jurassic to Lower Cretaceous age. This group
evolved out of some part of the Cenoceras complex in the Lias.
There is almost perfect homeomorphy between Permoceras
hitaimiense Ilaniel of the Middle Permian and Pseudonaidilus
geinitzi Oppel of the Upper Jurassic (Miller and Collinson,
1953). Each however represents similar adaptive types from
different root stocks and they are not related.

The fourth and last radiation centered on the sinuosity of
the suture is that of the Hercoglossinae and Aturinae. The
Aturinae with its single genus Aturia is a very specialized de-
velopment out of the Hercoglossinae. The specialization in this
case is not so much on a further elaboration of the suture but
in the development of the peculiar dorsal siphuncle. Within
the Hercoglossinae there is a beautiful developmental series
(usually treated as genera) showing gradual increase in sutural
complexity. The subfamily has its origin within the stable
evolving stoclc of the Xautilinae represented by Eutrephoceras
which is characterized by an involute smootli conch with straight
or slightly sinuous sutures. Cimoinia of the Hercoglossinae has
a slightly more individualized suture and is gradational with En-
trephoceras, as it is also with Hercoglossa which has very distinct
lobes and saddles. Angulithes is merely a sagittate dei-elopment
of these forms. Jn Aturoidea the sutural individuality is carried
slightly farther.

The Paracenoceratinae is a small subfamily of mainly Jurassic
age with four genera characterized by specialization of the venter.
The main stock of this subfamily, Paracenoceras, is nearly world-
wide in distribution and has by far the largest number of species.
The other genera, namely SomaUnautilus, Aulaconaufilus, and
Tithonoceras are more highly specialized forms with few known
species. Tn Paracenoceras the whorl section is subtrapezoid,
generally with a broad, sulcate venter. SomaUnautilus has angu-

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 351

lar ventral shoulders but a broad arched venter. Aulaconautihis
has four or more longitudinal ribs. In this respect it is homeo-
morphous to Aulametacoceras of the Permian and Triassic.
Tithonoceras has prominent rounded ventrolateral keels with a
sulcate venter. Paracenoceras and Somalinautilus are undoubt-
edly derived from distinct elements of the Cenoceras complex
but Aulaconautilus and Tithonoceras appear to be specialized
developments of Paracenoceras.

One of the must successful and diverse groups to stem directly
or indirectly from the Cenoceras complex is the Cymatoceratinae.
This subfamily comprising 10 genera ranging from the Juras-
sic to the mid-Tertiary is characterized by conchs bearing ribs.
This is the only group of post-Triassie nautiloids to have orna-
mentation, aside from Aulaconautilus of the Paracenoceratinae.
Radiation within the Cymatoceratinae is reflected in the shape
of the concn, suture and ornamentation. The main evolving
stock, Cinnatoccras, has an involute, rounded conch with only a
slightly sinuous suture. ]\Iost of the other genera of this sub-
family are thought to represent various specialized groups de-
rived from Cymatoceras. Paracyniatoceras has a more sinuous
suture, in fact much like that of Hercoglossa, and on even this
feature it is gradational with Ci/niatoceras. Those forms dif-
ferentiated on the basis of conch shape include Honinautilus,
Deltocymatoceras (n. gen. p. 438), Epicymatoceras (n. gen. p.
439) and Cymatonautilus. Herminautilus has a compressed in-
volute conch with a highly sinuous suture, that has deep ventral
and lateral lobes. Deltocymatoceras is a homeomorph of Afigu-
lithes with a sagittate whorl section. Epicymatoceras has an
evolute, highly compressed conch with a subrectangular whorl
section. Cymatonautilus has a concave venter and concave lat-
eral areas. Those genera differentiated on the basis of modifica-
tion of the ribbing pattern include Eucymatoceras, Anglonau-
tilus, Procymatoceras, and Syrionautilus. In Eucymatoceras the
ribs form prominent V-shaped salients on the venter and flanks.
In Ancjlonautilus there are coarse folds on the venter whereas
Procymatoceras has a rapidly expanding robust conch with
ribs mainly on the lateral areas. Syrionautilus has peculiar
.shaped ribs with wide interspaces like that of Proclyclonautilus
spirolobus of the Triassic. Cymatonautilus and Procymatoceras
are Jurassic developments, the remaining genera mostly Cre-

352 BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

taceous. In fact the most widespread and characteristic nautil-
oids of the Cretaceous are genera of the Cymatoceratinae.

Thinking of tlie Cymatoceratinae as representing a single
genetic unit with its origin within the Cenoceras complex, and
having as the basic common denominator the ribbing pattern, all
of the diverse morphologic types represented by the various
genera reflect a broad adaptive radiation which produced numer-
ous homeomorphs of other genera of the Nautilidae.

The most persistent stock evolved from the early Jurassic
Cenoceras complex is that of Eutreplioceras. This genus has a
world-wide distribution and ranges in time from the Upper
Jurassic to mid-Tertiary. It and Cymafoceras are the most com-
mon post-Triassic nautiloids. Some 90 so-called species, assigned
to this genus, have been described. The genotype has a tightly
involute, subglobular, smooth conch with a nearly straight
suture. This basic, very simplified, conch pattern became modi-
fied either by greater compression, depression or by slightly
looser coiling. The variants, however, on the basis of the record
available appear to be completelj" random in their chronologic
or geographic distribution. Each such variant has usually been
treated as a distinct species. The geologic record thus shows the
occurrence of nautiloids with the basic, simplified, globose conch
persisting from Upper Jurassic to mid-Tertiary time and with
numerous variants of this basic pattern.

Whereas the suture is generally straight or nearly so there are
gradational forms to Ciniomia. Data are not available as to
whether there is any particular chronologic or geographic rela-
tionship in regard to the sutural variants,' which appear to be
completely random as is the case with conch form.

The eutrephoceratids are thus a persisting generalized stock
which arose from the Cenoceras complex, and they show no par-
ticular adaptive trends throughout their history. They do show,
however, a certain variability in conch shape and suture. The
group has the longest range of any of the post-Triassic nau-
tiloids. It is from this persisting generalized form that the re-
maining more specialized genera and subfamilies arose. The
most important offshoots are the genus Nautilus and the whole
subfamily Hercoglossinae. The nature of the evolutionary pat-
tern from Enfrephoceras to Cimomia is not easy to decipher.
The only significant difference is the degree of sinuosity of the

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

353

suture, that in Eutrephoceras being straight, or nearly so, and
that in Cimomia having a distinct lobe and saddle on the lateral
areas. There are gradational forms between these two genera.
The time range of the two genera is approximately the same,
Cimomia being slightly younger in origin. Tt seems quite clear
that Cimomia arose from Eutrephoceras but whether or not there
was a single time and point of origin in the late Jurassic, or if

-13-
-12-

« 1

-||-

""^

-10-

-9-

-8-

-7-

ill

— b—

-5-

IpPplHiHp

ill

ill-iiiliiliii

-4-

■ /"

-3-

_ o

■. i ■'■ -i

— t

:r':\ "■ ^l

- ■" 1 ■

. . 1 .

"■■•■■•L-l-ilil

!;:i:k;::::::

^ -■■—■

::,;;-i:i::i;j;:-::i::i!iiiil

L.
Jur.

M.
Jur.

U.
Jur.

L.
Cret.

U.
Cret.

Paleo

Eo.

Oligo.

Mio.

Plio. Ptest

Fig. 4. Bar chart showing total number of genera of nautiloids present
in each series of the Mesozoic and Tertiary (light stippling) and number
of new genera appearing for the first time in each series (dense stippling).

354 BULLETIN : JIUSEUM OF COMPARATIVE ZOOLOGY

there were numerous times of origin, throughout the Cretaceous
and Tertiary cannot be told from the kind of data available.
The alternatives are a parallel evolutionary pattern of the two
genera or a consideration of these transitional forms of suture
(Cimomia) as a form genus including numerous distinct radia-
tions from the eutrephoceratid evolving stock.

Iterative patterns of evolution are recognized within the Car-
boniferous through Triassie nautiloids. With these nautiloids
three main, slowly-evolving, persisting stocks are recognized,
nameh- the Domafoceras-Grypoceras line, the Metacoceras-Mojs-
varoceras line, and the Liroceras-Paranautilus line. Each of
these stocks has a relatively large number of species (which is
merely an indication of relative abundance) and is widespread
geographically. Throughout their history they show only rela-
tively minor evolutionary changes. Each of these lines repre-
sents a stable, slowly-evolving, l)ut persisting parent stock of
their particular family groups. These lines are the evolutionary
reservoirs from which other, generally more specialized, groups
evolved. The offshoots most often are adaptations accentuating
one or more generalized characters of the parent stock. Most of
the offshoots are forms with few species and limited geographic
range. However, some of these offshoots became in themselves
virile, evolving stocks, each with a Avide adaptive range and
many species (e.g. Pleuronautilidae from the Metacoceras-Mojs-
varoceras line and the Clydouautilidae from the Liroceras-Para-
nautilus line ) .

The E ntrephoceras stock gave rise to four other small and
aberrant groups, Pseudocenoceras, Ohinautiliis, Carinonautilus,
and Nautilus. Pseudocenoceras is a compressed form somewhat
evolute and with an arched venter, flattened sides and steep
umbilical shoulders. Carinonautilus is a much compressed form
with a prominent rounded keel. It is a monotypic form from
Upper Cretaceous strata of South India. Ohinautilus is another
monotypic form recently described l)y Kobayashi from Oligocene
strata of Japan. It is a much compressed involute form with
a shallow but distinct furrow on the venter. This genus is still
very incompletely known since on the type and only available
specimen neither the suture nor position of the siphuncle is
preserved.

No fossil species are assigned to the genus Nautilus. No Plio-

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 355

oene or Pleistoeenp nautiloids are kno-\vn. Several species of
Eocene, Oligoeene, and Miocene age have been at times assigned
to Nouiiliis but all of these are here considered as having closer
affinities to the persisting eutrephoceratid stock than to modern
species of Nautilus.

TABLE 1

^ ~ ;~ ?. T c a

?- ^ C ^ :ij £i

Cenoceias x x

Eutrephoceras x x x x x x x

Pseudofonocc'.a : x x

Carinonautilus x

Obinantihis x

Xaiitihis X

Pseudfi^anides x x x

I'soiidonautilus x x

Paracenoeeias x x x

Aulaeouautilus x

Tithonoceras x

Somalinautilus • x x

Cyniatoi-eras x x x

Paraeyinatoeeras x x x

Pioeymatoceras x x

Cyniatonautilus x x

Anglonautilus x x

Eueyniatoeeras x

Syrioiiaiitilus x

Heminautilus x

Deltocymatoceras x

Epieymatoceras x

Cimomia x x x x x x

Angulithes x x x x x

Hercoglossa x x x x

Aturoidea x x x

Aturia x x x x

356 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

The geologic range of the geuei-a of post-Triassic nautiloids is
summarized on Table 1, and the bar chart of Figure 4 shows the
number of genera and number of new genera per series division
of the Jurassic, Cretaceous, and Tertiary. This bar chart clearly
brings out the increasing adaptive diversity of the nautiloids in
their resurgence after their near extinction in the late Triassic.
There is a gradual increase in numbers of new types through the
Jurassic. However, of the total of 11 genera in the Upper Juras-
sic, only 3 {Eutrephoceras, Paracymatoeeras and Cimomia) are
still destined for a long history. More than half of the total Upper
Jurassic fauna (6 genera) do not survive into the Cretaceous
and 2 genera {PscudonautUus and Paracenoccras) have their
main evolutionary plaj^ in the Upper Jurassic with only a few
species known from Lower Cretaceous. Thus the Jurassic faunas
are replaced in the Cretaceous by new phyletic lines, mainly of
the C^^llatoceratinae, with a much smaller portion from the
Nautilinae and Hercoglossinae. By the early Tertiary the eyma-
toceratid radiation is almost at an end, with a single surviving
species in the middle Tertiary. Only one new genetic unit, the
Aturinae, comes into the picture in the Paleocene ; this became
by mid-Tertiary time the most conspicuous element of the fauna.
Early Tertiary time was the heyday of the Hercoglossinae ; how-
ever, the origin of all members of this radiation is well down in
the Cretaceous. Wherever early Tertiary faunas are found,
members of the Hercoglossinae are the principal elements; the
great number of species of this subfamily for the early Tertiary
merely reflects the large amount of study concentrated on them
in recent years, especially by Miller.

Thus from this summary picture, post-Triassic nautiloid evo-
lution took place in three successive waves, one in the Jurassic,
another in the Cretaceous, and the last in the early Tertiary.
Each wave introduced new genetic lines which gradually re-
placed existing or previous lines. The Jurassic and Cretaceous
radiations are of approximately equal magnitude ; that of the
early Tertiary is greatly reduced. Whether or not Nauiilus
should be considered an additional radiation is a moot question.
Of the derivations from the eutrephoceratid surviving stock only
the Hercoglossinae had an extensive evolutionary history. The
other offshoots including Nautilus are either monotypic or repre-
sented b.y few species of no great geographic or stratigraphic

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

357

range. It seems higlily likely that Xautihis will follow this same
fate.

In most of the principal evolutionary trends of the jiost-Trias-
sic nautiloids the main evolving stock of each snhfamily is i-ejn-e-

-100-
-90-
-80-
-70-
-60-
-50-
-40-
-30-
-20-
-10-

L.
Jur.

M.
Jur.

U.
Jur.

L.
Gret.

U.
Gret.

Paleo,

Eo.

Cligo.

Mio.

Plio.

Pleist.

Fig. 5. Bar chart showing total munl)er of species of nautiloids for each
series of Mesozoic and Tertiary.

sented by the largest number of species, with the other genera
having only a fraction of that amount. Thus of the Xautilinae,
Cennceras and Enircphocevas have by far the largest numl)er of
species. In the Paracenoceratinae, it is Paracenoccras; in the

856

BL7LLETIX : MUSEUM OF COMPARATIVE ZOOLOGY

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

359

ci;

«1

?a. O

J-.

(/,

■W

cr,

crt

flfl

■^^

—

fii

■4-1

_r^

• r—

•—

»J

-^^

r-i

^T*

>— •

-*—

[/)

—

■Jl

r^.

►— '

:j;

5i-

—

P r*'

360 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Cymatoceratinae, it is Cyinatoceras; in the Pseudaganidinae it is
Pscudaganides, and in the Hereoglossinae it is Cimomia that has
the greatest number of species. The Aturinae has only the genus
Aturia. \\\ spite of the unsatisfactory nature of our understand-
ing of most nautiloid species, these named units do express
morphologic diversity, geographic range, and general abundance
in nautiloid faunas of these periods. These data likewise tend
to substantiate the thesis presented here for persisting, evolving
stocks of wide geographic distribution from which more re-
stricted adapted types evolved. On Figure 5 are summarized the
nninl)er of species per series of the Jurassic, Cretaceous, and
Tertiary. Except lor the large number of early Tertiary species,
interpreted here as a monographic high, this bar chart reflects
veiy well tlie liar chart on di .ti-iI)ution of genera during this
time span.

syste:^[atic paleoxtology

Family NAFTILTDAE d'Orbigny, 1840

This family is interprete;! here, to include all post-Triassic
nautiloid genera, which are placed in six subfamilies: Nautilinae
(I'Orbigny, Pseudagauidinae nov., Paracenoceratinae Spath.
Cymatoceratinae Siiath. Hereoglossinae Spath, and Aturi-
nae Hyatt. The overall evolutionary pattern of post-Trias-
sic nautiloids with the plastic Cenoveras complex of the early
Jurassic from which developed directly or indirectly several
distinct jihyletic trends produced a very homogeneous evo-
lutionary unit. Tlie subfamily units are adaptive trends inter-
preted for the most jjart on the basis of single characters — in
the Cymatoceratinae it is the presence of ribbing; in the Para-
cenoceratinae it is elaboration of the periphery ; in the Hereo-
glossinae and Pseudagauidinae it is elaboration of the suture;
and in the x\turinae it is the unique dorsal siphuncle and its
structure. None of these trends is really a major shift in the
evolutionary complex but each is more in the nature of elabora-
tion of genetic potentials in the evolving parent stocks. It thus
seems that the taxonomic ranks used here reflect the phylo-
genetic picture.

KUMMEL : P08T-TRIASSIC XAUTILOID GENERA 361

Subfamily NAUTTLIXAE d'Orbigny, 1840

The Nautilinae includes among its genera the earliest forms
and the latest and livinu- species of the Nautilidae. Thus in the
history of this subfamily we have the origin of the Xautilidae.
the persisting stock which gave rise to the remaining subfamily
units, and the few living species represented by the relic genus
Nautilus. The genera included in the Nautilinae are: Nautilus,
Cenoceras, Etitrcphoceras, Pseudocenoceras, Carino nautilus, and
Obinautilus. Cenoceras is the plastic evolving complex which
survived the great period of extinction at the end of the Triassic
and experienced a very intense radiation in the early Jurassic.
From the Cenoceras complex arose the Pseudaganidinae, Para-
eenoceratinae, and probably the Cymatoeeratinae. Eutrepho-
ceras is interpreted as a slowly evolving, long persisting, general-
ized stock also derived from the Cenoceras com])lex and from
which the remaining members of the Nautilinae were derived
])lus the Ilercoglossinae.

Carino nautilus and Ohinautilus are monotypic ; Pseudoceno-
ceras has 8 species, and Nautilus only 5 species (all Recent).
However, Cenoceras has 97 species and Eutrephoceras 90. This
great number of species (or named units) reflects the wide geo-
graphic range, morphologic diversity and relative abundance of
these two genera. Cenoceras is mainly confined to the Lower and
Middle Jurassic (with one Upper Triassic [Caruian] species),
and Eutrephoceras ranges from the Upper Jurassic into the
Miocene. Roth o-enera have world-wide distributions (Figs. 6,
7).

Genus CenOCEEAS Hyatt. 1883

Cenoceras Hyatt, 1883, pp. 300 301.

Cenoceras Hyatt, 189-1, p. aoO.

Digonioceras Hyatt, 1894, p]'. 548.549.

Nautilites Priiiz, 1906, p. 201.

Cenoceras Spatli, 1927a. pp. 20 24.

OiJhionautihis Spatli, 1927a, pp. 21, 24.

Sphaeronautilus Spath, 1927a, pp. 21, 24.

Cenoceras Flower and Kiimmel, 1950, p. 615.

Bisiphytes (Cenoceras) Kummel, 1954, p. 322.

T]ipe species. Xautihis intermedins d'Orbigny non Sowerbyr=X. nrbignyi
Prinz (1906, p. 213). By original designation. Type illustration repro-
duced on Plate 1, tigs. 1, 2.

362 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

The genus Cenoceras is here interpreted to include those nau-
tiloid species of the Liassic and Inferior Oolite that are part of
the evolving complex which survived the Triassic and which in
the great j)lasticity of the gi'oup reflect an extensive adaptive
radiation, it is from this very plastic adaptive unit that the
more stable, more clearly defined evolutionary lines evolved.
The nautiloids nearly became extinct at the end of the Triassic.
A single surviving line derived from the Syringonautilidae gave
rise to the sole surviving stock (Spath, 1927a, p. 28; Kummel,
l!)5;3b). This surviving stock entered an environmental vacuum
(insofar as nautiloids are concerned) in the earliest Jurassic.
The re-occupation of the diverse ecological niches available for
a homogeneous unit stock gave rise to many morphological types,
especially in regard to conch form. It is not possible at this time
to decipher the individual evolutionary lines within the mam
complex. There is a complete range of variation and gradation
in couch form, suture, position of siphuncle, ornamentation, etc.
To bring the taxonomy of these nautiloids to reflect the phylo-
genetic interpretation, all of the genera previously proposed for
aberrant types in the Liassic are placed in synonymy of Ceno-
ceras. There is no particular reason for retaining these genera
since their included species fit well within the range of variabil-
ity and evolutionary pattern of the genus Cenoceras.

Fig. 8. Cross sections of the couch of species of Cenoceras. Eedrawn
from Pia (1914, pi. 8 [5]). A, C. arari.s (Duniortier) after Dumortier,
diameter 210 mm.; B, C. arariformi.'i (Pia) after Pia, diameter 190 ram.;
C, C. adneticu.H (Pia) after Pia, diameter 137 mm.; J), C. dumeyhns
(d'Orbigny) after d'Orbigny, diameter 150 mm.; E, C. stoppani (Parona)
after Parona, diameter 142 mm.; F, C. haoonicua (Vadtisz) after Vadasz,
diameter 50 mm.; G, C. altisipJtitcN (Prinz) after Prinz, diameter 60 mm.;
n, C. Julianas (Fueini) after Fueini, diameter 20 mm.; /, C. siiiuHimu.'<
(Foord and Crick) after Pia, diameter 107 mm.; J, C. secernendu.^ (Pia)
after Pia, diameter 135 mm.; K, C. ornatus (Foord and Crick) var.
atanaten.sis (Pia) after Pia, diameter 158 mm.; L, C. parctoi (Gemmellaro)
after M. Gemmellaro, diameter 17.5 mm.; M, C. sp. ind. No. 2, after Prinz,
diameter 65 mm.; X, C. striatm (Sowerby) after d'Orbigny, diameter
110 mm.; 0, C. striatits (Sowerby) after Hauer, diameter SO mm.; P, C.
.ituri (Hauer) after Hauer, diameter 65 mm.; Q, C. semseyi (Prinz) after
Prinz, diameter 88 mm.; B, C. rohustus (Foord and Crick) after Foord,
diameter 200 mm.; S, C. geyeri (Prinz) after Geyer, diameter 80 mm.;
T, C. fourneti (Dumortier) after Dumortier and Pia, diameter 80 mm.

KUMMEL: POST-TRIASSIC NAUTILOID GENERA

368

^rx X

Fiajure 8

.S64 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

N^oineiic'latuiv of early .lurassic uautiloids has an involved
and ambignons history. As pointed ont by Spath (1927a, p. 20),
Hyatt's genus Ccnoceras (for the group of Nautilus intermedins
d'Orbigny non So-vverl)y") has been universally rejected, since
Hyatt left in Nautilus s.s. other members of the same group, such
as Nautilus striatus. However, Spath in his 1027 revision of post-
Triassie nautiioids accepted Cenoceras as a distinct unit on the
basis of Hyatt's (1894, p. 550) revised definition limiting Ccno-
ceras to forms with a trigonal ananepionic, a subquadragonal
metanepionic, and a dor,>al]y suleate nepionic stage. At the same
time Spatli readily recognized that the groups of Nautilus stri-
atus Sowerby and Nautilus intermedius Sowerby could not be
differentiated on the shape of the couch and they "agree in
ornamentation, position of the siphuncle, presence of an annular
lobe, and chiefiy in the course of the septal suture" (Spath,
]9:27a, p. 20).

Now for the group of Nautilus si riatus Sowerby, Spath revived
the genus Bsiplnjtes Montfoi't ( 1808). This genus has been dis-
cussed to a varied degree by Spath (1927a, pp. 19-24), Miller,
Dunlmr, and Condra (19;}3,"pp. 42, 43), Teichert (1940, p. 591)
and Kummel (1954, pj). o22-32')). Montfort listed as type of his
genus Bisipln/hs r<ticulai')(s (a then newly proposed species)
which came fi'oin S,)inl)(M-ni)u, Uurgundy (east-central France),
presumably from -hii-assic strata. As indicated by the generic
name, Montfort thought liis speciuicn had two siphuncles but the

Fig. 9. Cross sections of the conch of species of Cenoceras. Redrawn
from Pia (1914, pi. 9 [6]). A, C. toarccn.sii^ (d'Orbigny) after d'Orbigny,
diameter 240 mm.; B, C. jurensis (Qnenstedt) after Quenstedt, diameter
40 mm.; C, C. sp. iiid. Xo. (i, after Rosenberg, diameter 50 mm.; B, C.
demonensis (Gemmellaro) after M. Geuimellaro, diameter 40 mm.; E, C.
fiustriaeus (Haiier) after Hauer, diameter 230 mm.; F, C. aratus (Quen-
stedt) after Queitstedt, diameter 40 mm.; G, C. meyrati (Ooster) after
Ooster, diameter 80 mm.; K, C. terchratu.s (Dumortier) after Dumortier,
diameter 98 mm.; /, C. pseudorugosu^ (Pia) after Ooster, diameter 90
mm.; J, C. mayiani (Gemmellaro) after M. Gemmellaro, diameter 46 nun.;
K, C. riigosus (Buvignier) after Buviguier, diameter 75 nun.; L, C. asta-
coides (Young and Bird) after Dumortier, diameter 112 mm.; M, C.
intermedius (Sowerby), diameter 200 mm.; K, C. jourdani (Dumortier)
after Dumortier, diameter 130 mm.; 0, C. inopnatus (d'Orbigny) after
d 'Orbigny, diameter 85 mm. ; P, C. sp. ind. No. 8, after Vadasz, diameter
66 mm.; Q, C. friDicatiis (Sowerby) after Geyer, diameter 80 mm.

KUMMEL : l'OST-1 KIASSIC NAUTILOID GENERA

365

Figure 9

366 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

second "siplmiicle" is undoubtedly an indentation of the dorsal
lobe. The only illustration of Bisiphytes reticulaius is a rather
crude sketch (reproduced here on Plate 25, fig. 3). It is not
possible to recognize the specific features of the genotype of
Bisiphytes from Montfort's description and/or his illustration;
likewise the exact locality and horizon from which the specimen
was collected is not known. For these reasons the genus Bisi-
phytes should not l)e used (a recommendation set forth by
Teichert in 1940, p. 591, and in several written communications
to the writer). A petition to suppress the generic name Bisi-
phytes is being made to the International Commission on Zoologi-
cal Nomenclature. Abandonment of Bisipliytes leaves available
the genus Cenoceras — type Nautilus intermedius d'Orbigny non
Sowerby=iV. orhignyi Prinz. I am in perfect agreement with
Spath that the quadrate whorled Cenoceras cannot be separated
from the rounder whorled Bisiphytes. My previous use of Ceno-
ceras as a subgenus of Bisiphytes was an attempt to express the
homogeneity of these two groups but I am now convinced that
Teichert 's recommendation to suppress Bisiphytes, because of the
almost complete ambiguity surrounding the type species, will do
much to clarify the taxonomy of these Jurassic nautiloids.

Fig. 10. Cross section of the conch of species of Cenoceras and Pseuda-
ganides. Eedrawn from Pia (1914, pi. 10, [7]). A, C. quad rang ularis
(Pia) after Pia, diameter 103 mm.; B, C. halsamoerivellii (Parona),
after Parona, diameter 1.55 mm.; C, C. distefanoi (Gemmellaro) after
M. Gemmellaro, diameter 52 mm.; D, D. hrancoi (Gemmellaro) after M.
Gemmellaro, diameter 60 mm.; E, C. affimis (Chapuis and Dewalque) after
Chapuis and Dewalque, diameter 101 mm.; F, C. sp. ind. No. 5, after
Kosenberg, diameter 32 mm.; G, C. prnfundisipliites (Prinz) after Prinz
diameter 82 mm.; H, C. tricarinatus (Vadasz) after Vadasz, diameter
91 mm.; /, C. anomphalus (Pia) after d'Orbigny, diameter 90 mm.; J, C.
sp. ind. No. 9, after Vadasz, diameter 36 mm.; K, C. cliilensis (Hupp6)
after Moricke, diameter 99 mm.; L, C. amurcttii (Parona) after Parona,
diameter 80 mm.; M, C. hreislaclci (Parona) after Parona, diameter 93 mm.;
N, C. egregius (Pia) after Pia, diameter 207 mm. ; 0, C. semistriatus
(d'Orbigny) after d'Orbigny, diameter 165 mm.; P. Fseudaganides grave-
siana (d'Orbigny), diameter 240 mm.; Q, C. schwalmi (Prinz) after
Prinz, diameter 89 mm.; B, C. fischeranius (Foord and Crick) after Foord,
diameter 180 mm.; S, C. pertextus (Dumortier) after Dumortier, diameter
180 mm.; T, C. spreaficoi (Parona) after Parona, diameter 58 mm.; U, C.
pisanvs (Fucini) after Fucini, diameter 12 mm.

KUMMEL : POST-TKIASSIC NAUTILOID GENERA 367

Q o. a o. o

B C D ^^E F

^, ^.

I J

^ 0 0 ^, a

L M ^-^N 0 P Q

Z) V.

Figure 10

368 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

A precise morphological definition of Cenoceras is not possible
because of the wide variability and range in conch shape and
ornamentation. The suture is for the most part quite uniform
having only slight ventral and lateral lobes. The variability in
conch shape has been most admirably illustrated in a set of text
figures by Pia (1914) which are reproduced here as Text Figures
8, 9, 10. Interpreting conch form as an expression of adaptation,
these text figures clearly show the range of radiation in the
group. The uniformity of the suture and types of oruamenta-
tional patterns supports the conclusion that all these species are
part of a single genetic complex.

Spath (1927a) listed four other genera of Liassic nautiloids,
namely: Sphaeronautilus Spath, Digonioceras Hyatt, Ophio-
nautilus Spath, and Hercoglossoceras Spath. All these genera
are now placed in synonymy of other forms. Hercoglossoceras is
considered a synonym of PseudagankJcs. The remaining genera
{Spliaeronautilus, Digonioceras, and Ophionautiliis) are consid-
ered to be synonyms of Cenoceras.

The type species of Digonioceras Hyatt, 1894, is Digonioceras
rotundum Hyatt, non Crick, 1898, by original designation and
not Nautilus excavains J. de. C. Sowei'by (1826, pi. 529, fig. 1;
PL 9, figs. 3-5, of this report) as indicated by Spath (1927a, p.
21). The original description and illustration of D. rotundum
(Hyatt, 1894, p. 549) leaves much to be desired. The type speci-
men, which is in the M.C.Z., is a juvenile of one volution and con-
sists only of phragmocone. it measures 34 mm. in diameter;
the width and height of the most adoral whorl are 23 mm. and
17.5 mm. respectively. The whorls are depressed and ovoid in
cross section. The venter is broadly arched. The whorl sides
are convex and grade onto the venter and the umbilical wall with
no distinct shoulders to break the even curvature. The venter of
the second camera is bluntly angular or fastigate. The rounded

KUMMEL : POST-TRIASSIC NauTILOID GENERA

369

aspect of the venter is apparent on the third camera. Adorally,
the whorls become gradnally more depressed producing the broad
venter and narrowly rounded sides. The umbilicus is perforate,
measuring approximately 2 mm. in diameter. The suture has
very shallow ventral and lateral lobes, and no annular lobe. The
siphuncle is in a subventral position. The shell is smooth except
for fine growth lines that form a deep tongue-shaped sinus on
the venter.

b'i^'. 11. Cenoceras txcaiatii.s (Sowerby). Diagrammatic cross-section
of holotype represented by figures 1, 2, Plate 7. X 1.

Hyatt (1883, p. 288) included Nautilus excavatus, d'Orbigny
in Endolohus and considered that species to be the latest survivor
of that series. However, in 1894 (p. 549) he noted that there
were significant differences between the Carboniferous Endolo-
hus and the Jurassic species he placed in Digo7iioceras. My own
.study of Sowerby's type of Nauiilus excavatus in the B.M.N. H.
(figured here on Plate 7 and Text Figure 11) and Hyatt's type

370 BULLETIN : MUSEUM OF COMPARATRT: ZOOLOGY

of Digonioceras rotundum which is in the M.C.Z. lead me to con-
clude that it is not possible to tell whether or not these two
species are congeneric in Hyatt's sense. Digonioceras rotundum
is based on a juvenile specimen which is very generalized in its
characters. The locality and geologic horizon of I), rotundum
is not known, but Hyatt (1894, p. 549) writes, "It is obviously
from the Oolite but the locality is not kuoA\Ti. "

The large funnel-shaped umbilicus of Nautilus excavatus is the
principal character used by Hyatt and later Ijy Spath to set this
group aside as a separate generic category under the name
Digonioceras. Other Middle Jurassic species wiiich appear to be
related to Nautilus excavatus, at least in the general character
of the umbilicus, are N. smithi Foord and Crick, 1890, and .V.
dispansus Morris and Lycett, 1850. From the Lias, similar um-
bilical features are found in Nautilus nicyrati Ooster (1858)
(Fig. 9G) and in .V. tcrchratus Dumortier (1874) (Fig. 9H).
The umbilical shoulders of the latter species are flared and
angular; the umbilical wall is broad and sloping toward the
umbilicus. The group of Nautilus excavatus does not appear to
have given rise to any later stock of nautiloids and is here inter-
preted as merely one (of the many) adaptive types produced in
the early Jurassic within the Ci noceras complex. From an evo-
lutionary view'point the group is merely a part or expression of
the adaptive radiation or plastic phase found in the early Juras-
sic, and the group should be taxonomically treated as species of
Cenoceras.

Ophionautilus .Spath (1927), tyi)e species Nautilus hu)ionensis
Foord and Crick, 1890 (illustrated here on Plate 6, and Text
Figure 12), was differentiated on the basis of the very evolute,
widely umbilicate conch. The type species is from the Inferior
Oolite of England. In addition to the type species. Nautilus
zitteli Gemmellaro (1886) of the lower Dogger of Sicily, N.
austriacus Hauer (1856) (Fig. 9E) and N. schwalmi Prinz
(1906) (Fig. lOQ) from the Liassic of the Alps and Hungary,
respectively, belong to this species group. These species wliicli
have been or can be assigned to Ophionautilus are interpreted as
evolute adaptations of the Cenoceras complex. These evolute ele-
ments did not give rise to any of the succeeding more stable
evolutionary lines. Nautilus hurtonensis Foord and Crick has
an evolute. smooth, rather large conch. The whorl section is

KUMMEL: POST-TRIASSIC NAUTILOID GENERA

371

wider than hig-h. The venter is broadly rounded to slightly
flattened; the whorl sides are rounded as are the umbilical
shoulders. The umbilical wall is convex and steep. The umbilicus
is broad and deep, exposing much of the inner whorls. The
suture is only slightly sinuous with a shallow ventral and a
shallow lateral lobe that includes the whole whorl side. The onlv
markings on the shell are g-rowth lines and they form a deep
sinus on the ventei-. The siphuncle is more or less centrallv

Fig. 12. Cenoceras biuionen-sig (Foord and Crick). Diagrammatic cross
section of liolotype represented 1>y figures 1, 2, Plate 6, X 1.

located. The group of Nautilus hurtonensis includes the most
evolute adaptation of the Cenoceras complex but there is within
this complex a nearly complete gradational series in degree
of involution.

372

BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY

Whereas the group of Nautilus burtonensis includes the most
evolute elements of the Cenoceras complex, the most involute
form is Nautilus pisanus Fucini (1895, p. 828, pi. 13, fig. 5) from
lower Liassic strata of Italy for which Spath (1927a, p. 21)
established the genus Sphaeronautilus (Fig. lOf/). In its
arcestid-like shape Nautilns pisanus is indeed an extreme aber-
rant type. The conch is small and smooth. The suture has only
a slight lobe on the ventral and lateral areas. There is not known
at present a perfect gradational series from "typical" Ceno-
ceras to the extremely involute Nautilus pisanus. However, its
stratigraphic position and general morphological characters
make it logical to consider Nautilus pisanus as the extreme in-
volute development of the Cenoceras complex and the genus
Sphaeronautilus is placed in synonymy of Cenoceras.

More species have been described for Cenoceras than for any
other genus of post-Triassie nautiloids. Of the 96 species, two-,
thirds are Lias in age and the remaining third Middle Jurassic
in age. Cenoceras is a truly cosmopolitan form being known from
nearly all continental areas (Fig. 6). The species belonging to
Cenoceras with their age and geographic distribution are listed
below.

DISTEIBUTION OF SPECIES OF THE GENUS CENOCEBAS

Stratigraphic Geographic
Species Distribution Distriliution

C. (Khu'iicus (Pia) 1914

('. afflnis (CMiaiJuis aiul DewaUjue) 185li

C. altisiphites (Prinz) 190(3

C. amasianus (Gugeiiberger) 1928

C. amorettii (Parona) 1897

C. anompJialus (Pia) 1914

C. arariformis (Pia) 1914

C. araris (Dumortier) 1869

Jurassic

Austria

(Lias)

Jurassic

Luxembourg

(Lias)

Jurassic

Hungary

(Lias)

Jurassic

Turkey

(Lias)

Jurassic

Italy

(Lias)

Jurassic

France

(Lias)

•Turassic

Austria

(Lias)

Jurassic

Europe

(Lias)

KUMMEL : POST-TBIASSIC NAUTILOID GENERA

373

Species

Stratigraphic Geographic
Distribution Distribution

C. aratus (Quenstedt) 1846

C. arthaberi (Gugenberger) 1928

C. aatacoides (Young and Bird') 1828

C. austriacus (Hauer) 1856

C. baconicvf^ (Vadasz) 1911

C. hnlsamoorivellii (Parona) 1897

('. brudfordcnsis (Crick) 1898

r. brancui (Gemmellaro) 1884

C. hreislacJ.-i (Parona) 1897

C. burto7ieniiis (Foord and Crick) 1890

C. catonis (Gemmellaro) 1886

C. clvlensis (Huppe) 1854

C. demonensis (M. Gemmellaro) 1911

?6'. dispansus (Morris and Lycett) 1850

i . distefanoi (Gemmellaro) 1884

C. domeyJcus (d'Orbigny) 1842

C. egregius (Pia) 1914

C. ercycinus (Tagliarini) 1901

C. excavatus (J. de C. Sowerby) 1826

C. exiguus (Crick) 1898

Jurassic

Germany

(Lias)

Jurassic

Turkey

(Lias)

Jurassic

Europe

(Lias)

Jurassic

Alps

(Lias)

Jurassic

Himgary

(Lias)

Jurassic

Italy

(Lias)

Jurassic

England

(Bajocian)

Jurassic

Italy

(Lias)

Jurassic

Italy

(Lias)

Jurassic

England

(Bajocian)

Jurassic

Sicily

(L. Dogger)

Jurassic

Chile

(Lias)

Jurassic

Sicily

(Lias)

Jurassic

England

(Bathonian)

Jurassic

Italy

(Lias)

Jurassic

Chile

(Lias)

Jurassic

Austria

(Lias)

Jurassic

Sicily

(L. Dogger)

Jurassic

England

(Bajocian)

Jurassic

England

(Bajocian)

374

BULLETIN : JrUSEUM OF COMPARATIVE ZOOLOGY

Species

Stratigraphic Geographic
Distribution Distribution

C. exterehratus (Criek) 1898

C. fi-'icheramts (Foord and Crick) 1890

C. fourneti (Dumortier) 1874

C. geyeri (Prinz) 1906

C. hnllstattensis (Spengler) 193 9

C. imlayi (Kuniinel) 1954

C. impendens (Crick) 1898

C. inoniatu.s (d'Orbigny) 1842

E. intermeduis (J. Sowerby) 181 fi

C. jourdani (Dumortier) 1874

C. juUanvs (Fucini) 1895

C. jurensis (Quenstedt) 1858

C. Uneatus (J. Sowerby) 1813

C. lineolatus (Foord and Crick) 1890

C. Iiipheri (Kummel; 1954

C. lutatii (Gemmellaro) 1886

C. maUi.erhii (Terquem) 1855

C. manissad jinni (Gugenberger) 1928

C. mariani (M. Gemmellaro) 1911

C. marii (Gemmellaro) 1886

Jurassic

England

(Bajocian)

Jurassic

France

(Lias)

Jurassic

Europe

(Lias)

Jurassic

Austria

(Lias)

Jurassic

Austria

(Lias)

Jurassic

Alaska

(Bajocian)

Jurassic

England

(Bajocian)

Jurassic

Europe

(Lias)

Jurassic

Europe

(Lias)

Jurassic

Europe

(Lias)

Jurassic

Italy

(Lias)

Jurassic

Europe

(Lias)

Jurassic

England

(Bajociauj

Jurassic

England

(Lias and

France

Bajocian)

.Jurassic

Oregon

(Bajocian)

(U.S.A.)

Jurassic

Sicily

(L. Dogger)

Jurassic

Luxembourg

(Lias)

Jurassic

Turkey

(Lias)

Jurassic

Sicily

(Lias)

Jurassic

Sicily

(L. Dogger)

KUMMEL : POST-TRIASSIC XAUTILOID GENERA

375

Species

Stratigraphie Geographic
Distril lilt ion Distribution

C. mazzarensis (Tagliarini") 1901

C. meyrati (Ooster) 1858

C. multiseptaius (Foord and Oick) 1890

C. obesus (J. Sowerby) 1816

C. ohstructus (Deslongebanips) 1878

C. orbignyi (Prinz) 1906

C. nrnatus (Foord ;md Crick) 1890

C. nrnatus var. atanatcnsls (Pia) 1914

C. paretoi (M. Gemmellaro) 1911

C. Tperinflatus (Foord and Criok') 1890

C. pfirornaius (Crick) 1894

C. pe.rtexiii>i ( J)uni(ii'tli'r) 18(>7

C. pimniis (Fucini) 189o

C. po!>i(/'>nali.<: (.7. d.' C. Sowerby) 1820

C. postriutuK (Prinz) 1900

C. prof andisiphytes (Prinz) 1900

C. pseudolincalus (Foord and Crick) 1890

C. pseu(lori((/osus (Pia'^ 1914

C. psfiudotruncatuft (Crick) 1921

C. quadrangularis (Pia) 1914

C. robu.ifvs (Foord and Crick) 1890

Jurassic

Sicily

(L. Dogger)

Jurassic

Alps

(Lias)

Jurassic

England

(Bajocian)

Jurassi*

England

(Bajocian)

Jurassic

France,

(Bajocian)

England

Jurassic

Europe

(Lias)

Jurassic

England

(Bajocian)

Jurassic

Austria

(Lias)

Jurassic

Sicily

(Lias)

Jurassic

England

(Bajocian)

Jurassic

Western

(Bajocian)

Australia

Jurassic

France

(Lias)

Austria

Jurassic

Italy

(Lias)

Jurassic

England

(Bajocian)

Jurassic

Europe

(Lias)

J urassie

Austria

(Lias)

Jurassic

England

(Bajocian)

Jurassic

Alpine

(Lias)

Region

Jurassic

England

(Lias)

Jurassic

Austria

(Lias)

Jurassic

France,

(Lias)

Austria

376

BL LLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Species

Stratigraphie Geographic
Distribution Distribution

C. rotundum (Hyatt) 189-4

C. rotwndum (Crick) 1898

C. rugosu^ (Buvignier) 1832

C. schlumbergeri (Terqueiii) 1855

C. schmidti (Giebel) 1852

C. schwalmi (Prinz) 190ti

C. secernendus (Pia) 1914

C. semiornatus (Crick) 1898

C. .lemistriatus (d'Orbigny) 1843

C. semseyi (Prinz) 1904

C. simillimiis (Foord and Crick) 1890

C. smithi (Foord and Crick) 1890

C. spreaficoi (Parona) 1897

C. steiiimanni (Moricke) 1894

C. stoppanii (Parona) 1897

C. striutiis (J. Sowerby) 1817

C. 6-tu,ri (Hauer) 1856

C. subrotundus (Crick) 1898

C. subtruncatus (Prinz) 1906

C. terebraUis (Dumortier) 1874

J urassic

England

(Bajocian?)

Jurassic

England

(Bajocian)

J urassic

Europe

(Lias)

Jurassic

Luxembourg

(Lias)

Jurassic

Germany

(Lias)

J ura;-sic

Hungary

(Lias)

•J urassic

Austria

( Lias )

Jurassic

England

(Bajocian)

Jurassic

France

(Lias)

Jurassic

Hungary

(Lias and

L. Dogger)

Jurassic

England,

(Lias)

Austria

J urassic

England

(Bajocian)

Jurassic

Italy

(Lias)

Jurassic

Chile

(Middle)

J urassic

Italy

(Lias)

Jurassic

Europe

(Lias)

Jurassic

Alpine

(Lias)

Region

Jurassic

England

(Bajocian)

Jurassic

France

(Lias)

Jurassic

France,

(Lias)

England

KUMMEL : POST-TRIASSIC XAUTILOID GENERA

377

Species

Stratigraphie Geographic
Distribution Distribution

Jurassic

Sicily

(L. Dogger)

Jurassic

France,

(Lias)

Germany,

Jurassic

Austria

(Lias)

Jurassic

Europe

(Lias)

Jurassic

Sicily

(L. Dogger)

Jurassic

Sicily

(L. Dogger)

Jurassic

Sicily

(Middle)

C. thyrrenus (Tagliarini) 1901
r. ioarcensis (d'Orbigny) 1849
C. tricarinatus (Vadasz) 1911
C. truncatus (Sowerby) 1816
C. icaageni (Gemmellaro) 1886
C. zignoi (Gemmellaro) 1886
C.zitteli (Gemmellaro) 1886

Genus ErTREPHOCKKAS Hyatt, 1894

Eutrephoceras Hyatt, 1894, p. 555.
Eutrephoceras Eeeside, 1924, p. 2.
Eutrephoceras Eeeside, 1927b, pp. 6, 7.
Eutrephoceras Miller and Thompson, 1933, pp. 300-304.
Eutrephoceras Spath, 1927a, pp. 21, 24.
Eutrephoceras Spath, 1927b, pp. 425, 426, 428.
Eutrephoceras Stenzel, 1940, pp. 738-742.
Eutrephoceras Miller, 1947, pp. 26-39.
Eutrephoceras Miller, 1951, pp. 33-36.

The genus Eutrephoceras represents the long-lived and per-
sistent stock arising from the lower Jurassic Cenoceras complex
and is considered to be the root from which most of the remain-
ing Cretaceous and Tertiary uautiloids, excepting the Cymato-
oeratinae, are derived. It is most similar to Cimomia and Xau-
tilus. Cimomia arose from Eutrephoceras in the Upper Jurassic
or Cretaceous and is in itself the root stock of the Hercoglossinae.
Cimomia differs from Eutrephoceras in the greater sinuosity of
the suture, but many intermediate forms are known.

Thus the interpretation of Eutrephoceras presented here vis-
ualizes a persisting, slowly evolving stock consisting of generally
involute, smooth forms with straight or nearly straight sutures.
and a conch that is generally subglobular but can be quite com-
pressed or depressed. The rather wide variation in the shape of
the conch is illustrated iu Text Figure 13.

378

BULLETIN: MrSEr:\[ OF COMPARATIVE ZOOLOGY

Figure 1 3

KUMMEL : POST-TRIASSIC XAUTILOID GENERA 379

Probably the eoiumouest and most widespread genus of post-
Triassic nautiloids is Eutrephoceras. To date, approximately
89 species are recognized ranging in age from the Jurassic to
the JMiocene ; of these, 6 species are Jurassic in age, 39 Cretaceous
and 44 from the Tertiary.

Miller has recently described the American Tertiary species
of Eutreplioccras and his comprehensive diagnosis of the genus
is quoted here (1947, p. 27) : "Conch nautilicouic and typically
subglobular ; whorls reniform in cross section, broadly rounded
veutrally and laterally, and moderately deeply impressed dor-
sally. Aperture marked ventrally hy a broad shallow rounded
liyponomie sinus. Umbilicus small and inconspicuous; umbilical
shoulders low and rounded. Surface of conch smooth or essen-
t ially so. Septa moderately convex apicad ; sutures of t^-pical
forms slightly sinuous, but at least the external sutures of some
forms are essentially straight. An annular lobe is present in
some forms but not in the genotype. Siphuucle small, circular
in cro.ss section, and orthochoanitic in structure ; its position
varies considerably in the different species, but in no case is it
marginal.''

The large number of species of Eutrephoceras reflects their
general abundance and widespread distribution geographically
and stratigrajihically. The wide variability in conch shape,
degree of involution, and inflection of the suture suggest a wide
adaptive range for this genetic stock. Probably in no other

Fig. 13. Cross sections of the eoneh of A, Eutrephoceras cyclotus (Op-
l)el) from Zittel 1868, pi. 3, fig. 2a, diameter 117 mm.; B, E. doissieri
(Pictet) 1867, pi. 8, fig. 4b, diameter 63 mm.; C, E. nuhinjlatus (d'Orbigny)
1850, from d'Orbigny 1840, pi. 37, fig. 2, diameter 88 mm.; D, E. jonesi
Miller and Thompson, from Stenzel 1940, fig. 115, diameter 225- mm.;
E, E. piermntii (Sergio) 1933, pi. 2, fig. 4b, diameter 144 mm.; F, E.
Justus (Blanford), from Stoliczka 1866, pi. 93, fig. 2a, diameter 64 mm.;
G, E. dekayi (Morton), from Stenzel 1940, fig. 115, diameter 46 mm.; H, E.
fictormnum Teichert 1947, fig. 5, diameter 102 mm. ; /, E. reesidei Stenzel
1940, fig. 115, diameter 14.6 mm.; J, E. yieitiergicus (Eedtenbacher) from
•Schliiter 1876, pi. 48, fig. 4, diameter 90 mm.; K, E. dekayi (Morton^
from Stenzel 1940, fig. 115, diameter 85 nmi.; L, E. sloani Eeeside from
Stenzel 1940, fig. 115, diameter 120 mm.; M, E. carolinense Kellum, from
Stenzel 1940, fig. 115, diameter 20.7 mm.; :N', E. lentiformis (Stoliczka)
1866, pi. 93, fig. la, diameter 74 mm.; 0, E. cooTcanum (Whitfield), Stenzel
1940, fig. 115, diameter 220 mm.; P, E. laverdei Durham 1946, pi. 63, fig. 4,
diameter 61 mm.

380

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

generic group are the recognized species so gradational in charac-
ter and generally lacking in objectivity. Variability studies of
large populations would be extremely helpful in assessing the
many species groups.

Since EufrepJioceras is interpreted as the basic persisting stock
derived from the Cenoceras complex, its relationships to the
other genetic groups are discussed in detail under each of these
derived groups and need not be repeated here.

xl similar long-persisting, smooth, involute, generalized stock
is represented in the Permo-Carboniferous and Triassic by the
Liroccras-Poranautilus trend. From this particular trend was
derived the Upper Triassic Clydonautilidae which is a stock
adapted for sinuosity of the suture ; a homeomorphous de-
velopment of the Ilercoglossinae.

EutrepJwceras is world-wide in distribution and ranges in age
from the Upper Jurassic into the ^Miocene.

DISTRIBUTION OF SPECIES OF THE GENUS EUTREPHOCERAS

Species

Stratigraphic Geographic
Distribution Distribution

K. ahltenensu'i (Sehliiter) 187fi

E. aleetienne Reeside, 1927

E. allani (Fleming) 1945
E.alUoni (Michelotti) 1840
E. altifrons (Cliapnian) 1915
E. archiacianus d'Oibigny 1840
E. 'balcomben-sis (Cliapnian) 1915
E. hrllcropho)! (Lundgren') 1867

E. benyl Miller, 1947

E. Uanfordi (Douville) 1929
E. hoissieri (Pictet) 1867

E.houchardianus (d'Orbignyj 1840

E. bryani (Gabb) 1877

Cretaceous

Germany

(Upper)

Cretaceous

Montana

(Upper)

(U.S.A.)

Oligocene f

New Zealand

Miocene

Italy, Malta

Miocene

Australia

Cretaceoua

France

Miocene

Australia

Cretaceous

Sweden

(Danian)

Eocene

North

Carolina

(U.S.A.)

Eocene

Pakistan

Cretaceous

Switzerland,

(Lower)

Algeria

Cretaceous

France

(Albian)

Eocene

New Jersey

(U.S.A.)

KUMMEL : POST-TRIASSIC XAUTILOID GENERA

381

Species

Stratigraphic Geographic
Distribution Distribution

E. burl:arti (Castillo and Aguilera) 1895

E. butonensis (Martin) 1933
E. carolinense Kellum, 1926

E. centraHs (J. Sowerby) 1812

E. charpentieri (Leymerie) 1851

E. clemientinus (d'Orbigny) 1840

E.cooJcanum (Whitfield) 1892

E. crassus (Sehafhautl) 1863
E. cydotus (Oppel) 1865

E. dartevellei Miller 1951

E. darupensis (Schliiter) 1876

E. decipiens (Michelotti) 1861
E. dehayi (Morton) 1834

E. depres.suft (Binckhorst) 1861

E. dersertorum (Quaas) 1902

E. dietrichi (Zwierzyeki) 1914

E. douvillei Spath, 1927

E. dubaleni (Peyrot) 1932
E. euthymi (Pictet) 1867

E. expansum (J. de C. Sowerby) 1824

E. faxoense Hyatt, 1S94

E. felix (Chapman) 1915

Jurassic

Mexico

(Tithonian)

Oligocene

East Indies

Eocene

Xorth

Carolina

(U.S.A.)

Eocene

England,

Germany

Cretaceous

France

(Upper)

Cretaceous

England

(Albian)

France

Eocene

Xew Jersey

(U.S.A.)

Eocene

Bavaria

Jurassic

Moravia

(Tithonian)

Eocene

Landana,

West Africa

Cretaceous

(4ermany

(Senonian)

Miocene

Italy

Cretaceous

U.S.A.

(Upper)

Cretaceous

Belgium

(Maestrichtian)

Cretaceous

Libya

(Danian)

Cretaceous

East Africa

(Neocomian)

Jurassic

Eg>-pt

(Oxfordian)

Miocene

France

Cretaceous

France,

(Lower)

Switzerland

Cretaceous

England

(Cenomanian

)

Cretaceous

Denmark

(Danian)

Miocene

Australia

382

Bl'LLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Species

Stratigraphic Geographic
Distribution Distribution

E. ftamweus (Bom-lietti) 19-17

E. francomontaniis (Kului) 1939

E. geelo7i(/ensis (Foord) 1891

E. gosavicus (Eedtenbacher) 1873

E. hannai Vokes, 1937

EJ hallidayi (Waring) 1914

E. hendersoni (Etheridge) 1901

E.indicum (SpenglcM) 1910

E. izmnoensis Yokoyania, 1913
E. japonicus (Shimizu) 1926
E. javanus (Martin) 1879
E. johnsoni Miller, 1947

E. 'jonesi Miller and Thompson, 1933

E. lahcchi (d'Arehiac and Haime) 1854
E. laverdei Durham, 1946

E. leonei (Negri) 1934
E. mayJcsi Miller, 1947

E. marucoensis (Giovine) 1950

E. meta-fiettriuiitii (Douville) 1929

E. montanensi Kummel, 1954

E. montmoUim (Pictet and Campiche) 1859

E. neuhergicus (Eedtenbacher) 1873

E. oregonense Miller, 1947

Cretaceous

Tripolitania

(Maestrichtian)

Jurassic

Germany

(Upper)

Miocene

Australia

Cretaceous

Austria

(Upper)

Eocene

California

(U.S.A.)

Paleocene

California

(U.S.A.)

Cretaceous

Queensland,

(Lower)

Australia

Cretaceous

India

(Upper)

Tertiary

Japan

Eocene

Japan

Tertiary

Java

Paleocene ?

Alabama

(U.S.A.)

Paleocene

Alabama

(U.S.A.)

EoL-eno

Pakistan

Cretaceous

Colombia

(Aptian)

Eocene

Libya

Eocene

California

(U.S.A.)

Cretaceous

Argentina

(Hauterivian)

Cretaceous N. Africa

(Upper)

L. Tertiary Pakistan

Jurassic Montana

(Callovian) (U.S.A.)

Cretaceous Switzerland

(Albian)
Cretaceous
(Upper)
Eocene

France
Austria

Oregon
(U.S.A.)

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

383

Species

Stratigvaphic
Disti-il:)iition

Geograpliic
Distribution

E. ovoid.eum Crick, 1907

E. i)armensis (Deshayes) 1866

E. perlatus (IMorton) 1834

E. piersantii (Sergio) 1933

E. planoventer Stephenson, 1941

E. quadrilincatus (Favre) 1869

E. reesidei Stengel, 1940

E. regalis (J. de C. Sowcrby) 1843
E. regvJaris (Schafhuutl) 1863
E. restrict us (Griepenr-kerl') 1889

E. rrsupinaiits (Kedtenbacher) 1873

E. scDnfiJippoi (Sorrentino) 1932

E. simile Spath, 1953

E. sloani Eeeside, 1924

E. sphaericum (Forbes) 1846

E.staadti (Cossniann) 1902

E. stephensoni (Dickerson) 1914

E. stricteumtilicatus (Stchepinsky) 1943
E. subinflatvs (d'Orbigny) 1850

E. s tibia evigatum (d'Orbigny) 1840

E. subplicatum (Philippi) 1895

E.szoniaghi (Yogi) 1910
E. thomi Eeeside, 1927

E. tumescens (Franscher) 1895

Cretaceous Zululand
Eocene France

Cretaceous Alabama
(Upper) (U.S.A.)

Oligocene Italy
Cretaceous Tennessee,
(Upper) Texas

(U.S.A.)
Cretaceous Germany
(Upper)
Eocene Texas

(U.S.A.)
Eocene England

Eocene Bavaria

Cretaceous Germany
(Senonian)
Cretaceous Austria
(Upper)

Cretaceous Tripolitania
(Maestrichtian)
Cretaceous Graham

(Upper) Land
Eocene South

Carolina

(U.S.A.)
Cretaceous India
Eocene France

Paleocene California

(U.S.A.)
Oligocene Turkey
Jurassic France

(Kimmeridgian)
Cretaceous England,
(Upper) India,

Madagascar,

Libya
Cretaceous Argentina,
(Upper) Chile

Eocene Hungary

Cretaceous Montana
(Upper) (U.S.A.)

Eocene Austria

884

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Species

Stratigraphic

Geographic

Distribution

Cretaceous

South Africa

(Yalanginian;

Eocene

France

Eocene

England

Cretaceous

Poland

(Upper)

Tertiary

Italy

( Lower)

Eocene

Australia

Oligocenc

Italy

E. uitenhageiise Spath, 1930

E. iimhilicaris (Deshayes; 1835
E. urhanus (Sowerby) 1843
E. vastus (Kner) 1850

E. vicentinvs (Oppenheim; 1901

E. Victorian um (Teichert) 1943
E. vina-ssai (Yenzo) 1937

Genus PsEUDOCENOCEEAS Spath, 1927

Typt .species. Xai'.tiius largiUiertianvs d'Orbiguy, 1840, pi. 18 (by original
designation). Picsiotypp illusti'ated on Piate 10, figs. H, 4.

This genus can be diagnosed as follows : Conch involute, com-
pressed, smooth. AVhorl section sul)rectaugular, venter broad,
flattened, ventral shoulders rounded. Whorl sides flattened, sub-
parallel. Umbilical shoulders rounded, umbilical wall nearly
vertical. The suture is onlv slightlv sinuous, essentiallv straight
across the venter and with a broad, generally shallow, lateral
lobe. The siphuncle is subeentral in position, that is, closer to
the dorsum.

As stated by Spath (1927a, p. 24), "Pseudocenoceras is
easily distinguished from its Cretaceous contemporaries by
its steep umbilical edge, truncated whorl-shape, entomarginal
siphuncle and reclined septal edges."

The Cretaceous contemporaries of I'seiidocenoceras include
members of the Cymatoeeratinae, the Hereoglossinae, EutrepJw-
ceras, and Carinonauiilns. The smooth conch and simple suture
would ally Euirephocera^ and Pficudoccnoceras; all of the other
generic groups are specialized forms trending on quite different
patterns.

There are only 8 species of Cretaceous nautiloids that can be
assigned to Pseudocenoceras. and all are for the most part very
similar. The type species is a common form in Cretaceous strata
of England and France. Nautilus heryiasensis Pietet (1867)
from Lower Cretaceous strata of France appears to be closely
allied to Nautilus largilliertianus but has a more central siph-
uncle. Nautilus lehardyi Binckhorst (1861) is a quite aberrant

KUMMEL ; POST-TRIASSIC XAUTILOID GENERA

38o

form with subaugular ventral shoLilders aud a concave venter.
It is placed in Pseudocenoceras with question.

Pseudoceiioceras is a fairly common form in Cretaceous rocks
of Europe from England to the Crimea. One species, Nautilus
applanaius Wanner (1902), is from the Upper Cretaceous of
Libya. A list of the species of Pseudocenoceras, their age, and
geographic distribution is given below.

DISTEIBL'TIOX OF SPECIES OF THE
GEXrS PSEUDOCEXOCEEAS

Stratigraphic Geographic
Species Distribution Distribution

P. applanaius (Wanner) 1902

P. herriasensis (Pictet) 1867

P. campichei (Karakasch) 1WU7

P. ntoni (Sharpe) 1853

P. fialicianu^ (Alth) 1850

P. largilliertianus (d 'Orbignj- j 1840

PJ lehardyi (Binekhorst) 18()1

P.pirteti (Karakasch) 1907

Genus CaEIXOXAL'TILI'S Spengler, 1910
Type upccies. Carinonautihis ariyalurensis Spengler, 1910, p. 149, pi. 14.
.figs. 1 a (' (monotypic). Illustrated here on Plate 11.
The only data available on this monotypic genus are the
description and illustration of the type species. However, from
this source the following diagnosis is made. The conch is very
involute, compressed, and much higher than wide. The umbilicus
is small anci shallow. The whorl sides are broadly rounded and
converge toward the venter. A distinct furrow aligns the ventral
shoulder. The venter has a prominent rounded keel that on the
adoral part of the living chamber is much broader and is divided
by a median furrow. The umbilical shoulders are broadly
rounded. The widest part of the whorl section is just ventral of
the umbilical shoulders. The suture is onlv slightlv convex.

Cretaceous

Libya

(Upper)

Cretaceous

France

(Lower)

Cretaceous

Crimea

(Lower)

Cretaceous

Europe

(Upper)

Cretaceous

Po'.and

Cretaceous

Europe

(Cenomanian)

Cretaceous

Belgium

(Maestriehtian)

Cretaceous

Crimea

(Lower)

■ 186 BULLETIN : Ml'SEUil OF COMPARATH'E ZOOLOGY

The ouh- markings on the shell are growth lines -which form
a deep sinus over the venter. The position of the siphuncle is not
known.

At a diameter of 1.5 cm. the venter is not yet keeled but merely
sharply rounded. At 3 cm. the single prominent keel is present,
and at 6 cm. diameter the median furrow is well developed on
the widened keel.

The measurements of the type species are as follows (Spengler,
11)10, p. 149) :

Diameter 93 mm.

Height of last whorl 54 mm.

AYidth of umbilicus 5.5 mm.

Width of last whorl 32 mm.

Spengler (1910, p. 149) recognized the uniqueness of this
species and suggested that it was derived from Nautilus angustus
Blanford {=Ci))iumia) by sharpening of the ventral area. Con-
temi^oraneous forms include members of the Cymatoceratinae,
Hercogiossinae, Eutrephoceras and Pseudocenoceras. I- am in
essential agreement with Spengler on the phylogenetic position
of Carifw7iauiihis but would derive them from Eutrephoceras,
which is here considered the persistent evolving stock. There is
slight difference between the eutrephoceratid stock and the
cimomids. The eutrephoceratids show great variability in their
smooth, involute conchs and appear to be the only evolutionary
reservoir available to give rise to Carinonautilus.

This monoty])ic genus is known only from the Ariyalur
group (Campanian) of the Trichinopoly district of southeast
India.

Genus ObixaUTILUS Kobayashi, 1954

Tijpe .specie^. Obiiiaiitiluti pulchra Koliayashi, 19:')41). Type illustiatloii ro
produced on Plato 9, figs. 1, 2.

Kobayashi 's description of his new monotypic species and
genus is quoted below. The specimen is 78 mm. in diameter ;
last whorl 25 mm. in width. 46 nnn. in height and 35 mm. in
median height.

' ' Shell discoidal, compressed laterally, rapidly coiling and
quite involute ; umbilicus almost closed, if not completely ; flanks
a little convex, but nearl.v parallel in the inner half and inclined
abruptly near the umbilicus ; ventral sinus shallow but fairly

KUMMEL : POST-TRIASSIC NAUTILOID GEXERA 387

broad and provided with an obtuse peripheral carina on each
side which is siibangulated inside but outwardly merges with
the flank.

"Radial ribs narroAv, flattopped, separated by narrow and
shallow grooves and frequently branching distally by insertion
of fine grooves. These ribs and grooves distinctly bent forward
from the umbilicus, nearly straight or even slightly concave
backward in the broad flank, swing back a little near the venter
and moderately sinuated behind on the venter. The curvature of
these ribs is very similar to the growth lines of living Nautilus."

Kobayashi based the above description on a single well-
preserved specimen. The shell is very well preserved and he
(Kohayashi) notes "its deep and light bro^ai shades must be
the original colour pattern."' Unfortunately, neither the suture
nor position of the siphuncle was observed. The ribbing de-
scribed by Kobayashi appears to be accentuated growth lines
rather than the type of ribbing that is characteristic of the
Cymatoceratinae.

The most characteristic feature 0/ Otiyiauiilus is the shallow
I)ut distinct ventral furrow. Because of this modification of
the periphery, Kobayashi rightly pointed to possible affinities
with the Paracenoceratidae. However, I agree completely with
Kobayashi that OhinaiifUus should be considered a development
of the Xautilinae rather than derived from any member of
the Paracenoceratinae. Since the suture and position of the
.siphuncle are not known, this interpretation will have to remain
tentative until more data become available.

Ohinautil lis pulchra is known from one specimen collected at
a cutting between Aburatsu and Obi towns in the Province of
Hyuga (Miya^^aki Pref.), Japan, on the occasion of the high-
way operation. The locality is situated in the pre-Mioceue ter-
laiii where the Nichinan formation containing the Oligocene
Ashiya fauna lies unconformably beneath the ^Miocene Aburatsu
formation (Kobayashi, 1954b, p. 183).

Genus NAUTILUS Linne, 1758
7/7)' .sprcicK. Xdiitihis pompilius Linne.

The interpretation of the scope of the genus Nautilus has
changed considerably- since its formal introduction by Linne.

388 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

In the early period of paleoutological research on this group of
tetrabranchiate cephalopods nearly all species were placed in the
single genus Nautilus. Gradually as more and more genera were
proposed, Nautilus became a waste-basket genus for unassigned
forms. The greater amount of attention paid to Paleozoic nautil-
oids soon pre-empted the use of the genus Nautilus for species
of that age. The genus has persisted for most species of Mesozoic
and Tertiary nautiloids until rather recently. At the present, for
instance. Miller (1951, p. 32) restricts Nautilus to only the
Recent species and all of the Tertiary forms previously assigned
to Nautilus are placed in Eutrephoceras or Cimomia. In this
interpretation I am in full agreement. Eutrephoceras is the
persisting stock out of the Cenoceras complex in Upper Jurassic
time that shows no particular evolutionary trends from its time
of origin to the time of its apparent extinction in the late
Tertiary. It is the longest-lived single genetic stock of post-Tri-
assic nautiloids, and was ancestral to several other aberrant
groups, namely Pseudocenoceras, Carinonautilus, and Ohinau-
tilus. It is also ancestral to Cimomia of the Hercoglossinae which
developed into an extremely abundant and diverse evolutionary
complex. There appear to be no records of any Pliocene or
Pleistocene nautiloids. Aside from Aturia, all other species of
nautiloids seem to be more closely related to the eutrephoceratid
evolving stock than to modern Nautilus. The number of species
and specimens of Miocene nautiloids is still very few. However,
with the above interpretation Nautilus is thought to arise from
the eutrephoceratid stock at some time in the mid or late Ter-
tiary.

The biology and ecology of modern Nautilus has been very
ably summarized by Stenzel (1948, 1952). Similar comprehen-
sive discussions of the morphology, etc., can be found in Miller
(1947). The extensive literature on modern Nautilus is covered
in the bibliographies of Stenzel 's and Miller's papers.

Three species of living NautiJiis are generally recognized,
namely N. poynpilius (with the umbilicus practically obliterated
by the deposition of callus) N. macromphalus (with an open im-
perforate umbilicus), and N. umhilicatus (with an open finely
perforate umbilicus) (Miller 1947, p. 13). Several "varieties"
of N. pompilius are recognized but their significance and rela-
tionships are poorly known. To the above three species, Iredale

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 389

(1944) has added two new species, namely N. alumnus, from off
Queensland and New South Wales, and N. repertus from off
Western Australia.

Living NautUus is restricted to the southwest Pacific from
South Australia to the southern Philippines and eastward to
the Fiji Islands. Stenzel (1948, p. 84) gives a good summary of
the known records for live specimens of Nautilus.

Subfamily PSEUD AGANIDINAE nov.

This new subfamily is established for two genera of mainly
Jurassic age that include adaptive types characterized by great
sinuosity of the suture. These genera are Pseudaganides Spath,
and Pseudonautihis Meek which Spath (1927) had previously
included in his family Ilercoglossidae. As is discussed in more
detail below, Pseudaganides is considered to be a direct radiation
from the Cenoceras complex in the Lower Jurassic ; by the Upper
Jurassic it was a widely distributed form in the European
geosynclinal areas and show^ed a wide variation in conch form
and suture. The group did not survive into the Cretaceous.
Sometime during the Upper Jurassic an offshoot from Pseuda-
ganides developed, characterized mainly by greater elaboration
of the suture, that is, the formation of a deeper ventral lobe
and angular lateral lobes. Species of this genus range into the
Lower Cretaceous.

In the Carboniferous to Recent history of the nautiloids there
were four distinct adaptive trends built around elaboration of
the suture. As would be expected, there are many homeomorph-
ous forms between these adaptive trends. Each of these trends
had its origin in long-persisting, smooth, generalized stocks ^yitll
very simple, nearly straight sutures. The duration of each of
these trends is likewise somewhat limited. The Permian trend
is known only through a single species of the genus Permoceras
from Middle Permian strata of Timor. The Triassic radiatiou
is mainl}^ Upper Triassic (Carnian-Norian) in age. The Jurassic-
trend, represented by Pseudaganidinae, is Lower Jurassic to
Lower Cretaceous in age with the principal radiation in the
Upper Jurassic. The last such trend, that of the Hercogiossinae
and Aturinae, is Upper Cretaceous to mid-Tertiary in range with
the principal radiation in the Lower Tertiary.

390 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Genus PsEUDAGANIDES Spath, 1927

Pseudaganides Spath, 1927a, pp. 22, 25.
Hercoglossoceras Spath, 1927a, pp. 22, 26.
Nautilus (Pseudaganides) Jeannet, 1951, pp. 17-21.

Type species. Nautilus hutcliensis Waagen, 1873, pi. 3, fig. 4 (Plate 28,
figures 3, 4 of this report). Type by original designation.

When Spath (1927a, p. 22) established this genus the only
remark he made as to its characters was the following- statement :
"Pseudaganides, which comprises the early 'aganitici', like
Somalinautilus, has its origin in 'Cenoceras' and includes, for
example, the large suhsinuatus group of the Inferior Oolite, as
already mentioned." In his discussion of the type species,
Pseudaganides ktitchensis (Waagen), Spath (1927a, pp. 34-85)
mentions the following species as belonging to this genus : A'.
aganiticiis Schlotheim, N. scMotheimi Loesch {=Naatilus {Her-
coglossa) aganiticus Foord and Crick, 1890b, p. 394, fig. 1), and
N. ammoni Loesch; he also included with question N. girardoti
de Loriol and N. siihhiangulafus d'Orbigny {=Nautilus hiangu-
latus d'Orbigny, 1845, pi. 34, figs. 1-3).

There are approximately 40 species of Jurassic nautiloids with
"goniatitie" sutures. Besides the "aganitici" included by Spath
in Pseudaganides, he established the genus Hercoglossoceras
[type species — Hercoglossa {'gravesiana (d'Orbigny) var.')
kochi (Prinz)], a highly compressed form. Pseudonautilus was
established by Meek for Nautilus gcinitzi Oppel {in Zittel, 1868),
and is characterized by a pointed lateral lobe and a very deep
ventral lobe. A number of species from the Upper Jurassic
(recently discussed by Loesch, 1914), Spath tentatively assigns
to Hercoglossa (principally an Upper Cretaceous-Eocene de-
velopment) though he does mention that they probably repre-
sent an independent development.

My own studies lead me to conclude that all of the Jurassic
nautiloids with "goniatitic" sutures are part of a single evolu-
tionary line developed from some compressed elements of "Ceno-
ceras" and this line is quite independent of the later hercoglos-
sid development.

The type species of Pseudaganides — N. kutchcnsis — comes
from the Middle Chari group {anceps zone) of Kachh (Cutch)
Pakistan. The type measures 47 mm. in diameter, 28 mm. for

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

391

length of whorl, 27 mm. as width of the whorl, and the umbilicus
is 5 mm. across. AVaagen's illustration is reproduced here on
Plate 28, figures 3, 4. The form of the shell is involute with a
somewhat rectangular whorl section with flattened sides and
venter. The greatest width of the whorl is near the umbilicus.
The suture has a prominent lateral lobe and a saddle next to
the umbilicus. The suture is projected toward the venter au'l
has a shallow lobe on the venter. The shell has fine longitudinal
strigation and growth lines. The strigations disappear towards
the aperture.

Fig. li. Diagrammatic representation of sutures of A, Pseudaganidc.s
Jjrunhuheri (Loeseli) 1914, tig. oc; B, P. franconicus (Oppel) from Loescli
1914, fig. la; C, P. subMangidatus (d'Orbiguy) 1843, from d'Orbignv
1850. pi. 34, fig. 3; 1\ P. rogeri (Loescli) 1914, fig. 3a; E, P. pellerensis
(Vialli) 1937, fig. la; F, P. sinuafu.^ (Sowprby). from Vialli, 1937, fig. la.
All figures greatly reduced.

From the Carl)oniferous to the Tertiary, evolutionary trends,
formulated on a cfjmplication of the suture, have occurred re-

392 BULLETIN : MUSEUM OF (OMPARATIVK ZOOLOGY

peatedly. In most cases it can be established that such evolu-
tionary lines arose from ancestors with essentially uncomplicated
(nearly straight) sutures. Many of these evolutionary lines are
represented by a single or very few species. However, some
groups as the Triassic Clydonautilidac have large numbers of
species. In genera such as Clydonantdus and Procly do nautilus
there is a constancy in the pattern of the suture with, of course,
a degree of modification in shape and depth of the lobes, etc. ;
there is, however, a very striking difference or range of variation
in the shape of the conch (see Kummel, ll)53c, figs. 38, 40, 41).
This suggests that the genetic complex governing the nature of
the suture is less adaptive than the shape of the conch. Among
the Jurassic "agauitici" there is a certain homogeneity in the
suture, mainly in the pronounced lateral lobe, and in the pro-
jected ventral saddle with its shallow lobe. The differences en-
countered are more a matter of degree, with some element of the
suture being larger, deeper, or of slightly different shape (Fig.
14). The scope of variation is similar to that found in species
of the Clydonautilidac.

The early "aganitici" such as ''Hercoglossa" kochi (Prinz)
of the Lias, Nautilus subsinuatus d'Orbignj' of the Inferior
Oolite, N. crassisinuatus Crick also of the Inferior Oolite, N. pel-
lerensis Vialli of the Bajocian of north Italy, and N. (Inilii Gem-
Fig. 15. Cross sections of the conch of A, Pseudaganides pulchellus
(Jeannet) 1951, tig. 45, diameter -14 mm.; B, P. hrenlceVi (Jeamict) 1951,
fig. 37, diameter 50 mm.; C, P. oppeli (Zittel) 1868, pL 4, fig. lb, diameter
100 mm.; D, P. suhliungulatus (d'Orbigny) 1850, from d 'Oibigii.v, 1843,
pi. 34, fig. 2, diameter 92 mm.; E, P. rugerl (Loesch) 1914, fig. 31), height
50 mm.; F, P. .s('h}iekli (LuesL-h) 1914, fig. 2a, height 35 mm.; c;, P.
agdniticus (Selilotheim) from Jeannet, 1951, fig. 42, height 25 mm.; //, /'.
lodii (Prinz) 1906, fig. 5, height 77 mm.; /, P. frlchensis (Jeannet) 1951,
fig. 36, diameter 46 mm.; J, P. stramhergensis (Oppel) 1865, pi. 2, fig. 8a,
diameter 98 mm.; K, P. biunhuhiri (Loesch) 1914, fig. 51j, height 50 mm.;
living chamber; L, P. bninhiihoi (Loesch) 1914, fig. 5a, height 39 mm.;
phragmocone; M, P. ainuatus (Sovverby) from d'Orbigny, 1843, pi. 32.
diameter 188 mm.; N, P. gravesianus (d'Orbigny) 1843, pi. 38, diametei
270 mm.; 0, Pseudonautiius aturioidts (Pictet) 1867, pi. 11, fig. Ic, diametei
60 mm.; P, P. malhosl (Pictet) 1867, pi. 9, fig. 2b, diameter 100 mm.; Q.
Pseudaganides dumasi (Pictet) 1867, \)\. 10, fig. 2Ii, diameter (iO mm.

KUMMEL : rO.-T-TRlASSIC NAUTILOID GENERA

393

Figure 1 5

894 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

mellaro are all compressed, involute forms with the whorls
higher than wide. The sutures have pronounced lateral lobes and
are strongly projected towards the venter which they cross in
nearly a straight line. In nearly all other species of Fseudagan-
ides there is a slight lobe on the peripheral area. The lateral
saddle in the above early "aganitici" is quite variable. It is
very pronounced in iV. subsinuatus and N. crassisitmatus but
very low in N. pellerensis and N. diiilii. In each of these latter
two species the umbilical seam passes through about the middle
of the lateral saddle.

The sutures in most of the remaining ' ' aganitici ' ' of the Juras-
sic have a shallow lobe on the venter. This small ventral lobe
appears to persist in forms like N. ammoni Loesch, N. rogeri
Loesch, N. schncidi Loesch, etc., but in ^Y. schlosseri, after the
ventral lobe having been present, it disappears in the last few
sutures and is nearly straight or only slightly arched across
the venter (see Loesch, 1914, p. 115, fig. 8).

There is a rather large degree of variation in the shape of
the conch in the species here assigned to Pseiidaganides (Fig.
15). All of the species have nautiliconic conchs but vary in the
character of the ventral shoulders and in the width to height
proportions of the whorl section. In the type species, N. kut-
chensis, the whorl section is subrectangular with subangular
ventral shoulders, an arched venter, and flattened convergent
flanks. Subangular to acutely rounded ventral shoulders are
present in N. argoviensis Loesch, N. aiiimoni Loesch, N. klehels-
bergi Loesch, N. aganiticus Schlotheim, N. inbingensis Loesch,
N. helveticus Loesch, N. brunhubcri Loesch, N. frickensis Jean-
net, and N. snbbknigidaftts d'Orbigny. Well rounded ventral
shoulders and arched venters are found in X. pidcheUus Jeannet,
N. schneidi Loesch, A'. st)'anibergriisis ()i)pel, N. franconicus
Oppel and N. oppeli Zittel. Some species have slightly concave
venters, as in N. snbbiangulatus d'Orbigm- and X. rogeri Loesch.
In N. ammoni an:l N. schlosseri the venter becomes slightly con-
cave on the living chamber, but is arched on the phragmocone.
The width and height measurements of the whorl section in N.
kutchensis are approximately equal. This same ratio is present
in many other species of Pseiidaganides (e.g. N. pulchellus Jean-
net, N. rogeri Loesch, N. bninhnberi Loesch, .V. schneidi Loesch,
N. ammoni Loesch, etc.). More compressed species include N.

Kl'.M.AIEl. : I'OST-TKIASSIC XAUTILQID GENERA

39;

subbiaitgiilatns d'Orbigny, N. frickensis Jeannet, N. stramher-
(joisis Oppel. etc. Xautilus kochi has a highly compressed conch
Avith couvei-gent whorl sides and a rather narrow rounded venter.
It appeai-s to be merely an extreme form of Pseudaganides antl
does not Avarrant separation as a distinct genus. Hercoglos-
soccras established by Spath for N. kochi is here considered a
■synonym of Pseudaganides. Whereas the degree of convergence
of the whorl sides is greatest in N. kochi the degree of compres-
sion of the conch is found in several other species.

Loesch (1914) has described several species of Upper Juras-
sic nautiloids which are very similar in conch shape and suture
to typical HercogJossa of the late Cretaceous and early Tertiary.
That is, these species have well rounded A^entral shoulders and
arched venters. Spath (1927a) included them tentatively in
Hercoglossa l)ut also expressed the opinion that they "probably
foi'in an independent development." These Upper Jurassic
• ' h(n-coglossids " include among others: N. franconicus Oppel,
r\'. schneldi Loesch, N. rogcri Loesch, .V. hrnnhuheri Loesch, I\ .
roemeri Loesch, X. schwertschlageri Loesch, N. schlosscri Loesch,
N. stromeri Loesch, N. oppeli Zittel, and N. stramhergensis Op-
])(']. In all of these species, however, there is a shallow ventral
lobe which is not fouihl in the Upper Cretaceous-Tertiary Herco-
glossa. The great similarity of these two groups is thought to be
a case of homeomor|)hy. The true Hercoglossa develop from
Cretaceous CUmojiiia whereas the Jurassic "aganitici" are de-
rived from Liassic Cenoceras.

Pseudaganides is Avidely distributcnl in Jurassic strata through-
out Europe, and in Pakistan. Of a total of 37 species listed
below, 1 is from the Lias, 9 are from ]\Iiddle Jurassic strata, and
28 are known from Upper Jurassic formations.

DISTBIBUTIOX OF SPECIEK OF THE Gr:xr,S PSEUDAGANIDES

Species

Stratigi;ii)lii(' Geographic
Distribution Distriljutioii

P. aganiticus (Schlotheini ) IS'20

P. aniDioiii (Loesch) 1914:

P. argoviensis (Loesch) 191:^

Jurassic France

(Oxfordian)

Jurassic Europe

(Upper)

Jurassic Switzerlan

(Oxfordian

396

ni'LLETIX: ilUSEUM OF COMPARATIVE ZOOLOGY

Species

Stratigraphic Geographic
Distribution Distribution

P. bodeni (Loesch) 1914

P. bnnihvberi (Loesch; 1914

r. (•/«l/.s/^v (d'Orbigny) 184:^

r. crass itiiiiua tug (Crick) 1898

P.drepanensi^ (Tagliarinij 1901

P.duilii (Genimellaroj 188ti

P. franconicus (Oppel) 1865

P. frickensis (Jeannet) 1951

P. girardotl (de Loriol) 1903

P. glaber (Foord and Crick ) 1890

P. graves ianu,s (d'Orbigny) LS4;)
P.helveiicus (Loesch) 1914

P.klebelsbergi (Loesch) 1914

P. kocht (Prinz) 19u0

F.krenTceli (Jeannet) 1951

P. kutchensis (Waagen) 1873

P. ledonicus (de Loriol) 1903

P.oppeli (Zittel) 1868

P. pellereiisi^ (Yialli) 1937

P. portlandicus (Foord and Crick) 1890

P. pulchellus (Jeannet) 1951

Jurassic
(Oxfordian )
Jurassic
(Upper)
Jurassic
(Bajocian)
J urassic
(Bajocian)
Jurassic
(Middle)
Jurassic
(Middle)
Jurassic
(Tithouiau )
Jurassic
(Callovian)
Jurassic
(Oxfordian)
Jiirassic

.1 urassic

Jurassic

(Upper)

Jurassic

(Tithonian)

Jurassic

(Lias)

J urassic

(Upper)

Jurassic

(Upper)

Jurassic

(Oxfordian )

Jurassic

(Upper)

Jurassic

(:\nddle)

Jurassic

(Upper)

Jurassic

(Oxfordian)

France

Europe

France

England

Sicily

Europe

Switzerland

Smtzerland

England,
France
France
Switzerland

Czecho-
slovakia
Austria,
Hungary
Switzerland

Pakistan

France

Europe

North Italy

England

Switzerland

KUMMEL : POST-TRIASSIC XAUTILOID GENERA

397

Species

Stratigiaphir (Jeograpliic
Distribution Distribution

/'. roemeri (Loeseh) 1914

P. rogcri (Loeseh) 1914

P. r oyer I (de Loriolj 1872

P.schlosfteri (Loeseh) 1914

r. schlothcimi (Loeseh; 1914

P. schneiili (Loeseh) 1914

P. schutrtsclilugeri (Loeseh) 19lL'

P. strdrnbergpiisis (Oppel) 186.5

P. sironwri (Loeseh) 1914

P. suhbwnr/tilatim (d'Orbigny) 1850

P. nubshiuntus (d'Orbigny) ]8o0

P. tubingnusi.s (Loeseh) 1914

P.turl-criii (Gemmellaro) 188()

Jurassic East Europe

(Upper)

Jurassic East Europe

(Upper)

Jurassic France

(Upper)

Jurassic East Europe

(Tithouiau'i

Jurassic Lorraine

(Middle)

Jurassic Europe

(Tithonian)
Jurassic Europe

(Upper)

Jurassic Czecho-

( Tithonian) Slovakia
Jurassic Czeeho-

( Tithonian; Slovakia
Jurassic France

(Bathonian)
Jurassic England

(Middle)

Jurassic Germany

(Upper)

Jurassic Sicily

(Middle)

Geuus PSECDOXAI I'lLi ts Meek, 1876

P/ieadunaaUiu.s Meek, 1876, p. 491.
Paeudonautilus Spath, 1927a, pp. 20, 22, 26.
Pseudonautilus Miller aud Collinson, 1953, pp. 293-295.

T\ipc species. Nautilus geiiiitzi Oppel, in Zittel, 1868, p. 45, pi. 2, tigs. 17
(Plate 12, figures 1, 2 and Text-Figure 16 of this report).

This genus is well characterized by the extreme " goniatitic ' '
development of its suture. Only four species are assignable to the
genus and the following diagnosis has been compiled from speci-
mens in the M.C.Z. and the description and illustrations of
these species.

Conch involute, compressed, and smooth. Whorl sides flat-
tened, venter low and broadly arched. The ventral and umbilical
shoulders are rounded. The whorl section is higher than wide.

398 BULLETIN: MUSEUM OP^ COMPARATIVE ZOOLOGY

The umbilicus is very small and deep. The suture has a deep
narrow ventral lobe, a large V-shaped lateral lobe followed by
a rounded saddle with a small second lateral lobe just above
the umbilical seam. There is a deep pointed dorsal lobe. The
siphuncle is in a subventral position.

The most diagnostic character of Pseudoyiautilus is the sinuous
suture with the deep V-shaped ventral lobe. The four species
assignable to Pseudonautilus are : P. geinitzi Oppel, the type
species; N. aturioides Pictet (1867) ; N. dumasi Pictet (1867) ;
and N. malhosi Pictet (1867). Loesch (1914, p. 130-131) dis-
cussed in some detail various aspects of N. gemitzi and N. aturi-
oides. Nautilus geinitzi is from the Upper Jurassic (Tithonian)
of Moravia, Silesia, and France, and has also been recorded from
the Berriasian of Tunisia by Arnould-Saget (1953). Nautilus
aturioides is from the zone of Hoplites hoissieri, lowest Valan-
ginian (Loesch, 1914, p. 131). Nautilus malhosi Pictet has more
angular ventral shoulders than in the type species and likewise
the ventral and lateral lobes are rounded rather than pointed.
In respect to the suture, N . malhosi and N. dumasi are more or
less intermediate forms between typical Pseudaganides and N.
geinitzi, the type species of Pseudonautilus. The assignment of
these two species to Pseudonautilus is open to question. Nautilus
malhosi has been recorded from Lower Cretaceous formations in
central Europe (Pictet, 1867), Algeria (Pomel, 1889), and from
the Crimea (Karakasch, 1907).

One of the most remarkable cases of homeomorphy among
post-Carboniferous nautiloids is the similarity of Nautilus (Aga-
nides) hitauniensis Ilaniel from the Permian Bitauni beds
(Leonardian) of Timor and Pseudonautilus gei7iitzi Oppel (Fig.
16.) Miller and Collinson (1953) have recently discussed this
problem and erected the genus Pcrmoceras and the family
Permoceratidae for Haniel's si)ecies. They likewise recognized
the homeomorphous relationship of these two nautiloids. Among
Triassic nautiloids sinuous sutures are found in the Clydonautili-
dae, Gonionautilidae Siberionautilidae, and in the genus Clyme-
nonautilus of the Syringonautilidae. Sinuous sutures are, how-
ever, not common in Permian nautiloids. Among the forms
comprising the Grypoeeratidae, Grypoceras (Plummeroceras)
plummeri (Kunimel, 1953c) — a compressed evolute, quadrate
whorled form — has a suture with a deep ventral lobe and a broad

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

399

deep lateral lobe. The general shape of the conch except for the
degree of involution is rather similar in G. (Plummeroceras) and
Pennoceras. Within the complex of Permian nautiloids now
known it seems more logical to consider Permoceras an offshoot
of the Domatoceras-Grypoceras evolving stock and thus a mem-
ber of the Grypoceratidae.

Fig. 16. Diagrammatic representation of sutures and cross sections of
Permoceras hitauniense (Haniel) A, B, and Pseudonautilus geinitzi (Oppel)
C, B, all X 1. From Miller, and Collinson, 1953, fig. 1.

Subfamily PARACENOCERATINAE Spath, 1927

When Spath established this taxonomic unit (originally as a
family) he included in it Paracenoceras Spath, Tithonoceras
RetoAvski, ^omalinautilus Spath, Aulaconaiotilus Spath, Carino-
nautilus Spengler and with question Heminaiitilus Spath. Ex-

400 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

amination oi" topotypes of Heminautilus etheringtoni Durham
and the descriptions and illustrations of H. lalUerianus d'Or-
bignj^ (especially in Douville, 1916, pi. 17) lead me to conclude
that these species belong more properly in the Cymatoceratinae.
They have the characteristic ribbing", even though in some speci-
mens the ribbing is very weak or absent, and a hercoglossid suture,
more sinuous than the suture in most species of the Paraceno-
ceratinae. However, on the basis of the ornament alone this
genus should not be left in the Paracenoceratinae.

Carinonautilus is a monotypic form known only from the
Upper Cretaceous of south India. The paracenoceratids are
primarily a Jurassic radiation from the Cenoceras complex and
it is difficult to visualize a derivation of Carinonautilus from
this source. It is here allied to the Nautilinae and considered a
derivative of Eutrephoceras which is the long-persisting, slowly-
evolving stock ranging from the Upper Jurassic to the middle
Tertiary. Of the remaining genera, only Paracenoceras is wide-
spread and fairly common as far as Mesozoic nautiloids go. It
has by far the largest numlier of species ; of a total of 42 species
in this family 35 belong in Farace^ioceras. Tithonoceras is mono-
typic, Somalinautilus and Anlaconautilus have three or less
species each.

Whereas Spath did not diagnose this family at length, he did
state that it was "characterized by differentiation of the peri-
phery', generally associated with increase of the sinuosity of the
suture line, and a ventral lobe" (Spath, 1927a, p. 25). Para-
cenoceras has a broad, generally sulcate periphery. Tithono-
ceras has a broad flattened periphery with a median furrow
and the ventrolateral area projected into a prominent blunt keel
aligned by furrows. Somalinautilus has an arched venter with
angular ventral shoulders. Aulaconautilus has longitudinal ribs
on the peripheral area.

Both Paracenoceras and Somalinautilus are interpreted as off-
shoots from the Cenoceras complex. Both of these genera first
appear in the Middle Jurassic and it seems likely that they stem
from separate sources within Cenoceras. Aulaconautilus and
Tithonoceras are more specialized developments stemming from
Paracenoceras. The subfamily is largely confined to the eastern
hemisphere, being known from England, Europe, east and north
Africa, Arabia, and Pakistan. The only occurence of a member

KUMMEL: POST-TKIASSIC NAUTILOID GENERA

401

402 BULLETIN : :\IUSEUM OF COMPARATIVE ZOOLOGY

of this subfamily in the western hemisphere is Parace7ioceras in
Cuba. As is so characteristic of post-Triassic nautiloids the main
stock of a subfamily is generallj^ a highly variable form with
many described species and a wide geographic distribution. This
is well illustrated in the distribution of species of this subfamilv
(Fig. 17).

Genus PakaCENOCERAS Spath, 1927

Paracenoceras Spath, 1927a, pp. 22, 25.
Paraoenoeeras Si^ath, 1935b, pp. 224-225.
Paracenoceras Scott, 1943, pi). 88, 89.
Nautilus {Paracenoceras) Jeannet, 1951, pp. 10-17.
Paracenoceras Miller and Collinson, 1952, pp. 626-628.

Type species. Nautilus hexagonus J. de C. Sowerby, 1826, pi. 529, fig. 2
(illustrated here on Plate 8, fig. 1, and Plate 13, figs. 1, 2; Text Fig.
18). Type by original designation.

Paracenoceras is the only genus of the Paracenoceratinae that
is nearly world wide in distribution and includes a relatively
large number of species. Of the remaining genera of this sub-
family one is monotypic and the remaining two genera contain
no more than three species apiece. The type species — Nautilus
hexagonus J. de C. SoAverby — was previously only known
through the rather unsatisfactory description and woodcut illus-
tration (J. de C. Sowerby, 1826, pi. 529, fig. 2). Sowerby 's type
is illustrated on Plate 13, figures 1, 2, and Plate 8, figure 1 and
Text Figure 18.

The genus can be diagnosed as follows : Couch robust, involute,
smooth. Whorl section subtrapezoidal in outline. Whorl sides
flattened, convergent towards a broad venter that may be flat-
tened or truncate to distinctly sulcate. Ventral shoulders well
rounded. Umbilicus generally small, umbilical shoulders
rounded, umbilical walls steep. The shell is smooth except for
sinuous growth lines which form a deep sinus on the broad
venter. The suture is only slightly sinuous but with a distinct
ventral and lateral lobe and a small saddle at the umbilical
shoulder and wall. The position of the siphuncle is quite vari-
able but never in an extreme ventral or dorsal position. The
polished median section of Sowerby 's tj'pe specimen shows that
the siphuncle is beaded and with short funnels (Plate 8, fig. 1).
Similar beaded siphuncles have been also noted by Spath (1927a,

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 4()o

p. 23) in species of '' Bisiphytes," " Digonioceras/' and Cyma-
foceras.

Approximately 35 species of Paracenoceras have been recorded
most of which are from Upper Jurassic strata. Spath (1935b, p.
224) has described Paracenoceras prohexagonum from strata o.

Fig. 18. Paracenoceras hexagonuia \,ttowerby). Diagrammatic cross sec-
tion of holotype represented by figures 1, 2, Plate 13, and figure 1, Plate
8, X 1.

Bathonian? age. This is the oldest species known, and the
youno'cst species of Paracenoceras appears to be P. rJiodani
(Roux) from the Albian of Switzerland.

Sowerby's type of NaKfilus liexagonns, refignred here on Plate
13, figures 1, 2 and Text Figure 18, is a large phragmocone meas-
uring 95 mm. in diameter, 54 mm. in height of last whorl, and
67 mm. in width of the last whorl. The whorl sides are only
slightlv convex and converge towards a broad sulcate venter.

404 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

F'igure 19

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 405

The umbilical and ventral shoulders are well rounded. The
umbilicus is fairly large, deep, and funnel shaped. The umbili-
cal Avails are very steep. The large siphunele is subcentral in
position.

There is much A'ariation in shape of the whorl section among
the various species of Paracenoceras. The cross sections of 12
species are illustrated in Figure 19 to show the degree of varia-
tion. Sulcation of the venter of the adoral volution as on the
type species, is not present on several of the species which have
merely truncated venters or ventral areas slightly arched. Para-
cenoceras maroccnse Miller and CoUinson (1952) has a broad
sulcate venter at earh^ maturity but the venter becomes arched
with increasing size (Text Figure 19 G-I). The whorl sides also
show some variation from being slightly concave as in P. liexa-
(jonuni and P. marocense, to flattened or slightly convex in most
of the remaining species. Variations in suture are mostly a mat-
ter of degree. Sutures of several species of Paracenoceras are
illustrated in Figure 20 for comparison.

Paracenoceras costatum Scott (1943, p. 88, pi. 24, figs. 1, 3),
with prominent sigmoidal costae on the whorl sides, belongs in
the genus Procymatoceras. Scott (1943, p. 89) compared "Para-
cenoceras" costatum with P. juniarense (Waagen, 1873, p. 21, pi.
4, figs, la, b) which has sigmoidal ribs only on the whorl sides,
that terminate in large tubercles at the ventrolateral shoulder.
However, Spath (1927a, p. 32) questions the accuracy of
Fig. 19. Cross sections of the coneh of A, Paracenoceras calloviensis
(Oppel), from Waagen 1873, pi. 3, fig. 2b, diameter 65 mm.; B. P. suhhexa-
gonum Jeannet 1951, fig. 19, height i2 mm.; C, P. roUieri (Loesch) from
Jeannet 1951, fig. 29, height 27 mm.; D, Aulaeonautilus Mcarinntus Jeannet
1951, fig. 32, diameter 41 mm.; E, Paracenoceras icandaen,sis (Waagen)
1873, pi. 4, fig. 3b, diameter 85 mm.; F, Aulaeonautilus sexcarinatus
(Pietet) 1867, pi. 10, fig. lb, diameter 107 mm., the genotype of Aulaeo-
nautilus; G, H, I, Paracenoceras marocense Miller and Collinson, 1952,
fig. 1 {G, height 28 mm.; H, height 48 mm.; I, height, 72 mm.); J, P.
prohexagonum Spath 1935, fig. 46, height 44 mm.; K, P. hexagonum
(Sowerby) from Spath 1935, fig. 4d, height 56 nun., the genotype of Para-
cenoceras; L, P. dilatatus Jeannet 1951, fig. 8, height 140 mm.; J/, P.
Icumagunensi^ (Waagen) 1873, pi. 3, fig. lb, diameter 72 mm.; N, P.
hersnachcnsis Jeannet 1951. fig. 14, height 100 mm.; 0, P. enniunum
(Dacque), from Scott 1940, pi. 24, fig. 2, height 70 mm.; P, P. arduennensis
(Loesch, 1914) from Jeannet 1951, fig. 26, height 19 mm.; Q, Aulaeo-
nautilus picteti (Oppel), from Zittel 1868, pi. 3, fig. 36, diameter 95 unn.

406

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Waagen's illustration and points out that the specimens are
badly weathered. Spath quite rightly maintains that the affini-
ties of P. jumarense are very doubtful and the species can be
retained in Paracenoceras with question.

A list of the species assigned to Paracenoceras with their gen-
eralized geologic and geographic data is given below. i\Iost of
the species are from the Jurassic Tethyian geosynclinal area and
adjoining epicontinental seas of Europe. Several occurrences

lig. 20. Diagrammatic representation oi suiures oi A, Famcenoceras
dilatatus Jeannet 1951, fig. 8; B, P. hexagnnum (Sowerby), from Spath
1935, fig. 4e, the genotype of Paracenoceras; C, P. acMini Jeannet 1951,
fig. -25; D, P. marocense Miller and Collinson 1952, fig. lA ; E, P. rollieri
(Loesch, 1914) from Jeaunet 1951, fig. 29; F, P. hersnachoisis Jeannet
1951, fig. 14; G, Eeminaiitilus stantoni Scott 1940, fig. 179; H, Paraceno
ceras prohexagonum Spath 1935, fig. 4e; /, Heminautilus etheringtoni Dur-
ham 1946, fig. 3C; J, Heminautilus lallierianus (d'Orbigny) from Durham.
1946, fig. 3A. All figures greatly reduced.

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

407

in eastern Africa, Madagascar, Arabia, and in Pakistan form an
eastern extension of the European faunal distribution. The
onh* occurrence known in the Western Hemisphere is in Cuba
(Sanchez Roig, 1951). As noted above, most of the species of
Paracenoceras are Upper Jurassic in age.

DISTRIBUTION OF SPECIES OF THE GENUS PARACENOCERAS

Species

Stratigraphic Geographic
Distribution Distribution

P. acklini ( Jeannet) 1951

P . arduennensis (Loesch) 191i

P. bnintrutanae (Kuhn) 1936

P. calloviensis (Oppel) 1858

P. dilatatus (Jeannet) 1951

P. divesianus (Kuhn) 1936

P. dorsatus (Roemer) 1836

P. dorsoexcavatum (Parona and Bonarelli) 1897

P. ennianwm (Dacque) 1905

P. giganteus (d'Orbigny) 1825

P . gramulosum (d'Orbigny) 1843

P. herznachensis Jeannet, 1951

P. hexagonoides Spath, 1927

P. heragoninn (J. de Sowerby) 1826

P. jeannet i Sanchez Eoig, 1951

PJ jumarense (Waagen) 1873

Jurassic

SAvitzerland

(Callovian)

Jurassic

France

(Oxfordian)

Jurassic

Germany

(Upper)

Jurassic

Europe

(Callovian)

Jurassic

Switzerland

(Callovian)

Jurassic

France

(Upper)

Jurassic

Germany

(Upper)

Jurassic

France

(Callovian)

Jurassic

Somaliland

(Upper)

Jurassic

France

(Upper)

Jurassic

France

(Upper)

Jurassic

Switzerland

(Callovian)

Jurassic

Pakistan

(Tithonian)

Jurassic

Europe

(Upper)

Jurassic

Cuba

(Oxfordian)

Jurassic

Pakistan

(Bathonian)

408

BULLETIN : MUSEUM OF COMPAEATIVE ZOOLOGY

Species

Stiatigrapliic (geographic
Distribution Distribution

P. Icumagunense (Waagen) 1873

P. hitifrons (Zwierzycki) 1914

P. marncense Miller and Collinson, 1952

P. mjatsclil:uwa7iits (Loescb) 1914

P. miyreani (d'Orbigny) 1842

P. mullerriedi Sanchez Eoig, 1951

P. prohexagonum Spath, 1935

P.rhodani (Eoux) 1848

P. roUieri (Loesch) 1914

P.sattleri (Krenkel) 1910

P.schattenhcrfii (Kuhn) 1936

P. schusteri (Loesch) 1912

P. siruJus (Gemmellaro) 1868

P. sinuosn^ (Roemer) 1836

P. staffelbergensis (Kuhn) 1936

P. suhhexagorms ( Jeannet) 1951

P.voJgcnsis (Nikitin) 1888
P.u-andaense (Waagen) 1873

P. wilviae (Jeannet) 1951

Jurassic Pakistan

(Callovian)

Jurassic Tanganyika

(Upper)

Jurassic Morocco

(Upper) ?

Jui'assic Russia

(Upper)

Jurassic France

(Kimmeridgian)
Jurassic Cuba

(Oxfordian)
Jurassic British

(Bathonian?) Somaliland
Cretaceous Switzerland
(Albian)

Jurassic Switzerland

(Oxfordian)

Jurassic Tanganyika

(Kimmeridgian)
Jurassic Germany

(Upper)

Jurassic France

(Kimmeridgian)
Jurassic Sicily

(Upper)
Jurassic
(Upper)
Jurassic
(Upper)
Jurassic
(Callovian)
Jurassic Russia

Jurassic Pakistan

(Upper)

Jurassic Switzerland

(Callovian)

Germany

Germanj-

Switzerland

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 409

Genus AULACONAUTILUS Spath, 1927

Aulaconautilus Spath, 1927a, pp. 22, 25.
Nautilus {Aulaconautilus) Jeannet, 1951, p. 17.

Type species. Xaidilus sexcaririatus Pictet, 1867, PI. 10, figs, la-c (repro-
duced here as Plate 14, figs. 1, 2). Type by original designation.

The following diagnosis is made from study of the descrip-
tions and illustrations of the three species that can be assigned
to this genus. No specimens were available for study. Conch
involute, generally compressed, with greatest width of the
whorls just below the umbilical shoulders. AYhorl sides smooth,
convex and converging. Ventral shoulders rounded, venter
broad, low, and with longitudinal ribs. Umbilicus very small,
umbilical shoulders rounded. Suture sinuous with a shallow
ventral lobe, a broad deep lateral lobe with a saddle at the
umbilical shoulder. Position of the siphuncle not known.

Only three species can be placed in Aulaconautilus; these are
A. sexcarinatus (Pictet) the type species, A. picteti (Oppel),
and A. hicarinatus (Jeannet). There is variability in the conch
shape, suture, and ornamentation. The type species, A. sexcari-
natus, has eight longitudinal ribs on the venter (Plate 14, figs.
1, 2 and Text Figure 19i^), divided by shallow furrows. The
median furrow is the most prominent. The ribs at the ventro-
lateral edge are the weakest of the ribs on the venter. Aulaco-
nautilus picteti has only four longitudinal ribs all of equal
prominence ; the median furrow is the widest and deepest of the
furrows between these ribs. This species likewise is much more
inflated than the type species (Text Fig. 19^). Aulaconautilus
hicarinatus has a more depressed, subquadratic whorl section
with a broad flattened venter. There are four widely spaced
longitudinal ribs on the venter (Text Fig. 19D). The orna-
mentational pattern of longitudinal ribs on the venter, which
characterizes Aulaconautilus, is unique for post-Triassic nautil-
oids. A homeomorphous development is the Permian and Trias-
sic genus Aulametacoceras Miller and Unklesbay. The type of
this genus is Middle Permian in age (from Arizona). The only
other species assigned to it is Nautilus rectangularis Hauer of
Upper Triassic age. Aulametacoceras is considered a radiation
from the stable evolving Metacoceras complex (Kummel, 1953c).

The sutures of species of this genus, especially those of the

4:10 BULLETIN : MUSEUiM OF COMPARATIVE ZOOLOGY

type species, are very similar in plan and degree of sinuosity
to those of Pseudaganides. The other groups of the Paraceno-
ceratinae have sutures with A-ery shallow lol)es and saddles.

The three known species of Aidaconaniihis come from Upper
Jurassic strata of Europe.

Genus TiTHONOCERAS Retowski, 1894

TithoHOceras Ectowski, 189-4, p. 223.
Tithonoceras Crick, 1908, p. 14.
Tithanoceras Spath, 1927a, pp. 20, 22, 25.

Type species. Tithonoceras zitteli Retowski, 1894, p. 223, pi. 13, figs. 2a, b, c
(refigured here on Plate 15, figs. 1, 2).

This is another of the numerous mouotypic Mesozoic nautiloid
genera. The type species is from Tithonian (Upper Jurassic)
strata of the Crimea. The conch is evolute, compressed, and
smooth. The whorls are subrectangular in outline, being higher
than wide. The venter is broad, flattened, and with a shallow
median furrow. The ventral shoidders are inflated and form a
smooth keel-like ridge that is aligned on the venter and the
whorl sides by a furrow. The ventral furrow aligning the
ventrolateral keel is rather narrow but very distinct. The fur-
roAV on the whorl sides adjoining the ventrolateral keel is con-
cave and broad occupying nearly half of the whorl side. The
dorsal half of the whorl side is convex passing onto a broadly
rounded umbilical shoulder. The umbilicus is large, measuring
one-eighth the diameter of the conch. The shell is smooth except
for sinuous growth lines which are deeply concave on the venter.
The suture is very sinuous with ventral and lateral lobes. The
position of the siphunele is not known.

Proper evaluation of monot.ypic foi'ms is not entirely satis-
factory. Little can be said of TitJionoccras exce])t that it appears
to be a specialized offshoot of Paracenuceras. The general conch
form is unique among Carboniferous to Recent nautiloids.
There are no homeomorphous types of other ages known to me.
Tn its derivation from Paracenoccras there was merely greater
compression of the conch and further elaboration of the ventral
area.

The type and only known species of TiiJionoceras came from
Tithonian strata (Upper Jurassic) of the Crimea, Russia.

KUMMEL : POST-TRIASSIC NAUTILOID GEXERA 411

Genus SOMALIXAUTILUS Spath, 1927

Tiipe species. Xautilus mitiquus Dacque, 1910 i=X. hisulcutus Dacque, 1905,
p. 144, pi. 16, figs. 3a, b). Type by original designation. Type figure
reproduced here on Plate 14, figs. 3, 4.

The Paracenoceratinae are derived from Liassie cenoceratids
and are characterized by modified venters. The earliest species
of this subfamily are included in the genus Somalinautilns. At
this date only three species of Somalinautilns have been re-
corded. The type species is from Upper Jurassic (Kimmerid-
iiian) strata of Somaliland. The other species definitely assigned
to this genus are Xantilus fuscus Crick (1898, p. 122) and N.
clausus in Foord and Crick (1890b, p. 284) Jion d'Orbigny (1842,
p. 158, pi. 33).

From the description and illustration of the type species and
study of several specimens of Nautilus fuscus in the British
]\[uscum (Natural History), the following diagnosis is made.
Conch involute, depressed, rapidly expanding, whorls wider
than high. Venter slightly arched with subangular ventral
shoulders, with a distinct furrow on venter adjoining the ventral
.■-lioulder. The whorl sides are concave adjacent to the ventral
shoulders and convex towards the umbilicus. Umbilical should-
ers are sharply rounded and the umbilical wall very steep.
Umbilicus measures less than one-quarter the diameter of the
conch. Surface of the conch with sinuous growth lines and may
or may not have longitudinal striae. Suture forms shallow
ventral lolie. generally an angular saddle at the ventral shoulder
followed by a broad, shallow, lateral lobe, with a low saddle on
the umbilical wall. The position of the siphuncle in the type
species is not known but in Nautilus fuscus it is slightly below
the center.

Each of the generic groups of the Paracenoceratinae is very
distinct and Somalinautilus with its sul^angular ventral shoulders
and low arched venter is readily differentiated from the other
genera. Somalinautilus most probably arose directly out of the
Cenoceras complex and not from Paracenoceras. There are
several quadrate whorled forms of Cenoceras which could have
given rise to those Middle and Upper Jurassic species placed
in Somalinautilus. Species of this genus are known from Eng-
land, France, and Somaliland.

412 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Subfamily CYMATOCERATINAE Spath, 1927

This subfamily now comprises 10 genera of post-Triassic
nautiloids which have variable conch shapes and degree of
sinuosity of the suture but have in common an ornamentational
pattern consisting of ribs. Ornamentation is relatively rare in
post-Triassic nautiloids, excepting those forms within the Cyma-
toceratinae. There are no species with nodose ornamentation
known. Within this subfamily there is displayed a wide range
of adaptive radiation as seen in the extent of variation in
conch shape and suture. The genus Cymatoceras is the basic
adaptive type ; it has an involute, rounded conch with only a
slightly sinuous suture. Paracymatoceras is like Cymatoceras
in conch form but has a more sinuous suture with distinct lobes
and saddles. Syrionaiitilus, Anglonautilus, and Eucymatoceras
are differentiated on modifications of the ribbing plan. Cymato-
nautilus, Heminautiliis, Epicymatoceras (n. gen., p. 439), and
Delfocynmtoceras (n. gen. p. 438) are differentiated by their
modified conch shapes. Heminautilus and Epicymatoceras have
highly compressed conchs and flattened venters. The former is
very involute and the whorl section more convergent in plan.
The latter is more evolute and with a simple, only slightly
sinuous, suture. Deltocy matoceras has a subtriangular whorl
section and CymatonautUvs has an evolute conch with a concave
venter and wiiorl sides.

The only common and geographically-widely -distributed genus
is Cymatoceras. This fact is clearly expressed in the number of
species belonging to each group.

Cymatoceras 64 species

Parari/tytatoceras 6 species

Heminautilus 5 species

Procymatoceras 4 species

Eucymoioceras 3 species

Anglonautilus 3 species

Cymatonautilus 2 species

Syrionautilus 1 species

Epicymatoceras 1 species

Deltocymatoceras 1 species

Cymatoceras has a truly world-wide distribution. The other
genera are largely confined to Europe in the epicontinental seas

KUMMEL: POST-TRIASSIC NAUTBLOID GENEKA 413

adjoiuiiig the Tethyiaii geosyneline (Text-Figs. 21, 22). Eucy-
matoceras, Anglonautilus, Syrionaiitilus, Cymatonautilus, Epi-
c y mat oc eras, aud Dcltocymatoccras are known only from a belt
extending from England to the Crimea or Caucasus Mountains
in Kussia in sediments of the Jurassic and Cretaceous epicon-
tinental seas of and adjoining the Tethys proper. Procymato-
ccras is known from England, Pakistan, and Ethiopia; Hcmi-
naufilus and Parucymafoccras are known outside the European
Tethyian belt from North and South America, Ethiopia, and
Japan.

The known distribution of fossil nautiloids is so dependent
on factors of preservation and concentration of field collecting
that it is probably dangerous to make any too-sweeping conclu-
sions. However, the number of species does at least reflect the
relative abundance of the various genera. If we think of the
Tethys as an ancient circum-equatorial sea the Cymatoceratinae
are found to be largely confined to the borders of this seaway.
Many genera (e.g. Eucyniatoceras, Syrionnutilus, Cymatona\i-
tilus, Epicymafoceras, and Deltocyniatoceras) are confined en-
tirely to the European-Middle East portions of this circum-
equatorial sea.

Analysis of the distribution of the Cymatoceratinae on a time
basis throws some light on the general problem of their evolu-
tion. Procymatocenis and Cymaionautilus are known only from
Middle and Upper Jurassic strata. The species of these genera
are confined to the marginal seaways of the Tethys from Eng-
land to India. Whereas, to the best of my knowledge, the
species of these two genera have never been found in a single
stratum and locality, their geographic and time ranges do over-
lap.

In the Cretaceous, the genus Cyuiatoceras is widely distributed
throughout the period. Anglonautilus, Eucyniatoceras, and
Heminautilus are confined to the Lower Cretaceous. Anglo-
nautilus and Eiicymatoceras have approximately the same
geographic range from England to the Black Sea. Whereas
Heminautilus also occurs in this European-Mediterranean-belt
it is likewise known from Ethiopia, Arkansas (U.S.A.), and
Colombia. The Upper Cretaceous genera Epicymatoceras and
Deltocymatoceras occur together in the type Maestrichtian area
of Belgium, and Syrionautilus is confined to the Middle East.

414

BULIjETIN : MUSEUM OF COMPARATIVE ZOOLOGY

KlTMiMEL: POST-TRIASSIC NAUTILOID GENERA

415

5 i*

ai O

o ^

-5 a

13 =w
C

a o

be ;:
O r
<^ 3

«4-l

M .5

416 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

The wide range in variation of conch shape and suture within
the Cymatoceratinae produced many homeomorphs to other
members of the Nautilidae. The presence of the characteristic
ribbing- makes differentiation easy. The conch shape and suture
of Deltocymatoccras are closely similar to that of Angulithes of
the Hercoglossinae. The conch shape and suture of many species
of Cymatoceras and Eucymatoceras are nearly the same as that
found in many species of Eutrcphoceras and Cimomia. The
suture of Paracymatoceras is identical in plan with that found
in Hercoglossa.

This great variability in conch shape and suture leads me to
believe that the Cymatoceratinae form a phyletic group which
has as its main genetic base the ribbing character. Once estab-
lished, this genetic line ■ had its own adaptive radiation which
is expressed in the many conch shapes, rib patterns, and sutures
as it is in any other group of Mesozoic nautiloids. This subfamily
is the dominant group of Cretaceous nautiloids. Ribbing, on a
pattern somewhat like that in the Cymatoceratinae, occurred
only once before, and that was in a single species of Upper
Triassic nautiloid-Proclydonautilvs spirolohus (Dittmar). The
ribbing in this species is very fine and like that in Syrionautilus,
whereas in most other genera and species of the Cymatoceratinae
the ribbing is coarser. It is important to note that Proclydonau-
tilus spirolohus belongs to a family that characteristically has
smooth eonchs except for a few forms with nodes developed
during some stage of their ontogeny {C osmonautilus Hyatt and
Smith, 1905, and CaUaionautilus Kieslinger, 1924). The ribbing
patterns found in species of Pleuronanttlus are quite different
from those of the Cymatoceratinae.

Data are insufficient to decipher precisely the evolutionary
lines within the Cymatoceratinae. The evidence available
seems to point to Cymatoceras as the parent stock which gave
rise to most of the remaining adaptive types. The origin of the
subfamily must be directly or indirectly among the cenoceratid
complex of the Lower and Middle Jurassic. Even though Cyma-
tonautilus and Procymatoceras are the oldest forms known, they
are both rather specialized and probably not ancestral to Cyma-
toceras, which as far as we now know first appears in the Cre-
taceous.

Neocymatoceras Kobayashi (1954) was established for a single

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 417

specimen from the Oligoeene of Kjoisliu, Japan. Neocymatoceras
tsukushiense, the type species, is a typical eymatoceratid in conch
shape and rib pattern but has a strongly projected suture which
forms a rounded ventral saddle, a broad sweeping lobe occupying
nearly the whole whorl side, and with what appears to be the
beginning of a saddle on the umbilical area. Cymatoceratids
with sinuous sutures characterize Paracymatoceras. However,
in that genus the saddles and lobes are well defined ; that is,
there is always a well-defined saddle on or near the umbilical
shoulder. It is also true though that Paracymatoceras is grada-
tional with Cyynatoceras in regard to sutures. Those cymato-
ceratids with more sinuous sutures show the sutures to be more
strongly projected on the ventral area and in this way give
the sinuosity ; the lateral saddle at the umbilical area is usually
very weak. Cymatoceras pseudonegama Spengler (Fig. 24 C)
and C. semilobatus Spengler (Fig. 24 D) have such strongly
projected sutures on the umbilical area but not quite to the
extent found in the Japanese Oligoeene species. Similar strongly
projected sutures in very small juvenile specimens of Cyma-
toceras have been described and figured by Kummel (1953a)
from Lower Cretaceous formations of Texas. However, the
suture of Neocymatoceras tsiikushiense appears to me to lie
within the range of variation of the suture of Cymatoceras, and
the lack of any lateral saddle in the umbilical area disallows any
affinity with Paracymatoceras. The distinctness of the suture
in Neocymatoceras tsukushiense is only in the marked projection
on the peripheral area, a modification present in several Cre-
taceous species. This interpretation places the Japanese species
in the genus Cymatoceras, and Neocymatoceras is considered a
synonym of Cymatoceras. It should be emphasized that C.
tsukushiense is the only Tertiary Cymatoceratinae known.

Genus CYMATOCERAS Hyatt, 1884

Cymatoceras Hyatt, 1884, p. 301.
Cymatoceras Hyatt, 1894, p. 553.
Nautilus (Cyynatoceras) Spengler, 1910, p. 127.
Cymatoceras Spath, 1927a, pp. 21, 25.
Cymatoceras Miller and Harris, 1945, p. 2.
Cymatoceras Durham, 1946, p. 429.
Neocymatoceras Kobayashi, 1954a, pp. 18-20.

418 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Type species. Nautilus pseudoelegans d'Orbiguy, 1840, p. 70, pi. 8 (by
original designation).
Species of Cymatoceras are the most common and widely dis-
tributed nautiloids ol' the Cretaceous. The characteristic pat-
tern of ribbing and the slightly sinuous septa make identification
of the genus easy. A cast of the type specimen (which is in the
Museum of Natural History, Paris) is illustrated on Plate 16,
figures 1, 2. This same specimen has previously been illustrated
by Foord and Crick (1890a, p. 551, figs. 3a, b), and this same
figure was reproduced in Foord 's second Catalogue of the
Fossil Cephalopoda in the British Museum (Foord, 1891, p.
255, fig. 59 a, b). Foord and Crick (1890a, p. 552) quote a letter
from Dr. Paul Fischer expressing doubt as to the identity' of
d'Orbigny's type specimen. Fischer states: "The type of Nau-
iilus yseudoelegans is difficult to recognize. According to the
dimensions given in the original diagnosis (diameter 240 milli-
metres, thickness 160 mm. Paleont. Franc, p. 60) our specimens,
No. 4834D, which were sent to you are typical. [Foord and
Crick, page 552, state that two specimens were sent, both num-
bered 4834 D; the larger one is represented by their figure 3.]
Moreover, they come from the neighborhood of Vandeuvre

Fig. 23. Cross sections of the conch of A, Cynmtoceras clegans (,T
Sowerby) from d'Orbigny, 1840, pi. 19, fig. 2, diameter 190 mm.; B, C.
sliarpei (SchlUter) 1876; pi. 46, fig. 6, diameter 168 mm.; C, C. tourtiae
(Schliiter) 1876, pi. 46, fig. 2, diameter 88 mm.; 1), ('. rvnomunensis
(SchlUter), 1876, pi. 45, fig. 2, diameter 115 mm.; E, Pnicj/iiKilnci'rds siih^
truncatus (Morris and Lycett) 1850, pi. 1, fig. 22, diameter 72 mm., the
genotype of Procyviatocerns ; F, Frocymatoccras baheri (Morris and
Lycett) 1850, pi. 1, fig. la, diameter 82 mm.; G, Faracymafoceras asper
(Oppel) from Zittel, 1868, pi. 3, fig. la, diameter 83 mm., the genotype
of Faraeyiitatoceias ; II, Frocymnioceras.' intumescens (Waagx-n) 1873, ]il.
3, fig. 3, diameter 90 mm.; I, C. radiatus (Sowerby) from d'Orbigny, 1840,
pi. 14, fig. 2, diameter 188 mm.; J, C. neooomiensis (d'Orbigny) 1840, pi.
11, fig. 2, diameter 192 mm.; K, C. deslongchampsianum (d'Orbigny) 1840,
pi. 20, fig. 2, diameter 65 mm.; L. Eucymaiocerns requienianus (d'Orbigny)
1840, pi. 10, fig. 2, diameter 280 mm. {=N. pUcatus Fitton, 1835); M,
Cymatoceras necTcerianus Pictet, 1859, pi. 16, fig. lb, diameter 93 mm. ;
N, Deltocymatoceras rugatus (Fritsch and Schlonbach) 1872, pi. 15, fig. 2,
height 80 mm.; 0, Epicymatoceras vaelsensis Binekhorst, 1861, pi. 5c, fig.
2b, diameter 105 mm.; F, Cymatoceras albensis (d'Orbigny) from Pictet,
1859, pi. 17, fig. lb, diameter 95 mm.

KUMMEL: POST-TRIASSIC NAUTILOID GKNF.RA 419

Figure 23

420 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

(Department de I'Aube), where the species was found.

"But the drawing- given by d'Orbigny is faulty; first its
dimensions do not agree with the description; then the drawing
being reduced to one-third the natural size, the specimen which
it represents should have been at least 360 nun. in diameter;
besides, the ornaments are wanting in the drawing on a great
part of the last whorl. . . .

"Consider then the figure by d'Orbigny as only approximate.
D'Orbigny has restored a great many of his plates, for which,
in my opinion, he is very much to blame. . . .

"In the d'Orbigny collection no specimen is specially marked
as the type. But d'Archiac (Hist, du progres de la geologic, vol.
iv, p. 295) quotes N. pseudoelegans, and he has given to the
Museum a specimen marked iype. This specimen is small (greatest
diameter 150 mm.), and agrees neither with the figure nor with
the dimensions given in d'Orbigny 's original description.

"Perhaps the true type was contained in the collection of the
geologist Clement Mullet, who showed d'Orbigny over the local-
ity where N. pseudoelegans abounds?" In accordance with
the above statement Foord and Crick (1890a, p. 551) in the leg-
end of their figure 3, state that the drawing is from a specimen in
the d'Orbigny Collection in the Museum of Natural History,
Paris. However, Foord (1891, p. 255, fig. 59 a, b) labelled the
same figure as drawn from d'Orbigny 's type specimen.

The type specimen is a much inflated conch, very involute
and with a small umbilicus. The whorl section is wider than
high with a broadly rounded venter. Whorl sides are also
broadly rounded. The test bears prominent ribs that are sig-
moidal on the whorl sides and form a sinus on the venter. The
suture is only slightly sinuous with shallow ventral and lateral
lobes. The siphuncle lies below the center of the septa.

There are approximately 64 species of Cymatoceras recorded
to date. All of these are characterized by the ribbing and the
slightly sinuous septa. The shape of the whorl section is quite
variable in this genus (Fig. 23). The type species, C. pseudoele-
gans, has a depressed, globular conch ; one of the mo.st com-
pressed conchs is that of C. neocomiensis (d'Orbigny) (Fig.
23 J). Between these two extremes there are all gradations of

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 421

conch patterns with well rounded venters. Wtereas most of the
species are very involute, as in the type, there is a wide range
of variation to rather evolute species as C. patens (Kner, 1850)
and C. tskaltsithelensis (Rouchadze, 1931). Some of the species
develop sharply rounded to subangular shoulders. Cymatoceras
tourtiae (Sehliiter) has a subquadrate whorl section with sub-
angular ventral and umbilical shoulders ; the venter is low and
arched (Fig. 236'). The juvenile volutions, however, have broadly
arched venter with well rounded ventral shoulders (Sehliiter,
1876, pi. -±6, fig. 4). Cymatoceras deslongchampsianum (d'Or-
bigny) has an arched venter with rounded ventral shoulders but
subangular umbilical shoulders with a sloping, slightly arched
umbilical wall (Fig. 237i). Nautilus vaelsensis JMnckhorst (1861,
p. 15, and Sehliiter 1876, p. 57) is one of the most aberrant
cymatoeeratids known to me and is made the type of a new
genus (p. 43 9). The conch is very compressed and evolute. The
whorl sides are slightly convex with broadly rounded umbilical
shoulders and angular ventral shoulders; the venter is flattened.
The test bears the typical cymatoceratid ribbing. The characters
of the ventral area, that is, the flattened venter and the angular
ventral shoulders, are also found in Heminautilus. This latter
genus, however, is very involute and the whorl sides are generalh"
slightl}^ inflated near the umbilical shoulders. Cymatoceras
tourtiae also has the flattened venter and angular shoulders, but
the conch is inflated, involute and the whorl section subquadrate.

The position of the siphuncle in the type species (C. pseudo-
elegans) is below the center — that is, closer to the dorsum than
the venter (d'Orbigny, 1840, pis. 8, 9; Foord, 1891, p. 256). It
appears that this condition prevails in a majority of the species
of Cymatoceras (e.g. C. neocomiensis, C. radiatus, C. gahhi, C.
colomhiana, C. elegans, C. kayeanus, C. kossmati, and C. ceno-
manensis). The siphuncle lies above the center, that is closer to
the venter than the dorsum in C. atlas, C. tenuico status, and C.
hunstantonensis. A specimen of C. loricatus (Sehliiter) figured
by Griepenkerl (1889) has a near marginal (ventral) siphuncle.

The suture of C. pseudoelegans is only slightly sinuous with
shallow ventral and lateral lobes. Paracymatoceras differs from
Cymatoceras essentially only in the greater sinuosity of its
suture (Fig. 24). As noted by Miller and Harris (1945, p. 10)
these two genera are more or less gradational in respect to the

422

BULLETIX : MUSEUM OF COMPARATIVE ZOOLOGY

suture. Those specimens with more sinuous sutures, and thus
transitional to Paracymatoceras, include C. colombiana (Fig.

Fig. 24. Diagrammatic representation of sutures of A, Cymatoceras col-
onibiana Durham 1946, fig. 3G ; B, Paracymatocerufi texanum (Shumard)
1860, from Miller and Harris 1945, fig. ?.A; C, Cymatoceras pseudonegama
(Spengler) 1910, pi. 12, fig. 7d; D, C. semilohatiis (Spcngler) 1910, pi.
11, fig. 4b; E, C. hilJi (Shattuok) from Miller and Harris 1945, fig. 2A ;
F, C. kayeanum (Blanford), from Spengler, 1910, pi. 12, fig. 72; G, C.
tskaltsithelensis (Eouchadze) 1931, fig. 6; E, C. kos.'<mati (Spengler)
1910. fig. 7b; I, C. rirgatum (Spengler) 3 910, fig. 7f ; J. C. loeblichi Miller
and Harris 1945, fig. 2C. All figures greatly reduced.

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 423

24A), C. cenomanensis, C. tourtiae, C. patens, C. semilohatus
(Fig. 24 Z>), C. kossmati (Fig. 24 H), and C. virgatum (Fig. 24
/). The type species does not have an annular lobe, which ap-
l^ears to be the case in the majority of species of Cymafoceras.
However annular lobes in the center of the dorsal lobe are
present in ('. I'adiafiis, C. loehlichi, C. colomhiana, and C. tour-
tiae. Most species of Cymafoceras have broad ventral saddles but
some as C virgatum (Spengier) and C. tstxaltsithelensis (Rou-
cliadze) have very pronounced ventral lobes.

Cymatoceras is the most common, and has a larger number of
species than any of the other genera of the Cymatoceratinae. All
members of this subfamily have in common the characteristic
ribbing. The other genera are differentiated on the basis of spe-
cialization of the ribbing, shape of conch, or suture. Frocyjiuifo-
ceras has a depressed, rapidly expanding conch. M^ith vhorl sides
and venter flattened. Eucymatoceras has the inflated rounded
conch as in Cymatoceras pseudoelegans but the ril)s form Y-
shaped patterns on the venter and on the whorl sides. Paracyuw-
toceras is like Cymatoceras excei)t for the greater sinuosity of the
suture which has a well developed lateral lobe and a saddle at the
umbilical suture. Cymatonautilus has a depressed evolute conch
with concave lateral and ventral areas. Syrionautilus has acute,
widely spaced, and nearly straight ribs. Anglonautilus has a
depressed whorl section with large transverse undulations on
the venter. Hemi)wi(tihis has a much compressed conch, with a
flattened venter, and a xevy sinuous suture. Epicymatoceras
has a compressed, evolute conch with subquadrate whorl section.
Deltocyynatoceras has a subtriangular whorl section.

As stated above in the discussion of the subfamily Cymato-
ceratinae, Xeocymatoceras Kol)ayashi (19o4a) is thought to fall
within the range of variation of Cymatoceras and is placed in
.synonymy of that genus.

Cymatoceras is characteristic of the Cretaceous. Steuer (1921)
has described Cymatoceras perstriatus (Steuer) from yevy late
Jurassic and earliest Cretaceous strata of Argentina. Cymato-
ceras tsukushicnse (Kobayashi, 1954a i conies from Oligocene
strata of Japan. These two are the only non-Cretaceous species
of Cymatoceras known to the writer. Cymatoceras is world-
wide in distribution (Figs. 21, 22).

424

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

DISTEIBUTION OF SPECIES OF THE GENUS CYMATOCERAS

Species

Stratigraphic Geographic
Distribution Distribution

C.alhensis (d'Orbigny) 1850

C. altavensis (Pomel) 1889

C. andranofotsyense Colliguon, 1951

C.anguliferous (Schliiter) 1876

C. articulatus (Pulteney) 1813
C. atlas (Whiteaves) 1876

C? averilli (Andersou) 1938

r. bayfieldi (Foord and Crick) 1890

C. hifurcatum (Ooster) 1858

C. broitzemenfiis (Miiller and Wollenunui) 1906

CJcampbelli (Meek) 1861

C. carlottensis (Whiteaves) 1900

C. cenomamensis (Schliiter) 1876

C. colombiana Durham, 1946

C. compressns (Tavani) 1942

C. crehricostatus (Blanford) 1861

C. deslongchampsianum (d'Orbigny) 1840

C. eichwaldi (Karakasch) 1907
C. e.legans (J. Sowerby) 1816
C.elegantoides (d'Orbigny) 1840

Cretaceous

England,

France

Cretaceous

Algeria

(Neoeomian)

Cretaceous

Madagascar

(Albian)

Cretaceous

Germany

(Cenomauian)

Cretaceous

England

Cretaceous

England,

(Upper)

France

Cretaceous

California

(Lower)

(U.S.A.)

Cretaceous

England

(Senonian)

Cretaceous

Switzerland

Cretaceous

Germany

(Senonian)

Cretaceous

Vancouver

Island, B.C.

Cretaceous

Maple

(Upper)

Island, B.C.

Cretaceous

Europe

(Cenomanian)

Cretaceous

Colombia

(Aptian)

Oi-etaceous

Somaliland

(Albian)

Cretaceous

India

(Albian)

Cretaceous

Europe

(Neocomian-

Cenomanian)

Cretaceous

Crimea

(Lower)

Cretaceous

England

(Cenomanian)

Cretaceous

England,

(Cenomanian)

France

KUMMEL : POST-TRIASSIC NAUTELOLD GENERA

425

Species

Stratigraphic Geographic
Distribution Distribution

C. farmosus (Blanford) 1861

G.gabhi (Anderson) 1902

C. guilielmitelli (Ooster) 1858
Chilli (Shattuck) 1903

C. hunstantonensis (Foord and Crick) 1890

C. huxleyanus (Blanford) 1861

C. imhricatus (Crick) 1907

C. interstriaius (Strombeek) 1863

C. kayeanus (Blanford) 1861

C.kossmati (Spengler) 1910

C. loeblichi (Miller and Harris) 1945

C. loricatus (Sehliiter) 1876

C. madagascarensis (Yabe and Shimizu) 1924

C. manuanensi^ (Crick) 1907

C. mikado (Krenkel) 1910

C. nebrascense (Meek and Hayden) 1862

C. nedkerianus (Pietet) 1847

C. negama (Blanford") 1861

C . neocomiensis (d'Orbigny) 1840

C . neohispanicum (Burckhardt) 1925

Cretaceous

India,

(Senonian)

Madagascar

Cretaceous

California

(Lower)

(U.S.A.)

Cretaceous

Switzerland

Cretaceous

Texas

(Lower)

(U.S.A.)

Cretaceous

England

(Cenomanian)

Cretaceous

South India,

(Turonian)

Madagascar

Cretaceous

Natal,

Africa

Cretaceous

Germany

(Upper)

Cretaceous

India,

(Albian-

Madagascar

Cenomanian-

Turonian)

Cretaceous

India

(Albian)

Cretaceous

Texas

(Lower)

(U.S.A.)

Cretaceous

Germany

(Senonian)

Cretaceous

Madagascar

(Upper)

Cretaceous

Natal,

(Albian)

Africa

Cretaceous

Tanganyika

(Xeocomian)

Cretaceous

Montana

(Upper)

(U.S.A.)

Cretaceous

Europe

(Aptian,

Albian)

Cretaceous

India

(Albian)

Cretaceous

Europe,

(Lower)

Crimea

Cretaceous

Mexico

(Aptian)

426

BULLETIN : MUSELTM OF COMPARATIVE ZOOLOGY

Species

Stratigiaphic Geographic
Distiibutiou Distribution

C. occlusus Crick, 1907
C. patens (Kner) 1850

C. perstriatus (Steiier) 1921

C. pseudoathu (Yabe and Sliiniizir) lOill

C. psetidoelegans (d'Orbignyj 1840
C. psmidonegama Spengler, 1910

C. pseiuloneokomienae Shiniizii. 19ol

C.radiatus (J. Sowerby) 18'J2
C. sal'cdavtim Colliguon, 1949

C. saussvreanus (Pictet) 1847

C. scecqureniiis (Tavani) 19412

C. semilohaius Spengler, 1910

C. snniiindatus (Foord) 1891

C.sliarpei (Schliiter) 1876

C' sinuato-pUcatiis (Geiuitz) 1843

C.i^traticostatu^ (Crick) 1907

C.sucientie (Whiteaves) 1879

C. tenvicostatiis (Schliiter) 1870

C. tourtiac (Schliiter; 1876

C. Ukalt.siihrJensis (Rouchadze) 1931

C. tsulcushiensc (Kobayashi; 1954
C. iindulnfifnrmis Spath, 1927

Cretaceous

Zululand

< 'retaceous

Poland,

( Senonian )

Germany

Jurassic

( Upper )

Argentina

Cretaceous

(Lower)

Cretaceous

Japan

(Upper)

Cretaceous

Europe

Cretaceous

India

(Albian)

Cretaceous

Japan

(Aptian)

Cretaceous

Europe

( 'retaceous

Madagascar

(Albian)

Cretaceous

R^ritzerland

(Albian)

Cretaceous

Sonialiland

(Albian)

Cretaceous

India

(Albian)

Cretaceous

England

(Cenouianian)

Cretaceous

Germany

(Cenonianian)

Cretaceous

Germany

(Upper)

Cretaceous

Natal,

Africa

Cretaceous

Sucia Island

(Upper)

(iroup, B. C.

Cretaceous

(Germany

(Cenomaniairi

Cretaceous

(iermauy

(Cenonianian)

Cretaceous

Eussia

(Aptian)

Oligoceiie

Japan

Cretaceous

Tanganyika

(Xeocomian)

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

427

Species

Stratigiaphic
Distribution

Geographic
Distribution

C.ventroplicatus (Foord) 1891
C. virgatus (Spengler) 1910
C. woodsi (v. Hoepen) 1921

Cretaceous England

(Cenomanian)

Cretaceous India

(Albian")

Cretaceous South Africa

(Upper)

1927

Genus PaRACYMATOCEKAS Spatli,
Paniciimatoceras Spath, 1927a, pp. 21, 25.
Paracymatoceras Miller and Harris, 1945, p. 9.
Paracymatoccras Kummel, 19o3a, p. 4.
Tirpe species. Nautilus aspcr (Oppel) Zittel, 1868, pi. 3, lig. 1 (by original

designation). Illustrated on Plate 19, figures 1, 2 and Text Figure 23(t,

of this report.

Most of the other genera of the Cymatoceratinae are dis-
tinguished from Cymotoceras on differentiations of the conch
form or sculpture. Paracymatoceras differs from Cyrnatoceras
essentially only in the greater elaboration of the suture. It is also
generally agreed that in respect to the .suture Cymatoceras and
Paracymatoceras intergrade (Fig. 24). The conch form, degree
of involution, and shell sculpture agree perfectly with Cymato-
ceras. The type species — P. as per — which is of Upper Jurassic
age, has a rounded ventral groove on the most adoral part of
the living chamber (Pig. 236'). This feature is not present in
any of the five Cretaceous species of Paracymatoceras known to
date. Aside from the ventral groove in the type species, the
conch form in the other species assignable to Paracymatoceras
are completely similar to or within the i-ange of variation of
Cymatoceras.

Six species can be assigned to this genus i-anging in age from
Upper Jurassic through the Cretaceous. Four of the species are
Lower Cretaceous in age. The genus does not appear to be a com-
mon element in nautiloid faunas but it has been recorded from
India, Europe, Texas, and Mexico (Figs. 21. 22). Some juvenile
forms from the Lower Cretaceous PawpaAv formation of Texas
with perfectly smooth conchs have been recorded by Kummel
(1953a). These specimens are of interest in emphasizing the
difficulties in identifying weakly-ribbed Paracymatoceras and
differentiating these from Hercoglossa.

428

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

DISTRIBUTION OF SPECIES OF THE GENUS PAEACYMATOCERAS

Species

Stratigraphic Geographic
Distribution Distribution

Jurassic

Europe

(Upper)

Cretaceous

Crimea

(Lower)

Cretaceous

Durango,

(Aptian)

Mexico

Cretaceous

South India,

(Albian,

Madagascar

Maestrichtian)

Cretaceous

Texas

( Lower )

(U.S.A.)

Cretaceous

South

(Senonian)

India;

East Indies?

P.asper (Oppel) 1865
P . karpinskyi (Karakasch) 1907
P.? milleri Humphrey, 1949
P. rota (Blanford) 1861

P. texanum (Shumard) 1860
P.trichinopolitensis (Blanford) 1861

Genus PeOCYMATOCERAS Spath, 1927

Procymatoceras Spath, 1927a, pp. 21, 25.
Procymatoceras Miller and Harris, 1945, p. 11.

Type species. Nautilus stibtruncatus Morris and Lycett, 1850, pi. 1, fig. 2
(by original designation).

The genotype described and figured by Morris and Lycett
does not seem to be available. Several topotype specimens are in
the collections of the British Museum (Natural History) one of
which is illustrated on Plate 17 and the cross section of the same
specimen in Text Figure 25. From a study of topotype specimens
and the original description of the tj'pe species the following
diagnosis of Procymatoceras can be made.

The conch is robust, rapidly expanding, very involute to
occluded. Whorl section generally wider than high. On the
living chamber the venter and whorl sides are flattened. The
ventral and umbilical shoulders are well rounded. The whorl
sides converge towards the venter. The earlier volutions are
more rounded in outline. The surface of the conch bears sinuous
ribs that form a broad sinus on the venter. The ribs appear to
be fasciculate growth lines and may not be homologous with
those of typical Cymatoceras. The ribbing is most distinct on
the living chamber. The sutures are only slightly sinuous with
shallow ventral and dorsal lobes.

KUMMEL: POST-TRIASSIC NAUTILOID GENERA

429

When Spath (1927a) first established Procymatoceras he in-
cluded only the type species and P. haheri (Morris and Lycett)
both from the Middle Jurassic of Great Britain. The topotype
specimen of P. subtruncatus illustrated on Plate 17, and on Text

Fig. 25. Procymatoceras subtruncatus (Morris and Lycett). Diagram-
matic cross section of topotype represented by figures 1, 2, Plate 17, X 0.50.

Figure 25, clearly shows the robust involute character of the
conch and the subtrapezoidal outline of the whorl section.
Neither the suture nor the siphuncle are visible. Nautilus in-
tumescens Waagen has been referred with question to Procyma-
toceras. I agree with Spath (1927a, p. 33) that it is doubtful
whether Noetling's (1896) specimens from Baluchistan assigned
to N. intumescens Waagen are correctly identified. The only

4:JU UULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

other species that appears to belong to this genus is Paraceno-
ceras cosiatum Scott (1943).

Genus CymATONAUTILUS Spath, 1927

Cymatonautdiis Hpath, 1927a, p. l21.
Cymatonautilus Miller and Harris, 1945, p. 9.

Type species. Nautilus julii (Baiigier MS) d 'Orbigny, 1850, p. 328 (by
original designation).

The type species of this unusual Juj-assie cymatoceratid was
originally briefly diagnosed but not illustrated by d'Orbigny
(1850, p. 328) who gave Baugier the credit as author of this
species. However, Baugier 's name apparently was only a manu-
script name, and was never published by him. The first com-
plete description and illustration of Nautilus jidii was published
by Gueranger (1865, p. 189, pi. 3). Gueranger's figures arc
reproduced here on Plate 18, figui'cs 1-4.

This genus can be diagnosed as follows : Conch widely umbili-
cate, robust, whorls subquadratic, being slightly wider than
high. The whorl sides are flattened and with a broad lateral
groove. The venter is likewise flattened and with a median
groove. The conch bears sinuous ribs that ol:)liquely cross the
suture and form a deep ventral sinus. The suture forms a shal-
low ventral lobe and a broad concave lateral lobe. The siphuncle
is subcentral in position, being closer to the dorsum than the
venter. The type specimen came from strata of Callovian age
in France. Petitclerc (1926) has described and illustrated a
specimen of the same species from strata of Bathonian age in
France. Nautilus mojsisoric.n Neumayr (1870) from Upper
Jurassic strata of Austria most probably also belongs in Cyma-
tonautiJus.

The most distinctive- features of this genus are the wide
umbilicus and the grooved venter and whorl sides. Most species
of the Cymatoceratinae are involute forms.

Genus AXGLONAUTILUS Spath, 1927

Anglonautilus Spath, 1927a, pp. 21, 25.
Anglonautilus Miller and Harris, 1945, p. 8.

Type species. Nautilus undulatus J. Sowerby, 1813, pi. 40, upper figure (by
original designation).

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 431

Sowerby's type specimen does not appear to be in the British
Museum, but the specimen referred to by Foord (1891, p. 245)
labelled ^'Nautilus uiidulatus, M. C. 182" in Sowerby's hand-
writing is figured here, Plate 20, figures 1, 2, and Text Figure
26. This specimen is also numbered B.M.N.H. 66336. Dr. L. F.
Spath kindly informs me that this specimen is unlocalized, but

Fig. 26. Anglonautilus undulatus (Sowerby). Diagrammatic cross sec-
tion of adoral part of living chamber of topotype represented by figures
1, 2, Plate 20, X 1.

undoubtedly is a topotype from Nutfield, Surrey. This specimen
is incomplete consisting only of two camerae of phragmocone and
one-third whorl of living chamber. However, it shows clearly
the most diagnostic feature of the genus, that is the large, fold-
like undulations which are most prominent on the venter and
decrease rapidlj^ on the flanks. The suture forms a very shallow
ventral lobe and a l)road shallow lateral lobe. The siphuncle is
subcentral in position being nearer the dorsum than the venter.
Tn addition to the type species which is known from Aptian
to Cenomanian strata in England and various localities in
Europe {]ide Foord, 1891, p. 244), Nautilus suhalbensis SinzoA\
(1913) and N. hegudensis Kilian and Reboul (1915) should be
included in Anglonautilus. The former species is from Albian
strata in the Crimea and the latter from Hauterivian strata in
southern France.

Genus EuCYMATOCERAS Spath, 1927
Eucymatoceras Spath, 1927a, pp. 21, 22, 25.

432

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

E uoymatoceras Miller and Harris, 1945, p. 9.

Type species. Nautilus plicatus Fitton, 1835, p. 129 (by original designa-
tion).
The type specimen was not available for study but several
conspecific forms are in the British Museum. One of these speci-
mens from the Lower Greensand, Atherfield, Isle of Wight
(B.M.N.H. C335) is illustrated on Plate 21, and the cross section

Fig. 27. Eucymatoceras plicatus (Fitton). Diagrammatic cross section
of plesiotype represented by figures 1, 2, Plate 21, X 1.

in Text Figure 27. The only previous illustrations of the type
specimen are the original woodcut in Fitton (1835, p. 129)
and figures in d'Orbigny (1840, pi. 10; Nautilus requienianus
d'Orbigny 1840 =N. plicatus Fitton 1835) and in Uhlig (1883.
pi. 3).

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 433

The genus can be diagnosed as follows: Conch subglobular,
involute, umbilicus closed, whorl section rounded, broader than
high. Suture only slightly sinuous. The siphuncle is small and
subcentral. Surface of conch with prominent ribs that form a
deep angular V-shaped ventral sinus and on the whorl sides a
similar but asymmetrical salient.

The most diagnostic feature is the V-shaped pattern of the
ribs by which it can easily be distinguished from any other
genus of the Cymatoceratinae. At present only three species
of Eucymafoceras have been recorded, namely N. plicatus Fitton,
the type species, and N. sieveni Karakasch (1907, p. 30, pi. 2,
fig. 13; pi. 8, fig. 12) from Lower Cretaceous formations of the
Crimea; and N. stschiirouskii Milaschevitch, (1877, p. 125, pi. 1,
figs. 11, 11a) from the Lower Cretaceous of Russia. All of these
species have the depressed, subglobular, rounded conchs. From
the great variability in conch form in Cymatoceras it is doubtful
if the conch form in the known species of Eucymatoceras can be
considered a diagnostic character.

Genus SyrIOXAUTILUS Spath, 1927

Syrionautilus Spath, 1927a, pp. 21, 25.
SyrionautUus Miller and Harris, 1945, p. 11.
SyrionautUus Aynimelech, 1946, pp. 523-528; 1947a, p. 690.
Type species. Nautilus lihanoticus Foord and Crick, 1890, p. 404, fig. 6 (by
original designation).
All the specimens of Nautilus libanoticus in the British Mu-
seum (Natural Histor}^) upon which Spath established his
genus SyrionautUus are more or less crushed and distorted.
One of the syntypes with an attached mandible is figured here
on Plate 7, figure 3 (previously figured by Foord, 1891, p. 371,
fig. 82.). The general form of the conch is like that of Cymato-
ceras but the character of the ribbing is quite distinct. In Syrio-
nautilus the ribs are acute and separated by interspaces rather
exceeding their own width. The type of ribbing is very similar
to that of Proclydonautilus spirolohus of the Upper Triassic.
The form and spacing of the ribs in SyrionautUus are the dis-
tinguishing characters that serve to separate it from allied
genera of the Cymatoceratinae. The position of the siphuncle
and nature of the suture are not known. The type and only

434 BULLETIN : ilLWEUM OF COMPARATIVE ZOOLOGY

species so far referable to this genus comes from Senonian
formation of Syria.

Recenth' Avnimelecli (1946, 1947a) has published two short
notes on additional specimens of Syrionaiitilus lihanoficus from
Palestine. As with the types, this new material is also incom-
plete and generally crushed. However, the best and nearly com-
plete specimen was iigured by Avnimelech (1947a, pi. 15). Even
though this figured specimen is also compressed it clearly shows
Syrio7iaufilus to have an involute, compressed conch and arched
venter. In other words, it has a typical cymatoceratid conch but
differs in the character of the ribs.

Genus He^^LIXAUTILVS Spath, 1927

Heminautilus Spath, 1927a, pp. 22, 25.
Vorticoceras Seott, 1940, pp. 1074. 1075.
Heminautilus Scott 1943, p. 89.
Heminautilus Durham, 1946, ii. 432.
Plati/tiautihis Yabe and Ozaki, 1953, pp. 55-61.

Tijpe species. Nautilus saxbii Morris, 1848, p. 106 (woodcut) ; PI. 10, figs.
1, 2 of this report (by original designation).

When Spath established this genus he designated Nautilus
saxhii Morris (1848, p. 106, woodcut) as genotype and thought
this species was the same as X. JaUierimuts d'Orbigny (1841).
The latter species was first very In-iefl}' diagnosed but not illus-
trated by d'Orbigny in 1941 in the Revue Cuvierienne (p. 318).
and was listed the same year in his Paleontologie Francaise.
Terrains Cretaces (p. 620). Nautilus laUierianus was again
cited by d'Orbigny in his Prodrome (1850, t. 2, p. 112), and
N. saxhii was listed as a synonym. In none of these publications
is there an illustration and only the first (1841, p. 318) has a
very brief ddagnosis. A. laUierianus was next described and
illustrated from Switzerland by Pietet and Campiche (1859).
Douville (1916) gave a very complete description with several
figures of specimens of A. laUierianus from Aptian strata east
of Suez. Douville likewise placed A. saxhii in synonymy of A.
laUierianus. Foord (1891) recognized both of these species.

The type specimen of Nautilus saxhii is illustrated on Plate
10, figures 1, 2, and the cross section in Text Figure 28. The
only previous illustrations of this specimen are the woodcuts in
Morris (1848. p. 106) and Foord (1891, p. 311, fig. 68). Since

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 435

d'Orbigny's N. lallierianus has not been illustrated, and tho
description is very brief, it is impossible to tell whether it is
eonspecifie with N. saxbii. Nautilus lallierianus, Donville
(1916) is sufficiently distinct from N. saxhii Morris in the form
of the septa and ornamentation of the conch to keep these two
species separate. The plaster casts of the specimen of N. lallieri-
anus figured by Pictet and Campiche (1859, pi. 19, figs. 6a, 6c)
are in the Museum of Comparative Zoology. These show no
surface ornamentation whatsoever. The tricarinate character of
the venter is not as well preserved as the figures by Pictet and
Campiche would lead one to believe.

The genus TIeniinaiitilus can be diagnosed as follows: Conch
involute, very compressed, whorls much higher than wide. Whorl
sides flattened and strongly converge towards a narrow, flattened
or slightly concave venter. The ventral shoulders are angular.
The venter on the earlier volutions is rounded. The umbilical
shoulders are broadly rounded. Greatest width of the conch
is just ventral of the umbilical shoulders. The suture is rather
sinuous with a ventral lobe, a subangular saddle on the ventral
shoulder, broad deep lateral lobe and a narrowly rounded saddle
on the UDibilical shoulder. The si})huncle is subcentral in posi-
tion lying closer to the dorsum than the venter. The conch
bears sinuous ribs that curve strongly aborally over the ventral
half of the whorl side. The ribbing may be very weak and in fact
is absent on the type of Nautilus saxhii (which is probably a
juvenile).

Spath originally placed Heminautilus in his Paracenoceratidae
with question and made note of the possibility that it more
properly belonged in his Cymatoceratidae. The Paracenocera-
tinae are characterized by differentiation of the periphery, gen-
erally associated with increase in siiniosity of the suture lim-.
and a ventral lobe (S])ath, 1927a, p. 25). Plowever, the sutures
of Heminautilus are much more sinuous (Fig. 20). The Para-
cenoceratinae have smooth shells, devoid of ribbing. Tithoiio-
ceras, and Aulaconautilus have longitudinal ribs or carinae on
the ventral area but no sculpture on the flanks. Sinuous cyma-
toceratid-like ribbing is characteristic of Heminautilus. The
ribbing is well preserved on the specimens of H. lallierianus fig-
ured by Douville (1916, pi. 17, figures 2-6) and on //. rangei
Hoppe (1922, pi. 4. figures 2, 8). Among the specimens illus-

436

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

trated by Douville there is some variability in the strength and
patterns of the ribs. On the earlier volutions the ribs are low,
broad, and rounded and decrease toAvards the venter. The mature
specimen of Douville 's (1916) figure 5, plate 17, has a smooth
outer volution, whereas the larger specimen of figure 6a, b,
shows prominent ribs that are very broad and separated by
narrow concave interspaces. In Heminautilus etheringtoni Dur-
ham (1946) the ribs are most prominent on the living chamber

Fig. 28. Hemiimutilus saxbii (Morris). Diagrammatic cross section of
holotype represented by figures 1, 2, Plate 10, X 1.

which is about one-third of a volution in length. Topotype speci-
mens of Heminautilus etheringtoni kindly loaned to me by J.
Wyatt Durham show rather distinct ribbing on early volutions,
with a single large topotype smooth except for prominent growtli

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 437

lines. A crushed and weathered specimen, not referred to by
Durham (1946), from near the top of the basal Velez Sandstone
measures approximately 145 mm. in diameter and has very
prominent "cymatoceratid" ribs. The ventral area likewise
appears to have a median longitudinal rib. With the material
available it cannot be determined whether or not this specimen
is conspecific with //. ethcringtoni. They do not come from the
same horizon within the Aptian. No data are available on the
degree and extent of variation in the sculpture of this or most
species of nautiloids. It is very rare that really representative
populations can be collected.

Hemiriaufilns tijosiensis (Yabe and Ozaki) 1953, shows prom-
inent growth lines on tiie venter that form a deep V-shaped
I)attern. On the lateral whorl sides low radial folds with broader
intervals are present on the type (and only) specimen of this
species. These authors proposed a new genus - — Platynautilus
— for their new species, but it is clearly a representative of
Heminautilus and thus Platynautilus is placed in synonymy.

Heminautilus stantoni (Scott) has broad, low, rounded and
widely spaced ribs reaching their greatest prominence on the
middle of the flanks. Heminautilus saxhii, the type specimen
which is figured here on Plate 10, figures 1, 2, is perfectly
smooth but is probably not a mature specimen.

The suture in most species of the Cymatoceratinae are only
very slightly sinuous. Paracymatoceras has a well-developed,
deep, lateral lobe followed by a saddle on the umbilical shoulder.
The suture across the venter generally develops a ventral saddle,
not a lobe. Cyinatonautilus julii has a deep ventral lobe, and a
broad lateral lobe with the beginnings of a saddle on the umbili-
cal region. "Nautilus" vaelsensis (Binckhorst, 1861) is a very
compressed form with a subrectangular whorl section and a
wide, shallow umbilicus. The truncate venter and compressed
conch are very reminiscent of Heminautilus except in the dif-
ference of involution. The suture of ''Nautilus" vaelsensis has a
very shallow ventral lobe and a broad, shallow, lateral lobe.

The six known species of Heminautilus are from Lower Cre-
taceous formations in Arkansas, England, France, Switzerland,
Egypt, Palestine, Colombia, and Japan.

438

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

Stratigraphic

Geographic

Distribution

Cretaceous

Colombia

(Aptian)

Cretaceous

Europe.

(Xeocomian)

EgJTt,

Ethiopia

Cretaceous

Palestine

(Albian)

Cretaceous

England,

(Aptian)

France

Cretaceous

Arkansas

(Aptian)

(U.S.A.)

Cretaceous

Japan

(Aptian?)

DISTEIBUTION OF SPECIES OF THE GENUS HEISnNAUTILUS

Species
S. etheringtoni Durham, 1946

H.lallierianus (d'Orbigny) 1840

E.rangei (Hoppe) 1922

H. saxbii (Morris) 1848

E.stantoni (Scott) 1940

H. tyosiensis (Yabe and Ozaki) 195.3

Genus DeLTOCYMATOCERAS n. gen.

Type species. Nautilus leiotropis Schliiter, 187G, p. 175, pi. 48, figs. 1, 2
(Plate 22, figures 1, 2 of this paper).

The adaptive radiation within the Cymatoceratinae is ex-
pressed in patterns that are homeomorphous with evolutionarj-
trends in other phyletic lines within the Nautilidae. Modification
of rib patterns, conch shape, and suture are the more obvious
characters which express evolution within the subfamily. The
present genus is established for those cymatoceratids with a
more or less triangular outline of the whorl section. The type
species- — Nautilus leiotropis, and N. rugatus (Fritsch and
Schlonbach) are the only species known to date that can be
placed in this genus. The triangular outline of the whorl sectioji
serves readily to differentiate this genus from all others Avithin
the Cymatoceratinae.

The type species has a large, inflated, involute conch with
broadly arched and strongly convergent whorl sides. The venter
is marked by a slight rounded, smooth, keel-like ridge. There
are no distinct ventral shoulders, the ventral region grading
evenly onto the flanks. The flanks have prominent cymatocera-
tid ribs that bifurcate near the middle of the whorl sides. The
ribs do not cross the keel-like ridge on the venter. The suture
forms a narrow, somewhat pointed, ventral saddle, a broad,

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 439

leather shallow lateral lobe, and a prominent saddle on the um-
bilical shoulder. The position of the siphuncle is not known.

The distinctive subtriangular whorl section with the smooth
keel-like ridge on the venter set Deltocymatoceras leioiropis
apart from all other eymatoeeratids. In conch form and suture
pattern Ddlocymatoceras is homeomorphous with AnguUihcs
of the Hercoglossinae, being distinguished bv its ril)bing. Aiigu-
Jithes is most probably a development of Cimomia, a form with a
more rounded whorl shape. Deltocymatoceras is best visualized
as a sagittate derivative of Cjfmatoceras which show's much vari-
ation in conch shape.

The type species of Deltocymatoceras comes from Upper Cre-
taceous (Emseher-Mergel) of German}-. The original is in
the Museum of the University of Bonn {fide Schliiter, 1876, p.
175). Deltocymatoceras rugeitiis comes from Tapper Cretaceous
strata of Bohemia.

Genus EpirYMATOCER.AS n. gen.

Ti/pe species. Xaiitilus vaelsensis Bmckhor.st, 1861, p. 15, pi. 5c, figs. 2a, b, c
This new genus is established for a species of Upper Creta-
ceous nautiloid that has a most unusual conch shape for post-
Triassic nautiloids. The species has been described and figured
by Binckhor.st (1861, p. 15, pi. 5c, figs. 2a, b, c) and Schliiter
(1876. p. 177, pi. 51, figs. 3-4). Schliiter 's figure is reproduced
here on Plate 23, figs. 1, 2.

Xautilns raclscvsis has an evolute and greatly compressed
conch. The umbilical diameter is approximately 25 per cent
of the diameter of the conch. The whorl section is subquadrate,
being nearly twice as high as wide. The ventral shoulders are
angular, the venter narrow and flattened. The lateral flanks are
only slightly inflated. The umbilical shoulders are broadly
arched. Tlu^ suture inscribes a very shallow ventral lobe and
a shallow lateral lobe. The position of the siphuncle is not
known. The conch bears fine sinuous ribs that curve backwards
toward the ventral shoulder and form a slight sinus on the
venter.

Epicymatoceras is distinguished from all other genera of the
Cymatoceratinae by its compressed, evolute conch and flattened
venter. It is the ribbing which attaches this genus to the Cyma-

44U BFLLETIX : MFSEUif OF COMPARATRT: ZOOLOGY

toceratinae. Among the Cj^matoceratinae the only group some-
what similar is Heminautilus which is likewise compressed in
conch form and with a flattened venter. However, in Heminau-
tilus the whorl section is much more inflated near the umbilical
shoulders and the flanks more convergent. Likewise this genus
is very involute and with a more sinuous suture. Most of the
remaining genera and species of the Cymatoceratinae have more
involute and inflated conchs.

The presence of cymatoceratid ribbing readily differentiates
Epicymatoceras from other subfamilies of the Nautilidae. There
is some similarity in conch shape of Epicymatoceras with that of
Pseuclocenoceras of the Nautilinae. However, in that genus the
smooth conch has a well rounded ventral area ,steep umbilical
shoulders, and is more involute.

The type and only species of this genus has been identified
from Upper Cretaceous (Maestrichtiau) strata at Vaels near
Achen (Germany) and at Limbourg (Belgium) just south of
Achen.

Subfamily HERCOGLOSSINAE Spath, 1927

Spath (1927a) originally erected this family to include the
following genera: Pseiidaganides Spath of the Jurassic and
lowermost Cretaceous; Hercoglossa Conrad, Upper Cretaceous to
Eocene; Hercoglossoceras Spath, Liassic ; Pseudonautilus Meek,
Upper Jurassic; Paraturia Spath {=Aturoidea Vredenburgj,
Cretaceous to Eocene ; and Deltoidonautilus Spath, Eocene.
Later he (Spath 1927b) included Cimomia Conrad, and Stenzel
(1940) added his genus Woodringia. Restudy of this group
alters the above interpretation in many important details. The
Jurassic hercoglossids, namel}' Pseudaganides and Pseudonau-
tilus are interpreted as a distinct phjdetic line from the Cre-
taceous-Eocene trend involving Hercoglossa itself (see page 389).
Hercoglossoceras is considered to be a synonj-m of Pseudaganides.
It is merely an early, highly' compressed ''agauitiei" and is part
of the radiation from the cenoceratid complex which went in
for highly modified sutures (Pseudagauitinae). Adaptation to-
wards modified (highly sinuous) sutures is a trend that has
occurred many times in the evolutionary history of the nautil-
oids. Within the Upper Paleozoic few nautiloids are known with

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 441

highly sinuous sutures. The most extreme development is that
found in Permoceras hitauniensis (Haniel) from strata of
Leonard age in Timor. Grypoceras {Plummeroceras) plummeri
KuDimel (1953c) of the Grypoceratidae from beds of Leonard
age in Texas has a deep ventral lobe and lateral lobe. When
Miller and Collinson (1953) established the genus Permoceras
thev also erected the familv Permoceratidae for their monotypic
genus. It appears more likely that Permoceras is an aberrant
development from the Domatoceras-Grypoceras evolutionary line
and should be included in the Grypoceratidae.

In the Triassic, the first large-scale adaptation towards highly
modified sutures appears in the Clydonautilidae, Gonionautili-
dae, and Siberionautilidae all of the Upper Triassic. The sutural
patterns seen in these groups are closely similar to those of the
Pseudaganides-Pseudonautilus trend of Jurassic age and to the
Hcrcoglossa-Aturoidea trend of Cretaceous-Eocene age. The
details have been discussed previously (p. 389) and the evidence
seems quite strong that in the Mesozoic there were three separate
radiations built largely around modification of the suture, the
first in the Triassic, the second in the Jurassic, and the third
in the Cretaceous. A similar adaptive trend also took place in
the Cymatoceratinae. Since the basic sutural type must surely
be an adaptive character it is no more than to be expected that
repetition of similar forms will be produced. For the aboA'e
reasons Pseudaganides and Pseudonautilus are considered a
separate phyletic development and placed in the subfamily
Pseudaganidinae.

Deltoidonautilus Spath is considered to be a synonym of Aii-
gidithes Montfort (Kuramel 1953a) ). Spath placed AnguUthes in
the Nautilidae but it is here considered part of the Hercoglossi-
nae. Woodringia Stenzel (1940) is placed as a synonym of Herco-
glossa since the character upon which it was distinguished — a
slight ventral lobe in the ventral saddle — is of no more than
specific importance in other genera of this subfamily.

Thus, as now interpreted, the Hercoglossinae include the fol-
lowing genera: Hercoglossa Conrad, Cimomia Conrad, Angu-
Uthes Montfort, and Aturoidea Vredenburg. These genera are
all characterized by involute, smooth, conchs with modified
sutures. As has been shown by Miller (1947, 1949) on scA^eral
occasions this group of genera appears to form a linear series

BULLETIN: ML'SEUM OF COMPARATIVE ZOOLOGY

KUMMEL : PO^^T-TRIASSK XAUTILOID GENERA

443

444 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

of increasing sinuosity in the suture ranging from Cimomia to
Hercoglossa to Aturoidea. Angulitlies is probably a special de-
velopment out of Cimomia maintaining an approximate cimomid
tj'pe of suture but modifying the whorl section to a subtriangular
shape rather than rounded. Cimomia is readily derived from
Eutreplioceras with which it is more or less gradatioual. Since
the genera of this subfamily (with the exception of Angulitlies)
are separated on the basis of increasing sinuosity of the suture,
it is understandable that there should l)e gradational forms.
These gradational forms are found especially between Entrepho-
ceras and Cimomia and between Cimomia and Hercoglossa. It
has been generally assumed that each of these units had a single
time of origin and thereafter were distinct genetic stocks (Miller.
1949). An alternative interpretation is, however, jDOssible;
that is, that there were several times in the Cretaceous and
Tertiary when radiations towards sinuosity of the suture oc-
curred and what is normally included in the genus Cimomia is a
group of transitional species and not part of a single linear
trend. The kind and nature of the data available, however, is
not sufficient to test this suggestion. The eutrephoceratid-evolv-
ing stock is the parent stock of several radiations (generic
groups) of rather diverse morphology. The fact that the genetic
potential of the group shows variability in conch shape and
suture, is no reason why any number of these variants in suture
could not become in themselves geneticalh' distinct, to become
part of the species complex included in Cimomia.

All four of these genera occur in the Cretaceous and range
into the Eocene or Oligocene. Cimomia probably includes some
Upper Jurassic species. In numbers of species, Hercoglossa has
21, Cimomia 44, Angulithes 32, and Aturoidea 11. The relatively
large number of species is largely due to ''monographic" highs,
especially in Tertiary faunas.

Geographically, species of this subfamily are vevy wide-
spread. The distribution of the Cretaceous species of the sub-
family is illustrated on Figure 29, those of the Tertiary on
Figure 30. The maps show very well the general overlap in
range of all of the genera. None of the forms is sufficiently com-
mon at any of the localities indicated so that collections could
be made for detailed population studies.

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 445

Genus Ci:\rOMIA Conrad, 1866

Cim<»nia Conrad, 1866a, p. 102.
Cymomia Conrad, 1866b, p. 19.
Cymmomea Conrad, 1868, j). 732.
Cymomia Conrad, 1868, p. 732.
Cimomia Spath, 1927b, pp. 424-428.
Nautilus ( J avanoceras ) Martin, 1932, pp. 1, 2.
Cimomia Miller and Thompson, 1933, pp. 305-313.
Cimomia Stenzel, 1940, pp. 751-753.
Cimomia, Miller, 1947, pp. 39-49.
Cimomia Miller, 1951, pp. 38-47.
Cimoviia Haas and Miller, 1952^ pp. 329-338.

Type species. Naidilus hurtini Galeotti, 1837, from Eocene of Belgium, by
original designation. Figured here on Plate 24.

Most of the species of this genus are Tertiary in age (Paleo-
cene — Lower Oligocene) and the Tertiary species have in recent
years been thorouglhly discussed by Miller and Thompson
'(1933), Stenzel (1940)', Miller (1947), Miller (1951), and Haas
and Miller (1952). The very able diagnosis of the genus by
Miller (1947, pp. 39-40) is quoted here: "Conch subglobular to
subdiseoidal in shape and nautiliconic in its mode of growth.
AYhorls broadly rounded laterallj^ rounded veutrally, and im-
pressed dorsally. Umbilicus small : umbilical shoulders low and
broadly rounded. Surface of conch marked by growth lines but
does not bear ribs or nodes. Septa numerous and each suture
forms a broad very shallow broadly rounded or slightly retuse
ventral saddle, a broad shallow broadly rounded lateral lobe, a
narrower higher rounded lateral saddle (located near the umbili-
cal shoulder), a broad rounded lobe on the umbilical wall, and
a fairly prominent internal lateral saddle that extends to a
broad rounded dorsal lobe. Siphuncle small, circular in cross
section, and orthochoanitic in structure ; its position varies con-
siderably in different species, but in no case is it marginal."

Cimomia is in a sense a morphological transitional form, es-
pecially in its suture, between Eutrephoceras and Hercoglossa.
In Eutrephoceras the suture is essentially' straight, or only
slightly sinuous ; in Hercoglossa the suture has well defined lobes
and saddles. In both of these genera there are transitional forms
to Cimomia, reflected especially b}- the suture. That is, in
Cimomia there is great A-ariation in the intensitv of the lateral

lULLETix : .\i[:sei;m of comparative zoology

Figure 3 1

KUMMEL : POST-TRIASSIC XAUTILOID GENERA 447

lobe and saddle (Fig. 'S2) . There is likewise much variation in
the shape of the conch (Fig-. 31). I agree heartily with Miller
(1947, p. 41) that C. kuglen and C. vaughani which have slight
lobes in the ventral saddle should be retained in Ciniomin and
not separated from Hercoglossa, as Woodringia was, solely on
this character. In both these species the median lobe is not
developed until full maturity .

Most previous authors who have discussed this genus were
mainly concerned with the species which occur in Tertiary for-
mations. There are in addition to these, several species of Cre-
taceous and possibly Upper Jurassic age that should be assigned
to this genus. There is first of all Nautilus romeroi Ihering
(1903, pp. 195-196, fig. 1) of the "Etage Rocaneen" (Upper Cre-
taceous?) of the territory of Rio Negro in central Argentina,
previously mentioned by ]Miller and Thompson (1935, p. 570)
and Miller (1947, p. 41). Recently Miller (1951) has recorded
poorly preserved specimens which most surely belong to Cinio-
mia from Maestrichtian formation of Senegal. These specimens
are too poorly preserved to accurately identify specifically, but
there is no doubt as to their generic affinities. From Libya, in

Fig. 31. Cross sections of the couch of A, Hercoglossa harrisi Miller and
Thompson, from Stenzel 1940, fig. 116, diameter 100 mm. ±; B, H. ulrichi
(White) from Stenzel 1940, tig. 116, diameter 170 mm.; C, E? foriesianus
Blanford 1861, pi. 13, fig. 2, diameter 75 mm.; I), H. gardnerae Stenzel
1940, fig. 116, diameter 145 mm.; E, E. splendens Stenzel 1940, fig. 120,
diameter 73 mm.; F, Hercoglossa orUciilata (Tuomey) from Miller 1947,
lig. 12, height 174 mm.; G, E. mcglameryae Miller and Thompson, from
Miller and Collinson 1951, fig. IB, height 100 mm.; E, Aturoidea vieirai
Miller 1951, fig. 16B, height 130 mm.; /, A. parl-insoni (Edwards) from
Miller 1951, fig. 15A, height 180 mm., the genotype of Aturoidea; J, A.
paui'tfcr (Cope), from Miller 1947, fig. 18, diameter 228 mm.; E. Cimomin
vestali Miller and Thompson, from Stenzel 1940, fig. 118, diameter 126 mm.;
L, C. haltomi (Aldrich) from Miller 1947, fig. 8, height 104 mm.; M, C.
bticcinaeformis Haas and Miller 1952, fig. 2, height 29 mm.; N, C. vaughani
(Gardner) from Stenzel 1940, fig. 118, diameter 59 mm.; 0, C. turcicus
(Krumbeck) 1905, pi. 7, fig. 6b, diameter 120 mm.; P, C. haughti (Olsson)
from Miller 1947, fig. 9, diameter 60 mm.; Q, C. suhrecta Miller and Thomp-
son, from Stenzel 1940, fig. 118, diameter 120 mm.; R, C. angustus (Blan-
ford), from Stoliezka 1866, pi. 93, fig. 42, diameter 40 mm.

448

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

•D

Figure 32

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 449

North Africa, Wanner (1902, p. 143) has described Nautilus
jordani, from formations of Danian age, which is clearly a
species of Cimomia. Alberici (1940, p. 187) lists the range of this
species as Campanian to Danian. The lateral saddle next to
the umbilical shoulder in both C. romeroi and C. jordani is well
developed and more distinct than in many lower Tertiary species
such as C. septenicastrensis Haas and Miller from Lower Eocene
of British Somaliland and C. sahariensis (Keller) from Eocene?
of French Sudan. Another Cretaceous species which should be
placed in Cimomia is N. angustus Blanford (1861, p. 27, pi. 14,
figs. 1, 2) from the Ootatoor group (Cenomanian) of south
India. Stoliczka (1866, p. 209) records an additional specimen
of this species from the Ariyalur group (upper Senonian) of
south India. The type specimen of N. angustus is a much com-
pressed form with a narrowly rounded Tenter, and is the most
compressed species of Cimomia of which a record is available.
Stoliczka 's specimen (1866, pi. 93, fig. 4, 4a) is a more in-
flated form and more typical of most species of Cimomia. Speng-
ler (1910, p. 145) expressed doubt as to whether Stoliczka 's speci-
men is conspecific with that illustrated by Blanford. Even if these
two specimens are not conspecific, they both belong in Cimomia.
In both specimens the suture has the characteristic pattern found
in Cimomia.

rig. 32. Diagrammatic representation of sutures of A, EutrepJioceras
victorianum (Teichert) 1947, fig. 7; B, E. d.el-ayi (Morton) from Reeside,
1924, fig. Id.; C, Cimomia wyllei (Newton) from Haas and Miller, 1952,
fig. 5 ; i», EutrepJioceras sloani Reeside from Miller, 1947, fig. 6 ; E, Cimo-
mi<i landanensis (Vincent) from Miller, 1947, fig. 7A; F, Cimomia subrecta
Miller and Thompson, 1933, from Stenzel, 1940, fig. 119 (2); G, Cimomia
haltomi (Aldrich) from Stenzel, 1940, fig. 119 (1); H, Cimomia pseudo-
bouchardianum (Speugler) 1910, from Blanford, 1861, pi. 5, fig. 6; /,
Angulithes sow.erbiji (Wetherell) from Stenzel, 1940, fig. 124 (1); J, Cimo-
mia vestali Miller and Thompson, 1933, from Stenzel, 1940, fig. 119 (4) ;
K, Angulithes molli (Douville), from Miller, 1951, fig. 12A; L, Cimomia
macfadyeni (Haas and Miller) 1952, fig. 4A; M, Angulithes elliotti (Sten-
zel) 1940, fig. 124 (2); N, A. balceri (Teichert) 1947, fig. 3; 0, A. spathi
(Haas and Miller) 1952, fig. 7; P, A. biyogorensis (Haas and Miller) 1952,
fig. 6; Q, A. chudeaui (Douville) from Miller, 1951, fig. lOB; B, A. sp. from
Haas and Miller, 1952, fig. 8B. All figures greatly reduced.

450

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

The oldest species of Cimomia is thought to be Nautilus turci-
cus Krumbeck (1905, p. 137, pi. 14, fig. 6 a, b) from Upper
Jurassic Glandarienkalkes of Libj'a. This species has an inflated,
subglobular, smooth conch and a suture that is slightly sinuous
with a rather weak lateral lobe followed by a broad very shallow
saddle. As was pointed out above, graditional forms betAveen
Eutrephoceras and Cimomia are known ; N. turcicus can surely
be considered such a form. The shallow, but very distinct lateral
saddle lying above the umbilical shoulder shows a clear pattern
like that found in Cimomia, even though it is only weakly de-
veloped.

The distribution and age of the species of Cimomia are given
below. In addition to these it should be noted that Teichert and
Glenister (1952, p. 737) record the , presence of undescribed
species of Cimomia in beds of Upper Campanian to Lower
Maestrichtian age from the northwest basin of Western Au-
stralia.

DISTRIBUTION OF SPECIES OF THE GENUS CIMOMIA

Species

Stratigraphii"

Geographic

Distribution

Cretaceous

south India

(Upper)

Oligocene

Palestine

Eocene

British

Somalilaud

Eocene

Belgium

Cretaceous

England

(Cenoraanian)

Cretaceous

Bohemia

(Cenomaniau)

Tertiary

Pakistan

(Lower)

Eocene

Hungary

Cretaceous

Chile

Eocene

Bavaria

Cretaceous

East Africa

(Xeocomian)

Eocene

Pakistan

Paleocene

Alabama

(U.S.A.)

C. angu^tus (Blanford) 1861

C. blakei (Avnimelech) 1947

C. huccivMeformis Haas and Miller, 19.52

C. burtini (Galeotti) 1837
C.cantahrigiensis (Foord) 1891

C. colwnbinas (Fritsch and Schlonbach)

1872
C. cossmanni (Vredenburg) 1928

C . orassicoiicha (Yogi) 1908
C. d'orbignyanus (Forbes) 1846
C. eTlipticus (Schafhautl) 1852
C. exTpletiis (Zwierzycki) 1914

C. forbesi (D'Arehiac and Hainie) 1854
C. haltomi (Aldrich) 1931

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

451

Species

Stratigraphic Geographic
Distribution Distribution

C. heberti (Binckhorst) 1861

C. licsperia Miller and Downps, 19o0

C.himti Haas and Miller, 1952

Cintuscatenatus (Martin) 1932
C. imperialis (J. Sower by) 1812
C.jordani (Wanner) 1902

C. karkarensi.f Haas and Miller 1952

C. Jcugleri Miller, 1947

C. landanensis (Vincent) 1913

C . leonicensis (de Zigno) 1881

C. macfadyeni Haas and Miller 1952

(\ macro eephalus (Schat'hautl) 1863
('. niarylcondensis Miller and Thompson,

1933
C. molcattamenttis (Foord) 1891
C . negritensis (Olsson) 1928
C.parallelus (Sehafhautl) 1863
C. pernambucensis (Maury) 1930
C. phosphaiiciis (Bede) 1933
C. pHeudohoucliardianum (Spengler) 1910

C. pimlla Haas and Miller, 1952

C.romeroi (Iheriug) 1903

C. saJiariensis (Keller) 1932

C. seelandi (Peneeke) 1884

C. septemcastrensls Haas and Miller. 1952

C. si7idiensi^ (Vredenburg) 1928

Cretaceous

Belgium

(Maestrichtian)

Eocene

Washington

(U.S.A.)

Eocene

British

Somaliland

Eocene

Java

Eocene

England

Cretaceous

Libya

(Campanian-

Danian)

Eocene

British

Somaliland

Paleocene

Trinidad

Paleocene

Cabiuda,

Africa

Eocene

Italy

Eocene

British

Somaliland

Eocene

Bavaria

Eocene

Maryland

(U.S.A.)

Eocene

Egypt

Eocene

Peru

Eocene

Bavaria

Paleocene f

Brazil

Eocene

Morocco

Cretaceous

India

(Upper)

Eocene

British

Somaliland

Cretaceous

Argentina

(Upper) ?

Eocene 1

French

Sudan

Eocene

Austria

Eocene

British

Somaliland

Tertiary

Pakistan

(Lower)

452

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Species

Stratigr

aphic CTeographic

Distribi

Litiou Distribution

Paleoeene

Mississippi

(U.S.A.),

Trinidad

Paleocene

French

Sudan

Paleoeene

Senegal,

Africa

Jurassic

Libya

(Upper)

Paleocene

Texas,

Alabama

(U.S.A.)

Paleoeene

Mississippi

(U.S.A.)

Eocene

Sonialiland

C. suhrecta Miller and Thompson, 1933

C. sudanensis Miller, 1951
C. tessieri Miller, 1951
C. turcicus (Krumbeck) 1905
C. vaughani (Gardner) 1923

C. vestali Miller and Thompson, 1933
C.wyllei (Newton) 1925

Genus AXGULTTHES Montfort, 1808

Angulithes Montfort, 1808, p. 7.
Anguliihes Spath, 1927a, pp. 21, 24.
Deltoidonautilus Spath, 1927a, pp. 22, 26.
Deltoido nautilus Spath, 1927b, pp. 427, 428.
Deltoidonautilus Stenzel 1940, p. 759.
Deltoidonautilus Teichert, 1947b.
Deltoidonautilus, Miller, 1947, pp. 64-67.
Deltoidonautilus Miller, 1951, p. 51
Deltoidonautilus Haas and Miller, 1952, pp. 338-344.
Angidithes Kummel, 1953a, p. 8.

Type species. Nautilus triangularis Montfort, 1802 (by original designation
in Montfort, 1808, p. 7).

The generic name Angulithes remained unused by paleontol-
ogists until Spath (1927a) revived it and placed the genus in
the family Nautilidae. »Spath (1927a, p. 21) listed as genoleeto-
type 'Nautilites' triangularis, Montfort, in d'Orbigny, 1810.
plate 12. However, Montfort (1808, p. 7) specifically designated
as type of his genus 'NantiUtc triangulaire du Havre' (Histoire
des Mollusques, Buffon, edition de Sonniui, tom. 4, pag. 292,
planche 49, 1802). Montfort 's illustration of the type of Angu-
lithes triangularis (1802, pi. 49) is reproduced here on Plate 25.
figure 2. His 1808 illustration of this species is reproduced on
Plate 25, figure 1. Both of these illustrations clearly show the
triangular outline of the whorl section, the degree of involution,

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 453

and the subcentral position of the siphuncle. The sutures on
the type illustration (PI. 25, fig. 2) are undoubtedly oversimpli-
fied and wrong. Many specimens of this species from France
have been illustrated and recorded ; they all have a sinuous
suture with a wide lateral lobe and a small saddle near the
umbilicus. D'Orbigny (1840, p. 79) described and illustrated
Nautilus triangularis Montfort including in the synonymy both
of Montfort 's 1802 and 1808 citations. He likewise listed numer-
ous localities where X. triangularis had been found. It is not
known whether or not d'Orbigny had access to Montfort 's type.

The name combination of Angulithes triangularis Montfort
(1808) has been validly introduced. There may be a question
as to whether the species could be recognized from Montfort 's
text and illustrations. Even though the suture line as repre-
sented on Montfort 's (1802) illustration of Angulithes triangu-
laris is undoubtedly oversimplified I believe that the species is
recognizable.

At the same time that Spath (1927a) revived the name An-
gulithes, he established a new genus Deltoidonautilus, type
species. Nautilus sourrhyi Wetherell, 1836, in J. de C. Sowerby,
1843, p. 35, pi. 627, figs. 1-3. Angulithes Avas thought to be con-
fined to the Cretaceous and Deltoidonautilus to the Cenozoic.
While Spath (1927a, p. 26) did not diagnose his new genus, he
made the following comment: '' Deltoidonautilus in w^orl shape
resembles the Cretaceous Angulithes, and d 'Archiac, and Foord
(1891, p. 327) had compared Deltoidonautilus deluci (d 'Ar-
chaic) with Angulithes triangularis. The latter, however, with
its less sinuous suture line, is more closely allied, via Ang.
fleuriausianus (d'Orbigny) with the regular Xautilid stock that
produced Pseudocenoceras in the Ci'etaceous, as it had given
rise before to the less specialized Paracenoceras of the callo-
I'iensis type." The degree of variability in conch shape and
sinuosity of the suture are such that the two groups cannot be
distinguished (Figs. 32, 33). There is almost a continuous series
of species, through the Cretaceous to the Eocene, of involute
nautiloids with narrowly rounded to angular venters. Deltoido-
nautilus is placed as a synonym of Angulithes (Kummel, 1953a).

Angidithes can be diagnosed as follows: Conch very involute,
generally compressed, whorl sides slightl}^ convex, strongly con-
vergent, vente* narrowly rounded to angular. Suture moder-

454

BULLETIN: MUSEUM OP COMPARATIVE ZOOLOGY

KUMMEL : PORT-TRIASSIC NAUTILOID GENERA 455

ately sinuous with a narrowly rounded ventral saddle, a broad
lateral lobe, a narrow, rounded saddle, and a small lobe on the
umbilical shoulder. Siphunele small and located near the dor-
sum.

Foord (1891, p. 267) has noted that in the type species — A.
triangularis — the periphery can be alternately rounded and
sharply angular. Angulithes senegalensis (Douville) has a sub-
angular peripherj- on the earlier growth stages but at maturity
the venter is rounded (Miller, 1951, fig. 14).

Angulithes is in many respects transitional to both Cimomia
and Hercoglossa, ditifering mainly in its subtriangular whorl
section. It is open to question whether this generic group repre-
sents a continuous genetic line or trend or whether it may be
composed of numerous adaptive trends with this conch shape
stemming at several times from either Cimomia or Hercoglossa.
The oldest species of Angulithes known to date are A. arcuatus
(Deshayes) from the Albian of France and England and A.
mmiieri (Choffat) from Cenomanian (Cretaceous) of Portugal
(Choffat, 1886) and Peru (Schlag-intweit, 1912). Kummel
(1953a) has recorded an indeterminate species of Angulithes
from the Pawpaw formation, Tarrant County, Texas. The young-
est species is A. hakeri (Teichert, 1947, 1952) from Oligocene
strata of Australia. There are seven species of Angulithes re-
corded from the Cretaceous (Fig. 29). Most of these are known

Fig. 33. Cross sections of the couch of A, Angulithes sowerhyanus
(d'Orbigny) 1840, pi. 16, fig. 2, diameter 270 mm.; B, A. deluci (d'Archiac)
from Miller, 1951, fig. IIA, diameter 128 mm.; C, A. triangularis Moiitfort,
from d'Orbigny, 1840, pi. 12, fig. 2, diameter 200 mm., the genotype of
Angulithes ; D, A. fleurimisianus (d'Orbigny) 1840, pi. 15, fig. 2, diameter
192 mm.; E, A. fleuriausianus (d'Orbigny) from Stoliczka, 1866, pi.
94, fig. la, diameter 100 mm.; F, A. fleuriausianus (d'Orbigny) 1840, pi.
15, fig. 3, height 64 mm.; G, A. soiccrhyl (Wetherell) from Miller 1947.
fig. 13A, height 81 mm.; H, A. sp. (Haas and Miller) 1952, fig. 8B, height
38 mm.; 1, A. westphalicxis (Sehliiter) 1876, pi. 47, fig. 2, diameter ,369
mm.; J, A. tamulious (Kossmat) 1897, pi. 6, fig. 6, height 84 mm.; K, A.
elliotti (Steuzel) from Miller 1947, fig. 13B, height 110 mm.; L, A. chudeavi
(Douville), from Miller 1951, fig. lOA, height 56 mm.; M, A. rogeri (Miller)
1951, fig. 13, height 100 mm.; N, A. molli (Douville), from Miller 1951, fig.
12B, height 72 mm.

456

BULLETIN : :MUSEUM OF COMPARATIVE ZOOLOGY

from Upper Cretaceous formations of Europe. Species are also
known from south India, Peru and Texas (U.S.A.). Tertiary
species of Angulitlies are very widely distributed, being known
from North and South America, Europe, xifrica, Pakistan, and
Australia. Most of the Tertiary species are Eocene in age (Fig.
30).

DISTEIBUTION OF SPECIES OF THE GENUS ANGULITHES

Species

A.aegyptiacus (Foord) 1891

A. arcuatus (Deshayes) in Leymerie, 1842

A. bakeri (Teichert) 1947

A. iiyogorensis (Haas and Miller) 19.52

A.caheni (Miller) 1951

A. calviviontensis (Bede) 1948

A. cassmianiui (Foord and Crick) 1890

A.chudeaiii (Douville) 1920

A. deluci (d'Arehiac) 1854

A. elliotti (Stenzel) 1940

A.fifiuriausianus (d'Orbigny) 1840

A. galea (Fritseh and Sehlonbaeh) 1872

A.haughti (Olsson) 1928

A. hazaraeyisis (Das-Gupta) 1916

A. lamarcMi (Deshayes) 1824
A. lemoinei (Miller) 1951

A.moUi (Douville) 1920

A.munieri (Choffat) 1886

A. parabolicus (Schafhautl) 1863

Stratigraphie

Geographie

Distribution

Eocene

Eg^-pt

Albian

France,

England

Oligocene

Australia

Eocene

British

Somaliland

Eocene

Landana,

West Africa

Eocene

France

Eocene

England

Eocene

French

Sudan

Tertiary

Pakistan

(Lower)

Eocene

Texas

(U.S.A.)

Cretaceous

Europe

Cretaceous

Bohemia

(Upper)

Eocene

Peru

Tertiary

Pakistan

(Lower)

Eocene

France

Eocene

French

Sudan

Eocene

French

West

Africa

Cretaceous

Portugal,

(Cenomaniau)

Peru

Eocene

Bavaria

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

457

Species

Stratigraphic

Geographic

Distribution

Paleocene

Senegal

Eocene

France

Eocene

Senegal

Eocene

British

Sonialiland

Eocene

British

Somaliland

Cretaceous

France

(Upper)

Eocene

England

Eocene

British

SomalDand

Tertiary

Pakistan

(Lower)

Danian

India

Eocene

Togo

Cretaceous

France,

(Upper)

England

Cretaceous

Germany

(Senonian)

A.rogeri (Miller) 1951

A. rollandi (Lej-merie) 1846

A. senegalensis (Douville) 1920

A. singularis (Haas and Miller) 1952

A. somaliens'is (Newton) 1925

A. sowerhyanus (d'Orbigny) 1840

A. sowerbyi (Wetherell) 1836
A.spathi (Haas and Miller) 1952

A. subfleuriausianus (d'Archiac) 1850

A. tamulious (Kossmat) 1897
A. togoensis (Miller) 1951
A. triangularis Moutfort, 1802

A. westphalicus (Schliiter) 1872

Genus HerCOGLOSSA Conrad, 1866

EevGoglossa Conrad, 1866a, p. 101.
Hercoglossa Hyatt, 1883, pp. 270-271.
Enclimatoceras Hyatt, 1883, p. 270.
Hercoglossa Spath, 1927a, pp. 22, 25, 26.
H.ercoglossa Miller and Thompson, 1933, pp. 313-324.
Hercoglossa Stenzel, 1940, p. 743.
Woodrimgia Stenzel, 1940, pp. 753-759.
Hercoglossa Miller, 1947, pp. 49-64.
Woodringia Miller, 1947, pp. 67-69.
Hercoglossa Miller, 1951, pp. 47-51.

Type species. Xautilus orbiculaUtJi Tuomey, 1854, from Paleocene of Ala
bama, designated by Hyatt (1883), neotype designated by Miller and
Thompson (1933). Neotype refigurcd here on Plate 27, figs. 1, 2.

The most recent and comprehensiye study of this genus is
that by Miller and his diagnosis is quoted here (Miller, 1947,
p. 50) :

''Conch nautiliconic in its mode of growth and subglobular
to subdiscoidal in form. Whorls more or less flattened (but
typically very broadly rounded) laterally, rather narrowly

458

BULLETIN: MUSEUM OP COMPARATIVE ZOOLOGY

rounded ventrally, and deeply impressed dorsally. Umbilicus
small; umbilical shoulders rounded. Surface of the test smooth.
Septa fairly numerous, and each suture forms a broad very
broadly rounded ventral saddle, a deep rounded lateral lobe, a

Figure 34

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 459

similar lateral saddle, a shallower rather broad rounded lobe on
or near the umbilical wall, and a broad rounded internal lateral
saddle that extends to a deep rather narrowly rounded dorsal
lobe. Siphuncle small and orthoelioanitic in structure ; its posi-
tion varies considerably in the different species but in no case
is it marginal.'*

Hercoglossa is closely related to Cimomia from Avhich it dif-
fers in shape of the suture, to Angulithes from whicli it differs in
shape of the whorl section and from Aturoidca from which it
differs in shape of suture and i^osition of the siphuncle (Figs.
31, 32, 33, 34). Hercoglossa is derived from Cimomia and in-
corporates a more advanced or sinuous suture.

The genus Woodrmgia was established by Stenzel for W.
splendeyis Stenzel and W. simiensis (Vokes) which have a cimo-
niid type of conch in being subglobular, and a hercoglossid type
of suture except for a shallow lobe on the venter (Fig. 34).
Stenzel ditt'erentiates Woodringia on the depressed aspect of
the whorl section and the presence of a ventral lobe ; he con-
sidered Woodringia to be an independent oft'shoot from some
hercoglossid ancestor. "Whereas hercoglossids have in general
more compressed conchs than cimomids, there is much variation
in the shape of the whorl section (Figure 31). I do not consider
the shape of the conch in this particular case to be of any diag-
nostic value. As to the suture, the presence of the ventral lobe
in the ventral saddle is a modification that has occurred in
other stocks and has been held to be of not more than specific
value. Among species of Cimomia, for example, the ventral sad-
Fig. 34. Diagnunmatie i-epreseiitation of sutures ot" A, Aturoidea paiiei-
frx (Cope) from Miller and Thompson, 1935, pi. 65, fig. 2; B, A. rlrirui
Miller, 1951, fig. 16-A; C, HercogJossu gaidnerue Stenzel, 19-40, lig. 117
(1); I), H. diderrichi Vincent, from Miller, 1951, fig. 9; E, H. diderrich!
Vincent, from Stenzel, 1939, fig. 117 (2) ; F, H. harrisi Miller and Thompson.
1937, from Stenzel, 1940, fig. 117 (8) ; G, H. orbiculata (Tuomey) from
Stenzel, 1940, fig. 117 (3); H, H. splendens (Stenzel) 1940, fig. 121 (1);
/, H. mcf/lamfryae Miller and Thompson, 1933, fig. 2D; ,7, H. maracaiboen-
sis Miller and Collinson, 1951, fig. 2A; E, H. simiensis Vokes, 1937, from
Stenzel, 1940, fig. 121 (2) ; L, H. popenoei Miller and Downs, 1950, fig. 1;
M, E. danicu-s (Schiotheim) from Blanford, 1861, pi. 13, fig. 4. All figures
greatly reduced.

460 BULLETIN : MUSEUM OF COMPARATRT] ZOOLOGY

die is normallj^ well rounded, but in C. kugleri Miller and C.
vaughani (Gardner) a slight median lobe appears at late ma-
turity. Likewise two species of Aturia have ventral lobes whereas
the "normal" state is a rounded ventral saddle. In Atima
herryi Stenzel (1940, p. 764-770, pi. 40, figs. 1-3, text figs. 125
(7) and 126 (2)) the suture is "slightly wavy across the venter
so that there is a very shallow ventral lobe in the center flanked
on either side by a shallow ventral saddle." However, Miller
(1947, p. 88) suggested that the slight ventral lobe may be
adventitious in that he could not discern this lobe in any of
the other sutures. However, there is another species of Aturia,
A. coxi Miller, 1947 (^A. narica Cox 1927, non Vredenburg)
from Lower Miocene of Pemba Island in the Zanzibar Protec-
torate. In this species there is a distinct, rather deep, ventral
lobe. I have had the opportunity of examining this specimen
in the British Museum and agree _ with Cox that it belongs in
Aturia, and I agree with Miller that it is a distinct species.
In each of the above cases we are dealing with merely local
aberrant species.

Another generic group that displays the same type of varia-
tion in the ventral suture is Pseudaganides. In this genus
the "normal" condition is to have a slight lobe on the venter;
however, a few Upper Jurassic species as P. schlosseri (Loesch,
1914, pp. 114-118, fig. 8) develop a rounded ventral saddle on
the last mature sutures where previously a shallow lobe ap-
peared.

Two species of early Cretaceous naiitiloids with sinuous "her-
eoglossid" sutures are very difficult to evaluate. These are
Nautilus ricordeanus d'Orbignj'- and N. forhesianus Blanford.
The former species {N. ricordeafiiis d'Orbigny) was never illus-
trated by its author and only very briefly diagnosed. Subse-
quently Kilian (1915, pp. 17, 18) described and illustrated a
specimen from Aptian (Lower Cretaceous) formations of south-
eastern France which he assigned to N. ricordeanus. The speci-
men is small (44 mm. diameter), involute, compressed, with
rounded venter, and flattened sides. The suture is broadly
arched over the venter and forms a narrowly rounded V-shaped
lobe on the lateral sides. The remaining portion of the suture is
not visible in Kilian 's illustration. Nautilus forhesianus is from
Albian strata of south India (Blanford, 1861) and is likewise

KUMMEL : POST-TRIASSIC NAUTILOID GENERA

461

an involute, compressed form with an arched venter and the
suture has a broad, nearly straight course across the venter and
a V-shaped lateral lobe. The character of the suture of these
two species suggests affinity to Aturoidea more than to Herco-
glossa. However, the position of the siphuncle is nearer the
venter rather than nearer the dorsum as in Aturoidea. Cre-
taceous species of Hercoglossa and Aturoidea are indeed very
rare and any interpretation of these two species can only be
tentative pending discovery of new material. It seems that in
spite of the apparent specialization of the lateral lobe it
would be best to ally these two species, with question, in
Hercoglossa. The derivation of Aturoides from Hercoglossa has
been a long-accepted, and to me, a correct interpretation. The
oldest species of Aturoidea is A. serpentinus (Blanford) from
Senonian strata of south India. Nautilus ricordeanus and N.
forhesianus are thus here considered as part of the earliest
radiation from Cimomia but perhaps are not part of the main
evolutionary line but still congeneric with the principal group.
The species assigned to Hercoglossa with their time and geo-
graphic range are listed below. The geographic distribution is
noted on the maps of Figures 29, 30. Teichert and Glenister
(1952, p. 737) record the presence of an undescribed new cari-
nate species of Hercoglossa from Eocene strata from the north-
west basin of Western Australia.

DISTRIBUTION OF SPECIES OF THE GENUS HERCOGLOSSA

Species
H. danicus (Schlotheim) 1820

H.diderrichi Vincent, 1913

H.? forhesianus (Blanford) 1861

H. gardnerae Stenzel, 1940

ff. harrisi Miller and Thompson, 1937
H. imiominanda Fleming, 1945

H. lamegoi Oliveira, 1953

H. madagascariensis Collignon, 1951

Stratigraphic

Geographic

Distribution

Danian

Europe, Asia,

Africa

Paleocene

Senegal

Cretaceous

South

(Albian)

India

Paleocene

Texas

(U.S.A.)

Paleocene

Trinidad

Eocene

New

Zealand

Paleocene

Brazil

Cretaceous

Madagascar

(Uper)

462

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Species

Stratigrapliic Geographic
Distribution Distribution

H.mcglameryae Miller and Thompson, 1933 Paleocene
H. merriami Dickerson, 1914
H. orbieulata (Tuomey) 1854

H. pavlowi (Arkhanguelsky) 1904

H. peruviana Berry, 1923

H. popenoei Miller and Downs, 1950

HJ ricordeanus (d'Orbigny) 1847

H. simiensis Yokes, 1937

H. splendens (Stenzel) 1940

H. tuomeyi Clark and Martin, 1901

H.ulriehi (White) 1882

H. u-alteri Miller, 1947

n. warinai .Miller, 1947

Genus AturOIDEA Vredenburg, 1925
Aturoidea Vredenburg, 1925, p. 9.
Paraturia Spath, 1927a, pp. 22, 26.
Paraturia Spath, 1927b, p. 428.
Aturoidea Vredenburg and Cotter, 1928, p. 18.
Aturoidea Sehenck, 1931, p. 450.
Aturoidea Miller and Thompson, 1935, pp. 563-571.
Aturoidea Teiehert, 1943, pp. 258-261.
Aturoidea Haas, 1947, pp. 243, 244.
Aturoidea Miller, 1947, pp. 69-77.
Aturoidea Miller, 1951, p. 68.
Type species. Nautilus parhinsoni Edwards (subsequent designation by

Miller and Thompson 1935, p. 563). Figured here on Plate 28, figures

1, 2.

Paleocene

Alabama

(U.S.A.)

Paleocene

California

(U.S.A.)

Paleocene

Alabama

(U.S.A.)

Paleocene

Russia

Eocene

Peru

Paleocene

Baja,

California

(U.S.A.)

Cretaceous

France

(Aptian)

Paleocene

California

(U.S.A.)

Paleocene

Texas

(U.S.A.)

Eocene

Maryland,

Virginia

(U.S.A.)

Paleocene

Ark., Tenn.,

Ala., Miss.,

Texas

(U.S.A.)

Paleocene

Alabama

(U.S.A.)

Eocene

Trinidad

KUilMEL : POST-TRIASSIC XAUTILOID GENERA 46o

Miller (1947) has recently thoroughly reviewed the available
data on this interesting genus and I find myself in essential
agreement with most of his conclusions. His diagnosis of
Aturoidea is quoted here (Miller 1947, p. 70) :

"Conch sublenticular in shape and nautiliconic in its mode
of growth; all known forms are large. "Whorls compressed,
flattened laterally, rounded ventrally, impressed dorsally. Um-
bilicus closed ; umbilical shoulders rounded. Septa asymmetri-
cally sigmoidal ; each mature suture forms a very broad high
blunt ventral saddle, and on either side of it a deep narrow
asymmetrical narrowly rounded lateral lobe, a broad high
broadly rounded asymmetrical lateral saddle, a broad rounded
lobe with its center near the umbilical seam, and a broad rounded
saddle located on the side of the impressed zone and extending
to the large rounded V-shaped dorsal lobe. Siphuncle sub-
central or subdorsal in position (being nearer the dorsum than
the venter, but not in contact with the dorsum), and ortho-
choanitic in structure — the siphuncular segments are essentially
cylindrical in shape and the septal necks are relatively long,
being only slightly shorter than the connecting rings."

Only 11 species of Aturoidea have been recorded to date and
each species is known by one or very few specimens only, a fact
unfortunately applicable to most post-Triassic nautiloids. As
pointed out by ^Miller on various occasions Ahiroidea is very
closelj' related to Hercoglossa and Aiuria; in fact, it is more or
less intermediate between them. It ditfers from Hercoglossa in
the position of the siphuncle and in the character of the suture.
In Aiuria the dorsal (marginal) position of the siphuncle and
the infundibular flexures of the septa are distinctive. The main
variations in the suture are in the width and shape of the lateral
lobe.

In A. serpent inus and A. sclvweinfurfhi, both of Upper Cre-
taceous age, the lobe is well rounded and rather broad. However,
in A. fieirai from strata of questionable Upper Cretaceous age
the ventral flank of the lateral lobe has a distinctive flexure in
the lower part making the tip of the lobe narrowly rounded.
This same type of modification of the lateral lobe is present in
.4. matheivsonii of Paleoeene ? age. in A. paucifex of Eocene
age, in A. pilsbryi of Eocene age, and in A. parkinsoni, of

464

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

Eocene age. The lateral lobe is very narrow and pointed in A.
distans from the Eocene of Australia.

There are too few specimens or species of Aturoidea known
to recognize any distinct evolutionary patterns within the group.
Species of Aturoidea are known from formations of Cretaceous,
Paleocene, and Eocene age and have been recorded from Eng-
land, Austria, India, Africa, Australia, Peru, California, and
New Jersey (U.S.A.) (Figs. 29, 30). Teichert and Glenister
(1952) report undescribed specimens of Aturoidea from Upper
Maestrichtian age from western Australia.

DISTEIBUTION OF SPECIES OF THE GENUS ATUROIDEA

Strati^-aphic

Geographic

Species

Distribution

A. brunleehneri (Frauscher) 1895

Eocene

Austria

A. distans Teichert, 1943

Eocene

Australia

A. matheiosonii (Gabb) 1864

Paleocene

California

(U.S.A.)

A. olssoni Miller, 1947

Eocene

Peru

A.parhinsoni (Edwards) 1849

Eocene

England

A. paucifex (Cope) 1866

Eocene

New Jersey
(U.S.A.)

A.pilsbryi Miller and Thompson, 1935

Eocene

New Jersey

(U.S.A.)

A. schweinfurthi (Quaas) 1902

Cretaceous

(Campanian-

Maest.)

Libya

A.serpentinus (Blanford)

Cretaceous
(Senonian)

India

A.spathi (Vredenburg) 1928

Eocene

India

A. vieirai Miller, 1951

Cretaceous
(Upper?)

Angola

Subfamily ATURINAE Hyatt, 1894

Genus AtueiA Bronn, 1838

Type species. Naidilus aturi Basterot, by subsequent designation, Herrmann-
sen, 1846, and by virtual tautonjnny.
The genus Aturia, the sole member of the Aturinae, has re-
ceived more intensive study than any other group of post-Trias-

KUM]\TEI- : PO?^T-TRTASRIC NAUTILOID GENERA

465

466

BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

sic nautiloids. The most comprehensive discussion is that of
Schenck (1931), and recently Miller (1947) has described the
specimens of Aturia known from the Americas. Schenck (1931,
pp. 448, 449) gives a complete synonymy of the genus and this
need not be repeated here. Since Schenck 's paper, Stenzel
(1935, pp. 553-556) has proposed two new subgenera, A. (Nilatu-
ria) and A. (Brazatiiria) ; however, I agree with Miller (1947,
p. 81) that these subgeneric categories should not be used. Pre-
viously Ihering (1921, p. 76) had proposed the subgenus Aturia
(Sphenaturia) but this name is also suppressed (Schenck, 1931,
p. 450; Miller, 1947, p. 81).

This genus is characterized mainly by the infundibular adapi-
cal flexure of the septa around the dorsal marginal siphuncle.
The conch is usually more compressed and the suture more ad-
vanced than in other Tertiary nautiloids. There is very good
ontogenetic evidence of the origin of Aturia from Hercoglossa
through Aturoidea (Miller 1947, pp. 78, 79). The genus has
been so thoroughly discussed by Schenck (1931) and Miller
(1947) that I have nothing of significance to add.

Aturia is truly a cosmopolitan genus, nearl}^ world-wide in
distribution and ranges in age from lowest Paleocene (possibly
Upper Cretaceous) well into the Miocene (Fig. 35). The named
units, "species," of Aturia recorded so far are listed below.

DISTEIBUTION OF SPECIES OF THE GENUS ATURIA

Species

Stratigraphic- Geographic
Distribution Distributiou

A. aJabamensis (Morton) 1834

A.alaskensi.'i Schenck, 19ol

A. angustata (Conrad) 1849

A.atiiri (Basterot) 1825

Eocene

Atlantic

and Gulf

Coastal

Plain;

Mexico

Ollgoceue

Gulf of

Alaska

Oligocene-

Calif. ;

Miocene

Wash. ;

Oregon

(U.S.A.)

Mioceni'

France, Java

KUIMMEL : POST-TRIASSIC NAUTILOID GENERA

467

Species

Stratigraphic

Geographic

Distribution

Miocene

Soutlieast

Australia

Oligocene

France

Oligocene

Mississippi

(U.S.A.)

Eocene

Texas

ru.s.A.)

Eocene ?

Chile

(Tierradel

Fuego)

Eocene ?

Argentina

Eocene

England

Miocene

Western

Australia

Miocene

Pemba

Island,

Zanzibar

Oligocene ?

Cuba

Miocene

Florida

(U.S.A.).

Trinidad,

Venezuela,

Ecuador

Paleocene ?

California

(U.S.A.)

Eocene ?

Chile

Eocene

Italy

Eocene

Louisiana

(U.S.A.)

Eocene

Washington

Oligocene

(U.S.A.)

Miocene

Xew Zealand

Eocene

California

(U.S.A.)

Eocene

Germany

Eocene

Texas

(U.S.A.)

Eocene

Belgium

Miocene ?

S. Africa

Miocene

Angola

A. australis McCoy, 1867

A. basteroti Beuoist, 1888
A. ierryi Stenzel. 1940

A. hrazoensis Stenzel, 1935

A. hriiggeni Ihering, 1921

A. caroliameghinoi (Ihering) 1902
A. charlesworthi Foord, 1891
A. clnrkei Teichert, 1944

A. coxi Miller, 1947

A.cubaensis (Lea) 1841

A. Gurvilineafa Miller and Thompson, 1937 Miocene

A. dicTcersoni Schenek, 1931

A. felschi Ihering, 1921
A. formae Parona, 1899
A. garretti Stenzel, 1940

A. grandior Schenek, 1931

A. grangei Fleming, 1945

A . Tcerniana Anderson and Hanna, 1925

A.Tcoenei Gagel, 1928
A. laticlavia Stenzel, 1935

AJ linicentensis Vincent, 1907
Atitria lotzi Bohm, 1913
A. luculoevsiK Miller, 1938

468

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

Species

Stratigraphic

Geographic

Distribution

Eocene

New Zealand

Miocene

Italy

Eocene

California

(U.S.A.)

Oligocene

Pakistan

Eocene

Panama

Oligocene

Italy

Eocene,

Peru,

Oligocene

Ecuador,

Colombia,

Venezuela

Paleocene

Egypt

Eocene

Egypt

Miocene

Italy

Oligocene

Mississippi

(U.S.A.)

Eocene

Italy,

Hungary,

EgJTJt,

Palestine

Eocene

Somaliland

Miocene

Japan

Eocene

Texas

(U.S.A.)

Eocene

Texas

(U.S.A.)

Eocene

New Jersey

(U.S.A.)

Eocene

Japan

Eocene

England

A. mackayi Fleming, 1945
A. morrissi Michelotti, 1847
A. myrlae Hauna, 1927

A. narica Vredenburg, 1925
A. panamensis Miller, 1947
A. paronai Eovereto, 1900
A. peruviana Olsson, 1928

A. praeziczac Oppenheim, 1903
A. preaturi (Cuvillier) 1935
A.radiata Bellardi, 1872
A. ricliardsi Miller, 1947

A. rovasendiana Parona, 1899

A. somaliensis Haas and Miller, 1952
A. tolcunagai SMmizu, 1926
A. triangula Stenzel, 1935

A. turneri Stenzel, 1940

A. vaniLxemi (Conrad) 1847

A. yoJcoyamai Nagao, 1926
A.ziczac (Sowerby) 1812

ADDENDUM

The very stimulating monograph by Benavides (1956, Cretaceous System
in northern Peru. Am. Mus. Nat. Hist., Bull. 108, pp. 357-493, pis. 31-66)
was published when this paper was in page proof. The new genus of nauti-
loid proposed by Benavides, Lissoniceras (type, Xantilns mermeti Coquand)
is clearly a synonj-m of Angulithes as interpreted in this report.

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 469

BIBLIOGRAPHY

Agnesotti, Alda

1937. Su alcune forme di Xautilidi della Sirtica (Libia). Natura, vol.
28, fase. 3, pp. 109-123.

Albekici, Ekminia

1940. Contributo alia conoscenza dei Cefalopodi cretacici della Libia.
Ann. Mus. Lib. Stor. Xat., Tripoli, vol. 2, pp. 165-188, pi. 19.
Republished in 1st. di Geol., Paleont., e Geog. Fisica della R.
Univ. di Milano, Ser. P, Pub. 24, pp. 165-188, pi. 19.
Aldeich, T. H.

1931. Description of a few Alabama Eocene species and remarks on
varieties. Ala. Geol. Survey, ;Mus. Paper 12, pp. 1-21, pis. 1-6.
Alth, Alois

1850. Geognostisch-palaontologische Beschreibung der nachsten Umge-
bung von Lemberg. Xatur. Abh. Haidinger, vol. 3, pp. 172-248,
pis. 9-13.

Anderson, F. M.

1902. Cretaceous deposits of the Pacific Coast. California Acad. Sci.
Proc, ser. 3, Geology, vol. 2, no. 1, pp. 1-154, pis. 1-12.

1938. Lower Cretaceous deposits in California and Oregon. Geol. Soc.
America Spec. Paper 16, pp. 1-339, pis. 1-84.

, and G. D. Hanna

1925. Fauna and stratigraphic relations of the Tejon Eocene at the
type locality in Kern County, California. Calif. Aead. Sci., Occ.
Papers, vol. 11, pp. 1-249, pis. 1-16.

Aechiac, A.d ', and J. Haime

1853- Description des animaux fossiles du groupe nummuitique de
1855. I'Inde, precedee d'un resume geologique et d'une monogi-aphie
des nummulites. Paris, 2 vol., 373 pp., 36 pis.
Abkhangxjelsky, a. D.

1904. Depots paleocSnes de la region volgienne du gouvemement de
Saratov et leur faune. Materialien zur Geolgie Russlands, vol.
22, Lief. 1, pp. 1-207, pis. 1-12.

Aenould-Saget, S.

1953. Les ammonites pyriteuses du Tithonique Superieur et du Ber-
riasien de Tunisie centrale. Annales des Mines et de la geologie
(Rggence de Tunis) no. 10, pp. 1-132, pis. 1-11.

Avnimelech, Moshe

1943. Lower Tertiary Aturia from Palestine. Jour. Paleont., vol. 17,
pp. 115-119, 5 text figs.

470 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

1946. On Nautihis lihanoticus Poord and Crick, in the Senonian of

Palestine. Ann. Mag. Nat. Hist., 11th ser.. vol. 12, no. 92, pp.

523-528.
1947a. Supplement to the note on Xaiitilus lihanoticus Foord and Crick.

Ann. Mag. Xat. Hist., 11th ser., vol. 13, no. 106, p. 690, pi. 15.
1947b. Nautilus blaTcei, a new Oligocene mollusk from Palestine. Jour.

Paleont., vol. 21, pp. 476-479, pis. 70 71.
Basse, Eliane

1931. Monographie paleontologique du Cretace de la province de

Maintirano, Madagascar. Tananarive, Service des Miaes, Gou-

vernement General de Madagascar et Dependances, pp. 1-86, pis.

1 13.

Basterot, B. de.

1825. Description geologique du bassin tertiaire du sud-ouest de la
France, premiere partie comprenant les observations generales
sur les Mollusques fossiles, et la description particuliere de ceux
qu 'on rencontre dans ce bassin. Paris Soc. Hist. Nat., Mem., vol.
2, pp. 1-100, pis. 1-7.

Bede, p.

1933. Sur uue nouvelle espece de Nautile des terrains phosphates du

Sud Tunisien. Bull. Soc. Sci. Nat. Maroc, vol. 12, nos. 7-8,

pp. 281-287, 2 figs., (1932).
1948. Sur une nouvelle espece de Nautile, Nautilus calvimontensis nob.,

du Lutetien des environs de Paris. Soc. Geol. France, Bull. ser. 5,

vol. 18, pp. 219 225, pi. 9a.

Bellabdi, Luigi

1872. I Molluschi dei terreni terziari del Piemonte e della Liguria.
Pt. 1, pp. 1-264, pis. 1-15.

Benoist, E. a.

1888. Coquilles fossiles des terrains Tertiaires moyens du sud-ouest
de la France, description des ccphalopodes, pteropodes, et gastro
podes opisthobranches (Acteonidae). Actes Soc. Linneenne
Bordeaux, vol. 42, pp. 11-84, pis. 1-5.

Berry, E. W.

1923. A new Hei coglossa from the Eocene of Peru. Am. Jour. Sci., 5th
ser., vol. 6, pp. 427-431.

Binckhorst, J. T.

1861. Monographie des gasteropodes et des cephalopodes de la craie
superieure du Limbourg. Bruxelles et Maastricht, 1861. pp.
1-83, pis. 1-6; pp. 1-44, pis. 5b-d, 7, 8, 8a, 9 (Reprinted in 1873 j.

KUMMEL : POST-TRIASSIC XAUTILOID GENERA 471

Blanford, H. F. and F. Stoliczka

1861- The fossil Cephalopoda of the Cretaceous rocks of southern India ;

1866. Belemnitidae — Xautilidae (by H. F. Blanford). Aimnonitidae.

with a revision of the Xautilidae, etc. (by F. Stoliczka). Mem.

Geol. Surv. India (Palaeont. Indica), ser. 1, 3, vol. 1, pp. 1-126,

i-xii app., pis. 1-94.

BOHM, JOH.^NXES

1929. Eozane und miozane Versteinerungen aus Angola. Zeitschr.
Deutsche Geol. Gesell., vol. 81, pp. 450-458, pis. 15, 16.
BouLE, M., P. Lemoixe, and A. Thevenix

1907. Paleontologie de Madagascar, III Cephalopodes Cretaces d'en
virons de Diego Suarez. Ann. de Pal., vol. 2, fasc. 1.
BuacKHARDT, Carlos

1925. Faunas del Aptiano de Nazas (Durango). Inst. geol. Mexico,
Bol. 45, pp. 1-71, pis. 1-10.

BCVIGNIER, AilAXD

1852. Statistique geologique, mineralogique, minerallurgique et paleou
tologique du departement de la Meuse. Paris, 694 pp.. 32 plates
(in Atlas).

(3ASTILL0, A. DEL, and J. G. Agctlera

1895. Fauna fosil de la Sierra de Catorce, San Luis Potosi. Bol. de la
Comision Geologica de Mexico, I, pp. i-ix, 1-55, pis. 1-24.
Chapuis, M. F., and M. G. Dewalqlt;

1853. Descriptions des Fossiles des Terrains Secondaires de la Province
de Luxembourg. Bruxelles, pp. 1-303, pis. 1-38.

Chapmax, Frederick

1915. New or little known Victorian fossils in the National Museum,
Part XVII — Some Tertiary Cephalopoda. Roy. Soc. Victoria,
Proc, n. s., vol. 27, pp. 350-361, pis. 3-8.
1921. The specific name of the Australian Atiiria and its distribution.
Roy. Soc. Victoria Proc. n. s., vol. 34, pp. 12-16.
Choffat, Paul

1886. Recueil d 'etudes paleontologiques sur la faune Cretacique du
Portugal, vol. 1. Espeees nouvelles ou peu connues. Premiere
serie, pp. 1-40, pis. 1-lS.
Clark, W. B., and G. C. Martix

1901. Mollusca (of the Eocene of Maryland). Md. Geol. Survey.
Eocene, pp. 122-203, pis. 17-57.

Cdlligxox, M.

1949. Recherches sur les faunes Albiennes de Madagascar. I. L 'Albieii
d'Ambarimaninga. Ann. Geol. Serv. Mines, Madagascar, fasc.
no. 16, pp. 1-128, pis. 1-22.

472 BULLETIN: MUSEUM OF COMPAEATIVE ZOOLOGY

1951a. Beeherches sur les faunes Albiermes de Madagascar. V. L'Albien

superieur d 'Andranof otsky (Cercle de Manja). Arm. Geol. Serv.

Mines, Madagascar, fase. 19, pp. 7-40, pis. 1-7.
1951b. Le Cretace superieur d'Antonibe. Couches de passage du Cretace

au Tertiaire. Ann. Geol. Serv. Mines, Madagascar, fasc. 19, pp.

75-148, pis. 1-8.

CONEAD, T. A.

1847. Observations on the Eocene formation, and descriptions of one
hundred and five new fossils of that period, from the vicinity of
Vicksburg, Mississippi; with an Appendix. Acad. Xat. Sci.
Philadelphia, Proc, vol. 3, pp. 280-299.

1866a. Observations on Recent and fossil shells, with proposed new

genera and species. Am. Jour. Conch., vol. 2, pp. 101-103.

1866b. Check List of the Invertebrate Fossils of Xorth America; Eocene
and Oligoeene. Smithson. Misc. Coll., vol. 7, pub. 200, pp.
1-41.

1868. In Cook, G. H.: Synopsis of the Invertebrate Fossils (of the Cre-
taceous and Eocene of New Jersey). X. J. Geol. Survey. Ap-
pendix A, pp. 721-732.

Cope, E. D.

1866. An Aturia from the Cretaceous of X'^ew Jersey. Acad. Xat. Sci.
Philadelphia, Proc, vol. 18, pp. 3-4.

COSSMAN, M.

1901. Additions a la faune X'ummulitique d 'Egypte. Bull. Inst.
Egypte, ser. 4, no. 1, pp. 173-197, pis. 1-3.

1902. Catalogue niustre des coquUles fossiles de L 'Eocene des environs
de Paris, Appendice 3. Ann. Soc. Roy. Malacol. Belgique, vol.
36, pp. 9-110, pis. 1-7.

, G. PissARRO, and E. W. Vredexbueg

1909. The Mollusca of the Ranikot series; part I. — Cephalopoda and
Gastropoda. Mem. Geol. Surv. India (Palaeont. Indica), n.s.
vol. 3. no. 1, pp. 1-83, pis. 1-8.

Cox, L. R.

1927. Xeogene and Quaternary Mollusca from the Zanzibar Protector-
ate. Rept. on Palaeont. Zanzibar Protectorate, pub. by author-
ity of Govt, of Zanzibar, pp. 13-102, pis. 3-19.

1940. Cretaceous Mollusca described by R. Pulteney in the Second
Edition of Hutchins' History of Dorset (1813). Proc. Malaco-
logical Soc. London, vol. 24, pt. 3, pp. 121-128, pi. 7.

Ceagin, F. W.

1905. Paleontology of the Malone Jurassic formation of Texas.
L.S.G.S. Bull. 266, pp. 1-172, pis. 1-29 (incl. map).

KUMMEL : POST-TRIASSIC XAUTILOID GENERA 473

Crick, G. C.

1894. On a collection of Jurassic Cephalopoda from Western Australia.
Geol. Mag., n.s., (4), vol. 1, pp. 385-393, pi. 12.

1898. Descriptions of nevr or imperfectly kno-wn species of Xautilu^
from the Inferior Oolite, preserved in the British Museum (Nat-
ural History). Proc. Malacological Soc. London, vol. 3, pp. 117-
139.

1899. Note on Ammonites calcnr, Zieten. Geol. Mag., (4), vol. 6.
pp. 554-558, 2 figs.

1907. The Cephalopoda from the deposit at the north end of False Bay,
Zululand. Cretaceous Fossils of Natal, Pt. 3, No. 2, Third Eept.
Geol. Surv. Natal and Zululand, pp. 163-234, pis. 10-15.

1908. In Newton and Crick: On some Jurassic MoUusca from Arabia.
Ann. Mag. Nat. Hist., (8), vol. 2, pp. 10-24, pis. 2-3.

1915. Note on Xautilus vioTcattamensis, A. H. Foord, from the Eocene
of Egypt. Proc. Malacological Soe. London, vol. 11, pp. 286-290,
pi. 8.

1921. On Xautilus pseudotruncutus n.sp. from the Liassic rocks of
England. Proc. Cotteswold Natur. Field Qub (1920), vol. 20, pt.
3, pp. 245-247.

CUVILLIER, J.

1935. Etude complementaire sur la paleontologie du nummulitique
egyptien (prem. partie). Inst. Egypte Mem., vol. 28, pp. 1-81,
pis. 1-5.
Dacque, Edgak

1905. Beitrage zur Geologie des Somalilandes ; II Teil, Oberer Jura.
Beitr. Palaont. u. Geol. Osterr.-Ung. u. Orients, vol. 17, pp.
119-159, pis. 14-18.

1910. Dogger und Malm aus Ostafrika. Beitr. Palaont. u. Geol. Osterr.-
Ung. u. Orients, vol. 23, pp. 1-62, pis. 1-6.

Dana, J. D.

1849. U. S. Exploring Expeditions; Geology. 756 pp., Atlas, Ne'w
York.

Daquin, F.

1932. A propos de la signification stratigraphique de Xautilus danicus
Schlotheim. Actes Soe. Linn. Bordeaux, vol. 84, Extr. Proc. -verb.,
pp. 98-101, pis. 1, 2.
Das-Gupta, H. C.

1916. Palaeontological notes from Hazara. Jour. Proc. Asiatic Soc.
Bengal, n.s., vol. 11, pp. 253-257, pi. 16.

Deshayes, G. p.

1824. Description des coquilles fossiles des environs de Paris. Vol. 2,
Mollusques, pp. 1-814, pis. 1-101.

474 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

DiCKERSON, R. E.

1914. Fauna of the Martinez Eocene of California. Cn1if. T^niv. Pub.,
Dept. GeoL, Bull., vol. 8, pp. 61-180, pis. 6-18.

DiENER, Carl

1897. The Cephalopoda of the Lower Trias. India (reel. Survey (Palae
out. Indica) ser. 15, vol. 2, pp. 1-181, pis. 1-23.

Douville', Henri

1916. Les terraines secondaires dans le Massif du Moghara a I'est de

1 'Isthme de Suez, d'apres les explorations de M. Cougat-Bar-

thoux. Paleontologie. Mem. Acad. Sci. Inst. France, ser. 2, vol.

54, pp. 1-184, pis. 1 21.
1920. L 'Eocene au Soudan et au Senegal. Bull. Com. Etudes Hist. Sci.

Afrique Occidentales Fran^aises, no. 2, pp. 113-171, pis. 1-5.

1928. Les couches a Cardita beaumonti. Mem. Geol. Surv. India
(Palaeont. Indica), n.s., vol. 10, mem. 3, fase. 1, pp. 1-25, pis. 14.

1929. Les couches a Cardita heaummitl dans le Sud. Mem. Geol. Surv.
India. (Palaeont. Indica), n.s., vol. 10, mem. 3, fasc. 2, pp. 27-73,
pis. 5-11.

1933. Conlributions a la geologie de I'Angola; le Tertiaire de Loanda.
Bol. Mus. Lab. Min. Geol., Lisbon Univ., vol. 1, pp. 63-118, pis.
17

DUMORTIER, E.

1864- Etudes paleontologiques sur les depots jurassiques du bassin
1874 du Rhone. Pt. 1, Infras. Lias, 187 pp., 30 pis., 1864; Pt. 2, Lias-
Inferieur, 252 pp., 50 pis., 1867; Pt. 3, Lias-Moyen, 348 pp., 45
pis., 1869; Pt. 4, Lias-Superieur, 335 pp., 62 pis. 1874.

Durham, J. W.

1940. Aturia in the Upper Miocene of Java. Jour. Paleont., vol. 14,

pp. 160-161.
1946. Upper Aptian nautiloids from Colombia. Jour. Paleont., vol. 20,

pp. 428-434, pis. 62-65, 5 text figs.

Edwards, F. E.

1849. A monograph of the Eocene Mollusea; or Descriptions of shells
from the older Tertiaries of England, Part 1, Cephalopoda.
Palaeontographical Soc, vol. 2, pp. 1-56, pis. 1-9.

Htheridge, R., Jr.

1901. Additional notes on the paleontology of Queensland, Pt. 2.
Queensland Geol. Surv., Bull. 13, pp. 1-37, pis. 1-4.

Favre, Ernest

1869. Description des MoUusques Fossiles de la Craie des Environs
de Lemberg en Galicie. 187 pp., 13 pis.. Geneve et Bale.

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 475

Feruglio, E.

1936. Palaeontographia Patagonica : 1, Titoniano e Inf racretaceo della
Eegione del Lago Argentino. 2, Sopracretaceo della Eegione del
Lago Argentine. Mem. Inst. Geol. E. Univ. Padova, vol. 11.
pp. 1-192, pis. 1-20.

FiTTON, W. H.

1835. Ob.servations on some of the strata between the chalk and the

Oxford oolite in the southeast of England. Trans. Geol. Soe.

London, (2) vol. 4, pp. 103-388 (app. A. — ^ Descriptive notes

respecting the shells figured in pis. 11-23 by J. de C. Sowerby,

pp. 33.5-348).

Fleming, C. A.

1945. Some New Zealand Tertiary cephalopods. Trans. Eoy. Soc. New
Zealand, vol. 74, pp. 411-418, pis. 60-62.
Flower, E. H., and B. Kummel

1950. Classification of the Nautiloidea. Jour. Paleont., vol. 24, no. 5,
pp. 604-616.

FooRD, A. H.

1891. Catalogue of the fossil Cephalopoda in the British Museum
(Natural History), part 2, containing the remainder of the
suborder Nautiloidea, consisting of the families Lituitidae,
Trochoceratidae, and Nautilidae with a supplement. London,
pp. 1-407.
, and G. C. Crick

1890a. A revision of the group of Nautilus elegans, J. Sowerby. Geol.
Mag., U.S., decade III, vol. 7, pp. 542-552'.

1890b. Description of new and imperfectly-defined species of Jurassic
Nautili contained in the British Museum (Natural History).
Ann. Mag. Nat. Hist., 6th ser., vol. 5, pp. 265-291, figs. 1-18.

1890c. On some new and imperfectly defined species of Jurassic, Cre-
taceous, and Tertiarj' Nautili contained in the British Museum
(Natural History"). Ann. Mag. Nat. Hist., 6th ser., vol. 5, pp.
388-409.

Forbes, Edward

1846. Eeport on the fossil Invertebrata from southern India, collected
by Mr. Kaye and Mr. Cunliffe. Geol. Soc. London, Trans., 2nd
ser., vol. 7, pp. 97-174, pis. 7-19.
Frauscher, K. F.

1895. Nautilusse von Guttaring. Karnten Naturhist. Landes-Museums,
Jahrbuch, Heft 23, pp. 185-199, pis. 1-2.

Fritsch, a., and U. Schloxbach

1872. Cephalopoden der bohmisehen Kreideformation. Prag, pp. 1-52,
pis. 1-16.

476 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

FuciNi, Alberto

1895. Fauna dei calcari bianchi ceroid! con Phylloceras cylindrioum

Sow. sp. del Monte Pisano. Atti Soc. Tose. Sci. Nat. Pisa, Mem,,

vol. 14, pp. 125-351, pis. 6-13.
1919. Cefalopodi (Cretacei della Tripolitania). Carta Geol. Ital., vol.

18, Itin. geol. Tripolitania occid. Ing. D. Zaccagna con Append.

paleont., pp. 118-126, pis. 9, 10.

Gabb, W, M.

1864. Descriptions of the Cretaceous fossils (of California). Calif.

Geol. Survey, Palaeont., vol. 1, pp. 57-236, pis. 9-32.
1877. Notes on American Cretaceous fossils -with descriptions of some

new species. Philadelphia Acad. Nat. Sci. Proc, vol. 28, pp. 276-

324, pi. 5.

Gaqel, C.

1927. Uber eine grosse Aturia aus dem norddeutschen Eocan. Zeitschr.
deutsch. Geol. Gesell. Berlin. Monatsber., vol. 78, pp. 118-121.

1928. Die Tone von Jatznick und Friedland und ihr FossUiahalt.
Jahrbueh Preuss. Geol. Landesanst. Berlin for 1927, vol. 48, pp.

477-486, pis. 20-22.
Galeotti, M. H.

1837. Memoire sur la constitution geognostique de la province de
Brabant, en reponse a la question suivante : decrire la constitution
geologique de la province de Brabant, determiner avec soin les
especes minerales et les fossUes que les divers terrains renferment
et indiquer la synonymic des auteurs qui en ont deja traite.
Mem. couronnes 1 'Acad. Roy. Sci. Bel.-lett. Bruxelles, vol. 12,
pp. 1-192, pis. 1-4.

Gardner, Julia

1923. New species of Mollusca from the Eocene deposits of south-
western Texas. U. S. Geol. Survey, Prof. Paper 131, pp. 109-117,
pis. 29-33.

Gay, C.

1854. Historia fisica y politica de Chile. Vol. 8, Zoologia, 499 pp.
(plus atlas) Paris.
Geinitz, H. B.

1843. Die Versteiuerungen von Kieslingswalda, und Nachtrag zur
charakteristik des Sachsisch-bohmischen Kreidegebirges. Dres-
den und Leipzig, pp. 1-23, pis. 1-6.

1887. Ueber Nautilus alahamensis Morton, Naut. zicsac Sow. und
Nautilus lingulatus v. Buch. Neues Jahrb. Min., Geol. Pal., 1887,
vol. 2, pp. 53-56, pi. 3.

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 477

Gemmellaro, G. G.

1868- Studi paleontologici suUa Fauna del Calcare a Terebratuln
1876. janitor del Nord di Sicilla. Pt. 1, 56 pp., 12 pis.; Pt. 2, 92 pp.

16 pis.; Pt. 3, 36 pp., 5 pis. (Palermo).
1884. Sui fossili degli strati a Terehratula aspasia della contrada

Eocche Eosse presso Galati (prov. di Messina). Giorn. Sci. Nat.

Palermo, vol. 16, pp. 167-218, 7 pis.
1886. Sul Dogger inferiore di Monte San Giuliana (Erice). Bull. See.

Sci. Nat. Palermo, vol. 17, pp. 197-213.

Gemmellaeo, M.

1911. Sui fossili degli strati a Terehratula aspasia della contrada
Eocche Eosse presso Galati (prov. di Messina). Giorn. Sci. Nat.
Eeon. Palermo, vol. 28, pp. 203-246.

Geyer, G.

1886. Ueber die liasisehen Cephalopoden des Hierlatz bei Hallstadt.
Abh. K.K. Geol. Eeich., vol. 12, no. 4, pp. 213-287, pis. 1-4.

Giovine, a. T. J.

1950. Algunos cefalopodos del hauterivense de Neuguen. Eev. Asoc.
Geol. Argentina, vol. 5, no. 2, pp. 35-76, pis. 1-7.

Greco, Benedetto

1915. Fauna Cretacea dell'Egitto raccolta del Figari Bey; parte prima,
Cephalopoda. Palaeont. Italica, vol. 21, pp. 189-231, pis. 17-22
(1-6).
Griepenkerl, Otto

1889. Die Versteinerungen der Senonen Kreide von Koenigslutter im
Herzogthum Braunschweig. Palaeont. Abh., vol. 4, pt. 5, pp.
1-117, pis. 1-12.

Gue'ranger, E.

1865. Etude sur \' Ammo7iit.es discus, Sowerby, suivie de la description
du Xautilus julii Baugier. Ann. Soc. Linn. Angers, vol. 7, pp.
184-190, pis. 1-3 (1864?).

Gugenberger, Odomar

1928. Beitriige zur Geologie Kleinasiens mit besouderer Beriieksichti-
gung des anatolischen Lias. Sitzungsber. Akad. Wiss. Wien. Abt.
1, vol. 137, pp. 259-282, pis. 1, 2.

Haas, 0. ["M"] H.

1947. On the status of the name Aturoidea Vredenburg, 1925, (Class
Cephalopoda, Order Nautiloidea). Bull. Zool. Nomenclature, vol.
1, pp. 243-244.

Haas, O., and A. K. Miller

1952. Eocene nautiloids of British Somaliland. Amer. Mus. Nat. Hist.,
Bull., vol. 99, pp. 317-354, pis. 21-31.

478 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Hanna, M. a.

1927. An Eoeeue iuvertebrate fauna from the La Jolla quadrangle,
California. Calif. Univ. Pub., Dept. Geol. Sci., Bull., vol. 16, pp.
247-398, pis. 24-57.

Hauee, F. E. von

1856. Uber die Ceplialopoden aus deni Lias der nordostlichen Alpen.
Denkschr. k. Akad. Wiss. Wien, vol. 11, pp. 1-86, pis. 1 25.

Hayasaka, I.

1936. A fossil Nautilus from Taiwan. Taiwan Tigaku Kizi, vol. 7,
no. 7, pp. 65-67, 1 pi.

HOEPEN, E. C. N. VAN

1931. Cretaceous Cephalopoda from Pondoland. Ann. Transvaal Mas.,
vol. 8, pp. 1-48, pis. 1-11.

HOLZAPFEL, E.

1899. Die Cephalopoden des Domanik im siidlichen Timan. Mem. Cor
mite Geol., vol. 12, no. 3, pp. 1-56, pis. 1-10.
HoppE, Walter

1922. Jura und Kreide der Sinaihalbinsel. Zeitschrift Deutsch. Pala
stina-Vereins, vol. 45, pp. 97-244, pis. 3-6.

Humphrey, W. E.

1949. Geology of the Sierra de los Muertos area, Mexico (with de-
scriptions of Aptian cephalopods from the La Peiia formation).
Bull. Geol. Soc. Amer., vol. 60, pp. 89-176, 2 figs., 18 pis.

Hyatt, Alpheus

1883- Genera of fossil cephalopods. Boston Soc. Nat. Hist., Proc, vol.
1884. 22, pp. 253-338 [pp. 253-272 were published in Dec. 1883, pp.

273-338 in Jan. 1884].
1894. Phylogeny of an acquired characteristic. Prcji-. Amer. Philos.

Soc, vol. 32, pp. 349-647, pis. 1-14.

Ihering, H. vox

1902. On the moUuscan fauna of the Patagonian Tertiary. Proc. Amer.
Philos. Soc, vul. 41, pp. 132-137, pi. 19.

1903. Les mollusques des terrains Cretaciques Superieurs de 1 'Argen-
tine orientale. Mus. Xac Buenos Aires, An., vol. 9, (ser. 3,
vol. 2) pp. 193-229, pis. 1, 2.

19(t7. Les mollusques fossiles du tertiaire et du cretace superieur de

1 'Argentine. Mus. Nac Buenos Aires, An., vol. 14 (ser. 3, vol. 7),

pp. 1-611, pis. 1-18.
1921. Nautilidos del Terciario de Chile y de la Argentina (notaprelimi-

nar). Physis. Soc. Argentina Cien. Nat., Revista, vol. 5, no. 19, p.

76.

1923. Los nautilidos del Terciario Patagonico y Cliileno. Mus. Nac.
Hist. Nat. Buenos Aires, An., vol. 31, pp. 470-474, pi. 1.

KUMMEL: POST-TRIASSIC NAUTILOID GEXERA 479

Iredale, Tom

1944. Australian pearly Nautilus. Australian Zoologist, vol. 10, pt. 3,
pp. 294-298.

Jeannet, a.

1948. Sur la forme de I'ouverture ehez certains Nautiles Mesozoiques.
Bull. Suisse Miu. Petr., vol. 28, pp. 178-187.

1951. Stratigraphie und Palaeoutologie des oolithisehen Eisenerzlagers
von Herznach und seiner Umgebung. Beitr. Geol. Schweiz,
Geotech. Ser., Lief 13, vol. 5, pp. 1-240, pis. 1-107.

Kakakasch, N. I.

1907. Le Cretace inferieur de la Crimee et sa faune. Trav. Soc. Imp.
Xatur. St. Petersburg, vol. 32, livr. 5, pp. 1-482, pis. 1-28.

Keller, Alexandre

1932. Mission saharienue Augieras-Draper, 1927-1928; sur une nouvelle
espeee de Xautile eocene du Soudan Fran^ais. Bull. Mus. Xatl
Hist. Xat., Paris, ser. 2, vol. 4, pp. 564-567.

Kellum, L. B.

1926. Paleontology and stratigraphy of the Castle Hajnie and Trent
marls in Xorth Carolina. U. S. Geol. Survey, Prof. Paper 143,pp.
1-56, pis. 1-11.

KiLiAN, W., and P. Eeboul

1909. Les cephalopodes neocretaces des lies Seymour et Snow Hill.

Schwedisehen Siidpolar Expedition, 1901-1903, vol. 3, Lief 6,

pp. 1-75, pis. 1-20.
1915. Contribution a 1 'etude des faunes paleocretaeees du sud-est de

la France. Mem. Carte Geol. France, pp. 1-296, pis. 1-15.

Kner, E.

1850. Versteinerungen djs Kreidemergels von Lemberg und seiner Uni-
gebung. Xatur. Abh. Haidinger, vol. 3, pp. 1-42, pis. 1-5.

KOBAYASHI, T.

1954a. A new cj-matoeeratid from the Palaeogene of Xorthern Kj-ushu

in Japan. Jap. Jour. Geol. Geog., Trans., vol. 24, pp. 15-21, pis.

4, 5.
1954b. A new Palaeogene paracenoceratoid from Southern Kyushu in

Japan. Jap. Jour. Geol. Geog., Trans., vol. 24, pp. 181-184, pi.

22.

KoENEN, A. vox

1883. Die Gastropoda holostomata und tectibranchiata. Cephalopoda
und Pteropoda des Xord-deutschen Miocan. Xeues Jahrb. Min.,
Geol.. Pal., Beil.-Bd. 2. pp. 223-367, pis. 5-7.

480 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

KossMAT, Franz

1897. The Cretaceous deposits of Pondicherri. Ree. Geol. Surv. India,
vol. 30, pt. 2, pp. 51-110, pis. 6-10.

Krenkel, E.

1910. Die untere Kreide von Deutsch-Ostafrika. Beitr. Palaont. Geol.

Osterreichs-Ungarns und des Orients., vol 23, Heft 4, pp. 201-250,

pis. 20-23.
1915. Die Kelloway-Fauna von Popilani in Westrussland. Palaeonto-

graphica, vol. 61, pp. 191-362, pis. 19-28.

Krumbeck, Lothar

1905. Die BracMopoden und Molluskenfauna des Glandarienkalkes.
Beitr. Palaont. Geol. Osterreieh-Ungarns und des Orients, vol.
18, pp. 65-162, pis. 8-14.

1906. Beitrage zur Geologie und Palaeontologie von Tripolis. Palaeon-
tographica, vol. 53, pp. 51-136, pis. 7-9.

KUHN, OSKAR

1936. Uber die Gruppe des Nautilus giganteus d'Orb. Palaeont.

Zeitschr., vol. 18, pp. 302-306.
1939. Uber einen neuen Nautilus aus dem mittleren Malm. Palaeont.

Zeitschr., vol. 21, no. 3, pp. 161-163.

KUMMEL, BeRNHARD

1953a. Lower Cretaceous nautiloids from Texas. Breviora, No. 19, pp

1-11, pis. 1, 2.
1953b. The ancestry of the Family Nautilidae. Breviora, No. 21, pp.

1-7, pi. 1.
1953c. American Triassic coiled nautiloids. U. S. Geol. Survey Prof.

Paper 250, pp. 1-104, pis. 1-19.
1954. Jurassic nautiloids from western North America. Jour. Paleont.,

vol. 28, pp. 320-324, pis. 34, 35.

Lea, Issac

1841. Notice of the Oolitic formation in America, with descriptions of
some of its organic remains. Am. Phil. Soc, Trans., n.s., vol. 7,
pp. 251-260, pis. 8-10.

Leymerie, M. a.

1846. Memoire sur le Terrain a nummulites (Epicretace) des Corbieres

et de la Montagne Noire. Mem. Soc. Geol. France, ser. 2, vol. 1,

pp. 337-373, pis. 13-17.
1851. Memoire sur un nouveau type Pyreneen parallele a la Craie

proprement dite. Mem. Soc. Geol. France, 2nd ser., vol. 4, pp.

177-202.

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 481

LOESCH, K. C.

1912. Eine fossile pathologische Nautilusschale. Neues JaJbrb., vol. 2,
pp. 90-102, pi. 7, figs. 1-2.

191-4. Die Nautilen des weissen Jura; Erster Teil. Palaeontographica,
vol. 61. pp. 57-146, pis. 10-15.

LOBIOL, P. DE

1903. Etude sur les Mollusques et Brachiopodes de I'Oxfordien su-
perieur et moyen du Jura IMouien. Mem. Soc. Pal^ont. Suisse,
vol. 30, pp. 77-160, pis. 6-19.
, E. EoYER, and H. Tombeck

1872. Description geologique et paleontologique des etages jurassiques
superieurs de la Haute-Marne. 542 pp., 26 pis., Paris.

LoTZ, H., J. BoHM, and W. Weissermel

1913. Geologische und palaontologische Beitrage zur Kenntnis der
Luderitzbuchter Diamantablagerungen, Beitr. Geol. Erforsch.
deutsehen Schutzgeb., no. 5, pp. 1-111, pis. 1-4.

LXTNDGREX, B.

1867. Palaeontologiska lakttagelser ofver Faxekalken pa Limhanm.
Lunds Universitets Ars-skrift. 1866, vol. 3, pp. 1-31, pi. 1.

Maktix, J. K. L.

1932. Ein eocaner Nautilus von Java, Dienst. Mynb. Ned. Indie,
Meded., no. 20, pp. 1-2, pis. 1, 2.

Martin, K.

1879- Die Tertiarsehichten auf Java. Leiden, 1879-1880, pp. 1-164;
1880. Allgemeiner theil pp. 1-51; Anliang pp. 1-6, pis. 1-28, 1 map.
1889. Die Fauna der Kreideformation von Martapoera. Samm. geol.
Eeichsmus. Leiden. L Ser., vol. 4, pp. 126-196, pis. 15-221.

1933. Eine neue Tertiare Molluskeufauua aus dem Indischen Archipel.
Leidsche Geol. Meded., vol. 6, pt. 1, pp. 7-32, pis. 1-5.

Maury, C. J.

1930. O Cretaceo de Parahyba do Norte. Serv. geol. min. Brasil, Men.,
u. 8, pp. 1-305, pis. 1-35.

McCoy, Frederick

1876. Prodromus of the palaeontology of Victoria; or, figures and
descriptions of Victoria ; or figures and descriptions of Victorian
organic remains, decade 3. Geol. Survey Victoria, pp. 1-40, pis.
21-30.

Meek, F. B.

1861. Descriptions of new Cretaceous fossils collected by the North-
western Boundary Commission, on Vancouver and Sucia Islands.
Philadelphia Acad. Nat. Sci. Proc, vol. 13, pp. 314-318.

482 BULLETIX : MUSEUM OF COMPARATIVE ZOOLOGY

1876. Eeport ou the invertebrate Cretaceous fossils of tlie Upper Mis
souri country. U. S. Geol. Surv. Terr., vol. IX (Hayden), 629
pp., pis. 1-45.
and F. V. Hayden

1862. Descriptions of new Cretaceous fossils from Nebraska Territory.
Philadelphia Acad. Xat. Sei. Pror-., vol. 14, pp. 21-28 (186.3).

MiLASCHEVITCH, K. S.

1877. Paleoutological sketches. I. On some fossils of Cretaceous in
Crimea. Bull. Soc. Imp. Nat. Moscou, vol. 52, pp. 65-128, pi. 1.

MiCHELOTTI, G.

1847. Deserii^tion des fossiles des terrains miocenes de 1 'Italia septen-

trionale. Xatuurk. Verb. Holland Maatseh. Wetensch., Haarlem,

vol. 3, pp. 1-40-B, pis. 1-17.
1861. Etudes sur le miocene inferieur de 1 'Italic, Xatuurk. Verb.

Holland ]\laatseh. . Wetensch., Haarlem, vol. 15, pp. 1-183, pis.

1-16.

Miller, A. K.

1935. The "Paleocene" nautiloid cephalopods of Laudana, Portuguese

West Africa. Jour. Paleont., vol. 9, pp. 167-173, pis. 19, 20.
1938. Cephalopods from the Quimbriz Eocene. Comm. Serv. Geol.

Portugal, vol. 20, pp. 38-43, pi. 9.
1941. An AUiria from the Tonga Islands of the Central Pacific. Jour.

Paleont., vol. 15, pp. 429-431.
1947. Tertiary nautiloids of the Americas. Mem. Geol. Soc. Amer., no.

23, pp. 1-234, pis. 1-100.
1949. The last surge of the nautiloid cephalopods. Evolution, vol. 3,

pp. 231-238.

1951. Tertiary nautiloids of west-coastal Africa. Ann. Mus. Congo
Beige, ser. 8, sci. geol., vol. 8, pp. 1-SS. p!s. 1 r.l.

, and C. CoLLiNSON

1952. A new Jurassic nautiloid species from Morocco. .lour. Paleont.,
vol. 26, pp. 626-629, pi. 85.

1953. An aberrant nautiloid of the Timor Permian. Jour. Paleont.,
vol. 27, pp. 293-295.

, and I. Crespix

1939. An Aturia from the northwest divisions of western Australia.

Jour. Paleont., vol. 13, pp. 79-81, pis. 13, 14, 2 text figs.
, and H. E. Dowxs

1950. Tertiary nautiloids of the Americas: Supplement. Jour. Paleont.

vol. 24, pp. 1-18, pis. 1-10.
, C. O. Dunbar, and G. E. Condra

KUMMEL: POST-TRIASSIC NAUTILOID GENERA 483

1933. The nautiloid eephalopods of the Peunsylvanian system in the

mid-contiuent region. Neb. Oeol. Survey, ser. 2, Bull. 9, pp. 1-240.

pis. 1-24.

, and W. M. Furnish

1938. Aturias from the Tertiary of Mexico. Jour. Paleont., vol. 12, pp.

149-155, pi. 25.

, and E. A. Harris

1945. North American Cjnnatoceratidae (Mesozoic Nautiloidea). Jour.

Paleont., vol. 19, pp. 1-13, pis. 1-6.

, and 3L L. Thompson

1933. The nautiloid cephalopods of the Midway group. Jour. Paleont..

vol. 7, pp. 298-324, pis. 34 38.
1935. ThL' nautiloid genus Aturoidea in America. Jour. Paleont., vol.

9. pp. 563-571, pis. 65-66.
1937. Beitrage zur Kenntnis tropisch — Amerikanischer Tertiarmol

lusken; VI, Some Tertiary nautiloids from Venezuela and Trini

dad. Eclog. Geo!. Helvetiae, vol. 30, pp. 59-73. pis. 7-10.

MOBERG, J. C.

1884. Cephalopoderna i Sveriges Kritsystem, I, Sveriges Kritsystem
systematiskt Framstaldt. Sveriges Geol. Undersokning, ser. C,
no. 63, pp. 1-45, pis. 1, 2.

1885. Cephalopoderna i Sveriges Kritsystem, II, Artbestkrifning.
Sveriges Geol. Undersokning, ser. C, no. 73, pp. 1-64, pis. 16.

MONTFORT, DENYS DE

1802. Histoire naturelle, generale et particuliere des Mollusques, ani-
maux sans vertebres et a sang blane ; Ouvrage f aisant suite
a 1 'histoire naturelle generale et particuliere, composee par
Leclerc de Buff on, et redigee par C.S. Sonnini, membre de
plusieurs Societes savantes. Paris, vol. 4, pp. 1-376, i)ls. 42-50.

ISOS. ('onehyliologie systematique et classification methodique des co-
quilles; off rant leurs figures, leur arrangement generique, leurs
descriptions caracteristiques, leurs noms; ainsi que leur synony
niie en plusieurs langues. Vol. 1, Paris. 409 pp., 100 pis.

MORICKE, W.

1S94. Versteineruugen der Lias und Uuteroolith von Chile. Xeues
Jahrb. Min. Geol., Pal., Beil.-Bd. 9, pp. 1-100, pis. 1-6.

]\IORKIS, J.

1848. Description of a new species of Nautilus from the Lower Green
sand of the Isle of Wight. Ann. Mag. Nat. Hist., ser. 2, vol. 1,
pp. 106-107.

484 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY
, and L. Lycett

1850. A monograph of the Mollusca from the Great Oolite, chiefly from
Minehinhampton and the coast of Yorkshire. Pal. Soc. Mon., 4
parts, pp. VIII, 130 + 147 + 129, 45 pis.

MOETON, S. G.

1834. Synopsis of the organic remains of the Cretaceous group of the
United States, illustrated by nineteen plates ; to which is added
an appendix, contaLahig a tabular view of the Tertiary fossils
hitherto discovered in North America. Philadelphia, pp. 1-88
(and pp. 1-8, the appendix), pis. 1-19.
MuLLEE, G., and A. Wollemann

1906. Die Molluskenfauna des Untersenon von Braunschweig und
Ilsede. II. Cephalopoden. Abh. K. preuss. Geol. Landesanstalt.,
n.s. no. 47, pp. 1-30, Atlas, pis. 1-11.

Nagao, Takumi

1926. Aturia yoTcoyamm, nov. nom. from the Paleogene of Kyushu.
Sci. Rept. Tohoku Imp. Univ., ser. 2, (geol.), vol. 9, pp. 29-32,
pis. 9-11.

Negri, Luiqi

1934. Gasteropodi, Scafopodi e Cefalopodi paleogenici della Sirtica e
del Fezzan orientale. Missione Scientifica della Reale Accademia
d'ltalia a Cufra (1931 — IX), vol. 3, pp. 131-181, pi. 14.

Neumaye, M.

1870. IJber einige neue oder weniger bekannte Cephalopoden der Ma-
crocephalen-Schichten. Jahrb. K.K. geol. Eeichsanst., vol. 20,
no. 2, pp. 147-156, pis. 7, 8.

Newton, B. B.

1905. Eocene shells from Nigeria. Ann. Mag. Nat. Hist., ser. 7, vol.
15, pp. 83-91, pi. 5.

1919. On a sandstone cast of Aturia aturi (Basterot), from the
Miocene of Western Australia. Malac. Soc. London, Proc, vol.
13, pp. 160-167, pis. 5, 6.

1925. Eocene Nautili from British Somalilaud. Glasgow Univ., Hun-
terian Mus. Geol. Dept., Mon. 1, pp. 165-170, pis. 17, 18.
, and G. F. Harris

1894. A revision of the British Eocene Cephalopoda. Malac. Soc. Lon-
don, Proc, vol. 1, pp. 119-131, pi. 10.

NiKITIN, S.

1881. Der Jura der Umgegend von Elatma. Nouv. M6m. Soc. Imp. Nat.
Moscou, vol. 14, pt. 2, pp. 85-135, pis. 1-6 (8-13).

NOETLING, F.

1896. Fauna of the Kellaways of Mazar Drik. Geol. Survey India
(Palaeont. Indica), Mem., ser. 16, vol. 1, pt. 1, pp. 1-22, pis. 1-13.

KUMMEL : POST-TRIASSIC NAUTILOID GENERA 485

1897. Fauna of Baluchistan. Fauna of the Upper Cretaceous (Mae
strichtian Beds of the Mari Hills) Geol. Survey India. (Palaeont.
Indica) Mem., ser. 16, vol. 1, pt. 3, pp. 1-79, pis. 1-2.3.

Odum, Hilmar

1926. Studien over Daniet i Jylland og paa Fyn. Danmarks Geol.
Unders0gelse, ser. 2, no. 45, pp. 1-306, pis. 1-6.

Oliveira, Paulo Erichsen de

19.53. Invertebrados Fosseis da Formacas Maria Farinha. I. Cephalo-
poda. Divisao da Geol. e Mineral. Bol. 146, pp. 1-33, pis. 1-4.

Olsson, a. a.

1928. Contributions to the Tertiary paleontology of northern Peru.
Part I, Eocene Mollusca and Brachiopoda. Bull. Am. Paleont.,
vol. 14, no. 52, pp. 1-154, pis. 1-26.

1930. Contributions to the Tertiary paleontology^ of northern Peru.
Part 3, Eocene Mollusca. Bull. Am. Paleont., vol. 17, no. 62, pp.
1-96, pis. 1-12.

1931. Contributions to the Tertiary paleontology of northern Peru:
Part 4, the Peruvian Oligocene. Bull. Am. Paleont., vol. 17, no.
63, pp. 1-164, pis. 1-21.

OOSTEE, W. A.

1858. Catalogue des Cephalopodes FossUes des Alpes Suisses. 3rd part,
Cephalopodes Tentaculiferes, Nautilides. pp. 1-20, pis. 8-12.

Oppenheim, Paul

1900- Die Priabonaschichten und ihre Fauna, im Zusammenhange mit
1901. gleichalterigen und aJialogen Ablagerungen vergleichend be-

trachtet. Palaeontographica, vol. 47, pp. 1-348, pis. 1-21.
1903. Ueber die Fossilien der Blattermergel von Theben. Sitzungsber.

K. Bayer. Akad. Wiss., Math.-Phys, CL, vol. 32, pp. 435-455,

Pl. 7.

1906. Zur Kenntniss Alttertiarer Faunen in Agypten; Lief. 2, Die
Bivalven zweiter Teil, Gastropoda und Cephalopoda. Palaeonto-
graphica, vol. 30, div. 3, pp. 165-348, pis. 18-27.

1916. Die eoeane luvertebraten — Fauna des Kalksteins in Togo, im
Zusammenhange mit anderen Tertiarablagerungen Afrikas ver-
gleichend betrachtet. Beitr. Geol. Erforsch. deutseh. Schutzgeb.,
no. 12, pp. 1-126, pis. 1-5 (1915).

Okbigny, a. d'.

1840- Paleontologie Francaise. Terrains Cretaces, vol. 1, (for dates see
1847. Geol. Mag., pp. 223-225, 1899), 662 pp., pis. 1-148 (in Atlas).

1841. Descriptions de quelques esp^ces de Mollusques fossiles de France.
Rev. Zoologique par la Soc. Cuvierienue, Annee 1841, pp. 318-319.

1842. Voyage dans I'Amerique meridionale. Vol. 3, Partie 4, Paleon-
tologie, Paris. 188 pp., 22 pis.

486 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

1842- Paleontologie Franeaise. Terrains Jurassiques, vol. 1 (for dates

1851. see Geol. Mag., pp. 223-225, 1899), 642 pp., pis. 1-236 (in Atlas).

1850. Prodrome de Paleontologie stratigraphique universelle des ani

maux mollusques et rayonnes faisant suite au cours elementaire

de Paleontologie, Premier vol., Masson, Paris. 394 pp.; deuxiense

vol., 428 pp.

Parona, C. F.

1897. I Nautili del Lias Inferiors de Saltrio. Bull. Soc. Malae. Ital.,

vol. 20, pp. 7-20, pi. 1 (1895).
1899. Note sui Cefalapodi Terziari del Piemonte. Palaeont. Italica,
vol. 4, pp. 155-168, pis. 12, 13.

■ — , and G. Bonarelli

1897. Sur la faune du Callovien inferieur (Chanasien) de Savoie. Meni.
Acad. Sci. Savoie (4), vol. 6, pp. 35-211, pis. 1-11.

Penecke, K. a.

1884. Das Eoean des Krappfeldes in Karnten. Sitzungsber. K. Akad.
Wiss., Wien, vol. 90, pt. I, pp. 327-371, pis. 1-5.

Petitclerc, Paul

1926. Note sur la presence du Nautilus julii Baugier, dans le Bathonien
des environs de Belfort; suivie de la description d'especes nou-
velles ou interessantes du Callovien des Deux-Sevres, de Maine-
et-Loire, de 1 'Ardeehe, etc. ; et en£n de 1 'enumeration de certaines
ammonites rares ou peu connues du Cretaee des Basses-Alpes.
Imprimerie Marcel Bon, Vesoul, 1926.

Peyrot, M. a.

1932. Conchologie Xeogenique de I'Aquitaine, tome 6 Gastropodes.
Actes Soc. Linn. Bordeaux, vol. 85, pp. 5-71, (pis. 11-18 pub.
1932, vol. 84).

PiA, Julius

1914. Untersucliungen iiber die Liassiselien Nautiloidea. Beitr. Pala-
ont. Geol. Osterreieh-Ungarns uud des Orients, vol. 27, i)p. 19 86,
pis. 4-10.

PiCTET, F. J.

1867. Etudes paleontologiques sur la faune a Terehratula diphyoide.t
de Berrias (Ardeehe). Melanges Paleont., vol. 1, pt. 2. pp. 43
131, pis. 8-28.

, and G. Campiche

1858- Description des fossiles du terrain cretaee des environs de
1860. Sainte-CroLx. Geneve. 380 pp., 43 pis.

, and W. Euux

1847- Description.s de."! mollusques fossiles qui se trouvent dans les

KUMMEL : POST-TRIASSIC NAUTILOID GEXERA 487

1853. gres verts des enviious do Geneve. Mem. Soe. Phys. Hist. Xat..

Geneve, vol. 11, pt. 2, pp. 257-412; vol. 12, pp. 157 287; vol. 13,

pt. I, pp. 73-173, 489-538, pis. 1-51.
POMEL, A.

1889. Les eephalopodes Xeoeomiens de Lamoriciere. Materiaux pour

la Carte Geologique de I'Algerie. I serie. Paleont. No. 2. pp.

1-96, pis. 1-14.

Prinz, J.

1906. Die Xautiliden in der Unteren Jura-Periode. Ann. Hist. Xat.
Mus. Xat. Hung., vol. 4. pp. 201-243. pis. 3, 4.

Qua AS, A.

1902. Die Fauna der Overwegischicliten uml der Blatterthoue in der
libysehen Wiiste. Palaeontographica, vol. 30, pt. IT. jip. 153 336.
pis. 20-33.

QUENSTEDT, F. A. V.

1845- Petrefaetenkunde Deutschlands, etc. ^'ol. I, Cephalopoden. Lief.
1S49. I, pp. 1104. (1845); II, pp. 105-184, (1846); III, pp. 185 264.
(1847); IV-Y. pp. 265-472, (1848); VI. pp. 473 580. (1849).

Ravn, J. P. J.

1 902. Molluskerne i Danmarks Kridtafle.iringer, 11. Seaphopoder, Gas
tropoder, og Cephalopoder. K. Danske Videusk. Selsk. Skr., ser.

6, Xaturvidensk. Math, Afd.. vol. 11. no. 4, pp. 205-270, pis. 1 5.

IvKDTEXBACHER, AXTON"

1 873. Die Gephalopodenfauua der Gosauschicliteu in dcu uordostliclien
Alpen. Abh. K.K. Geol. Eeieh.^anstalt, vol. 5, pt. 5, pp. 91-140.

pis. 22 30.

Reeside, J. B., Jr.

1924. A new nautiloid cephalopod, Eutrephoceras sloani, from the

Eocene of South Carolina. U. S. Xat. Mus., Proc, vol. 65, art. 5,

pp. 1-4, pis. 1-3
1927a. Cephalopods from the lower part of the Cody Shale of Oregon

Basin, Wyoming. U. S. Geol. Survey. Prof. Paper 150, pp. 1 19,

pis. 1-8.
19271 1. The cephalopods of the Eagle sandstone and related forr.iatitui^

in the western interior of the United States. I'. S. Geol. Survey,

Prof. Paper 151, pp. 1-87, pis. 1 45.

Retoavski, O.

1>^94. Die tithonischen Ablageruugen von Theodosia ; ein Bcitrag zur
Palaontoiogie der Krim. Moseou Soc. Imp. Xat. Bull., n.s., vol.

7. pp. 206-301, pis. 9-14 (1893).

RoEilER, F. A.

1836. Die X'ersteinerungen der norddeutschen Oolithen-Gebirges. 218
pp., 16 pis., (in Atlas).

488 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

RONCHETTI, C. E.

1947. Appendice alia revisione dei Gen. Nautilus, Indoceras, e Bacndites
del Neoeretacico della Libia. 1st. Geol. Paleont. Geog. Fisica,
Univ. Milano, ser. P, pub. 50, pp. 1-4, fig. 3.

ROUCHADZE, J,

1931. Notice siir les Belemnites et les Nautilus de I'Aptien de Koutais.
Bull. Mus. Georgie, vol. 6 (1929-1930), pp. 125-138, pis. 1-3.

Sanchez Roig, Mario

1951. La Fauna Jurasica de Vinales. Ann. Ciencias Medicas, Fis. Nat.,
Havana, vol. 89, pt. II, pp. 47-94, pis. 1-28.

SCHAFHAUTL, K. E.

1852. Der Teisenberg oder Kressenberg in Bayern. Neuea Jahrb.

Mineral. Geogn. Geol. Petrefakten-Kunde. Jahrgang 1852, pp.

129-175, pis. 3, 4.
1863. Siid-Bayems Lethaea Geognostica, — Der Kuessenberg und die

siidlieh von ihrn gelegenen Hochalpen, geognostisch betrachtet

in iliren Petrefacten. Leipzig, 487 pp., 86 pis., 7 maps (in Atlas)

SCHENCK, H. G.

1931. Cephalopods of the genus Aturia from western North America.
Calif. Univ. Pub., Dept. Geol. Sci., Bull., vol. 19, pp. 435 491,
pis. 66-78.

SCHLAGINTWEIT, 0.

1912. Die Fauna des Vracon und Cenoman in Peru. Neues Jahrb.,
Beil.-Bd. 33, pp. 43-136, pis. 5-7.

SCHLOTHEIM, E. F.

1820. Die Petrefactenkunde. Cephalopoda. Pp. 45-88.

SCHLUTEB, C.

1876. Cephalopoden der oberen Kreide. Palaeontographiea, vol. 24,
pt. II, pp. 123-263 (1-144), pis. 36-55.

Scott, G.

1940. Cephalopods from the Cretaceous Trinity group of the south-
central United States. Texas Univ. Bull. 3945, pp. 969-1106, pis.
55-68.

1943. Palaeontology of Harrar Province, Ethiopia Part 4, Jurassic
Cephalopoda and a Cretaceous Nautilus. Am. Mus. Nat. Hist.
Bull., vol. 82, art. 3, pp. 55-94, text figs. 1-23, pis. 10-25.

Sergio, Venzo

1933. II cattiano di Monte Brione presso Riva del Garda e la sua
nuova fauna. Studi Trentini Sci. Nat. Trento, vol. 14, fasc. 3, pp.
188-213, pis. 1-2.

KUMMEL : POST-TRIASSIC NAUTILOID GEXERA 489

Sharpe, D.

1853. On the age of the fossiliferous sands and gravels of Farringdon
aud its neighbourhood. Quart. Jour. Geol. Soc. London, vol. 10,
pp. 176-198, pis. 5-6.

Shattuck, G. B.

1903. The Mollusea of the Buda limestone. U. S. Geol. Survey Bull.
205, pp. 1-36, pis. 1-25.

ShIMIZU, S^iBUEO

1926. On two species of Nautiloidea from the Tertiary of Japan. Sci.
Kept. Tohoku Imp. Univ., ser. 2 (geol.), vol. 9, pp. 25-27, pi. 8.

1931. The marine Lower Cretaceous deposits of Japan, with special
reference to the Ammonite-bearing zones. Sci. Kept. Tohoku
Imp. Univ., ser. 2 (geol.), vol. 15, no. 1, pp. 1-40, 4 pis., 1 text fig.

Shumard, B. F.

1860. Descriptions of new Cretaceous fossils from Texas. St. Louis
Acad. Sci. Trans., vol. 1, pp. 590-610.
SiNZOW, I.

1913. Beitrage zur Kenntnis der unteren Kreideablagerungen des Nord-
Kaukasus. Trav. Mus. Geol. Imp. Sci. St. Petersburg, vol. 7, pp.
93-117, pis. 4-6.

SOKOLOV, D. X.

1928. Fossiles Mesozoiques de la Bolschezemelskaja Tundra et de Kash
pur. Trav. Mus. Geol. Leningrad, vol. 3, pp. 15-62, pis. 3, 4.

SORRENTIXO, StEFAXO

1932. Cefalopodi Maestrichtiani della Tripolitania. Bol. Soc. Geol.
Italiana, vol. 51, pp. 184-209, pis. 4-6.

SowERBY, J., and J. de C. Sowerby

1812- Mineral Conchology. 7 vols., pis. 1-337 (1812-1822) by J.
1846. Sowerby; pis. 338-648 (1822-1846) by J. de C. Sowerby.

Spath, L. F.

1921. Cretaceous Cephalopoda from Zululand. Ann. S. African Mus.,

vol. 12, pt. 7, no. 16, pp. 217-321, pis. 19-26.
1927a. Eevision of the Jurassic cephalopod fauna of the Kachh (Cutch).

Mem. Geol. Surv. India (Palaeout. ludica), n.s., vol. 9, no. 2,

pp. 1-84, pis. 1-7.
1927b. On the classification of the Tertiary nautili. Ann. Mag. Nat.

Hist., ser. 9, vol. 20, pp. 424-428.
1930a. The Eotriassic invertebrate fauna of East Greenland. Medd. oni

Gronland, vol. 83, pp. 1-90, pis. 1-12.
1930b. On the Cephalopoda of the Uitenhage beds. Ann. S. African

Mus., vol. 28, pp. 131-157, pis. 13-15.

-tyU BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

1934. Catalogue of the fossil Cephalopoda in the British Museum (Nat
ural History). Part 4, The Amnionoiflea of tlic Trias. London,
l)p. 1-521, pis. 1-18.

1935a. Additions to the Eotriassic invertebrate fauna of East Green-
land. Medd. om Granlaud, vol. 98, pii. 1-115, pis. \--2o.

1935b. The Mesozoic palaeontology of British Sonialiland. X, Jurassic
and Cretaceous Cephalopoda. Geol. Palaeont. British Somali-
land, pt. 2, pp. 205-228, pis. 24-25.

1953. The Upper Cretaceous cephalopod fauna of Graham Land.
Falkland Is. Dependencies Survey, Sci. Eept. no. 3, pp. 1-60,
pis. 1-13.

Spexgler, Erich

1910. Uutersuehungen liber die siidindisehe Kieideformation. Die

Nautiliden und Belemniteu des Trii-liinopolydi8trikts. Beitr.

PalJiont. Osterreich Ungarns und Orients, vol. 23, pp. 125-157,

pis. 11-14 (26-29).
1919. Die Gebirgsgruppe des Plassen und Hallstattcr Salzberges ini

Salzkammergut. Jahrb. Geol. Eeich., Jalirgnnj; 1918, vol. 68,

pp. 285-474, pis. 14-18 (1-5).

Stchepixsky, V.

1943. L'Oligocene mariu dc Gaziantep (Turquie meridiouale). Maden
Tetkik ve Arama, Ankara, sene 8, sayi 2/30, pp. 223-248, pis.
1-4.

Steinmann, G.

1895. Die Cephalopoden der Quiriquinaschichten. Neues Jahrb. Beil.-
Bd. 10, pp. 64-94, pis. 4-6.

Stexzel, H. B.

1935. Xautiloids of the genus At aria from the Eocene of Texas and

Alabama. Jour. Paleont., vol. 9, pp. 551-562, pis. 63, 64.
1940. Tertiary nautiloids from the Gulf Coastal Plain. Texas Univ.

Publ. No. 3945, pp. 731-794, pis. 35-42,
1948. Paleoecology of Tertiary nautiloids. Rept. Comm. on a Treatise

on Marine Eeol. and Paleoecol., Div. Geol. Geog., Natl. Res.

Council, no. 8, pp. 96-97.
1948. Ecology of living nautiloids. Kept. Comm. on a Treatise of

Marine Ecology and Paleoecol., Div. Geol. and Geog., Nat. Ees.

Council, pp. 84-90.
1952. Living Nautilus. Treatise on Invertebrate Paleontology, Pt. H,

Cephalopoda, Nnntiloidea, pp. HI — 1119, Preprint.

KUMMEL: POST-TKIASSIf XAUTlLulD GENERA 491

Stephknsox, Iv. W.

lOil. Thp larger invertebrate fossils of the Xavarro group of Texas
(exc-lusivo of corals and crustaceans and exclusive of the fauna
of the Escondido formation). Texas Univ., Pub. 4101, pp.

]-«41, pis. 1 !t.l.

.Stetee, Alejandro

19i:i. Estratos Juracicos Argentinos. Aetas Acad. Nac. Cienc. Cor
doba, vol. 7, pp. 3.3-128, pis. 1-24.

Strombeck, a. v.

1863. Ueber die Kreide am Zeltberg bei Liineburg. Zeitschr. deutsch.
geol. GeselL, vol. 15, pp. 97-187, pi. 4.
Tagliarixi, F. p.

1901. Monografia sui Nautili tlel Dogger inferiore di Monte San
Giuliano (Erice). Gior. Sci. Xat. Econ., Palermo, vol. 23, pt. I,
pp. 186-203, pis. 1(5.

Tavani, G.

1942. Paleoutolugia della .Somalia, AEolluscki del Cretaceo della
Somalia. Palaeontographica Italica, vol. 32, suppl. 4, fasc. 2,

pp. 7-47, pis. 1 5 (10-14).

Teichert, C.

1940. Coutributious to nautiloid nomenclature. Jour. Paleont., vol.
14, pp. 590-597.

1943. Eocene nautiloids from Victoria. Proe. Roy. Soe. Victoria, vol.
55. pt. 2 (n. s.j, pp. 257-264, pi. 11.

1944. The genus Aturia in the Tertiary of Australia. Jour. Paleont.,
vol. 18, pp. 73-82, pis. 14-16.

1947a. Notes on Eocene nautiloids from Victoria. Mining Geol. Jour.,

vol. ;!, Mu. 1, pp. 40-42.
1947b. New nautiloids from the older Tertiary of Victoria. Mining

Geol. Jour., vol. 3, no. 2, pp. 48-51.

, and B. C. Cotton

1949. A new Aturia from the Tertiary of South Australia. Rec.

South Australian Mus., vol. 9, pp. 255-256, pi. 21.

, and B. F. Glenister

1952. Fossil nautiloid faunas from Australia. Jour. Paleont., vol. 26,

pp. 730 752, pis. 104-108.

Tf.rquem, O.

1855. Paleontologie de I'etage inferieur de la formation Liasique
(le la province de Luxembourg, Grand-Duche (Hollande) et de
llctlange, Departement de la Moselle. Mem. Soc. Geol. de
France. 2nd ser.. vol. 5. no. 3, pp. 219-343, pis. 1-15.

492 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY

TuoMET, Michael

1854. Descriptions of some new fossils from the Cretaceous rocks of
the southern states. Acad. Nat. Sci. Philadelphia, Proc, vol. 7,
pp. 167-172.

Uhliq, Victok

1883. Die Cephalopodenfauna der Wernsdorfer Schichten. Denksch.
Akad. Wiss. Wien, vol. 46, abt. 2, pp. 127-290, pis. 1-32.

Vacek, M.

1886. ijber die fauna der Oolithe von Cap. S. Vigilio. Abh. K. K.
Geol. Eeich. Wien, vol. 12, no. 3, pp. 57-212, pis. 1-20.

Vadasz, M. E.

1911. Die Juraschichten des Siidlichen Bakony. Eesultate Wiss.
Erforsch. des Balatonsees, vol. 1, Pal. Abh. 3, pp. 1-89, pis. 1, 2.

Valenciennes, Achille

1841. Nouvelles recherches sur le Nautile flanibe. Paris Mus. Natl.
Hist. Nat., Archiv., vol. 2, pp. 257-314, pis. 8-11.

Venzo, S.

1937. La fauna Cattiana delle glauconie Bellunesi. Mem. Inst. Geol.
E. Universita di Padova, vol. 13, pp. 1-207, pis. 1-12.

VlALLI, ViTTERIO

1937. Ammoniti giurassiche del monte Peller. Trento, Mus. Storia
Nat. Venezia Tridentina, Mem. an 7, vol. 4, pp. 99-148, pis. 1-3,
5 figs.

Vincent, Emxle

1907. Description d'une espeee de cephalopode tetrabranchial nouvelle,
du Landenien de Belgique {At^iria linicentensis, E. Vincent).
Ann. Soc. Eoy. Zool. Malac. Belgique, vol. 42, pp. 321 323, fig. 1.

1913. Contribution a la paleontologie des Falaises de Landana (Bas-
Congo) ; Mollusques. Mus. Congo Beige. Ann., sec. 3, vol. 1,
pp. 1-46, pis. 1-6.

VOQL, V.

1908. iJber Eozane Nautilidcn. Foldtani Kozlony, vol. 38, pp. 568-582,
636-649.

1910. Neuere Beitrage zur Kenntnis der alttertiaren Nautiliden.
Centralbl. Min., Geol. Pal., Jahrg. 1910, pp. 707-710.

VOKES, H. E.

1937. Nautiloid cephalopods from the Eocene of California. Jour.
Paleont.. vol. 11, pp. 3-9, pis. 1, 2.

KUMMEL: POST-TBIASSIC NAUTILOID GENERA 493

Veedenburg, E. W.

1925. Descriptions of Mollusca from the post-Eocene Tertiary forma
tion of northwestern India; Cephalopoda, Opisthobranehiata,
Siphonostomata. Mem. Geol. Surv. India n. s., vol. 50, pt. 1, pp.
1-350, pis. 1-13.

, and G. de P. Cotter

1928. A supplement to the Mollusca of the Eanikot series. Mem.
Geol. Surv. India (Palaeont. Indica), n. s.. vol. 10, no. 4, pp.
1-75, pis. 1-9, and corr. slip.
Waagen, W.

1873. Jurassic fauna of Kutch. The Cephalopoda. Mem. Geol. Surv.
India (Palaeont. Indica), ser. 9, vol. 1, pt. 1, pp. 1-22, pis. 1-4.
Wanner, J.

1902. Die Fauna der obersten weissen Kreide der libysehen Wiiste.
Paleontographica, vol. 30, Abt. 2, pp. 91-152, pis. 13-19.

Waring, C. A.

1914. Eocene horizons of California. Jour. Geol., vol. 22, pp. 782-785.
Weaver, C. E.

1931. Paleontology of the Jurassic and Ctetaceous of west-central
Argentina. Mem. Washington Univ. (Seattle), vol. 1, pp. 1-594,
pis. 1-62.

Wetherell, N. T.

1836. Observations on some of the fossils of the London clay, and
in particular those organic remains which have been recently
discovered in the tunnel for the London and Birmingham Rail-
road. London and Edinburgh Phil. Mag. Jour. Sci., ser. 3, vol.
t 9, pp. 462-469.

White, C. A.

1884. On the nautiloid genus Enclimutoceras Hyatt and a description
of the type species. U. S. Geol. Surv. Bull. 4, pp. 16, 17, pis. 7-9.

Whiteaves, J. F.

1876. On some invertebrates from the coal-bearing rocks of the Queen
Charlotte Islands collected by Mr. James Richardson in 1872.
Canada Geol. Survey, Mesozoic fossils, vol. 1, pt. 1, pp. 1-92,
pis. 1-10.

1900. On some additional or imperfectly understood fossils from the
Cretaceous rocks of the Queen Charlotte Islands, with a revised
list of the species from these rocks. Canada Geol. Survey,
Mesozoic fossils, vol. 1, pt. 4, pp. 263-307, pis. 33 39.

Whitfield, R. P.

1892. Gastropoda and Cephalopoda of the Raritan clays and Greensand
marls. U. S. Geol. Surv., Mon. 18, pp. 1-402, pis. 1-50 [also
issued as N. J. Geol. Surv., Paleont., vol. 2].

494 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY

Yabe, H., and H. Ozaki

19n3. A new type of Cretaceous nautiloids from Tyosi Peninsula,
Kwanto Eegion, Natl. Sei. Mus. (Tokyo), Bull. 32, pp. 55-61.

, and S. Shimizu

1924. A new species of Nautilus, X. (Cymatoceras) pseudo-atlas Yabe
and Shimizu, from the upper Cretaceous of Amakusa. Japanese
Jour. Geol. Geogr., vol. 3, pp. 41-43, pi. 5.

YOKOYAMA, M.

1913. On two new fossil Cephalopoda from the Tertiary of Izunio.
Jour. Geol. Soc. Tokyo, vol. 20, pp. 33-35 (1-3), pis. 7, 8.

Young, G., and J. Bird

1828. A geological survey of the Yorkshire Coast. 2nd ed. Whitby.
324 pp., 17 pis.

ZlETEN, C. H., V.

1830. Die Versteinerungen Wiirttembergs, 12 pts., 102 pp., 72 pl.s.
(Stuttgart). For dates of publication Fee Crick, 1899, p. 554.

ZiTTEL, K. A. VON.

1868. Palaeontologische Studien iiber die Granzschichten der Jura
und Kreide-Formation im Gebietc der Karpathen, Alpen und
Apenninen, I. Abtheilung, Die Cephalopoden der Stramberger
Schichten. Mus. K. Bnyer. Staates Palaeont. Mitt, (begonnen
von Oppel, forgesetzt von Zittel), vol. 2, pp. 33-118, pis. 1-24
in an atlns.

ZWIERZYCKI, J.

1914. Die Cephalopodenfauna der Tendagurusehichten Deutsch-Ost
afrika. Arehiv. Biontologie, Gesell. Xaturforsch. Freunde Bev
lin, vol. 3, pt. 4, Teil 3, pp. 7-96, pis. 1-10. •

1915. Fossielei! Afkomstig van eenige Vindplaatsen op Sumatra. Jaat
Mi.inwezen Neder. Oost-Indie, vol. 42, pp. 101-129 (1913).

PLATES

PLATE 1. CENOCEKAS ORBIUNVI

Figures Pago

1, 2 Cenocevan orhignyi (Prinz ) 361

Two views of type species of the genus Ccriuccra.s (=A'aw-
tilus inter 'medius d'Orbigny — non Sowerby). From d'Or-
bigny. X 1.

PLATE 1

PLATE 2. CENOCERAS TBECHMANNI

Figures Page

1-3 Cenoeeras trechmanni (Knmmel » 349

Three views of holotype from Upper Triassic, Carnian,
Bed C, Otomita, Hokonin Hills, New Zealand. B.M.N.H.
C21947. X 1.

<»J

''x^''-''^'

V\/

^fy

PLATE 2

PLATE 3. CENOCERAS STRIATUS

Figures Page

1, 2 Cenoceras striatus (Sowerby) 362

Two views of plesiotype, X 0.60, from Lower Lias, locality
unknown. B.M.N.H. 43852 (Sowerby Collection).

PLATE 3

PLATE 4. CENOCEEAS SIMILLIMUS

Figures Page

1, 2 Cenoeeras simillimvf; (Foord and Crick) 362

Two views of holotype, X 1, from Lias, Oharmouth, Dor
set, England. B.M.N. H. 39887 (Harrison Collection).

PLATE 4

PI.ATH 5. OENOCERAS TNORNATFS

Figures Pago

1. 2 Ccnoci rtiK innnuitus (d 'Orbigny) 3H2

Two views of i)lesiotyi)e, X 0.75, from Inferior Oolite,
Burton Bradstock, Dorset, England. B.M.N. II. C2843
(Baber Collection).

PT.ATF 5

PLATE 6. CENOCERAS BURTONENSIS

Figures Pnge

1, 2 Cenoccras burtnncnsis (Foord and Crick) ;J71

Two views of liolotype, X 0.75, from Inferior Oolite,
Burton-Bradstock. Dorset. Eng;land. B.M.N.H. 2841 (Ba-
ber Collection ) .

"^-

PLATE 6

PLATE 7. CENOCERAS AND SYRTOXArTILUS

Figures Page

1, 2 Cenoceras excavatus (Sowerby) 369

Two views of holotype, X 1, from Inferior Oolite, Dorset.
B.M.N. H. 4.3855 (Sowerby Collection).

3 Syriotmutiliis Ubanoiicus (Foord and Crick) 433

Ventral view of syntype, X 1, with mandible on surface,
from Upper Cretaceous, Sahil Ahr.a, I^ebanon, Syria.
B.M.N.H. 83663 (Lewis Collection).

Previously figured in Cat. Foss. Ceph., B.M., I't. '1, p. 371,
fig. 8'2 ; mandible figured, Ann. Mag. Xat. Hist. [(i|. vol. 5
(1890;, p. 405, fig. 6c; and Cat. Foss. Ceph. B.M., Pt. 2,
p. 305, fig. 67c.

PLAIE 7

PLATE 8. PAEACENOCERAS AND CENOOEEAS

Figures Page

1 Paracenoccrds hfXdf/onum (Sowerby; 4Q2

View of polished section of holotype from Corallian Oolite,
locality unknown. See Plate 13 for viewt? of external fea-
tures. B.M.X.H. 43S56 (Sowerby Collection MC 529f2).
X 1.

2 Cenociras cjiavatux (Sowerby ) ;{(>y

Mew (if polished section of holotype from Inferior Oolite,
Dorset. See Plate 7 for views of external features.
B.M.N.H. 43855 (Sowerby Collection). X 1.

PLATE 8

PLATE 9. OBINAUTILUW AND CENOCEKAW

Figures Page

1, 2 ObinaulUus jnilclira Kt)li;t vashi 386

Two views of liolotype from Palaeogene Nichinan forma-
tion at a cutting between Obi and Aburatsu in Province
Hyuga, Japan. X 1. From Kobayashi.

3-6 Cenoctrax roi nudum (Hyatt ) 3(j0

Four views of liolotype of IHgoniorrras Hyatt. Locality
and horizon unknown; Hyatt labels it ''obviously from the
Oolite." X 1. M.C.Z. 5:217.

•C

':%

^^^

PLATE 9

PLATE 10. TIEMINAUTILUS AND PSEUDOCENOCERAS
Figures Page

1 , 2 llemiiumUlu.s Mixbii (Morris ) 434

Two views of holotype, X 1, from Lower Greensand,
Atherfield, Isle of Wight. B.M.N. H. 47019 (Siixli.y Collec-
tion).

3, 4 raeudoeenoctras Idrylllierluinu.s (d 'Orbiguy) 384

Two views of plesiotype, X 1, from Craie Chloritee, Rouen,
Er:tiu-e. B.M.N.H. C1016.

/WM

ZJ.. I

M . if ■

/ /'

^^■- \'\ V

PLATE 10

PLATE 11. CARINONAUTILUS ARIYALURENSTR

Figures Page

1-3 Carinonautilus ariyalurensis Spengler 385

Three views of holotype. X 1, from Ariyalur group (Up
per Cretaceous), Trichinopoly district, India. From
Spengler.

^-J

PLATE 11

PLATE 12. PSETTDONAUTILUS GEINITZT

Figures Page

1, 2 Psevdonautiltts gcinitzi (Oppel) 397

Two views of the holotype, X 1, from Stramberger beds
(Upper Jurassic) of Stramberg, Moravia. FroTu Zittel.

fi;r^3S^*f*^-^

^5*^^..

'■~-*-*^Ta<^gss?r--

PLATE 12

PLATE l.-.. PAR.A(^EN0CEKA8 HEXAGONUM

Figures Page

1, - ParacenoceraH hcxac/onum (Sowerby) 402

Two views of liolotype, X 1, from Corallian Oolite, locality
uuknown. R.M.N.H. 43856 (Soworhy roUection. MV
52i)-f2).

PLATE 13

PLATE 14. AI^LACONAUTILUS AND SOMAT.TNAUTTLUS

Figures Page

1, 2 AvlaronautiluN sr.rcarwafii.s (Pictet) 4()fl

Two views of holotypc, X 0.75, from beds with Terc-

bratulo dipliyoideft of Berrias (Ardeche), France. From

Pictet.
n, 4 Snmalinaiitilvs anticpnis (Daf-que) 411

Two views of holotype, X 1, from ITpper .Turassie strata

of SomalilaTid. From Daeqiie.

^mmm^~.

\i!.

PLATE 14

I'LATE 15. TITHONOCERAS ZITTELI

Figures Pago

1 . L' Tithonooerati zitteli Retowski 410

Two views of holotype, X 1, tioni I'lijier JuiMssie strata
of Crimea. From Retowski.

.*- -

PLATE 15

I'LATE Ui. CVMATOCERAS PSEUDOELEGANS

Figures Page

1, 2 Cymatoceru.s puc/ndoeleyans (d 'Orbiguy j 418

Two views of cast of holotype ? (see text page -ilS) in
Museum of N.,t,„.j,i History. Paris. X 0.5. B.M.N. 11.
C526(i.

'^^

•^

PLATE 16

PLATE 17. PEOCYMATOCERAS SUBTRUNCATUS

Figures I'agf

1,2 Procymatoceras nubtruncutus (Morris and Lycett) 428

Two views of topotype, X 0.5, from Great Oolite, Min-
chinhampton, Gloucestershire, England. B.M.N. H. 21275.

PLATE 17

PLATP: ]8. CYMATONAUTILUS .FULII

Figures Pago

1-4 Cymatoriautilus juiii (,d'Url)igny j 430

Four views of type species from strata of (Jallovian age
near Montreuil-Bellay and OhaufPoui-, Fraiuc. P^om
Gueranger.

,,5^*H*r

PLATE 18

PLATE 19. PABACYMATOCERAS ASPER

Figures Page

1, 2 Paracymatoceras asper (Oppel) 427

Two views of holotype, X 1, from the Stramberger beds
(Upper Jurassic) of Stramberg, Moravia. Proni Zittel.

'^^V.^-;. :;'<'V-- -'^ -'^
it^fV ^- :■■ . - c. v./ . %

.?A^-

PLA'I'K 19

PLATE 20. ANGLONAUTILUS UNDHLATT'S

Figures Page

1, 2 AnglonauUlus undulatns (Sowerby) 430

Two views of topotype, X 1, from Lower Greensand, Nui-
field, Surrey (fide L. F. Spath, p;Msonal communication).
B.M.N.H. 6(53315. (Sowerby Collection MC 182).

-r.-^^^^K-^as^

>■'> , ■•

CvJ

_2S-:-

PLATE 20

PLATE 21. EUCYMATOCERAS PLICATUS

Figures Page

1, 2 Eucymatoceras plicatus (Fitton) 431

Two views of plesiotype, X 0.73, from Lower Greensand,
Atherfield, Isle of Wight. B.M.N.H. C335.

C«J

N.;^

■i\

PLATE 21

PLATE 22. ])ELTOCYMATOCERAS LEIOTROPIS

Figures Page

1, 2 Ddtacjiiiiatorcids Ic'ioiropia (.Schliite'r') 437

Two views of type specimen, X 0.4. frotri Tppcr Vvi.'
tMceous I'oiin.-iticii of Essen, Gei-ni;in.v. From Scliliiter,
1876.

" ' .%#*,. ^".si'r^

FLATK 22

I'LATK L'.l. KHlCVMATOCEJtAW VAEJ^SENSIW

Figures Page

1, 2 Epicyniatoctra.s vatLsc.nnis (Biru-khoisf ) 43§

Two views of type species, X ], from Upper Cretaceous
(Maestriehliaii) formation at l>imlioing, Belgium. From
Schliiter, 187(i.

-^r*^5<M^

PLATE 23

PI^ATP: 24. ("iMOMIA BUKTINI

Figure Page

1 (.'vmo'tnM burlini (Galeotti) 445

lieproiliu-tlon of original illustration of holotype, from the
Eocene n(>;n Rruxelles, Belgium. X 0.0. From Burtin.

PLATE 24

PLATE 25. ANGULITHES AND "BISIPHYTES"

Figures Page

1, 2 An(julUh( li triangulari.!< Moiitfoit 452

Figure 1, from Montfort 1808, p. (i ; Figure 2, from Mont

fort, 1802. pi. 49.
3 ' ' Bisiplnites" reticulatux Montfort ;5(il

View of holotype of the genus Bisiphytes from beds of

uncertain age near Somberuon, France. X 1.

PLATE 25

PLATE 26. ANGULITHES TRIANGULAEIS

Figures Page

1, 2 Angulithes triangularis (Montfort) 452

Two views of plpsioty])e, X 0.(55, from Craie Chloritee
(Cenomanianj Escragnolles, France. B.M.N.H. 37097
(Astier Collection).

v^^_

PLATE 26

PLATE 27. HERCOGLOSSA ORBIOULATA

Figures Page

1, 2 HcrcogJossa orbiciilata (Tuouiey ) 457

Two views of neoholotype from the Clayton foniijitit)ii on
Pine Barren Creek, about 5 miles nortli of Alleiiton, Ala-
bama, X 0.5. From Miller.

CVJ

■>t

>'4

^-<^:

PLATK 27

PLATE 28. ATUROIDEA AND PSEUDAGANIDES

Figures Page

1. 2 Atnroidca parkinsoni (Edwards) 462

Lateral view (X 0.5) and septal view (reduced) of holo-
type, from London Clay at Harwich, England. From
Edwards.

3, 4 Pseiulaganides TcutclienMs (Waagen') 390

Two views of holotype from beds with Perisphinetes
obtusicosta of Wanda, Kachh (Cutch) Paki-stan. From
Waagen.

PLATE 28

Harvard MCZ Libra

liiliniiii

3 2044 066 303 868

Date Due

oef-3-n992

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Bulletin of the Museum of Comparative Zoology at Harvard College"

See other formats