0;

<=o

o
m

EVOLUTION OF
MAMMALIAN MOLAR TEETH

BIOLOGICAL
STUDIES AND ADDRESSES.

READY.

I. EVOLUTION OF MAMMALIAN MOLAR
TEETH.

JN PREPARATION.
II. ADAPTIVE RADIATION OF MAMMALS.

III. EVOLUTION OF THE VERTEBRATES.

IV. RECTIGRADATIONS IN EVOLUTION.
V. SOME GREAT NATURALISTS.

VI. SEVEN FACTORS OF EDUCATION.

fk,
BIOLOGICAL STUDIES AND ADDRESSES. VOL I

EVOLUTION OF
MAMMALIAN MOLAR TEETH

TO AND FROM THE TRIANGULAR TYPE

INCLUDING COLLECTED AND REVISED RESEARCHES ON

TRITUBERCULY AND NEW SECTIONS ON THE FORMS

AND HOMOLOGIES OF THE MOLAR TEETH IN

THE DIFFERENT ORDERS OF MAMMALS

BY

HENRY FAIRFIELD OSBORN

Sc.D., LL.D., D.Sc.

DACOSTA PROFESSOR OF ZOOLOGY IN COLUMBIA UNIVERSITY

CURATOR OF VERTEBRATE PALAEONTOLOGY IN THE AMERICAN MUSEUM

OF NATURAL HISTORY

EDITED BY

W. K. GREGORY, M.A.

LECTURER IN ZOOLOGY IN COLUMBIA UNIVERSITY

JUtu llork
THE MACMILLAN COMPANY

LONDON: MACMILLAN AND CO. LTD.
1907

to

GLASGOW: PRINTED AT THE UNIVKKS1TV PK1 -.s
BY ROBERT MACLEHOSE AND CO. LTD.

PREFACE

THE older odontography or description of teeth treated each type as
a distinct and perfect form in itself. Curler's Odontographie '/<•--.•
Mammiferes, (licliel's Odontographie, Owen's Odontography are examples
of more or less comprehensive treatises in the pre-evolutionary spirit.
They antedate the discovery of what may be called the ' new odonto-
graphy,' which is based upon the unity of dental type, upon the
evolution of the teeth of mammals from a common reptilian prototype,
and which treats of each form in relation to its origin, its descent, its
gradual complication, and the laws of analogous evolution or independent
production of similar forms.

The new odontography centres around the ' tritubercular theory ' of
Cope. This theory had wider acceptance ten years ago than it has
to-day : there has been a strong reaction against certain features of it on
the part of many of the most able anatomists. This is partly due to
misunderstanding, partly to the fact that all the evidence has never
been fully marshalled, partly to the discovery of new embryi (logical
and palaeontological evidence which may disprove certain features of
the theory : but chiefly to the fact that some of the most decisive
and convincing palseontological evidence in support of the theory has
not been clearly advanced. It is hoped that this collection of the
contributions of the writer, with additional observations, illustrations,
and data, and with a discussion of various other theories and criticisms
will serve to convince the reader and student that the new odonto-
graphy in its general principles rests upon an adequate basis of evidence
and, while subject to modification in many details, marks a turning
point in the science of the teeth.

vi PREFACE

»

The present volume treats only the primary evolution of the molar
and premolar teeth of mammals, and is thus more restricted in scope
than the admirable ' Dental Anatomy ' of Tomes which covers the teeth
of the vertebrates generally.

The writer is especially indebted to his assistant Mr. W. K. Gregory
for invaluable assistance not only in bringing together and rearranging
the essays and figures, but for many original suggestions, and for the
critical resmnt of the opposing views which is set forth in the last
chapter.

HENRY F AIRFIELD OSBORN.

COLUMBIA UNIVERSITY AND

A M KRICAN MUSEUM OF NATURAL HISTORY,

September, 1907.

CONTENTS

PAGl

INTRODUCTION,

CLASSIFICATION OF THE MAMMALIA ADOPTED IN THIS BOOK, n

CHAPTER I.

THE TRANSITION FROM THE REPTILIAN TO THE MOST PRIMITIVE
MAMMALIAN TEETH.

1. The Upper Triassic Mammals Dromatherinm and Microconodon, - 18

2. A new classification of the Mesozoic Mammals, - 21

3. Illustrations of the chief Triconodonta and Trituberculata, 24

4. The origin of the tritubercular type sought in 1888 among the

Mesozoic Mammalia, 31

CHAPTER II.

FIRST OUTLINE OF TRITUBERCULAR EVOLUTION IN MAMMALS, 36

CHAPTER III.

TRITUBERCULY IN RELATION TO THE HUMAN MOLAR TEETH AND
THE PRIMATES.

1. Ontogenetic development of the teeth, 48

2. "The history of the cusps of the human molar teeth," p. 55. The

concrescence theory, p. 57. Mechanical relations of the upper
and lower teeth, p. 60. Evidence that the upper human molars
were triangular, p. 62

viii CONTENTS

PAGE

CHAPTER IV.

TRITUEERCULY IN ITS APPLICATION TO THE MOLAR TEETH OF THE
UNGULATES OR HOOFED MAMMALS. COMPLETION OF THE
NOMENCLATURE.

1. Disadvantages of previous systems of nomenclature of the molar cusps, 66

2. Methods of analysis of molar elements. Nomenclature of the molars

of Ungulates, - 66

3. Application of the theory of trituberculy to the Perissodactyla, p. 72.

The horse molar, p. 72. The rhinoceros molar, p. 72.

CHAPTER V.
TRITUBERCULY : A REVIEW DEDICATED TO THE LATE PROFESSOR COPE, 74

Polybuny, p. 74. Comparison of nomenclatures, p. 76. Tritubercular
homologies, p. 77. The early stages of sexituberculy, p. 80. Lower
molars, p. 80. The nomenclature of the molar cusps and crests, p. 82.
The evolution of the Ungulate molar, p. 83.

CHAPTER VI.
CHRONOLOGICAL OR GEOLOGICAL SUCCESSION OF MOLAR TYPES.

1. Reptilian Ancestors of Mammals in the Trias, -

2. The Triassic Mammals, 94

3. Mammals of the Jurassic, 94

4. Upper Cretaceous Mammals, 95

5. Basal Eocene Mammals, 98

6. Lower Eocene Mammals, 98

Resume of Geological Succession of Types, 99

CHAPTER VII.

THE ORIGINAL OR PRIMITIVE STRUCTURE OF THE MOLAR TEETH IN

THE DIFFERENT ORDERS OF MAMMALS, 100

The major classification of the Mammalia, p. 100. Protodonta, p. 101.
Allotheria or Multituberculata, p. 101. Monotremata, p. 105. Mar-
supialia, p. 108. Molar teeth of Multituberculates, Marsupials (?), and
Placentals in the Upper Cretaceous, p. 115. Insectivora, p. 117.
Cheiroptera, p. 129. Carnivora, p. 131. Evolution of carnasial teeth in
Creodonta and Fissipedia, p. 135. Convergence of upper carnassials in
Creodonta and Fissipedia, p. 1 38. Fissipedia, p. 1 42. Rodentia, p. 1 44, and

CONTEXTS ix

PACK

Simplicidentata, \>. 14."). Duplicidentata, ]». 148. Tillodontia, )>. 151.
Superorder I'ngnlata, p. H>3. Aml>l\p<>da, p. Mil. Condylarthra, p. K!N.
Artiodactyla, p. 171. Perissodaetyla, p. 171. Chalicotheroidea or An-
rylopnda, ]>. 1S4. Hyracoidea, }>. IS."). Proboscidea ]>. isfi. Sirenia,
p. 18S. South American Ungulates, p. 189. ( 'etacea, p. 19O. Zeug-
. 191.

CHAPTER VIII.

EVOLUTION OF THE PREMOLAKS, 193

Premolars in primitive M annuals, p. 193. Adaptation <>f prenmlars,
|i. 191. Various upper premolar types, p. 194. Cusp addition in the pre-
inolars, p. 19"). Superior premolars, ]). 19"). Inferior premolars, p. 198.

CHAPTEK IX.
OBJECTIONS AND DIFFICULTIES, AND OTHER THEORIES.

I. That the tritubercular type is not primitive :

1. The plexodbnt or progressive simplification theory of Ameghino, - 201

•2. Objections by Fleischmann and Malm answered by Scott, 204

3. The primitive polybuny theory, 205

II. That the Cope-Osborn theory of the origin of the superior molars

is incorrect, 208

1. (Aisp homologies founded on embryogeny, p. 208. Order of em-
bryonic cusp-development in Insectivora, p. 210. Summary of
Woodward's conclusion, p. 212. Ontogenetic order according to
Marett Tims, p. 213. The premolar analogy, theory, p. 215.
Palseontological difficulties in the premolar analogy theory, p. 217.
Gidley's restudy (1906) of Jurassic Mammals supports embryo-
geny and the premolar analogy theory, p. 219. Addendum: the
" trituberculy " of Zalambdodonts a secondary acquirement or
u pseudo-trituberculy," p. 225. Conclusion, p. 227.

CHAPTEK X.
RECTIGRADATIONS IN THE TEETH, 229

Homoplasy as a law of latent or potential homology, p. 229. Homologv,
p. 237. Lankester's reply to the preceding article, p. 238.

BIBLIOGRAPHY, 240

INTRODUCTION.

THE teeth are the hardest of the tissues, and, unlike other tissues, are
not improved, but on the contrary constantly worn away and finally
destroyed, by prolonged use. Yet they are also the most progressive
of the tissues. For example, in the evolution of the horse, no other
system of organs undergoes so profound a change as that of the teeth.

The fitness which they present for every possible mode of capturing
and eating fills volumes of the older works on descriptive anatomy and
teleology, and is even now constantly disclosing new and fascinating
subjects for the study of adaptations.

Because of their hardness the teeth are the most generally and
perfectly preserved of all fossilized organs ; hence they are the especial
guides and friends of the palaeontologist in his peculiar field of work
from imperfect evidence. Thus it happens also that the palaeontologist
has been obliged to study the teetli in more detail even than has been
done by the comparative anatomist or zoologist.

All this use of the teeth for teleological, descriptive, and taxonornic
purposes is, however, entirely aside from the main purpose of the present
volume, which is, to set forth the mode of evolution cruelly of the
complex crowns of the molar teeth of mammals, how the main types
originated, and how they can be compared with each other and with
those of reptiles. The evolution is not, however, traced to its final
modifications in the elaborate hypsodont and tubercular types, but only
so far as the formation of the fundamental patterns.

That comparison can be made on a grand scale most attractive to
the student, both of homologies and analogies, is a recent discovery.
It dates back only to the year 1883, when what may be called the
law of trituberculy was discovered by Cope.1 No harmony existed in
our ideas and descriptions of the grinding teeth of mammals previous to
that time, although Huxley in his discussion of the teeth of the Insec-
tivora had anticipated the discovery of such a harmony. Many of
the details and of the broader outlines of the law were either touched

1 American Xatwalist, April 1883, pp. 407-408.
A

EVOLUTION OF MAMMALIAN MOLAR TEETH

upon or fully discussed by the great comparative anatomist Cope, who
dwelt, however, chiefly upon the basal Eocene stage. In 1887 the
present writer took up the subject in the earlier Mesozoic stages of
the evolution of the mammals, and in later years pursued it into the
later Eocene and all subsequent Tertiary stages.

The theory was everywhere welcomed as a decided advance on
the old odontology and odontography, in which there was no unifying
principle either of homologies or of nomenclature. It has been applied
more or less throughout the whole class of mammalia, first in palseonto-
logical, then in embryological and anatomical lines ; but thus far only
to a rather limited extent by zoologists, or students of living mammals,
and, so far as the writer knows, still less by anthropologists. It has
been critically examined, discussed, and either partly or wholly adopted,
or entirely rejected as unproven, by various authors.

The chief contributors to the development and critical examination
of the theory are the following :

From the pakeontological standpoint, Cope, Osborn, Schlosser, Scott,
Earle, Kutinieyer, Lydekker, v. Zittel, Ameghino, Goodrich, Wortman,
Smith Woodward, Gidley.

Among the zoologists who have either advocated or opposed the
theory are Lankester, Newton Parker, Fleischmann, H. Allen, Forsyth
Major, Dybowski, Winge, Sims, Beddard.1

Among anatomists, Schwalbe and Dwight have accepted the theory,
while Howes failed to find sufficient evidence for it.

The most influential opponents of certain features of the theory
are those who have examined it embryologically, namely, Kiikenthal,
Rose, Taeker, Leche, M. F. Woodward.1

Four distinct principles have been developed in connection with
the general theory, as follows :

I. FIRST PRINCIPLE. THE PRIMITIVE TRITUBERCULAR TYPE.

The discovery of the oldest fauna of the age of mammals, or Tertiary
period, near the Puerco Canon in Northwestern New Mexico, was
announced by Professor Edward D. Cope in 1879. These small and
strange fossil mammals exhibited a general similarity in all the molar or
grinding teeth, even among animals which evidently had great diversity of
feeding habits. This similarity consisted in the possession of three main
tubercles on the crowns of loth upper and lower molars, disposed in tri-
angles* This was evidently a primitive type of molar tooth, and in 1883

1 These names are placed in the general order in which the criticisms appeared. (See
pp. 200-227.)

*[In addition to the main tubercles there were often smaller cuspules, "styles "and
" talons," but the upper molars of the Carnivores and Ungulates were all roughly speaking
" tritubercular," the lower molars all " tuberculo-sectorial " or of plainly derived types.]

INTRODUCTION

Professor Cope 1 appropriately named it the tritubercular ////"'.- By com-
parison with the teeth of more recent animals, the further conclusion
was reached that the tritubercular f ///><• n-as ancestral to many if not to all
of the higher types of molar teeth. This is one of the most important
generalizations ever made in mammalian comparative anatomy ; it
outranks in importance the proof of the primitive pentadactyl nature
of the feet of hoofed animals. In the writer's opinion, the evidence in
favour of it is so overwhelming that primitive trituberculy is no longer
an hypothesis or a theory, but an established fact. In the accompany-
ing table it will be seen that those orders in which generalized ancestral
forms are positively known to possess tritubercular molar teeth (I)
and those orders in which primitive tritubercular teeth or some of
the immediately derivable types are occasionally observed, although
the line of descent has not actually been traced (II), far outnumber
those (III) in which we must reason by analogy, because we have as
yet no positive light on the descent of the teeth.

I.

Mammalian orders in which
ancestral forms are posi-
tively known, which exhibit
tritubei'cular or plainly de-
rived types of molars.

Edentata TYeniodonta

(Ganodonta) (p. 151).
Insectivora primitive

(Mesozoic) (pp. 26-30).
Insectivora (p. 117).
Marsupialia Polyproto-

dontia (cf. Oligocene

Perat/ierium, p. 1U9).
Carnivora Creodonta (p.

131).
Oarnivora Fissipedia (p.

135).

Tillodontia (p. 151).
Primates (p. 157).
Amblypoda (e.g. Panto-

lambda p. 165).
Condylarthra (e.g. Proto-

gonodon, p. 169).
Artiodactyla (e.g. Trigono-

lestes, Homacodon, p.

171).
Perissodactyla (e.g. Hyra-

coth eriv.m, Systemodon^

p. 174).

II.

Orders in which tritubercular
or plainly derived types
of molars are occasionally
observed although the line
of descent has not actually
been traced.

Marsupialia Diproto-

dontia (p. 109).
Rodentia Duplicidentata

(p. 148).
Rodentia Simpliciden-

tata (p. 146).

Ancylopoda (cf. Schizo-

tln'rium, Chalicotherium,

(p. 184).

Zeuglodontia (p. 191).
Hyraeoidea (p. 185).
Proboscidea (e.g. Mceri-

therui.m, p. 186).
Sirenia (cf. Halitherium

veronense, p. 189).
Toxodontia and other

South American Ungu-

lata (p. 189).

Ill

Orders in which we have as
yet no positive light on the
descent of the molars.

All other Edentates (p.

152).

Monotremata (p. 107).
Multituberculata (p. 101).
Cetacea3 (p. 190).
Carnivora Pinnipedia (p.

143).

1 American Natural i*t, April, 1883, pp. 407-408.

2 N. B. The lower molars generally show also a talonid or heel.

3 Zeuyhdon not included.

4 EVOLUTION OF MAMMALIAN MOLAR TEETH

This main principle of the tritubercular theory has been widely but
by no means universally accepted. Able comparative anatomists, chiefly
Ameghino, Rose, Forsyth-Major, have urged against it that a molar
tooth with a number of tubercles is still more primitive. According to
this rival view, commonly known as the polybuny theory (TroXvs,
many, ftowos, a hillock), the tritubercular type, even in the Cretaceous
period, was the result of a secondary suppression of some of the numerous
original tubercles. This opposing view deserves fair consideration and
a clear statement (see p. 205).

This first generalization as to primitive trituberculy in the later
Mesozoic and Tertiary periods must be clearly distinguished from the
following second generalization as to the earlier Mesozoic origin of
trituberculy.

II. SECOND PRINCIPLE. THE ORIGIN OF THE TRITUBERCULAR TYPE
FROM THE SINGLE REPTILIAN CONE.

Professor Cope's active and inquiring mind did not, however, stop
at this point. He asked himself, " If the oldest Tertiary mammals
exhibited a common tritubercular type, what do the mammals which
lived during the Mesozoic period, or age of reptiles, show as to the
origin of the tritubercular type ? " He had neither the time nor the
material to enter thoroughly into this inquiry ; but he looked into
Owen's memoir on Mesozoic mammals far enough to clearly perceive
that the tritubercular type developed during the long Mesozoic period,
from an ancestral reptilian type of tooth consisting only of a single
cone or cusp. Therefore he made a second generalization : that the
tritubercular type sprang from a single conical type by the addition of lateral
denticles. " I have already shown," he says,1 " that the greater number
of the types of this series have derived the characters of their molar
teeth from the stages of the following succession. First, a simple cone
or reptilian crown, alternating with that of the other jaw. Second,
a cone with lateral denticles. Third, the denticles to the inner side of
the crown forming a three- sided prism, with tritubercular apex, which
alternates with that of the opposite jaw."

III. THIRD PRINCIPLE. CUSP ADDITION OR DIFFERENTIATION.

In the second generalization is, however, involved a third, which has
also been the subject of wide difference of opinion, namely, the successive
addition of new denticles, cuspules or smaller cones on the sides of the
original reptilian cone ; this may be simply known as the cusp addition

1 Oriyin of the Fittest, p. 347.

INTRODUCTION 5

principle ; it is a process analogous to budding or outgrowth in other
tissues.

An opposed theory, advanced by Ameghino, Rose, and Kukenthal is
that of concrescence, namely, that from an original large supply of conical
reptilian teeth in the primitive longer jaws, and from the multiple
succession or replacement of such teeth, the cones were clustered or
grouped, by fours or more, and thus arose respectively the quadri-
tubercular, and multitubercular or polybunic types, the tritubercular
and triconodont stages being secondary (Ameghino).

IV. FOURTH PRINCIPLE. EEVERSED UPPER AND LOWER TRIANGLES.

Still another or fourth principle, entirely distinct from the fore-
going, is involved in the sentence quoted above, " Third, the denticles to
the inner side of the crown [in the lower jaw] forming a three-sided
prism, with tritubercular apex, which alternates with that of the
opposite [or upper'] jaw." This principle involves the theory which rests
upon strong but perhaps not altogether conclusive pakeontological
evidence (pp. 32, 43, 217) that in the lower molars the reptilian cone, is
external and the two denticles internal, while in the upper molars the reverse
is the case, namely, the reptilian cone is internal and the denticles are
external. This principle, if a true one, enables us to establish a kind of
serial homology between the main primary cones and secondary denticles
or cusps of the upper and lower teeth respectively. Osborn expressed
such an homology in a system of nomenclature (protocone, paracone,
metacone, etc.), which Professor Cope welcomed and accepted.

According to this principle, the evolution and relation of both the
upper and lower molars are those of a pair of reversed triangles in
every stage above the protodont and triconodont ; thus, it might be
simply known as the trigonal theory ; but since it was based by Cope
and the writer entirely upon the evidence afforded by the Mesozoic
molar teeth, it may be more strictly termed the ' PAL/EONTOLOGICAL
THEORY.'

As applied to the upper molars, this theory and the homologies it
involves with the lower molars have been far more vigorously and
generally opposed than either of the other principles ; in fact, the chief
weight of opinion has now gathered against it from three different
classes of positive evidence, namely, embryological, anatomical, and
palfeontological by comparison with premolar evolution, also from the
negative argument that the evidence at hand among the Mesozoic
mammals does not demonstrate the principle in the upper teeth.

The theory opposing the pakeontological theory in morphological
contrast first sprang from embryological evidence, and may, therefore,

6 EVOLUTION OF MAMMALIAN MOLAR TEETH

be known as the ' EMBRYOLOGICAL THEORY.' It is another instance of
the apparent conflict between palseontological and embryological evidence
as witnesses of the ancient order of development. The final verdict,
therefore, will be most interesting. Briefly, the embryologists, especially
Eose, Taeker, Kiikenthal, M. F. Woodward, Marett Tims, have shown
that in the upper molars the cusp which Cope and Osborn assert to
be the oldest, is, on the contrary, later in development than the paracone,

Order of Cusp Development

Ja I a ontology

FIG. 1. Order of cusp development as attested by Embryology (fide Rose) (left-band
numerals) and by Paleontology (fide Osborn) (right-hand numerals). Typical (synthetic)
dentition of Man, from Selenka. (The paraconid is not represented in human molars.
See p. 59 and Fig. 38.)

and therefore (?) not the oldest palseontologically. The oldest upper cusp,
according to the ' embryological theory,' is the antero-external cusp
(paracone), and this is homologous with the reptilian cone ; from this
there is a slowly evolving triangle in the upper molars . according to
the same authorities the reptilian cone is, however, central or apical
in the lower teeth, embryology quite 'agreeing with palaeontology. From
this it would follow that the only serial homologies which can properly
be established are those between the lower and upper reptilian cones.
The embryological theory, in brief, is to the effect that while in the
lower molars the central reptilian cone remained external and the two

INTRODUCTION t

denticles arose internally, forming a triangle, in the upper molars the
reptilian cone remained at the antero-cxternal angle and the two denticles
arose on the inner and posterior sides.

It will thus be seen that the difference between the palseontological
and embryological theories is radical. The latter finds not only strong
support but a beautiful illustration by analogy in the normal modes
of evolution of the simple premolar teeth into the complex molai
type (see pages 194-200). It is developed in what may be termed the
' PREMOLAR- ANALOGY ' theory.

The accompanying figure clearly demonstrates the fundamental
difference between these theories. There is no middle ground between
them. If the premolar analogy be correct, the Osbornian cusp homo
logics of the upper teeth are of less value, but the homological iimm -//r/W///v
should be retained for convenience because it has found its way so largely
into literature.

Although the writer has devoted a vast amount of time and study
to the development of the ' palseontological theory,' and is personally
inclined to believe in its tenability, there is no denying the great force
of the objections which have been urged against it and the need of
substantial proof from the early Mesozoic mammals, even in addition
to that which is freshly adduced in the present work. The arguments
pro and con are, therefore, partly stated in course of these collected essays,
and especially brought together in the new chapter IX. at the close.

ORIGIN OF THE TRIANGLE.

An important feature of the fourth principle is the position and mode
of origin of the cusps of the triangle.

Four different modes of origin have been suggested :

The first suggestion is that the triangle originated as such, the
denticles appearing from the outset on the inner and outer sides
respectively of the lower and upper cones.

The second suggestion is that the denticles first arose on the anterior
and posterior sides of the reptilian cone, and were secondarily rotated
inwards and outwards respectively ; the latter is known as the ' cusp
rotation ' theory, or rather hypothesis.

Osborn l set forth both the above modes guardedly in the following
language :

i£ There can be no doubt that the cusps seen upon the inner face
of the inferior molars of this genus [Spcdacotherium] are homologous
with the para- and ineta-cones, and there are several facts which

1 "The Structure and Classification of the Mesozoic Mammalia," .l»>n\ A<-nd. X«t. Sci.,
Phila., Vol. IX., No. 2, July, 1888, p. 243.

8 EVOLUTION OF MAMMALIAN MOLAR TEETH

support Cope's hypothesis 1 that they represent a stage of inward
rotation of cusps which were at an earlier stage in the same fore and
aft line with the main cusp. These are, that in Phascolotherium the
lateral cones are seen to be slightly internal to the main cone, so that
their median slopes descend upon the inner face ; in Tinodon, of a
later geological period, this disposition is slightly more pronounced ; in
Mcnacodon it is still more marked, but less so than in Spalacotherium.
These genera, although evidently in two different lines of descent, afford
the desired transition stages. The Spalacotherium [Peralestes] molar as
seen from above ~ has a striking resemblance to the anterior sectorial
triangle of the Stypolophus or Didymictis molar of the Puerco. It is,
in fact, sub-triangular, the superior molars probably having the lateral
cones rotated outwards, so that the upper molars form an alternating
series, the ridges connecting the main and later cones acting as
sectorial blades." Again, in the " Evolution of Mammalian Molars to
and from the Tritubercular " type, Osborn pointed out (Amer. Naturalist,
Dec., 1888, p. 1075) that "it has been assumed by Cope and the
writer (op. cit., p. 243) that the para- and meta-conids were first
formed upon the anterior and posterior slopes of the protoconid and
then rotated inwards, but it is also possible that they were originally
formed upon the inner slopes." There is thus evidence for cusp
rotation, but it is not an essential part of the tritubercular theory,
because, as above stated, the denticles may have arisen on the inner
and outer sides of the cone from the outset (see p. 33).

The third suggestion is that after the main cone had been estab-
lished the lateral cusps or denticles arose as dngules on the broad
external cingulum of the upper molars and from the broad internal
cingulum of the lower molars. This hypothesis, suggested by Osborn
(Mcsozoic Mamm., p. 245) from a study of the molar teeth of the
Jurassic Amblotheriidiv, has been supported by the observations of
Gregory. The comparison of the molar teeth of such forms as
Amblotherium, Phascolestes, Dryolestes (see pp. 29, 30), lends support to
this view, which is more fully discussed on page 33 (footnote).

The fourth suggestion or hypothesis is the newest ; it springs from
embryological evidence (Woodward, Tims) and from another interpreta-
tion of the palseontological evidence (Gidley). It ends with the idea
that the oldest cone in the upper molars is the paracone (of Osborn)
on the outer side of the crown, from it extends inward a broad ledge
like a heel which finally rises up and secondarily forms the prominent
protocone (of Osborn). According to this hypothesis the paracone is
the primary (or reptilian cone), the protocone is secondary or a deriva-

111 The Creodonta," American Naturalist, 1884, p. 259.
2 Owen, The Mesozoic Mammalia, Plate I., Fig. 32c.

INTRODUCTION 9

tive cone. This hypothesis accords with the embryological theory and
the premolar analogy theory.

It is quite possible that the trikibercular or triangular stage arose
independently in different groups of animals, by two or possibly three
different modes of origin, as outlined in the four sugg* \stions above
advanced, on the principle of convergence of similar forms from dis-
similar beginnings. Tin's, however, does not invalidate the theory of
the passage of the majority if not of all the higher mammalia through
the tritubercular stage, however arrived at.

SUMMARY.

To sum up, it must be clearly re-stated that the four great principles
of molar evolution do not stand or fall tocjctln-r. The first or primitive
trituberculy principle is now almost undeniable for the majority of
mammals ; entirely apart from the disputed question of the original
homology of the cusps of the upper and lower teeth, there is no
question whatever as to the beautiful and almost incredible homo-
logies between the cusps of the molar teeth in the most diverse orders

pa*

FIG. 2. Two divergent derivatives of the tritubercular molar pattern. Right-hand
figure, a grinding molar of the modern horse ; left-hand figure, a sectorial molar of a Flesh-
eating mammal (Oxynena) of the Eocene Period.

of mammals. The upper carnassial of the Carnivora and the upper
molars of the Equida?, for example, are types adaptively so far apart,
it is small wonder the older odontologists did not even suspect the
existence of hornogeny or common derivation, through which we can now
compare cusp for cusp. The reptilian cone origin theory is next in
order of demonstration and acceptance : it has recently gained strength
by the very general admission that the Theriodont reptiles are at
least nearly ancestral to the mammals. The cusp addition theory also
finds more advocates at present than the concrescence theory, and rests
upon indisputable evidence. Finally, the greatest conflict of evidence
exists between the Cope-Osborn palpeontological and the embryological
plus premolar-analogy theories of the homologies of the upper and
lower cusps (see pp. 208-227).

10 EVOLUTION OF MAMMALIAN MOLAR TEETH

ARRANGEMENT OF THIS VOLUME.

The writer became interested in trituberculy while attempting to
monograph the Mesozoic mammals ; in course of this work many new
ideas came out, and many other ideas arose, apparently new, which
the writer subsequently found by more extensive reading had already
occurred to Professor Cope, and had been expressed in out-of-the-way
and unlooked-for places.

The writer thus had the opportunity of fully developing the Mesozoic
origin of trituberculy , a part which Professor Cope had been obliged
to leave in the stage of suggestion. Subsequently the writer took up
the molar teeth of the monkeys and lemurs, and then of the hoofed
animals, which strangely enough form a perfect morphological succes-
sion of types, and thus enjoyed the opportunity of working out \vhat
might be called the secondary and tertiary addition, suppression, or
modification of cusps as illustrated in the modern Carnivores and
Ungulates. These papers are republished here in chronological order
with editorial notes. All corrections and insertions by the Editor are
indicated in [ ] brackets. All figure references are to figures as arranged
and numbered in this volume. The original pagination of the reprinted
essays is not given.

Certain of the more purely philosophical and biological questions,
as distinguished from the anatomical, that is, matters of causation and
of evolution theory, are touched upon at one or two points in these
pages, but are more fully treated in another volume of these collected
papers, entitled Rectigradations in Evolution (cf. pp. 228-239).

CLASSIFICATION OF THE MAMMALIA ADOPTED IN

THIS BOOK

(See especially pages 91-192).

CLASS MAMMALIA LINNJEUS

Sub-class 1. PROTOTHERIA Gill (egg laying mammals)
Infra-class. ORXITHODELPHIA De Blainville

? Order : *Protodonta Osborn (Of uncertain systematic position)

(p. 18.)

Dromatkerium ) .,

,,. 7 >(Tnassic)

Microconot/oii }

Order: Monotremata Geoffrey Saint Hilaire (p. 105.)

Ornithorhynch t/.s
Echidna

? Order : * Allotheria Marsh (Multituberculata Cope), of un-
certain systematic position (p. 101.)
Examples
Playiaulax
Poli/mastodon
Tritylodon
Microlcstes

Sub-class 2. EUTHERIA Gill (Viviparous mammals)

Infra-class 1. DIDELPHIA De Blainville (Metatheria Huxley)

? Order : *Triconodonta Osborn (Of uncertain systematic posi-
tion, usually regarded as carnivorous
Marsupials) (p. 21.)
Examples
Amphilestes
Triconodon

Order: Marsupialia Illiger (Pouched mammals) (p. 108.)

Sub-order : Polyprotodontia Owen
Examples

Opossums (Didelphys)
Dasyures
Thylacynes, etc.

The asterisk (*) denotes an entirely extinct order.

12

EVOLUTION OF MAMMALIAN MOLAR TEETH

f.

05

03
£

03

.s

So
t§

O

Sub-order: Diprotodontia Owen (p. 109.)

Examples
Kangaroos
Phalangers, etc.

Infra-class 2. MOXODELPHIA De Blainville (Placentalia of authors,

Eutheria Huxley)

Order : * Pantotheria Marsh (Trituberculata Osborn, Insectivora

Primitiva Os.) (p. 22.)
Examples
Amp hither ium
Dryolestes
Amblothcrium

Order: Insectivora Cuvier (p. 117.)

Examples
Moles
Shrews

True Hedgehogs (Erinaceus)
Tenrecs (Gentetes), etc.

Order : Dermoptera Illiger

" Flying-lemur " (G-alcopithecus)
Order: Cheiroptera Blumenbach (p. 129.)

Sub-order : Megacheiroptera Dobson
Pteropodidse (fruit bats)

Sub-order : Microeheiroptera Dobson

Examples
Vespertilios
Vampires
Nose-leafs, etc.

Order: Ferse Linnaeus (Carnivora of authors) (p. 131.)

* Sub-order : C r e o d o n t a Cope (p. 1 3 2.)

Examples
Patriofelis
Hycenodon
Mcsonyx

Sub-order : F i s s i p e d i a Flower (Carnivora Vera Flower)

(p. 135.)
Examples
Cats (Felidse)
Civets (Viverridie)
Hysenas (Hyamidae)
* Extinct.

CLASSIFICATION OF THE MAMMALIA

13

03

c3
f— (

a

.2

—

to

-p

rH

O

o

Dogs (Canid;r)
Eaccoons (Procyonida?)
Bears (Ursidn-)
Mustelines (Mustelid;e)

Sub-order : Pinnipedia (p. 143.)

Eared or Fur seals (Otariidae)
Walruses (Odobsenidse)
Earless seals (Phocida?)

Order: *Proglires Osborn (of uncertain relationships) (p. 145.)

Example
Mixodectes

Order: Rodentia Vicq d'Azyr (p. 144.)

Examples

Squirrels (Sciurid*)
Beavers (Castoridae)
Eats (Muridae)
Porcupines, Cavies, etc. (Hystriconiorpha)

Order: *Tillodontia Marsh (p. 151.)

Examples
Tillotherium

Estlwnyx

Order: *T3eniodonta Cope (Ganodonta Wortman, possibly

related to the Edentates) (p. 151.)
Examples
Hcmiganus
Psittacotherium
Calamodon

Order : Edentata (Vicq d'Azyr) (Xenarthra Gill) American

Edentates (p. 151.)
Examples

Ant bears (Myrmecophagidai)
Ground Sloths (Megatheriidte)
Sloths (Bradypodidte)
Armadillos (Dasypodida?)
Glyptodonts (Glyptodontidae)

Order : Pholidota Weber

Manis (Scaly Anteater)
Order: Tubulidentata Flower (p. 152.)

Oryctcropus Aard-A"aark
* Extinct.

14

EVOLUTION OF MAMMALIAN MOLAR TEETH

S

13
<D

CO

a

03

O

O

cd

CD

<4H
O
O

•s

o
O

Order: Primates Linmeus (p. 157.)

Sub-order: *Mesodonta Cope (American leinuroids)
Examples

Notharctus
Anaptonwrphus

Sub-order : L e m u r o i d e a Mi v a r t

Examples
Lemurs
Galagos

Aye-Aye (Cheiromys)
Tarsier, etc.

Sub- order : A n t h r o p o i d e a M i v a r t

I. South American or Platyrrhine monkeys

Marmosets (Hapalidse)
Capuchins, etc. (Cebidte)

II. Old World or Catarrhme monkeys

Baboons, Apes (Cercopithecidse)
True Monkeys (Semnopithecidse)
Anthropoid apes (Simiidse, Gibbon, Gorilla, Chim-
panzee)
Man (Hominida?)

( Order: *Condylarthra Cope (p. 168.)

Examples
Protogonodon
Euprotogonia
Phenacodus

Order: *Amblypoda Cope (p. 164.)

Examples
Pcriptychus
Pantolambda
Coryi^hodon
Uintatlicrium

Order : Artiodactyla Owen (" even-toed " hoofed mammals) (p.

171.)
Examples
Pigs (Suidce)

Hippopotami (Hippopotamidse)
Deer (Cervidos)
Giraffes (Giraffidse)

* Extinct.

S

o

o

o

r-j

O

CLASSIFICATION OF THE MAMMALIA 15

Prongliorn Antelope (Antilocapridse)
Sheep

Goats

- Bovidse

Antelopes

Oxen

Camels (Camelidre)

Oreodonts (Oreodontidse)

Tragulines or Chevrotains (Tragulidse)

Order: Perissodactyla Owen ("odd-toed" hoofed mammals) (p.

174.)

Examples

Tapirs (Tapiridie)
Horses (Equidae)
Palffiotheres (Paleeotheriidte)
Titanotheres (Titanotheriidse)
Lophiodonts ( Lophiodontidae)
Rhinoceroses ( Khinocerotida? )

? Order : *Ancylopoda Cope (probably an early Eocene offshoot

of the Perissodactyla) (p. 184.)
Examples
Macr other ium
Chalicotherium

Order: Proboscidea Illiger (p. 186.)

Examples
Mceritherium
Dinothcrium
Mastodon
Elephas (Asiatic and African Elephants, Mammoths)

Order: Sirenia Illiger (p. 188.)

Examples

Manatees (Trichechid;e)
Dugongs (Halichoridse)
Ehytina (Pihytinidte)

Order: * Embrithopoda Andrews (Barypoda Andrews)
Arsinoithcrium

Order: Hyracoidea Huxley (p. 185.)

Dassies (Hi/rax, the Coney of the scriptures)
Itfegalohyrax, etc.

* Extinct.

16

EVOLUTION OF MAMMALIAN MOLAR TEETH

2

o

c

o
O

CO
CD

O

c3

4^>

CD

o

O

O

Super-order: NOTUNGULATA (Both) Scott (South American

Ungulates) (p. 189.)

Order I. *Toxodontia (Owen) Scott
Sub-order 1. *Toxodontia Scott
Examples
Toxodon
Nesodon

Sub-order 2. *Typotheria Zittel
Examples

Pachyrucos

Icochilns

Typotherium
Sub-order 3. * H o m a 1 o d o t h e r i a Scott

Homcdodon totherium

Order II. * Astrapotheria (Ameghino)
Astrapotherium

Order III. *Litopterna Ameghino (p. 189.)
Proterotherium
Macrauchenia

Order: Pyrotheria Ameghino (short-footed South American

Ungulates)
Pyroiherium

Order: * Zeuglodontia Van Beneden (p. 191.)

Examples
Protocetus
Zeuglodon

Order: Odontoceti Gray (toothed whales) (p. 190.)
Squalodonts (Squalodontida?)
Fluviatile dolphins (Platanistidse)
Dolphins (Delphinid^e)

Belugas ) /-rx -i i • • i \

&. . I (Delphmapteridae)

Narwhals J

Beaked whales (Ziphiidse)

Sperm whales (Physeteridse)

Order : Mystacoceti Gray

Whalebone whales (Balsenidae)
Plight whales (Balccna)
Humpbacked whales (Megaptera)
Finbacked whales (Balcenoptera)

* Extinct.

TERTIARY SECTIONS

17

GREAT PLAINS SECTION

PLEISTOCENE

PLIOCENE

OREGON,, SECTION^

=. ^E»fmvs*_f=-b--^E MIOCENE

~

JMERYGQCHCERUS

L EPTA UCTJ EN I AX- -_•

v_

QREODCXEt:

DICERATHERlUr*

OLIGOCENE

TITAMOTHERIUM-

UTAH SECTION

_ QlP-l>

WYOMING SECTION

OOOOQOOOOOOOO O O O O O OOOOO

(joooooaoooooooo ooooooo

OOOOOOOOOOOOOO O O OOOOO

EOCENE

trMAT6THERIUM

,r

NEW MEXICO SECTION

MONTANA SECTION'-':

.• ....... . .. .

CRETACEOUS

Fio. 2d. — Diagram showing the chronological and stratigraphic succession of the Cretaceous,
Tertiaiy and Pleistocene formations of the western states, in which fossil mammals are found.

B

CHAPTER I.

THE TRANSITION FROM THE REPTILIAN TO THE MOST PRIMITIVE
MAMMALIAN TEETH. BEGINNINGS OF THE NOMENCLATURE.

1.

THE UPPER TRTASSIC MAMMALS, DROMATHERIUM AND MICROCONODON.

[Reprinted.]

(Read before the American Philowphical Society, April 15th, 1887. Published in the

Proceedings, 1887, p. 109.)

THE mammalian jaws : discovered by Professor Emmons in the Upper
Triassic beds of North Carolina, and ascribed to a single genus. Droma-
tlicrium, were recently examined by the writer and found to belong to
separate genera. The type mandible of Dromathcrium is preserved in the
Williams College Museum, and differs widely from the mandible preserved
in the Museum of the Philadelphia Academy. These differences have
already been pointed out,2 but require to be more fully stated, as both
Professors Marsh and Cope have expressed doubts as to the distinct
separation of these genera. The accompanying lithographic figures bring
out the characteristic features of these mandibles much more fully than
in the pen drawings which accompanied my earlier description.

In many respects these genera agree with each other, and stand
separate from the Jurassic mammals of both England and America.
There is, first, a considerable diastema behind the canine, a very rare
feature in the division of Mesozoic mammals to which these genera
belong, although always present in the division to which Plagiaulax and
its allies belong, viz., the sub-order Multituberculata Cope.

J[The chief reason for considering these jaws mammalian is that they are composed of a
single bone, there being no evidence of the separation into dentary, articular, and angular
elements, as in the jaws of reptiles. H.F.O. November, 1904.]

z Proreediiiij/j of the Academy of Natural Sciences of Philadelphia, 1886, p. 359. I find
upon a second examination of Prof. Emmons' original figure, that I unintentionally criticised
it too severely in the former article, p. 359. While far from accurate, the figure is not so
misleading as I at first supposed.

TEETH OF THE MESOZOIC MAMMALS

19

Dromaiherium has three premolars and seven molars, Imt the number
in Microconodon is quite uncertain, as only four of the series are preserved.
The molars agree in one particular, which separates them widely from
other Mesozoic genera, viz., in the imperfect division of the fangs. This
division is indicated merely l>y a depression at the base of the crown, as
in the genus Dim<'tro<I<>ii. among the Theromorph * reptiles.

p

FIG. 3. No. 1. Microcoiwdoii tenuirogtris. The outer surface of the right mandibular rarnus,
enlarged. The two premolars preserved are the first and third, with the fang of the second
between. The space behind the third was occupied either by a fourth premolar and the first
molar, or by the first and second molars. The molars preserved are, therefore, either the second
and fourth, or the third and fifth. The dotted outlines are purely conjectural.

la. The same, natural size. I/). The fonrthtor fifth molar, much enlarged.

No. 2. Dromathcrium syli-cstrc. The inner surface of the left mandibular ramus. enlarged.
2a. The same, natural size. 26. The second molar, much enlarged.

ABBREVIATION'S. — o. Angle ; c. canine : en. condyle ; cr. coronoid ; i. incisors ; m;i. mylohyoid
groove t ; »i. molars ; p. premolars.

Ill all other respects these mandibular raini differ widely. The Micro-
conodon ramus is two-thirds the length of that of Di'mn/////, rum/ ; it is
flattened and slender, with a nearly straight lower border beneath the
molar alveoli, and a characteristic depression of the border which possibly
represents the angle of the jaw, as in Prof. Chven's genus PC/V/////.S (Fig. 16.)
The coronoid process is low and the vertical diameter of the jaw at this
point is very narrow. This ramus offers a great contrast to that of

*[Dimetrodon is now classified with the Sphenodon-like reptiles. — ED.]
t[Now believed to be the groove for Meckel's cartilage. — ED.]

20 EVOLUTION OF MAMMALIAN MOLAR TEETH

Dromathcrium, which is very stout and convex with a thick lower border,
projecting widely from the matrix, an elevated coronoid process, while the
curvature of the lower border is unbroken by any downward projection.
If these differences may be given merely a specific value, and attributed
in part to the fact that the Microconndmi jaw is seen upon the outer
surface, and that of Dromathcrium upon the inner surface, let us compare
closely the teeth in the two genera. Unfortunately the canine and
incisors of the Microconodon ranius are wanting. We first observe that
the premolars of Dromathcrium are styloid and procumbent ; if erect
they would rise above the level of the molars ; they have no trace of a
cingulum. In the other genus the premolars are subconical, and, although
erect, they do not reach the level of the molar tips ; they show a faint
posterior cingulum, and the third premolar has the same evidence of a
division of the fang which is seen in the molars, while in Dromatlurium
there is no trace of such a depression, but a distinct groove on the
postero-internal face of the tooth reaching nearly to the summit. The
molars of Dromathcrium are narrow and loft}-; the general pattern of the
crown consists of a single main cone with a high anterior and lower
posterior accessory cusp upon its slopes; but these cusps are very irregular
in disposition. For example, in the second molar there are two anterior
cusps ; in the third molar the posterior cusp is nearly as large as the
main cusp : in the fifth molar there is a trace of a postero-external cusp ;
in the last molar both the anterior and posterior cusps are distinctly bifid
at the tip. In Microconodon, on the other hand, the molars are com-
paratively low and broad, with a low anterior and higher posterior
accessory cusp ; these cusps are regular and very prominent ; there is also
a well-marked posterior cingulum, which cannot be distinguished in the
corresponding molars of the other genus.

Although the two posterior molars are wanting in Microconodon, the
rise of the coronoid probably marks the position of the last molar ; taking
this estimate of the posterior point of the molar-premolar series and
comparing it with the length of the series in Dromathcrium, we find that
while the ramus of one genus is only two-thirds the length of the other,
the total space occupied by the molar-premolar series is very nearly the
same. Estimated in another way, the molar-premolar series of Micro-
conodon is a little less than one-half the entire length of the jaw (y6^),
while that of the other genus is exactly one-third the length of the jaw.
This discrepancy is due to the difference in the proportions of the molars;
in one genus they are low and broad at the base, in the other they are
unusually high and compressed.

It is difficult at present to assign any systematic position to either of
these genera. Dromatherium is entirely unlike any known mammal,
fossil or recent. The form of the molars is extremely primitive both in

TEETH OF THE MESOZOIC MAMMALS 21

respect to the incomplete separation of the fangs and the remarkable
variations in the number and size of the accessory molar cusps. In fact
the molars appear to be in what may lie called an experimental stage of
structure. The accessory cusps are sometimes large and distinct, as in
the third true molar ; sometimes minute as needle points, as in the
second molar. The incomplete separation of the fangs is a reptilian
character, which when correlated with the styloid prernolars and recurved
canine-incisor series, places Dromafheriiim very remote from any of the
known Mesozoic mammals. Microconodon, on the other hand, is of a
somewhat more recent type, the premolars have the trace of a low
posterior heel, and the molars have that regular tricuspicl division of the
crown which is first observed in the genus Ainpliilestes (Fig. 5) of the
English Lower Jurassic and characterizes a large number of the Jurassic
mammals.

2.
A NEW CLASSIFICATION OF THE MESOZOIC MAMMALS.

[In order to make the following section clearer we insert here a brief
classification of the Mesozoic (Lower and Upper Jurassic) mammals of
the orders Triconodonta and Pantotheria (Trituberculata). — ED.]

A. Infra-class Marsupialia.

1. ORDER: TRICONODONTA OSBORN.
1. FAMILY : TRICONODONTID.E MARSH.

Probably carnivorous pro-Marsupials. Molars with three stout crest
cusps, the anterior and posterior cusps derived from the crown, and
a .strong internal cingulum, rising into anterior, median and posterior
prominences. Opposition of upper and lower molars subtrenchant.
Postcanine teeth 7 to 11, usually 7. Angle of jaw inflected. Coronoid
broad, recurved.

a. Subfamily Amphilestince Osborti.

Lower Jurassic. Anterior and posterior cusps of molars much lower
than middle cusps, all three cusps being in the same fore-and-aft line.
Postcanine teeth 9-11. Angle distinct in outside view. Mandible rather
slender.

Genus Amphilestes. Lower Jurassic (Stonesfield Slate), England.
Figs. 4 (No. 1), 5.

22 EVOLUTION OF MAMMALIAN MOLAR TEETH

I. S'libfamily Triconoclontince Osborn.

Upper Jurassic.* Anterior and posterior cusps of molars progressively
increasing in size, finally equalling the middle cusps. All three cusps in
line. Postcanine teeth 7. Angle not seen in outer, side view. Canines
erect and piercing. Mandible stout.

1. Triconodon Upper Jurassic,* England (Middle Purbeck Beds).
Figs. 11, 11 a.

2. Priacodon (probably a synonym of Triconodon), Upper Jurassic
(Como Beds), Wyoming. Fig. 8.

3. Triacanthodon, Upper Jurassic * (Middle Purbeck Beds), England.
Fig. 7.

c. Subfamily Phascolotheriinm Osborn.

Lower and Upper Jurassic. Anterior and posterior cusps of lower
molars progressively shifting to inner slopes of central cusps. Postcanine
teeth 7. Premolars and molars alike (or premolars greatly reduced in
number). Angle not seen in outer side view.

1. Phascolotherium, Lower Jurassic (Stonesfield Slate), England.
Figs. 4 (No. 3), 6.

2. Tinodon, Upper Jurassic * (Como Beds), Wyoming. Fig. 10.

INCERT^E SEDIS.
d. Subfamily Spcdacotheriince Osborn.

Upper Jurassic. Molars with large central cusp and two smaller
€usps somewhat internal to it. Premolars and molars unlike (or pre-
molars not greatly reduced in number). Angle not seen in outer side
view. Postcanine teeth 7-10. Mandible slender.

Spalacotherium, Upper Jurassic """ (Middle Purbeck Beds), England.
Fig. 11.

Menacodon, Upper Jurassic (Como Beds), Wyoming. Fig. 9.

Peralestes, Upper Jurassic * (Middle Purbeck Beds), England.

B. Infra-class Placentalia.

ORDER : PANTOTHERIA MARSH (TRITUBERCULATA OSBORN).

Probably insectivorous pro-placentals, angle of mandible not inflected,
lower molars primitively of the tuberculo-sectorial type, with a central
external and three internal cusps, pad, med, hyd, the latter probably
derived from the internal cingulum. Upper molars subtriangular, the

* [Throughout this book the Purbeck Beds of England and the Como or Atlantosaurus
Beds of Wyoming are called Upper Jurassic, but English and Continental geologists now
regard them as Lower Cretaceous in age. — ED.]

TEETH OF THE MESOZOIC MAMMALS 23

trigon reversing the pattern of lower molars, the accessory cusps being
on the outer side of the main or internal cusp, probably also derived
from the cingulum. Molars progressively styloid and piercing, often
with recurved tips. Lower incisors typically procumbent, spatulate,
p3 and especially p4 typically larger than m^ erect and piercing.
Postcanine teeth usually 11 (p i, in 7).

1. Family Amphitheriidcc Owen.

Lower and Upper Jurassic. Lower molars tuberculosectorial bifanged,
incisors more erect. Condyle low, coronoid broad, angle well rounded
below, sharply depressed. Postcanine teeth 9-12. Incisors more erect.

Amphitherium, Lower Jurassic (Stonesfield Slate), England. Figs.
15, 17.

Amphitylus, Lower Jurassic (Stonesfield Slate), England. Fig. 4
(No. 2).

PeramiLs, Leptocladus, Upper Jurassic (Purbeck Beds), England. Figs.
18, 26.

2. Family Amblotheriidce Osborn or Stylacodontidce Marsh.

Upper Jurassic. Molars progressively styloid and piercing, finally
with single fang, canine with single fang. Coronoid more slender.
Angle small, continuous with lower contour of mandible. Mandible
graceful and slender. Incisors procumbent, spatulate. P3, p4 larger
than m1. Postcanine teeth 11-12.

1. Amblotherium (Upper molars = ? Peralestes) Upper Jurassic (Pur-

beck Beds), England. Figs. 11, 23.

2. Peraspalax, Upper Jurassic * (Purbeck Beds), England. Figs.
4 (No. 9), 22.

3. Achyrodon, LTpper Jurassic* (Purbeck Beds), England. Fig. 24.

4. Phascolestes, Upper Jurassic* (Como), Wyoming. Figs. 31-34.
Dri/olestes (probably a synonym of Phascolestes).

5. Laodon, Upper Jurassic* (Como Beds), Wyoming. Fig. 30.

6. Stylodon (Upper molars = ? Kurtodoti). Fig. 29.

Stylacodon, Upper Jurassic * (Como), Wyoming (a synonym of
Stylodon?)

7. Asthenodon, Upper Jurassic* (Como), Wyoming. Fig. 35.

3. Family Paurodontidce Marsh.

Upper Jurassic. Molars feebly tuberculo-sectorial, that is, with
accessory cusps poorly developed. Molars bifanged. Postcanine teeth 6.
Mandible very short and stout.

*See note on page 22.

24

EVOLUTION OF MAMMALIAN MOLAR TEETH

Paurodon Marsh. Upper Jurassic* (Como), Wyoming. Some approach
toward these distinctive characters is made by Achyrodon of the Amblo-
theriidie (Stylacodontidffi), from early members of which family Paurodon
may have descended. Fig. 26.

4. Family Diplocynodontidw Marsh.

Upper Jurassic. Upper molars with large crushing protocones ;
lower molars with very broad basin-shaped talonids.

Diplocynodon, Upper Jurassic* (Como Beds), Wyoming. Fig. 20.
Docodon, Upper Jurassic* (Como Beds), Wyoming. Fig. 21.
Enneodon, Upper Jurassic * (Como Beds) Wyoming.

3.

ILLUSTRATIONS OF THE CHIEF TRICONODONTA AND TRITUBERCULATA.

FIG. 4. The types of the British inesozoic mammals, representing the natural size.
1. Amphilestes. 2. Amphillterium. 3. Phascolothcrium. 4. Triconodon mordax. C. Peramus. 7.
Spalacotherium. 8. Peralestcs. 9. Pcraspalax. 10. Leptocladus. 11. Amblotherium. 12. Phus-
colestes. 13. Achyrodon. 14. Stylodon. 15. Athrodon. 10. Bolodon. IS. Plagiaulax minor. 10.
Stereognathus.

FIG. 5. Amphilestes. Lower Jurassic, England.
Internal view, enlarged.

Fia. 6. Phascolotherium. Lower Jurassic.
England. Internal view, enlarged.

Fro. 7. Iriconodon (Triiicanthodon). Upper Jurassic, England. External view, enlarged.

* See note on page 22.

TEETH OF THE MESOZOIC !MAMMALS

Fir-. 8. Triconodon (Priacodon). Upper Jurassic, America. Internal view. x|. After Marsh.

FIG. 9. Mena.coi.lon. External and internal views, xf. Upper Jurassic, America. After Marsh.

FIG. 10. Tiiiodon. Upper Jurassic, America.
Internal view. xf. After Marsh.

FIG. 11. Spahtrntlicriiiiii. Upper Jurassic,
England. Internal view, enlarged.

FIG. 11«. Upper and lower teeth and lower jaw of Triconniln,! j\ /-us from the Purbeck Ueds.
Upper Jurassic (Lower Cretaceous), England, x IT.

Figs. 7-lla. Jurassic Triconodonta. Lower Teeth and Jaws. (All enlarged.) For
the natural size of Figs. 5, 6, 7, 11 see Figure 4.

26

EVOLUTION OF MAMMALIAN MOLAR TEETH

5**sJI

J ~

r

• * V

*«• ' .. *-./••

.;

772.3,

FIG. 12. Superior molars of Ptralcstts Owen. Upper Jurassic, England. Right side.
External, oblique and crown views. These upper molars were probably associated with lower
molars of the type seen in Spalae^Hn.riiim (Fig. 11). /«(*, nietastyle; pr, pr, pr, main internal
cusps believed to be protocones ; ?pa, ?iw., smaller external cusps believed to be para- and meta-
cones.

'lit

FIG. 13. a, Kurtodon. Superior molar series of the left maxilla, viewed upon the wearing
surface, b, Amblotheriuni soriciwiui, inferior molar series, viewed from above. It1 A. (Peraspalax)
talpoides. A lower molar viewed upon the internal face. c. The same. A lower molar viewed
from above, d, Ac/iyrodon nanus. A lower molar viewed from above. Much enlarged.

TEETH OF THE MESOZOIC MAMMALS

27

l';xtci'iia[ or maxillary licic.

i.of.

A

'fine 6C s r 4 c 3 c a i_s

r r f f ' ^•4Ki= — - — '**"••&.•-"*—

c ' £ i c ^r-

'^•:^u|

fi -mt z pr 3 4 fir 5 i / mr

->— -^

. , , ^

m

*

-,'^ ^

Painted or internal view .

FIG. 14. Superior molai-s of Dryoh-stcs Marsh. Upper Jurassic, Wyoming. A. Series of the
left side, external and crown views. B. Series of the right side, external, crown and internal
views. Yale Museum, c, c, c, external and internal cingula. i. o. f. infraorbital foramen. Other
abbreviations as in Fig. 12. (Cf. Fig. 2072 and p. 220.)

Jurassic Triconodonta ? (Fig. 12) and Pantotheria (Trituberculata). Upper Teeth.

and Jaws. For scale see Fig. 4.

Fin. 1">. AmphitJierium prevostii. Lower Jurassic, England. Internal view, enlarged. After Goodrich.

FIG. 16. Peramus. Upper Jurassic, England. External view, enlarged.

Fir;. 17. Inferior molars of Amphitln i-ium /n; roxlii. A, inner view of a molar of the right side ;
-B, outer view of a molar of the left side (restored), enlarged. After Goodrich.

Figs. 15-17. Jurassic Pantotheria (Trituberculata). Lower Teeth and Jaws.

For scale see Fig. 4.

28

EVOLUTION OF MAMMALIAN MOLAR TEETH

FIG. IS. a. Peramus(Sp«,lacotherium) minus Owen. Internal view of left mandibular ramus.
6. P. (LeptocJadus) dubius Owen. External view of left mandibular ramus. c. P. tenuirostris
Owen. Outer face of anterior portion of left ramus. Also, second molar of Amphitherium
Prevostii Owen, internal view, and second molar of P. minus, enlarged from fig. 1 a above;
internal view, pi-, protoconid ; pa, paraconid ; me, metaconid ; hy, hypoconid ; nig, Meckelian
groove. Much enlarged.

FIG. 19. Enneoilon, Family Diplocynodontidse. Upper Jurassic, America. External view. x-j.

After Marsh.

d-

FIG. 20. Diplocynodon, Family Diplocynodontidse. Upper Jurassic, America. External view. xf.

After Marsh.

FIG. 21.— Ztocodow, Family Diplocynodontidse. Upper Jurassic, America. Internal view. xf.

After Marsh.

Jurassic Pantotheria or Trituberculata. Lower Jaws. Amphitheriidse (Fig. 18)
and Diplocynodontidse (Figs. 19-21). For scale see Fig. 4.

TEETH OF THE MESOZOIC MAMMALS

29

FIG. 22. Pcraspalax. Upper Jurassic,
England. Internal view, enlarged.

I \(t

^ .(/

C. 23. Amlloth<:rium. Upper Jurassic,
England. Internal view, enlarged.

Flo. '24. Achyrodon. Upper Jurassic,
England. Internal view, enlarged.

FIG. 25. Phn.<nil:xt<s. Upper Jurassic,
England. Internal view, enlarged.

FIG. 26. Po.urod.on. Upper Jurassic, America. Internal view. xf. After Marsh.

FIG. 27. Pc,-ani>'.< (Leptocladus). Upper
Jurassic, England. Internal view, enlarged.

FIG. 2S. Sti/lodoii. Upper Jurassic, England.
External view, enlarged.

FIG. 29. Stylodon (Stylacodori). Upper Jurassic, America. External view. Xy. After Marsh.

FIG. 30. Laodon. Upper Jurassic, America. Internal view. XT. After Marsh.

Jurassic Pantotheria or Trituberculata. Lower Jaws. AmblotheriidaB or Stj'la-
codouticlae and Paurodontidae. For scale of Figs. 22, 25, 27, 28, see Fig. 4.
Fi gs. 26, 29 are three times natural size.

30

EVOLUTION OF MAMMALIAN MOLAR TEETH

PIG. 31. Phascolestes vorax. Upper Jurassic, N. America. Oblique crown view. xj.

FIG. 32. Phascolestes (Dryolestcs) priscus. Upper Jurassic, America. Internal view. f[ x f-.

After Marsh.

FIG. 33. Phascolestes (Dryolestes) vorax. Upper Jurassic, America. Internal view. x-f-.

After Marsh.

FIG. 34. Phascolestes (D)'i/olestcs) vorax. Upper Jurassic, America. External view, xf--

After Marsh.

FIG. 35. Asthcnodon. Upper Jui-assic, America. External view. ~ X f. After Marsh.
Figs. 31-35. Jurassic Pantotheria or Trituberculata. Lower Jaws. Amblotheriidse.

TEETH OF THE MESOZOIC MAMMALS 31

4.

THE ORIGIN OF THE TKITUP-EI;CI:LAI; TYPE SOUGHT (1888) AMONG

THE MESOZOIC MAMMALIA.

[Extract from Memoir entitled "The Structure and Classification of the Mesozoic Mammalia."
Jonrn. Acad. Nat. S<->. Pfiifa., Vol. IX., No. 2, July, 1888.]

(«) If, as now seems probable, the derivation of the mammalian molar
from the single reptilian cone can be demonstrated by the comparison of
a series of transitional stages between the single cone and the three-cone
type, and from the latter to the central tritubercular type, the separate
history of each cone can certainly be traced throughout the series in its
various degrees of modification, development, and degeneration. The
remarkable part played by the tritubercular molar has been unfolded by
the discoveries and writings of Cope. It is undoubtedly the ancestral
molar type of the Primates, the Carnivora, the Ungulata, the Cheiroptera,
the Insectivora, and of several, if not all, of the Marsupialia. For
example, we can trace back the quadritubercular btmodont, or parent
ungulate type, to the tritnlto'culai- ; this to the type with three cones in
line, which we may call the triconn<l<>nt. type, and this in turn to the
Tiaplodont l reptilian crown. A nomenclature may be suggested for these
cones, with reference to their order of development and primitive position,
to keep clearly before the mind their homologies during secondary
changes of form and position. The primitive cone may be called the
protoconc ; upon the anterior and posterior slopes of which appear,
respectively, the paracone and iininrniir. After the tritubercular crown is
produced, by the rotation of the lateral cones, inwards in the lower jaw
and outwards in the upper jaw, the Jit/poconc, or heel, is developed, giving
us the tubercular-sectorial molar. Exclusive of the MnHiliili-rcnlfita and
of Stereognathm (Fig. 51, No. 5) this is the most advanced stage of
molar development thus far found in the mesozoic period.

The protocone of Dromatherium (Fig. 3, No. 2) is prominent and con-
stant through the molar series while the para- and meta-cones are irregular
in size and position, always close to the main cone and in several teeth
either splitting into two needle-like cusplets or bifid at the tip. Alto-
gether, they are in what appears to be an experimental stage of develop-
ment. Microconodon (Fig. 3, No. 1), however, from the same strata, has
well defined para- and meta-cones which are widely separated from the
main cone, the crown presenting the pure triconodont type. This
reoccurs in Amplii.l<'*t<'s (Figs. 4, 5), of the lower Jurassic, and Tricomxlnn
(Figs. 4, 7) of the upper Jurassic. In this series we are struck by the

1 See Cope : "The homologies and origin of the types of molar teeth of the Mammalia
Educabilia," Jour. Phila. Ami:/., 1874.

32 EVOLUTION OF MAMMALIAN MOLAR TEETH

gradual increase of size and prominence of the lateral cones until they
are upon the level of the main cone and sub-equal to it, this increase
being accompanied by a marked elongation of the crown so that the three
molars of Triconodon occupy a greater proportion of the jaw than is taken
by the seven molars of Dromatherium. This unmodified triconodont type
is very rare in the more recent mammalia. It persists in the lower jaw, at
least, of Dissacus from the Puerco, and in the lower molars of Thylaciiin*,
the upper molars presenting an internal heel.*

(&) In his paper upon the Creodonta,1 Cope observed that the Spala-
cotJierium molars (Figs. 4, 11) represents a stage of transition between the
triconodont and tritubercular molars. There can be no doubt that the
cusps seen upon the inner face of the inferior molars of this genus are
homologous with the para- and meta-cones,t and there are several facts
which support Cope's hypothesis that they represent a stage of inward
rotation of cusps which were at an earlier stage in the same fore and aft
line with the main cusp. These are, that in Phascolotherium (Figs. 4, 6)
the lateral cones are seen to be slightly internal to the main cone so that
their median slopes descend upon the inner face; in Tinodon (Fig. 10),
of a later geological period, this position is slightly more pronounced ; in
Menacodon (Fig. 9) it is still more marked, but less so than in
Spalacothenum (Fig. 11). These genera, although evidently in two
different lines of descent, afford the desired transition stages. The
Spalacotlierium molar as seen from above 2 has a striking resemblance to
the anterior sectorial triangle of the Stypolophus or Didymictis molar of
the Puerco. It is in fact sub-triangular, the superior molars probably
having the lateral cones rotated outwards, so that the upper and lower
molars form an alternating series, the ridges connecting the main and
lateral cones acting as sectorial blades.

The question now arises whether the Stylacodon- (Fig. 29) molar
represents the next higher stage of development, viz., the tubercular-
sectorial molar in which the anterior triangle is followed by a low heel.
And if so has the Stylacodon type passed through the stages of inward
rotation of the lateral cusps ? The superior aspect of the Stylacodon
molar presents an anterior triangle with the long styloid cone forming the
apex and connected by divergent ridges with the anterior pair of cusps ;
behind these is a third cusp not connected by a ridge with the styloid
cone. In the upper jaw the three cusps are external and the single cone
internal, these relations are reversed in the lower jaw. We cannot well

* [This condition is now believed to be not primitive but secondary in both the genera
mentioned. — Eu.]

1 " The Creodonta," American Naturalist, 1884, p. 259.

t[But see page 33* where the homologies of these cusps with the paraconid and meta-
conid of Trituberculates are doubted. — ED.]

2 Owen, The Mexozoic Mammalia, Plate I., Fig. 32c.

TKKTH OF TUK MESOZnlC MAMMALS

avoid the inference that the M,//<ir<,t/t,n lower molar is a specialized
tubercular-sectorialj that the styloid external cusp, which until Marsh's
discovery of Drii<>l<-*ti'* (Figs. 31-34) was rognnlrd as the single summit
of the crown, is the protocone while the anterior pair of internal cusps
represent the paracone and metacone, followed by a third element, the
hypocone or heel. This is further confirmed l>y the transition to the
simpler Spalacotherium type seen in the molars of Asthcmnlnn (Fig. 35)
in which the hypocone is entirely wanting while the remainder of the
crown is closely similar to that of Stylacodon. The internal cusps
present many degrees of development in different members of the
Stylacodontidse ; in Lna<lon (Fig. 30) they are much less prominent than
in T)/'//olcstes, the heel being also inconspicuous. While the relations of
the four cones composing the Stylacodont crown strongly suggest the
tubercular-sectorial molar there is one matter of doubt in the way of
the derivation of this tooth from the Spalacotherium type (Fig. 11);
that is, the position of the fangs. In Spalacotherium and Menacodon
(Fig. 9) the fangs are paired and placed beneath the para- and
metacones. In the Stylacodonts the external fang is directly beneath
the protocone ; the question is, does this represent the anterior or
posterior, or an additional fang ?

4. The molars which have been considered thus far show directly
or indirectly the triconodont type, i.e., the presence at some stage
of their evolution of the central and two lateral cones. In the
Amphitheriidfe (Figs. 15-21) it is clear that the main cone and
the lesser one, upon its anterior slope, represent the protocone
and paracone but it is uncertain whether the basal cusp, seen for
example upon the external face of the Diplocynodon molar (Fig. 20)
is homologous with the metacone or hypocone. The latter alterna-
tive excludes the development of the metacone or the passage
of these genera through a triconodont stage, and implies a con-
siderable separation of the Amphitheriidse from the stem of the
two families already considered. The former involves the supposition
that the metacone has metamorphosed into a heel.* The most primitive
molar in this family is seen in Eitnt'ixlon1 (Fig. 19). The crown has
an obtuse recurved protocone, more like that of a premolar ; upon
the anterior slope is a rudimentary paracone which affords the only
means of distinguishing the molars from the premolars. The posterioi
slope terminates in a low extended heel. This molar pattern largely

* [The questions and difficulties here stated are largely resolved by the hypothesis that
in the Triconodontidce the so called para- and metaconids are direct outgrowths from the
molar crown, while in the Amphitheriidse and Amblotheriidse the para-, meta- and
hypoconids have been derived from the internal cingulum. See page 8. — ED.]

1 Marsh, "American Jurassic Mammals," Am. Jour. Sc., April, 1887, PI. X.
Fig. 4.

C

34 EVOLUTION OF MAMMALIAN MOLAR TEETH

confirms the second of the above alternatives, viz., that this heel is to
l>e compared to the hypocone of the tubercular-sectorial crown. Further
confirmation is seen in the fact that this heel is not above the level
of the internal cingulum, as in the metacone of all the trieonodonts,
but is continuous with the broad shelf-like projection of the internal
cingulum, which is well represented in the internal aspect of the
Dipliii-i/inxlnii (Fig. 20) molars. The concave internal slope of the
protocone descends into this shelf and the cingulum rises at the margin
into numerous crenations, which cannot properly be called cusps. The
Diplocynodon (Fig. 20) molar presents a decided advance upon that
of Emieoilon in the development of the paracone, which is much more
prominent. In Amphitherium (Figs. 4, 15, 17), the paracone is
subequal to the protocone in several of the molars, and the heel is
on the level of the internal cingulum, from which, according to Owen,
there arise one or two small cusps.1 Internal cusps which develop in
this manner are from the first separated from the external cusps by a
longitudinal valley instead of being united with it by divergent ridges,
and cannot therefore at any stage possess a sectorial blade, such as
is more or less distinctly developed in the Spalacotherium and Stt///t-
codon molar.

5. It follows also that the triangle of cusps presented by the
Pcraspalax molar (Fig. 22) cannot, with probability, be considered as
representing a tritubercular stage, and that the Amphitheriidse furnish
the key to the mode of derivation of the internal cusps of the
molars of the Peralestida? (Fig. 12). The inferior molars of Pcraspnl«.'
and Pan ri»l on (Fig. 26) are apparently very similar, consisting
of a prominent external cone, and two internal cusps followed by a
third cusp at the end of the crown. As pointed out in the synopsis of
molar types, this internal surface strongly suggests the Dryolestcs pattern
(Figs. 31-34), but may be clearly distinguished by the absence of trans-
verse ridges and the presence of a longitudinal valley between the
cusps instead of a transverse valley opening inwards. The internal
cusps have probably, therefore, arisen from the internal cingulum,'2
but these molars do not seem to be a later development of the
Amphitherium type (Fig. 1-5), because both the paracone and meta-
cone are wanting, the main cone showing no trace of the lateral cusps

1 As previously stated the writer has not personally examined the internal surfaces
of the molars of this genus.

2 Numerous instances of the origin of molar cusps from the cingulum might be
cited. One of the most important is seen in the transition from a tritubercular to a
quadritubercular superior molar by the addition of the postero-internal cusp which is
primitively a cingule ; this was first demonstrated by Dr. Harrison Allen, op. cit.
Mivart (Jour, of Anat. and Pht/fs., Vol. II., p. 138) shows how the four cusps of the
Insectivore molars are frequently fortified by additional cusps from the cingulum.

TEETH OF THE MESOZOIC MAMMALS 35

upon its slopes.* The superior molars of Peralestes,^ however, when
viewed from above, present one large internal and two smaller external
cusps disposed in a triangle opening outwards, and as this is the general
disposition of superior cusps of the tritubercular type, we must admit
the possibility that the smaller cusps do represent the para- and nieta-
cones in a stage of inward rotation not accompanied by the production
of the sectorial blades, for this is by no means an essential feature
of the tritubercular molar. The history of the derivation of the
molars of the Peralestida; must, therefore, be left in some doubt ;
while the balance of evidence points to a line of development similar
t«i that in progress in the Amphitheriidse, although the line of descent
appears to be different.

Eeviewirig this study of the molars the following are the principal
deductions : (a) The molars of all the mesozoic mammals of this
group present one main cusp which is either so central or so pro-
minent that it may be considered homologous with the single reptilian
cone or protocone. (6) In one line of genera two lateral cusps, the
para- and meta-cones, appear upon the anterior and posterior slopes
of the protocone. This is a central and frequently repeated stage
of evolution. It gives rise to two lines of molar development ; in the
first, the para- and meta-cones are retained in the same fore and
aft line, as the persistent triconodont type, but increase greatly in
size ; in the second, they are rotated inwards as the tritubercular
type, which finally acquires a heel, (c) In a second line of genera
the paracone appears upon the anterior slope of the protocone, but
the metacone is not developed, being replaced by a basal talon or
hypocone which extends inwards to form the internal cusp. (V) In
a fourth line of genera neither the para- nor meta-cones are developed
upon the sides of the protocone, but they are replaced by basal
cusps derived from the cingulum.

* [Per contra the conditions seen in Amphitherium may have been derived from those
seen in PeraspaJax. — ED.]

t [Supposed to correspond with the lower molars of the Spdlacotherium type, Fig. 11.—
ED.]

CHAPTKK II.

FIRST OUTLINE (1888) OF TRITUBERCULAR EVOLUTION

IN MAMMALS.

[Reprinted directly from the paper entitled "The Evolution of Mammalian Molars to
and from the Tritubercular Type,1'1 The American Xatiirali*t, December, 1888.]

THE dentition in the recent Mammalia is so diverse that the most
sanguine evolutionist of fifteen years ago could not have anticipated
the discovery of a common type of molar, in both jaws, as universal
among the Mammalia of an early period as the pentadactyle foot,
and as central in its capacity for development into the widely
specialized recent types.

The tritubercular molar, discovered by Professor Cope in the Puerco,
is exactly such a type, and may be considered with the pentadactyle
foot as playing a somewhat analogous role in mammalian history, with
this important difference — the unmodified pentadactyle foot was pro-
bably inherited direct from the reptiles, and its subsequent evolution,
with a few exceptions, has been in the direction of the greater or
less reduction of primitive elements towards special adaptation, as, to
borrow an extreme illustration, in the transition from Plicnacoclm with
26 elements in the manus to Eym* with only 12 such elements. On
the other hand, the tritubercular tooth was not inherited, but in all
probability developed within the mammalian stock, from a hypothetical
form with almost, if not quite simple conical molars, implanted by single
fangs, in a nearly homodont series.2 No such primitive type of mamma-
lian dentition is actually known, although Dromatherium approximates it ;
but the apparent reversion to this type among the Cctacca, and apparent

JRead in the geological section of the British Association at Bath, September, 1888.
Read in abstract by Professor Cope, National Academy of Sciences, at New Haven,
Nov., 1888.

2 See Author, "Structure and Classification of the Mesozoic Mammalia," Jour.
Phila. Academy, 1888, p. 240.

TRITUBERCULAR EVOLUTION IX MAMMALS 37

retention of it in the E<l<'nl,it«,1 * support all the independent evidence
upon this point derived from the Mesozoic Mammals.

The principle of growth was the regular addition of new parts to
the simple cone, not at random, but according to a certain definite
order which apparently progressed independently in different phyla,
through a series of subtritnbercular stages until trituberculy - was
attained.

The tritubercular molar consists essentially of three cusps, forming
what may be called the primitive triangle, so disposed that the upper
and lower molars alternate. This, when attained, formed a central
stage from which the great majority of recent molar types have
diverged by the addition, modification, and reduction of cusps: we
must except the Monotremes, the Edentates, and possibly the Ceta-
ceans, although there is considerable evidence that the cetacean
molars were once of the triconodont type.3t Among extinct orders,
the Multituberculata (P/i./r/in/'/n.,', Triti/loclon, etc.) must also be excepted
from this series and discussion.

The almost universal predominance of trituberculy in the early
geological periods, is very significant of the uniformity of molar origin.
Of twenty known Mesozoic genera,4 all except three 5 J show trituber-
culy in some of its stages. In the Lower Eocene, eighty-two Puerco
species, representing twenty-six genera and five orders (Creodonta, Tillo-
dontia, Lemuroidea, Condylarthra, Amblypoda), only four species have
quadri tubercular teeth, all the remainder are tritubercular.6 Prof.
Kiitimeyer has recently pointed out the predominance of this type
in the nearly parallel Egerkingen beds. The contemporary Cernaysien
fauna in the collection of Dr. Lemoine at Ptheims, recently examined
by the writer, shows exclusively tritubercular molars or their deriva-
tives. By the Middle Eocene, the lines of divergence towards the

:See Oldfield Thomas, "The Homologies and Succession of the Teeth in the Dasyu-
ridre," Phil. Trans., 1887, p. 458.

* [Evidence is presented on pp. 151, 190, that the haplodonty of certain Edentata and
of the Cetacea is secondary. — ED.]

2 A term first employed by Riitimeyer, " Ueber Einige Beziehungen zwischen den
Saugethierstammen Alter und Neuer Welt," Abh. d. schweiz. pat. Gesellsch., Vol. XV.,
1888, p. 54.

3 See Brandt, Die Fossilen u. Subfoss. Cetaceen Europas, Taf. XXXII., Figs. 4-9.

t[But this in Zeur/lodon was probably a secondary derivation from the tritubercular
type. See p. 191.— ED.]

4 The list given by the writer (op. cit., p. 247) is found to contain several synonyms.
See "Additional observations upon the Structure and Classification of the Mesozoic
Mammalia," Proc. Phila. Acad., Nov., 1888, p. 292.

5 Dicrocynodon (Diplocynodon), Docodon, Enneodon, Marsh.

I [The author refers only to the orders Protodonta, Triconodonta, Pantotheria. — ED.]

''Cope, "Synopsis of the Vertebrate Fauna of the Puerco Series," Am. Phil.
Soc., 1888, p. 298.

38 EVOLUTION OF MAMMALIAN MOLAR TEETH

existing types of molars were well advanced, but trituberculy persisted
in the dentition of several orders, in which it is found to-day (Lemu-
roidea, Insectivora, Carnivora, and many Marsupialia).

It follows that it is quite as essential for the comparative anatomist to
thoroughly grasp the meaning and history of each of the component cusps
of the tritubercular molar and of their derivatives, as it is to perfectly
understand the elements of the maims and pes. For, the homologies of
the cusps can now be determined almost as certainly as those of the
digits. Take a human molar, for example, every component tubercle
has its pedigree, and it can be demonstrated, almost beyond a doubt,
which of these tubercles is homologous with the single reptilian cone.
The writer recently (op. cit., p. 242) proposed the adoption of a dis-
tinct nomenclature for the different cusps of the tritubercular molar,
and offered a series of terms for the primary cusps based as far as
possible upon the primitive position and order of development, and in
most instances in accord with their secondary position. This nomencla-
ture can be extended to the secondary cusps in the sexitubercular
superior and quinquetubercular inferior molars. The terms now in
general use are based, for the most part, upon the secondary or
acquired position, and in no instance upon the homologies of the cusps
in the upper and lower molars, or even in corresponding molars of
different genera, thus involving much confusion. For example, the
antero-internal cusp of the lower molar of Mioclcenus is not homo-
logous with the antero-internal cusp of Hyopsoclus, nor with the antero-
internal cusps of the upper molar of either genus.

The present contribution is based principally upon the writer's
studies among the Mesozoic Mammalia, and, with some additions, upon
Professor Cope's numerous essays upon the tritubercular type in the
Tertiary Mammalia.1

Four propositions may be laid down for discussion :

(1) That trituberculy was acquired during the Mesozoic period, in
a series of stages beginning with the single cone and attaining to the
primitive sectorial type in the Jurassic period.

(2) The majority of Mesozoic mammals showed trituberculy in
some of its stages. Present evidence goes to show that the remaining,
or aberrant types, if such existed, did not persist. The majority of
the persisting forms of later periods were derived from the forms with

1 Professor Cope's essays abound with discussions and notes upon the origin and
succession of the tritubercular type. (See collection in Origin of the Fittest.) He
has outlined the transition from the single cone to the tritubercular crown (p. 347) ;
tubercular sectorial (p. 246) ; the quadritubercular type (p. 245 and p. 359) ; the Spala-
cotherhim molars as a transition to the f>-ifnl>< rcular (p. 259). The acquisition of the
superior and inferior quadritubercular molar (p. 361). The prediction of the discovery
of Carnivora with triconodont molars (p. 365), and of the simple tritubercular type in
both jaws (p. 362).

TK1TUBERCULAR EVOLUTION IN .MAMMALS :59

simple tritubercular molars, of earlier periods. It follows that tritu-
berculy was an important factor in survival.

(3) The definite homologies of the primary and to some decree
of the secondary cusps in the upper and lower molars can be
established.

(4) The mode of succession of tooth forms favours the kinetogenesis
theory advanced by Ryder and Cope.

There are three general observations to be made :

First. In attempting to complete the history of each of the cusps
we naturally find that the paleeontological record is not sufficiently
perfect to admit of our following a certain type along a single phylum
back to the primitive type. We must at the outset proceed upon
the principle of similar effects, similar causes. For example, since
the history of the development of the intermediate tubercles- in the
superior molars of the Lemumidea (Pseudolemuroidea, Schlosser) is
perfectly clear during the Wasatch and Bridger epochs — it is safe
to infer that the intermediate tubercles of the Ungulate molars, which
are fully developed in the underlying I'uerco, had the same history.
Second. There are in each period aberrant ////><>• which embrace either
incomplete or degenerate tritubercular stages, i.e. a high specialization
in which the past record is obliterated, or, finally, stages in non-
tritubercular lines of development. Thin/. In the parallel evolution
of trituberculy in different phyla we find that the progression is by
no means uniform. In every geological period in which the fauna
is well known we observe progressive genera which outstrip the others
in reaching a certain stage of molar development, contrasted with
jx-i-xixft'iif types which represent arrested lower stages of development,
while between them are the central types which represent the degree
of evolution attained by the majority of genera. The latter may be
said to constitute the stage which is characteristic of the period.

Tin xt(itjt-s of Irihini't'i-nli/ may now be defined as seen in different
types in their order of succession :

I. Haplodont Type (Cope).1 A simple conical crown. The fan^
usually single and not distinguished from the crown. This type has
not as yet been discovered among the primitive Mammalia.

A. Protodoni Sub-Type.2 The crown with one main cone, and
lateral accessory cuspules ; the fang grooved. There is some cpiiestion
as to the advantage of distinguishing this as a type, for it stands

1 " The Homologies and Origin of the types of Molar Teeth in the Mammalia
Educabilia," Journ. Phila. Acad., 1874. The term Homodont was previously applinl
to this type by Riitimeyer, " Odontographie der Hufthiere, etc," }'> rhr. d. Xalnrl'or--<-h.
Gesellsch. in Basel, Band III., 1S63, p. 563. ]ii the writer's opinion this term has
acquired a special significance as applied to a whole series of teeth, viz., the reverse
of " heterodont," and may well be retained in this sense.

2 Osborn, op. tit. , p. 222.

40 EVOLUTION OF MAMMALIAN MOLAR TEETH

intermediate between types I. and III. Example, Dromaiherium of
the American Triassic (Fig. 3).

II. TricoiHu/nitf Type (Osborn, «i>. cit., p. 242). The crown elongate,
trificl, with one central cone and two distinct lateral cones. The
fang double. Example, Tricoimdim (Fig. 8).

III. Tritubercular Type (Cope). The crown triangular, surmounted
by three main cusps, the central cone placed internally in the upper
molars and externally in the lower molars. Example, the lower
molars of Spalacot/K'i-ium (Fig. 11) and Asthenodon (Fig. 35). This
type is rare in its primitive condition as above defined.

The upper and lower molars are alike in types I. and II. : in
type III. they have a similar pattern, but with the arrangement of
the homologous cups reversed. These types are all primitive. In.
the following sub-types, the primitive triangle forms the main portion
of the crown, to which other " secondary " cusps are added, the homo-
logies of which in the upper and lower molars are somewhat doubtful.
Parallel and with an intimate relation to the addition of the secondary
cusps, is the division of the tritubercular into a secodont and buno-
dont series, according to the assumption of a purely cutting or
crushing function. In departing from the primitive type, the upper
and lower molars diverge in structure, and the homologies of the
secondary cusps in each are somewhat doubtful.

Lower Molars.

A. Tiilinrnhi)' Scdorial, Sub-type (Cope). («) The primitive triangle
elevated and its cusps connected by cutting crests ; a low posterior
heel. (b) This type embraces a qmnquetubercular form in which
'the heel consists of two cusps, an internal and external.1 (c.) In
the Bunodont series it develops into the quadritubercular form, by
the loss of one of the primitive cusps,

Upper Molars.

B. Trilnln'iri<l«-r. (a). The primitive triangle in the secodont
series purely tricuspid. (b) This embraces a quinguetubercular form
in which " intermediate " tubercles are developed, both in the Seco-
dont and Bunodont series. (c) In the Bunodont series a postero-
internal cusp is added, forming the sexitubercidar molar.

NOMENCLATURE OF THE CUSPS. As above stated, there is no doubt
about the homologies of the three "primary" cusps (protocone, para-
cone, metacone) in the upper and lower molars. They may be given
the same terms, with the arbitrary suffix id, to distinguish the

Naturalist, April, 1883, p. 407.

TRITUBERCULAR EVOLUTION IN MAMMALS 41

lower cusps. The first "secondary' cusps (hypocone, hypoconid),
which are added to the upper and lower molars of the primitive
triangle, modify the crown from a triangular to a quadrangular shape.
and hence may be considered homologous. The three additional
secondary cusps (protoconule, metaconule, entoconid) evidently have
no homology with each other.

TERMS NOW IN USK PROPOSED TERMS. ] ABHKKV.

Upper Molars.

Antero-internal-cusp, ....... Protoconc. pr.

Postero- „ „ or 6th cusp Hypoconc. hy.

Antero-external ........ Paracone. pa.

Postero- ,, ,,....... Metacone. me.

Anterior Intermediate cusp, Protoconule. pi.

Posterior ,, „ Metaconule. ml.

Lower Molars.

Antero-external cusp, ....... Protoconid. prd.

Postero- „ ,, Hypoconid. hyd.

Antero-internal cusp, or 5th cusp, .... Paraconid. pad.

Intermediate or antero-internal cusp (in quadrituber-

cular molars), ....... Metaconid. med.

Postero-internal cusp, ....... Entoconid. end.

Evolution of the Cusps. The cusp evolution in the Mesozoic period
has been fully discussed by the writer (op. cit., pp. 240-4) and in
the Tertiary period, by Professor Cope, so that only a brief resume
is necessary here. In Dromatherium (Fig. 3), from the upper Triassic,
the oldest mammalian type known, with the exception of Microlcstes,
the molars have a main protoconicl with several minute lateral cuspules,
differing in size in the different teeth, but in general giving a trifid
appearance to the crown. The molars of the contemporary Microconodon
(Fig. 3) also have impaired fangs, but distinctly trifid crowns, with
the anterior and the posterior cusps, or para- and meta-conids, upon
the slopes of the protoconid. This Triconodont type reappears, with
the addition of a cinguluni and paired fangs, in Am i>Ii listen (Fig. 5)
and Phascolotlicrium (Fig. 6) of the lower Jurassic and persists in
Triconodon (Figs. 7, 8) of the upper Jurassic. In this succession we
observe especially the relative subsidence of the protoconid and upgrowth
of the para- and meta-conids. Contemporary with Amphilcstcs is the
classical genus Amphitherium (Fig. 15). A recent examination of
the type specimen by the writer revealed the very interesting fact
that the molars of this genus are probably of the primitive tubercular-
sectorial type — the oldest known example. Only the paraconid,
metaconid and hypoconid have been observed heretofore, but one can
see the tip of the main external cusp between the internal pair.

^am much indebted to my colleagues Professors Macloskie and Winaus for assiM
ance in the selection of these terms.

42 EVOLUTION OF MAMMALIAN MOLAR TEETH

This pattern is repeated, with a considerable elevation of the heel,
in Peramus (Fig. 18) of the upper Jurassic.1 Neither of the two
foregoing are of the primitive heelless tritubercular type which is
apparently found in Spalaeotherium (Figs. 11), also upper Jurassic,
and in the nearly related if not synonymous Peralestcs (Fig. 12).
Contemporary with the above, are numerous genera of the Stylodon
order; among these, Astlteiioihn (Fig. 35) is of the primitive tri-
tubercular type without the hypoconid, all the remainder present
various modifications of the tubercular-sectorial.

This covers our knowledge of trituberculy in the Mesozoic period.
No bunodont forms are known — they were probably developed during the
Cretaceous, for a few are found well developed in the Puerco. In the
Secodont series many of the types do not widely depart from those seen
in the Jurassic, but the Bunodont series are universally characterized by
the initial or advanced development of the proto- and nieta-conules in the
upper molars and the appearance of the cntocouid upon the inner side of
the hypoconid below.

The Principles govern in// Cusp Development. It is remarkable to note
in how many particulars the actual succession of molar development in
the Mesozoic period coincides with the theoretical scheme of origin of
trituberculy proposed by Cope 2 and supported by Wortman 3 several
years ago. At that time Spalaeotherium and the genera now embraced
under the Triconodontidee were the only Mesozoic mammals whose molar
structure was fully known, and the views of these authors were partly
speculative and partly deductive from recent dental anatomy.

Two hypotheses may be advanced to explain the evolution of the
tritubercular type. The first is that the type has been acquired by the
selection of accidental variations in the production of new cusps and
modelling of old ones. The second is, that the interaction of the upper
and lower molars in the movements of the jaws has resulted in local
increase of growth at certain points, resulting first in new cusps, then in
a change of position and of form in the cusps. Both hypotheses are open
to numerous objections and are by no means mutually exclusive, but the
whole subject is so complicated as to require a separate treatment. The
balance of evidence in tritubercular evolution seems to favor the second
or kinetogenesis theory — as apparently witnessed in two laws of cusp
development :

I. The primary cusps first appear as cuspules, or minute cones, at the

1 This genus includes also Leptocladus (Inhiiix Owen, and Spalaeotherium minus
Oweii (Fig. 18). See Proc. Phila. Acad., Nov., 1888, p. 292.

2 " The Evolution of the Vertebrata Progressive and Retrogressive," American

t, April, 1885, p. 350.

3 "The Comparative Anatomy of the Teeth of the Vertebrata," 1886, p. 418.

TIIITUBERCULAR EVOLUTION IN MAMMALS 43

first points of contact between the upper and lower molars in the vertical
motions of the jaws.

II. The modelling of the cusps into new forms, and the acquisition of
secondary position, is a concomitant of interference in the horizontal
motions of the jaws.

The second law applies especially to the evolution of the molars after
the acquisition of the tritubercular stage, and has been ably proposed and
supported by Ryder,1 principally in its application to recent types of
teeth. The first, although not heretofore distinctly formulated, is partly
founded upon facts and principles advanced by Cope, and applies chielly
to the stages which have been discussed in this essay.

GO?

PIG. 36. Molars of opposite jaws in normal mutual relation. 1. Dclphinug. 2. Vromathcrium.
3. Triconotlon. 4. Spalacotkeriuin (lower), Pcralcstts (upper). 5. Vii;-,-,-tt cv.s. 6. Miodacnus.

l V.s.

During the homodont mammalian or sub-mammalian molar stage, the
jaws were probably isognathous* and the simple cones alternated as in the
Delphinidse (Fig. 36, No. 1). The first additions to the protocone appeared
upon its anterior and posterior surfaces. The growth of the para- and
meta-conids involved anisognathism,2 for we find in the later triconodonts
that the lower molars closed inside of the upper (Triconodon, Fig. 36,
No. 3). There are several transition forms, such as Tinodon (Fig. 10)
and Menacodon (Fig. 9) between the primitive triconodont type and
Spalacoth&rium (Fig. 11), and it has been assumed by Cope and the writer
(<>p. cit., p. 243) that the para- and meta-conids were first formed upon
the anterior and posterior slopes of the protoconid and then rotated
inwards, but it is also possible that they were originally formed upon the
inner slopes. In the complemental formation of the upper and lower

1 "On the Mechanical Genesis of Tooth Forms," Proc. Phila. Acad., 1878, p. 4.1.

"[From recent discoveries among the South African Theriodonts it seems more
probable that even in the ancestral reptile-mammals the upper jaws bit outside of the
lower jaws and teeth, i.e. the jaws were anisognathous. The isognathism of the Dolphins
is probably secondary. — ED.]

2 As employed by Ryder (op. cit., p. 45): "So as not only to indicate respectively
parity and disparity in transverse diameter of the crowns of the upper and lower molars,
but also the parity or disparity in width transversely, from outside to outside," etc.

It is clear that in the homodont condition, with the teeth simply piercing the food, the
greatest comminution (of the food) is effected by isoguathism ; in the triconodont stage, the
jaws must be anisognathous to close upon each other, but the tritubercular stage admits a
return to isognathism by the alternation of the triangles.

44 EVOLUTION OF MAMMALIAN MOLAR TEETH

triangles the jaws remained nearly isognathous (Fig. 36, Xo. 4). There
is no evidence as to the origin of the hypoconid, which as a rule preceded
the hypocone, as it was developed very early. In the Stylacodontidse
(Figs, 22-35), PJiascolestes, Amblotherium, etc., the crowns rapidly increased
in transverse diameter, and, in some genera (Kurtodon, Fig. 13) they so
i';ir lost the tritubercular aspect that, but for the connecting form
Asthenodon (Fig. 35), we might hesitate to place them in this series. The
key to the further evolution of the crown is seen in the bunodont series
during the lower Eocene period.

The superposition of the lower and upper molar patterns brings out
many interesting facts. First, even in the complex crowns of the buno-
dont molars the primitive triangles retain their primitive alternating
arrangement. Second, the jaws are somewhat anisognathous. Third, in
support of the first law of cusp development, we observe that the proto-
comile and metaconule are developed at the points of contact with the

PIG. 37. Diagram of quadritubercular molars of both jaws iu normal mutual relation ; the
superior molars in double lines ; the inferior in black.

ridges which extend from the hypoconid, and, secondly, that the hypocone
appears at the point where the paraconid abuts against the hypocone. It
follows from a comparison of numerous species of Pelycodus and Miocluenus
that as the hypocone develops the paraconid recedes, as first observed by
Cope ; a fact difficult to reconcile with the kinetogenesis theory. In this
manner the inferior primitive triangle is broken, the upper molars develop
into the sexitubercular, the lower into the quadritubercular type.

The complemental development of the upper and lower molars in the
known genera of successive horizons is approximately displayed in the
subjoined table. The Eocene list of genera will be greatly reduced,
especially in the tritubercular-sectorial type, when the upper and lower
jaws are found associated, and it must be clearly understood that the sub-
types o, b, c, in this table, are closely related by transition forms. In
fact, in carnivorous forms, the extreme secodont and bunodont types are
frequently seen side by side, as in the first and second inferior molars of
Didymictis. The chief distinction between these two series is the greater
development of the secondary cusps and the almost invariable loss of the
paraconid in the latter; this is effected by the broader surfaces of contact
in the bunodont crowns. In the secodont series, on the other hand, the

TRITUBERCULAE KVOUTIOX IX MAMMALS

45

3 ^

. *

J do

|

«

S

J.

0

p

1— 1 T)<

be

3

9

<

J

,.*

Ari • ^

3

^ — ^ rf

•d

. 33 •£ 3

£) 1^ Jj ~ Q

O5 —

o

S

'x s S

O

co O''2

•a

OQ

^

3

^

W
CO

. c5 . d

0> S-. 0 £
£ O X D

3
>
+3

H

•_

D

O
Q

•5 "^ -5 *°

00

O

5- c-

^

z

0

ffl

i_§ _gjj

EH

x-^

[Q

d 11 lp

-* -M

3

' 1 1*

ci
ft

M

O

S

•- ^ -3

« "^ <•; .S

f>

. ,= = ^T

73 3 05 O
•4-3 a -*^

fM iO ^
CM

§

IO

.

EH EHJC

.S

-t-t

a

3

H

05

. t-' i — '

ir " - "

o - .
K « ^

'A

= 11 |1

I CM

0 " I-

O

Q
O

o

1 £

Y1

bo '' P i

.5 °

w

r^ fl ^

M

2 -7? S

*

o

"rH "tn

x - a,

c S i o
~ 3 o"3

r-t rc —

0 d « c

«f^ 1 3

.3 .2

3 a w

H H

(D o O r^

bo-0 q 3

53 01 bo 3

-tj

be ^J o 60

3

<- , -3

•r1 "*? " .2

•S-2 2 9

1 5

•

^^ 2 3 N

S

PH

«^H ^ " CO

l*ji

•

1 ' i i i

^~* q5 -2 o
ij'o -ffl "«

3 S

g Sj 0 S

P. 5

3 d o .™

c, 5

r;

•*< ?! O *«

0 J

oT ' N ' iT iT '

.S °

,

— O> CC 'S

"S o "S ?

oT

« a> g o" S I

W^P rS" ^ ^
•S PH ^ ^ ^ *^

!H — i ^ "*

a a a

0 00

«H S-l 1*

£

iJ in S ^

&H fe fe

Q

^^ ^^ Q? ^H C^ Q,» ^

^ -, ^-

"

rs ^ 5- f j~-

^^ ^ ij^r^ ^ «

46 EVOLUTION OF MAMMALIAN MOLAR TEETH

development of the secondary cusps is subordinated, and the metaconid is
almost invariably suppressed.1

Ada-pis and Anaptomorphus are examples of sub-types a, c, associated;
i'<>r it frequently happens that the paraconid atrophies without a complete
enlargement of the hypocone. A study of the diagram demonstrates,
however, that the association of sub-types & and c is impossible. The
recent monkeys [Primates] Tarsius and Loris afford a good illustra-
tion of the association of tritubercular, quinquetubercular, sexitubercular,
and quadritubercular molars.

The subsequent evolution of the molars in different orders was
variously characterized, first, by loss of the primary cusps, e.g. the meta-
conid in the Carnivora, the paraconid in the Ungulata. Second, by the
loss of some of the secondary cusps, e.g. the proto- and meta-conules in
the Artiodactyla.2 '' Third, by the metamorphosis in the form of the cusps.
This subject has been fully treated by Kutimeyer, Kowalevsky, Cope,
Schlosser, and others.

The Relation of Trituberculy to the Persistence of Mammalian Phyla —
The above table shows somewhat indefinitely, but none the less positively,
the general progression of the Mammalia to and from the primitive
tritubercular type. As already stated, even with our present very
limited knowledge, certain stages appear to have been characteristic of
certain periods, as follows : the triconodont in the lower Jurassic ; the
primitive tritubercular and tubercular sectorial in the upper Jurassic :
the secodont and bunodont sub-types of trituberculy, predominated in the
Puerco ; in the Bridger, the Perissodactyl ungulates had mostly passed
beyond into the lophodont.and symborodont types, and the Artiodactyls
were approximately in the stage of sub- type c ; but the Lemuroidea,
Creodonta, Insectivora, etc., were, almost without exception, tritubercular.

There can be little doubt that, parallel with the tritubercular forms,
in each period, there were aberrant or degenerate types, but it is difficult
to determine which these are. Many Mesozoic types which the writer
formerly considered aberrant, have now proven to be tritubercular.3 The
upper Jurassic genera included under the Diplocynodontida? (see Marsh,
Amer. Jutmi. Sc., April, 1887, p. 338) are apparently aberrant. There

1 See Cope, "Origin of the specialized Teeth of the Carnivora," Am. Naturalist,
March, 1879.

2 Schlosser, " Beitrage zur Kenntniss der Stammgeschichte der Hufthiere," Morph.
Jahrh., 1886, p. 123, has especially drawn attention to the probability that the Artiodactyla
were derived from sexitubercular forms.

*[Many Artiodactyl families, e.if. Trigonolestida?, Leptochoaridre, Dichobunidse, Anthraco-
theriidaa, Elotheriida;, retain the protoconule. The loss of the metaconule is only apparent,
for it is represented in the enlarged postero-internal cusp, analogous in position to a
cingulum-hypocone, in the molars of most Artiodactyla. — ED.]

3 See "Additional Observations upon the Structure and Classification of the Mesozoic
Mammalia," Proc. Phila. Acad., Nov., 1888.

TRITUBERCULAR EVOLUTION IN MAMMALS 47

are several degenerate types among the Puerco and Wasatch Creodonts.
such' as Dissacus and J/c.s(* ////./. But there is a striking proof of the
superiority of the tritubercular molar in the fact that, according to our
present knowledge at least, the Jurassic mammals possessing aberrant or
degenerate molar types did not persist into the Puerco, nor did such types
in the Puerco persist into the Bridger. There is some doubt as to the
persistence of the sub-tritubercular stages ; the writer formerly considered
the T/t ///<«•//// ii* molars as triconodont ; but Mr. Lydekker has called
attention to the probability that the metaconid has disappeared and been
replaced by a heel as in the sectorial teeth of the Carnivora. The dis-
appearance of the degenerate types may be attributed to the general
principle that rapid ' specialization and loss of parts leads ultimately t<>
extinction, by depriving the animal of the means of adaptation to new
conditions or surroundings. The mechanical superiority of the trituber-
cular type over every other has been repeatedly demonstrated in its
plastic capacity of adaptation to the most extreme trenchant and crushing
functions.

CHAPTER III.

TRITUBERCULY IN RELATION TO THE HUMAN MOLAR TEETH

AND THE PRIMATES.

1.
ONTOGENETIC DEVELOPMENT OF THE TEETH.

[Reprinted from a paper in the Anatomisch.es Anzeiger, Jahrg. VII. (1892), 8vo, Jena,
pp. 740-747, entitled "The History and Homologies of the Human Molar Cusps
(a review of the contributions of Dr. A. Fleischmann, Dr. Julius Taeker, and
Dr. Carl Rose)".]

THE embryonic development of the cusps of the molar teeth in the
Mammalia has lately been discussed in two very interesting papers by
Taeker1 and Rose,2 and the homologies of the upper and lower cusps have
been investigated by Fleischmann.3 The work of the latter is based upon
the comparative study of recent types of molars, and the author reaches
the conclusion that the system of homologies proposed by Cope4 and
expanded by Osborn 5 between the upper and lower molars is erroneous.

Taeker's paper is chiefly devoted to the study of the succession and
embryonic form of the molar cusps in different Ungulates : he supports
by ontogeny the view based upon palaeontology that the ancestral cusps
were conical ; he shows that in the lower molars the ontogenetic order of
development corresponds with the phylogenetic order as traced by Cope
iimong the fossil forms, but that in the upper molars the ontogenetic order
does not correspond with the primitive phylogenetic succession as traced

1 " Zur Kenutnis der Odontogenese bei Ungtilaten," Dorpat, 1892.

2 " Uber die Entstehung und Formabanderungen der menschlichen Molaren," Anat.
Anz., 1892, Nr. 13 u. 14.

3 " Die Grundform der Backzahne bei Saugetieren und die Homologie der einzelnen
Hocker," Berlin, 1891.

4"The mechanical Causes of the Development of the Hard Parts of the Mammalia,"
Journ. of Morphology, 1889. Also earlier papers.

•""'The Structure and Classification of the Mesozoic Mammalia," Journ. Acud. Nat.
Sc. Phila., 1888, p. 240; also: "The Evolution of Mammalian Molars to and from the
Tritubercular Type," American Naturalist, 1888.

Ti;rrrr,Ki:<n.Y ix I-KI.MATKS 49

by Cope. Upon the whole, however, he finds a very striking parallelism1
between embryogeny and phylogeny both as to the form and succession of
the cusps (see Table mi p. Til ).

Rose's paper is also of great value in proving that in Homo and
It'idc/jilii/*, representing two widely separate classes, the embryonic history
of the lower molars approximately repeats the ancestral history: he
independently supports Taeker in the conclusion that the upper molar
cusps do not repeat the ancestral order assigned by the Cope-Osborn
theory, he therefore agrees with Fleischmann that we have mistaken the
history and homologies of the upper molar cusps, and suggests very
courteously that the Osborn nomenclature should be transposed to cor-
respond with the embryological order ; he further advances the original
theory that the mammalian cusps have arisen not by addition to the
single reptilian cone, but by the fusion of a number of cones together. I
will first consider the main principles involved in these papers, and then
mention some of the less important special points. The following table
exhibits the correspondence and contrast between the phyletic and
embryonic succession, as well as the homologies and order of appearance,
according to the Cope-Osborn theory.

From this table the striking parallelism between ontogeny and
phylogeny in the lower molars is brought out. Also the contrast between
the early appearance of the hypoconid both in phylogeny and ontogeny
and the late appearance of the Jn/pocone phylogenetically and ontogeneti-
cally,

1. That iJie primitive form of mammalian molar was a siiif/le cone to
which all the other cusps have been successively added. I may first take
up the different theory of cusp origin proposed by Rose, and observe that
whatever support it may receive from embryology is offset by the over-
whelming evidence of palaeontology. In figure 38. I have epitomized the
slow transformation of the single-fanged conical reptilian tooth (1), such
as we see persisting in the Cetacea,* into the low-crowned human lower
molar (8). The first departure towards the development of lateral cusps
is seen in the triassic Dromatherium (2) ; the second is in the contem-
porary Micron i, iddini (3) : the third is in the Jurassic Spalacofhcri/nn (4) ;
in the fourth (Amphitli&rium, Jurassic) (5), we see the three cusps of the
primitive triangle and the first cusp of the talon, hi/'1. In Miacis of the
lower Eocene (6) the figures of the internal and crown views of the three
molars show how the primitive anterior portion (trigonid) of the crown
was reduced to the level of the posterior portion (talonid) while retaining
all of its cusps. In the next figure (7) we see the lower molars of the

1 See Osborn, " Odontogenesis in the Ungulates," A met: Aat., 1892, p. 621. A fuller
review of Dr. Taeker's paper.

*[More probably secondary in Cetacea. See pp. 79, 190.— Ei>.]

D

i (LI B R AR YJ3QJ

• \

50

EVOLUTION OF MAMMALIAN MOLAR TEETH

oldest monkey or lemur known, Anaptomorphus, which illustrate the
loss of the antero-internal cusp or paraconid, pa'1, — this is present as a
rudiment in ml and m2, but has disappeared in ?/?3. This accounts for
the history of all the cusps in the human lower molar. Thus in the rich
series of Mesozoic l and lower Eocene Mammals we can observe the actual
rise, succession, and decline of all the six cusps, and do not require any
new hypothesis to explain their appearance.

Dr. Rose supports his fusion hypothesis by a reference to the Multi-
tuberculates (p. 101); he could hardly have made a more unfortunate

me"

en

FIG. 38. Evolution of the cusps of the Human Lower Molar. [1. Simple conical, reptilian
tooth. 2. Dromatherium. 3. Microconodon. 4. Spalacnthd-ii'tn. 5. Amphith<.riv.in. 6. Minds.
7. Anaptomorphus. 8. Homo.]

choice, because between the little pauci-tubercular Microlfxtf* of the upper
Triassic and the multi-tubercular Neoplayimdax of the lower Eocene we
can follow the successive addition of tubercles with ease. I expect soon
to demonstrate that the molars of this aberrant group were also of
tritubercular, i.e. haplodont origin.*

It is a striking fact that all t the molars of the Triassic and Jurassic
periods are distinguished by one conic cusp much larger and more
prominent than the others which are smaller and upon a lower level.
What are the positions and homologies of this cusp in the upper and
lower teeth '.

1 See the Memoirs of Owen and Osborn upon the Mesozoic Mammalia.
*[See, however, the later views expressed on page 105, foot note. — ED.]

t[Even in MirroleNtes one cusp is higher than the others. (See Fig. 48, p. 102.) But
too much importance should not be attached to this fact. — ED.]

TRITUBERCULY IX PRIMATES

51

-r .-.

C

oi

O

so'

u
OJ

•o

u,

0

Ungulatei

Paracom

-3 -3

a! P
- £

OH
hH

«

CJ

'•n

0

_bb

"S

u
ha

~CS
'ft
00

acone

3 ^

O

PH

\~r

^~t—

o

f-,

o o>

^

Q_

-4-1

s
s

£

>" ^H
p i f^

w'

O

s

o
O

03

CQ

SERVED BY

H

Q

Primates2.

Paracone

Protocol! e
Metacone

i

o
o

W

03

O3

0

•o

•.wp.iQ

i— *

CM C>7

-t

05

PH

<1

><

K

W

W

0

Q

a

0

>

03

| ;

, ;

GENETIC

a
cc

pq

0

X

<

Human

03

Si

<

^

1 *

bo

Q

O

J

>H

PLH

pa

&3
P
03

0

u
0)
T3

CP

J

O

g

Pi

w

" PH

a

o

6

-f£

cfi

h

a

* 2

OJ

3

GENETI

O
t5
0)

hfl

c8
ft

1

o

o
O

-)-3
0

£

Paracol
Metaco

0

a,
>>

^

0

O

v— — y '

o

EH

*>»

«-.

W
-j-.

z

0

£

uapio

-

s

«

3

E

^

£J

9

H

'oD

jj

H

OQ

g

"« •

o -^

-

o

o

'bio's
,2'S

^ a

k3 01^

O^ l-

( K,

f-

&^J

O

.B •§ Si

=

% a h

• S a

J u/

) ^

Q)

C 2

0

PH

H O W

o
o
o

5 M

* >•. o

o

o i 6 i

-*-i ~ ^

o '" - •*-"

2 <s ^ a

cu P-I M w

1 1 1 1 1

d "^ ^~^ •*— > *^"^

2 01 >^ >-.

£ S M

r4

2
3

"S U ^ 15 g IS

| g So | lc 'i

^ .2 C^S '^H r^

- ~ ' H i f- + ^H

-— [t2

ll

as

a

63 E-

IB

— H

j3

-^

^"^

(— H

CO
h

B

pi

a>

(U

<<

4J

be

^LJ

^ll

02

"~^

nj
O

^

Ol

^

C

<5

o
PH

**

•^

o>
50
cu

J

S

-a

2

HI

&

—

S3

*~^

X

H

H

1^3

_

^3

~

0

^

0

otoconi

raconid

0

0

^

flj

0
o

0

toco n ic

'poconv

03

J

hH

a)

«*!

oa

PH

nd

PH

s.

S

K

^

w'

r- =«

0

0)

5
o
o

.5

a

o!

o .H

* 'T* .-^

0

01

.5

H

1-3 O

52 EVOLUTION OF MAMMALIAN MOLAR TEETH

2. That the protocone is invar/<tl>/// the nntd-ior lateral (antero-externai)
citxji n/ tin' lower rmtliti's «n<l f//<- anterior Unt/nn] (antero-internal) cusp vn
the H /trier molnr*. The former part of this proposition is now almost self-
evident. It is absolutely proven in such a series as we see in Figure 38,
and is now corroborated by the embryological researches of Taeker and Rose.

As to the present position or homologue of the reptilian protocone in
the upper mammalian molar there is relatively, I admit, more room for
doubt, mainly for the reason that fossil upper jaws are very scarce. If,
as held by Fleischmann and Rose, the antero-externai cusp is the protocone,
then the whole system of homologies held by Cope and myself falls to the
ground. Let us look at the evidence :

First : In the numerous upper jaws of Triconodon (Figs. 7, 8) of the upper
Jurassic, the main cusp is always the middle one of the three, correspond-
ing with the large middle cusp of the lower molars which we know to be
the protocone. Second : In the upper molars of Spalacotherium *
(Jurassic), in which the lower molars are of the simplest tritubercular
type (Fig. 36, No. 4), the most prominent cusp by far is the antero-internal,
supporting my view. Third : In all the Amblotheriidae t of the upper
Jurassic there is a triangle of cusps in both upper and lower molars, in
each the apex is formed by the most prominent styliform cusp, this is
antero- external in f/ir lower molars ami antero-internal in the /<j>j>/',' molars.
Is it at all probable, at this early period, when the protoconid is still the
most conspicuous cusp in the lower molars that a corresponding cusp of
the same form, but reversed position, invariably found in the upper
molars is not homologous ? According to the Fleischmann-Rose view it
is not, but the main lower cusp is homologous with one of the spurs of
the main upper cusp. Fourth : There are other important grounds of a
mechanical nature, Starting with the study of modern, instead of the
oldest fossil forms, Fleischmann has, I believe, reached not only an
erroneous conception of the homologies of the separate cusps, but of the
equally important homologies in the functional regions of the upper and
lower crowns. In each we may distinguish two regions :

The elevated primitive triangle (trigon) with a primitive cutting,
piercing or sectorial function.

The depressed heel (talon), with a primitive crushing or grinding
function.

In the earliest stages the upper and lower molars were simple
triangles of cusps, as in the modern Cape Mole, Chrysochloris. J The lower
molar had the apex (protoconid) turned outwards and the base (para-

* [Regarded as a synonym of Percdestes. (See p. 35 and Fig. 12.) — ED.]

t [Represented by Kurtodon (Fig. 13, p. 26), Dryolestes (Fig. 14, p. 26).— ED.]

J [According to the views expressed on pages 124, 126, 227, the form of the molars
in Chrysochloris is entirely secondary. — En.]

TKITri:KI!('lTLY IX PKI.MATKS

53

and meta-conids) turned inwards, wliile the upper molar had these
relations reversed. As shown in the accompanying diagram (Figure 39),
the opposition of these triangles makes a perfect cutting mechanism, and
as (Jope has shown this is effective at every stage of development. If
the protocone were at the outer angle of the upper molars, it is impossible
to conceive of an effective series of intermediate stages.*

The first step t towards the crushing function is the development of

trig art-

Lower Molar.

Upper and Lower Molars opposed.

Upper Molar.

FIG. 39. Key to Plan of Upper and Lower Molars in trituberculate mammals. Each tooth
consists of a triangle (triium) with the protocoue (pr) at the apex. The apex is on the inner side
of the upper molars and on the outer side of the lower molars. [In C the upper molar is too
far internal and posterior. The protocone of the upper molar should fit into the basin of the
talonid of the IUWLT molar. See Figs. 60, 208, 20!>.— ED.]

the hypoconid upon the incipient talonid which is later reinforced by two
more cusps, the entoconid and hypoconulid. Tin/* f//c entire //a/ or talonid
is complete upon the lower molars before it commences to develop upon tin-
upper molars, as shown in Anaptomorphm (Fig. 36, No. 7, and Fig. 40,
No. 9, Fig. 130) as well as in the ontogeny.

Upon the upper molars the talon is only developed in bunodont types,
such as the Primates and Ungulates, to still further increase the crushing
area of the crowns; it always arises as seen in the Primates (Fig. 40), by
a slow upward growth from the cingulum, opposite the protocone. In
its early stages the hypocone, hi/, always resembles the early stages of the
hypoconid and conclusive proof of its talon-like character is seen among
the Condylarthra (Haploconus), in which it appears as a wide separate
heel. So far, therefore, from the truth of Fleischmann's supposition that
the upper molars have one more element (the " entomere ") than the
lower, exactly the reverse is the case, for the lower molars early acquire
much the greatest extension of the talon, while retaining all the elements
of the trigon.

* [In opposition to this, however, see the views of Gidley iu Fig. 208. — ED.]
t[This paragraph appears misleading. See pp. 61*, 68*, 82*. — ED.]

54 EVOLUTION OF MAMMALIAN MOLAR TEETH

A beautiful illustration of the fundamental pattern of trigon and talon
in the upper human molars is shown in Hose's figure 4 of the molar of a
six months' child (Fig. 42). The protocone makes the apex, and is con-
nected by two spurs with the two external cusps, the space between
which is comparatively open as in the primitive forms.

Thus the homology of the antero-internal cusp of the upper molar
with the protocone is well supported by palaeontology and by dental
mechanics, but how shall we meet the embryological counter-evidence
established by the agreement between the independent investigations of
Rose and Taeker ?

This is also, I believe, explained by a study of the fossil forms. As
we have seen in the most primitive types the protocone was the most
prominent cusp in both jaws, but in course of later development of the
upper molars, during the Cretaceous and Eocene periods, the protocone was
depressed to the level of the paracone and metacone (see the primitive
Carnivora, Creodonta, and Insectivora). On the other hand, in the lower
molars the protoconid retained its relatively prominent position and size.
If the ontogenetic development of the lower molars corresponds with the
ancestral order, it is probably because the relative primitive position of
the cusps was conserved ; whereas in the upper molars, in which there is
less correspondence, it was lost. I find in the lower Eocene Ungulates
that the paracone and metacone are more important cusps than the
protocone. So far as the fossil Primates of the lower Eocene are known,
we find the protoconid is the most prominent cusp in the lower molars,
while in the upper molars the protocone is less prominent than the
external cusps. Rose's argument really turns therefore upon the expecta-
tion that foetal development should repeat ancestral history of the
Cretaceous period ! As the flattened form of the crown is from the start
a Ca?nozoic type, we should hardly expect the order of cusp succession to
invariably revert to a Mesozoic type. While not thoroughly convincing,
there is a great deal of force in this way of meeting the embryological
data.

Nomenclature. Rose (p. 400) apparently mistakes the homologies of
the lower molar cusps of man, for he has overlooked the fact that the
primitive anterior lingual cusp, or paraconid, has degenerated in the
Primates (excepting in a few Lemurs) while it persists in Diddphys, It
is not seen in the human lower molar at all. Its declining stages mark
the loss of sectorial function and can be readily followed in the lemurs,
and fossil monkeys ; as shown by Cope and myself it degenerates while
the hypocone in the upper molars develops. It follows that the anterior
lingual in man is the metaconid, while in Didelpliys it is the paraconid
and the mid-lingual is the metaconid. The posterior lateral cusp is
undoubtedly the hypoconid. Rose proposes the term " pentaconid " for

TUTUBERCULY IX I'll I MATKS 55

the distal or posterior intermediate cusp (///'', Fig. 38, No. 8). The term
is inappropriate, because this is not the fifth but the sixth cusp when \\v
reckon the paraconid. It is analogous to the intermediate; tulicivli-s of
the upper molars — I have therefore suggested the term " hypoconulid "
for it; this cusp is almost universal among lower Eocene Mammalia; in
the last lower molar it forms the additional lobe ; it is found strongly
developed in many of the higher Primates.

Hose (p. 406) expresses the belief that the typical form of primate
molar was quadritubercular as opposed to Cope's view that trituberculy in
human dentition is a reversion to the Lemurine type. The study of the
fossil forms as well as of any complete zoological series can leave no doubt
that the quadritubercular form is a comparatively recent acquisition.

In conclusion, I would refer both these authors to the types of molar
teeth found among the Mesozoic Mammalia. It was while studying the
rich collection in the British Museum that I became convinced of the
force and universal application of the tritubercular theory proposed by
Cope.

AMERICAN MUSEUM OF NATURAL HISTORY,
NEW YORK, July 18th, 1892.

2.

THE HISTORY OF THE CUSPS OF THE HUMAN MOLAK TEETH.

Address before the New York Institute of Stomatology, April 19th, 1895.
[Reprinted, under the title given above, from the International Dental Journal. July, 1895.]

I wish to congratulate the members present upon the formation of
this Institute of Stomatology. It seems to me to mark one of the
stages in the remarkable specialization of human knowledge when, at
the present time, it is proposed to devote the work of an entire
society to the scientific study of the mouth parts, as I understand
your object to be ; and I also gather from the fact that you have
asked me, as a comparative anatomist, to deliver an address this
evening, that you look at the subject in two ways, — from the stand-
point of applied or practical science and from the stand-point of theory.
It is on the theoretical side that I should like to bring before you
this evening the history or origin of the cusps of the human molar
teeth.

We take up this skull of an Eskimo, and you will observe that
the teeth (Fig. 40, No. 11) are slightly worn, and that the molars

56

EVOLUTION OF MAMMALIAN MOLAR TEKTH

have four cusps.1 Half a century ago this would have been con-
sidered as something ultimate, simply as an adaptation to human
diet ; but now that we have come to understand the doctrine of
evolution, we ask ourselves, What is the meaning of these cusps ?
what is their history ? what is their origin ? Now, these four cusps
which are present on the four corners of the teeth might be explained
by evolution in three ways. We might imagine that the crown of
the tooth was originally a low rounded summit, and that on the
summit these four cusps appeared at each angle ; no one has
advocated this. Or we might imagine that they represent the coming
together of a number of tips of pointed teeth, such as we see in the
jaw of this member of the dolphin family ; this is the theory which
has been recently advanced in Germany, and it has been called the
" cusp concrescence " theory. Or, again, we might imagine that these

me.

me.

me.

Kii;. 40. Evolution of the Human Upper Molars, '.i. Aiiiiiito,iiii,-/i/nix. a lower Eocene monkey.
10. An upper Eocene monkey. 11 and 12. Human: 11, Esquimaux; 12, Negro. See addition of
"talon," /i if, ti> ''trigon" composed of pit, jn-, i,n.

cusps have originated by a gradual addition to the sides of a primitive
single cone ; this we call the " cusp differentiation " theory, or the theory
of cusp addition, in distinction from concrescence. The differentiation
theory is supported by Cope, by myself, and others in this country.

Now, suppose an evolutionist were to trace back the history of the
monkeys and of other animals among their fossil ancestors, he would
find that the further back his researches extended the more simple
the types of the teeth would be ; he would find that the teeth
of the oldest types of ancestral mammals have a simple conical
form, the form that is preserved in the teeth of the whales and the
dolphins of the present day (Fig. 42* A, p. 64), or in the Edentates
as represented in the group to which the sloth and the armadillo of
South America belong.*

We have the same type of conical tooth preserved in the human
canines, and if we turn from the teeth of man, in which the canine
has almost entirely lost its original laniariform or flesh-tearing shape,
to that of the lower monkeys, we see that the canine is really a

1 E. D. Cope, "On the Tritubercular Molar in Human Dentition," Jour, of Mor-
phology, July, 1888, p. 7.

*[Iii both groups the simple conical form is now believed to be secondary (pp. 79, 151,
191). — En.]

TI-triTMKKCl'LY IX 1'KIM ATI-> •> <

pointed tooth: so that we may draw a surest ion tVoni this fact
that all the teeth of the series at one time were pointed.

It is moreover true that wherever we tind these pointed teeth
they are present in the jaw in large numbers, sometimes sixty or
seventy on one side and usually running far back into the mouth,
and it is this fact which led to the suggestion of the theory of
" concrescence " in the formation of molar teeth.

The

You might at this stage be not inclined to take this " concrescence
theory " seriously, but my address has been suggested largely by the
fact that it lias been taken very seriously by some well-known
anatomists in Germany : as seen in the position of Professor Schwalbe,1
in a recent article, in which he reviews the entire literature in
regard to the formation of teeth published during the past fourteen
or fifteen years, and concludes that in the concrescence theory
and the differentiation or cusp addition theory the evidence is so
evenly balanced that he cannot decide between them. It is, therefore,
a question xnl judiu1, and worthy of the attention of odontologists.
As to the source of this theory, it was proposed simultaneously by
two Germans, both of whom claim the credit of originating it. One
is Dr. Carl Rose, a physician of Freiburg, a man of fine powers of
research and great energy, since he has, during the past few years,
issued in rapid succession a series of valuable papers on the embryo-
logical development of the teeth, which place him in the front rank
of students of this subject in this decade. The other is Professor
W. Klikenthal, of Jena, whose views sprang principally from the
study of the teeth of whales. While these two writers are in doubt
as to which should enjoy the precedence, I find, in correspondence
with my friend Dr. Ameghino, of the Argentine Republic, also
originally a physician and now a distinguished palaeontologist, that
he promulgated this theory as far back as 1884. In a work which
he published at that time, entitled Filogenia he says : ''' For the
reasons we are about to give it is evident that all mammals which
possess compound teeth have in past periods possessed a very
much larger number of teeth, but of quite simple conical form,
like those of the modern dolphin. The most primitive mammals
must also have had a number of very elevated teeth, but it is dim-
cult at the present time to determine how large this number was.
Nevertheless, if we take as an example a mammal in which the
dentition is complete, as in the Macrauchenia z or in the horse, and

1 " Ueber Theorien tier Dentition," Anatomischer Anzeiger Central-Mat t, 1894.

2 This is one of the peculiar extinct South American hoofed animals.

58 EVOLUTION OF MAMMALIAN MOLAR TEETH

if we reduce the number of its compound teeth, we find that the
most remote ancestors of these forms must have possessed more
than one hundred and fifty teeth. This number is certainly not
exaggerated, because Priodon, the giant tatusia (armadillo), a mammal
in an already quite advanced stage of evolution, possesses nearly one
hundred simple teeth, and in the dolphin this number rises from
one hundred and fifty to one hundred and seventy." I read this
to show that if there is any truth in the concrescence theory, Dr.
Ameghino partly deserves the credit for it. Moreover, we learn from
Schwalbe that the same theory was advanced by Professor Gaudry
in 1878, and still earlier by Professor Magitot in 1877.

Now let me illustrate, in a very simple manner, what is meant
by the theory of concrescence and how we can imagine that the
human molars have been built up by bringing together a number
of isolated teeth. Placing a number of conical teeth in line, as
they lie in the jaw of the whale, they would represent the primitive
dentition. In the course of time a number of these teeth would become
clustered together in such a manner as to form the four cusps of
a human molar, each one of the whale-tooth points taking the place
of one of the cusps of the mammalian tooth, — in other words, by a
concrescence, four teeth would be brought into one so as to constitute
the four cusps of the molar crown. Vertically succeeding teeth might
also be grouped.

Now, what evidence is there in favour of this theory, and what
is there against it ? First, there is this, that all primitive types of
reptiles from which the mammalians have descended and many
existing mammals, as we have noted, have a large number of isolated
teeth of a conical form ; secondly, we find that by a shortening of
the jaw, the dental fold or embryonic fold, from which each of the
numerous tooth-caps is budded off' in the course of development, may
be supposed to have been brought together in such a manner that
cusps which were originally stretched out in a line would be brought
together so as to form groups of a variable number of cusps accord-
ing to the more or less complex pattern of the crown.

What may be advanced against this theory ? This, and it is
conclusive to my mind : we find at the present time that cusps,
quite similar in ail respects to each of the cusps which form the
angles of the human molar, are even now being added to the teeth
in certain types of animals, such as the elephant, whose molar teeth
cusps are being complicated now or until very recent times. Then
we find in the Mesozoic period certain animals with tricuspid teeth.
Xow, according to the theory of concrescence these teeth ought not
to show any increase of cusps in later geological periods ; but as

TRI'lTDKRCULY IN I'll I. MATHS 59

we come through the ages nearer to the present time \ve find that
the successors of those animals show a very much larger number
of cusps. How is this increase of cusps to be accounted for? II;is
there been a reserve store of conical teeth to increase the cluster?
No. Must obviously, to every student of the fossil history of cusps
there is no reserve store, but new cusps are constantly rising up on
the original crown itself by cusp addition.

However, do not let me give you the impression that these
researches of Eose and Kukenthal are not of the greatest value
and interest : we shall see later on how the very facts of embryology
which are advanced by Dr. Carl IJose in support of his hypothesis
can be turned against him and used to support the differentiation
theory.

I have no doubt many of you have observed, in the examination of
human lower molars, that occasionally instead of having four cusps
they have five. The fifth cusp always appears in the middle of the
heel, does it not, or between the posterior lingual and the posterior
buccal ? You find this in the monkeys and in many other mammals,
but I know of no record of the ancient anterior lingual reappearing.

So we see that the human lowrer molar tooth with its low, quadri-
tubercular crown has evolved by addition of cusps and by gradual
modelling from a high-crowned, simple-pointed tooth. Xow this, and
I say it with great confidence, is what has actually taken place.
It has not come about by bringing together single reptilian cones ; it
has been simply by the addition of one cusp after another to an
original single reptilian cone until there were six cusps, and then,
in the adaptation and fitting of the lower teeth to the upper, one
of the cusps has disappeared. This cusp was the primitive anterior
lingual, or, in comparative anatomy, the paraconid (Fig. 38, No. 8).

Now let us follow the history of the upper teeth and see why
the " primitive anterior lingual," or paraconid, in the lower jaw has
disappeared.

You are constantly in your practice, observing that one tooth in
the lower jaw gets into the way of another tooth and has to be pushed
out of place in order to place its opponent in the upper jaw into
its proper position. This is exactly what Nature has done ; Nature
has abandoned that lower cusp simply because, in the simultaneous
transformation of the upper teeth from a three-cusp to a four- cusp
type, there was no room for it.

60 INVOLUTION OF MAMMALIAN MOLAR TKKTH

Mechanical Relations of tlic Upper an/I Loire r Teeth*

Let us examine the upper teeth. \Ye must say, in the h'rst
place, that our evidence here is not nearly so complete, because a
lower jaw, from its thin nature, is more apt to be presented fossil
than an upper jaw : so that in the older rocks we meet with ten
lower jaws to one upper ja\v, and we cannot get the same evidence
as to the history of the upper jaw that we have of the lower ; but
although we are not able to trace the history of the upper teeth
with the same accuracy or degree of certainty, we have every reason
to think it was the same. We find the upper teeth shaped like a
triangle, as in Figs. 12, 13, 14, so we may imagine that the same
triangle which was formed in the lower jaw was formed in the upper
jaw, with this important difference, that in the upper jaw the base of
the triangle was turned outward, whereas in the lower jaw the base
of the triangle was turned inward (Fig. 36, No. 4).

What I mean by this is illustrated in the accompanying figure
(Fig. 41, A — J), which is an epitome of the whole history. The
upper teeth are represented solid, the lower teeth as hollow circles.

In A we see a row of single cusps, the lower somewhat inside
of the upper. In B the lateral cusps are added. In C they are
enlarged. In I) the cusps are pushed outward and inward into
triangles. In E a spur is added on the lower molar triangle, which
in F and G grows out into a broad heel. In H and / a spin-
appears upon the upper molar triangle, and in J this causes the
lower molar triangle to lose its anterior cusp. Nature has corrected
any possible interference between these triangles in a simple manner
by turning the base of the triangle of the upper molars outward
towards what you call the buccal side. In the lower jaw, on the
other hand, the base of the triangle is turned inward to the lingual
side, so that finally we have the two triangles alternating, coming
together as in D and making a beautiful cutting mechanism ; because
if any food gets in between these triangular shears the food tends to
press these teeth forward and backward, therefore crowding the teeth
more closely together and tending to tighten and improve the shear,
whereas if the teeth were placed in line, as in C, and food were
to get in between, the effect would be to crowd the two jaws apart
and lessen the exact cutting power of the shear.

Now we see that we can compare the lower and upper triangles
to each other. How about the heels or spurs, and why were
they developed ? They were developed because these animals required

* [Other and more recent views as to the mechanical relations of the evolving tooth
parts are presented on pp. 61*, 68*, 82*, Figs. 208, 209.— ED.]

TKITri'.KlKTLY IN PRIMATES

61

crushers as well as cutters*: they required to break up their food,
and consequently a crushing surface was developed in each heel. lu
the course of time the animal gave up its cutting and tearing
function, and in all the group of animals to which man belongs it

A

000

B

VAVAV

E

F

\<w,

J

FIG. 41. Mechanics of Cusp Addition (diagrammatic). Compare with shaded drawings in
Figure 3$. A, the conical stage* (No. 1); B, C, the triconodont stages* (Nos. 2, 3); D, the first
triangular stage t (No. 4) ; E, !<', 6, the triangular upper molar, the lower molars, with triangle
and heel (Nos. 5. (5, 7, !>) ; H, I, upper and lower molars, with triangle and heel ; /, human type,
upper molar*, with four cusps, triangle, and heel (Xos. 10, 11, 12); lower molars, with five cusps,
antero-internal cusp having disappeared (No. 8). [Compare this diagram with that of Mr. Gidley
on p. 20s. It is now regarded by Professor Osborn as erroneous in several particulars. — En. ]

* Nutc that the upper teeth (Mack) bite outside the lower teeth (see Fig. 221).

t Note that the protocont's bite inside uf and between the lower teeth (see Fig. 221).

acquired a purely crushing function, as seen in the teeth of the
baboon. As that became necessary, the next step was to transform
the entire upper tooth into a crusher as well as the lower, and to

* [While the anterior and posterior aides of the upper and lower triangles no doubt
formed the principal cutting surfaces, yet the principal piercer and crusher in the upper
molars is the internal tip of the triangle, namely, the protocone, which opposes the talonid
below. The hypocone is an accessory crusher developed to oppose the trigonid and the
space back of the preceding lower molar. — ED.]

62 EVOLUTION OF MAMMALIAN MOLAR TEETH

till out all the spaces between them, so that a square lower tooth
would abut against a square upper tooth, as in J, and this was done
by simply adding a heel to this tooth. Now, what would that heel
come against in /? It would come against the anterior cusp of the
lower triangle ; therefore that cusp had to be removed, so when
the upper heel was developed this lower cusp was removed and the
lower molar, which had six cusps, presented only five ; then the
second lingual cusp was pushed forward as in J, and the tooth
was transformed into a quadritubercular molar.

Evidence that the Upper Human Molars were Triangular.

How do we know that is so i We have some conclusive evi-
dence of it in other animals of the group to which man belongs.
Beginning with the lemurs, the lowest type of monkeys, and entirely
separate in many respects from the higher types, we find almost
without exception that the upper teeth are triangular, there being
no posterior cusp, so that Fig. 40, Xo. 9, accurately represents a tooth
of the lemurs, and it also represents the tooth of the true monkeys
which we find in the Eocene period ; in other words, all monkeys or
all primates (the group to which man belongs) had at the outset
this triangular upper molar. Then earlier or later in the Eocene or
Miocene the spur began to be developed which transformed a three-
cusp tooth or a triangular tooth into a quadritubercular tooth. That
spur became enlarged and finally, in civilized races of men, we have
a tooth of this form as the prevailing type of tooth. These stages
are shown in Fig. 40.

Xow, we might say that the evidence is not perfectly satisfactory,
because we have no positive reason for believing that the human
teeth were derived from such a type as this ; they may have come
along another line of descent, and for that reason we have to show
here, through the kindness of one of the members of the dental pro-
fession in this city, the teeth of an Eskimo (Fig. 40, Xo. 11), which,
as Professor Cope has pointed out, differ from the teeth of all negroes, all
Indians, and all the lower races of men, in presenting in a much
clearer manner the primitive triangular arrangement of the cusps that
characterize the lemurs. A friend has just been telling us what
very few of us knew, — that the Eskimos do not chew their food :
they simply swallow7 it whole or gulp it down ; and their food
consists largely of blubber. Blubber does not form much resistance
to the teeth, and, whether as a mechanical or an inherited effect
of the lack of resistance of soft food through many generations

O v C?

of blubber-eating Eskimos or not, the teeth of these Eskimos are

Ti:rrrr,KKrriA i.\ PRIMATKS 63

exceptionally //•//////, rr//A//'. This fact was pointed out l>y Professor
Cope in his article entitled. " Leniurine Reversion in Human Dentition."

Up to a certain point in their evolution the molar teeth of all
mammals followed exactly the same route.* It follows that if we once
grasp the principles of cusp addition upon this triangular ground
plan we can compare the cusps of the molars of man with those of
any other mammal. In the teeth of the bear, for example, the
homology is very obvious indeed. But in the teeth of the cat the
homologies can only be determined when we procure the ancestral
forms of cats, for in the evolution of the large sectorials many
cusps have degenerated. Some years ago, when I had fully demon-
strated the truth of Cope's theory by my own studies, I saw the
importance of using a set of standard terms for the cusps. These
have since been almost universally adopted by comparative anato-
mists, but have not, as yet, I believe, made much headway among
human odontologists. They are, as follows, as applied to the human

teeth :

UPPER MOLARS.

Anterior palatal . . . Protocone )

Anterior buccal . . . Paracone , Primitive triangle, or " trigon."

Posterior buccal .... Metacone ^

Posterior palatal .... Hypocone Primitive heel, or "talon."

LOWER MOLARS.

Anterior buccal Protoconid In- •,.•

. , , . }- Primitive triangle, or " trigonid.

Anterior lingual . . . Metaconid J

Posterior buccal . . . Hypoconid )

Posterior lingual .... Entoconid > Primitive heel, or " talonid."

Posterior mesial .... Hypoconulid j

When we understand that all the teeth of all mammals have this
key, this tritubercular key, we c'an unlock the comparisons through
the series and point out the homologies.

There is further evidence in support of the theory of cusp addition
which I will now briefly mention. It is that brought forth by tin-
very investigations of Dr. Carl Rose, which he has used to support
the concrescence theory. We should expect, in the embryonic jaw
that the calcification of the tooth-germ would be very significant,
because we know that the embryonic structures in their development
follow the order of addition or evolution. The order of evolution is,
to a certain extent, repeated in embryonic development. How is it
with the teeth ? Dr. Rose has given a most exact account of the
mode of calcification of the tooth-germ within the jaw ; this is also
now to be had in the form of wax models, prepared by Professor
Zeigler, of Freiburg.

1 Journal of Morphology, Vol. II., 1888, pp. 1-24.

* [At the present time this statement seems very doubtful. — -ED.]

64

EVOLUTION OF MAMMALIAN MOLAR TEETH

To begin with the lower molars, the dental cap in the jaw forms
a broad, saucer-like surface, and then at the corners of that cap

Fi<;. 4-J. Developing Upper and Lower Molars of a four weeks old child. Note the resem-
blance in the upper molar to the molar of the Eocene Primate in Fig. 40 No. 10. After Rose.

A

B

C

FIG. 42'. The three Primary Forms of Molar Teeth, secondarily attained in three modern
forms. A, Haplodont, of the Dolphin. B, Triconodunt, of the Leopard Seal, Ogmorhinus
l< /itvni/.i: C, Tritubercular, of the Cape Mole, Chrywhlorin (see p. 125).

calcified points appear (Fig. 42*). In what order (Fig. 1) do they
appear I The order is shown in the following table :

COMPARISON OF EVOLUTION AND EMBRYONIC DEVELOPMENT.

UPPER MOLARS .

LOWER MOLARS .

Order by "Cusp
Addition' Theory."

1. Anterior palatal.

/ Anterior buccal.

\ Posterior buccal.
4. Posterior palatal.

2.

1 . Anterior buccal.

2. Anterior lingual.

3. Posterior buccal.

4. Posterior lingual.

5. Posterior mesial.

Order of Embryonic
Development.

1. Anterior buccal.

2. Anterior palatal.

3. Posterior buccal.

4. Posterior palatal.

1. Anterior buccal.

2. Anterior lingual.

3. Posterior buccal.

4. Posterior lingual,

5. Posterior mesial.

TRITUBKRCULY IX PRLMATKS (55

In the lower molar teeth the order of calcification is precisely
the order of evolution, — in other words, the anterior buccal was
the first to evolve, representing the reptilian cone ; it is also the
first to calcify. The anterior lingual is the second in age, and also
the second to calcify. The third and the fourth cusps calcify almost
simultaneously. So we find that the order of embryonic development
exactly repeats the order of historical development, and in every way
presents the strongest kind of confirmation of the theory of cusp
formation which we have been discussing. But this you see is not
exactly the case in the upper molars. Nevertheless, out of eight
cusps in the upper and lower molars considered together, six cusps
calcify in the order in which they were successively added to the
single reptilian cone.

Gentlemen, I trust that I have not in this address taken you
too far afield. I have reached a conclusion on this subject which
could be elaborated in much greater detail. In closing, I would
like to refer to the work of Dr. J. L. Wortman, who is here this
evening, and who was for some years a collaborator with Professor
Cope in Philadelphia, and who in association with Professor Cope
had quite a share in the establishment of the " tritubercular or cusp
addition " theory. This theory is now a rival to the " concrescence "
theory ; and, while it may not seem a matter of great importance, if the
concrescence theory may not seem one we ought to take seriously, still,
iu view of the attention which it has gained in Germany, it is
time that we produce and bring forward the unimpeachable evidence
which we get of the history of these teeth from the rocks, the solid
evidence from the geological formations, the evidence of comparative
anatomy, which, as we have just seen, is so far supported by the
evidence of embryonic development.

BIBLIOGRAPHY.

Works of reference in addition to those cited above :

Rose, " Ueber die Entwickelung und Formabanderung der menschlichen Molaren,"
Anatomisch&r A»:<'i(/er, Band VII., 1892.

Kukenthal, '' Ueber den Ursprung und die Entwickelung der Saugethier-
/iihne," Jenaische Zeitschrift fiir Naturwissenschaft, Band 28, 1893.

Osborn, "The Evolution of Mammalian Molars to and from the Tritubercular
Type," American Naturalist, 1888, p. 1067.

"The History and Homologies of the Human Molar Cusps," Anatomischer
Anzeiyi-r, VII., 1892, pp. 740-747.

Cope, " The Mechanical Causes of the Development of the Hard Parts of the
Mammalia," Journal of Morphology, III., 1889.

E

CHAPTER IV.

TRITUBERCULY IN ITS APPLICATION TO THE MOLAR TEETH OF
THE UNGULATES OR HOOFED MAMMALS. COMPLETION <>K
THE NOMENCLATURE.

1.

DISADVANTAGES OF PREVIOUS SYSTEMS OF NOMENCLATURE OF TIIH

MOLAR CUSPS.

[Extract from article entitled "The Nomenclature of the Mammalian Molar Cusps,''

The American Naturalist, October 1888.]

IN view of the evidence for the almost universal presence of the tri-
tubercular stage in the present or past history of the upper and lower
molars, I have already advocated a distinct nomenclature for the different
cusps which compose this molar and its derivatives, up to the stage of the
acquisition of six tubercles in the upper molars and five in the lower.
This is the final stage in which the tubercles remain distinct. The
nomenclature now in general use is based, for the most part, upon the
secondary or acquired position, and in no instance, so far as I know, upon
the demonstrable homologies of the cusps in the upper and lower jaws.
Compare, for example, the molars of Miodcenus and Hyopsodus. By those
familiar with Cope's writings upon this subject, it will be recognized at
once that the antero-internal cusp of the lower molar of Miodcenus is not
homologous with the antero-internal cusp of the upper molar of the same
genus, nor is it homologous with the antero-internal cusp of ihe lower
molar of Hyopsodus.

•2.

METHODS OF ANALYSIS OF MOLAR ELEMENTS, NOMENCLATURE OF THE

MOLAIIS OF UNGULATES.

[Extract from Osborn and Wortman, " Fossil Mammals of the Wahsatch and Wind
River Beds Collection of 1891," Bull. Ame.r. Mm., Nat. Hist., Vol. IV., No. 1, Oct. 20th,
1892, pp. 84-93.]

In October, 1888, a table of nomenclature for the cusps of the molar
teeth of mammalia was published in the American Naturalist. } The

1 Osborn, " The Nomenclature of the Mammalian Molar Cusps," op. cit., p. 927.

T]![Tr.BKl;(TIA IX UNGULATES 67

terms were carefully chosen with reference to the gradual rise of these
cusps from the single cone of the reptilian type, through the tritubercular
to the sexitubercular stages.1 They have since been wholly or in part
adopted by Cope, Scott, Lydekker, Schlosser,2 Flower, and lately by
liiitimeyer.3 The tritubercular stem form has been recognized by
Doderlein and Fleischmann, but these authors have employed various
Greek symbols for the cusps. The latter has opposed the adoption of
similar terms for the main cusps of the upper and lower molars, upon the
ground that Cope and myself have mistaken the homologies : this
objection would be fatal to a uniform system of nomenclature for
the upper and lower cusps if it could be sustained, but a compre-
hensive survey of the Mesozoic trituberculates, especially of the
Amblotheriidse and Spalacotheriidse, leaves no doubt that the. anivm-
ciixji in fin /m/r/1 mo/a/-* «n<l tin antero-internal cusp in fin
i' molars of the mammalia, are homologous -with fin- i-<j//i/i«,i com an>/
,'ncli <>flit:r ; these cusps are invariably the most prominent, and are
always styliform in primitive types ; they always form the apices of the
primitive crown ; they persist in almost all mammals, while one or all of
the later cusps may disappear.

This cardinal point established, it will be a great gain for palaeontology
and comparative odontology when the further truth is recognized that
fin jio^iliilitics of modification of ti/pc l/i tin* molars are limited, that
essentially similar types of teeth are evolved independently over and
over again, and that in course of what Schlosser has well termed
' modernization ' we find such diverse orders as Primates, Ungulates,
Insectivores, Marsupials, Rodents, all exhibiting the same laws of dental
modification, and the same or similar ' secondary ' cusps, crests and
peripheral styles.

Except in the Cetacea and Edentata, these modifications centre
around the simple tritubercular crown, which seems to possess unlimited
capacity of adaptation by the development of some parts and degeneration
of others, by changes of form and position, and by the addition of
secondary cusps.

The first step is to distinguish and separate clearly the primary and
secondary regions of the primitive crown, for originally they have absolutely
different functions ; the part first developed in both upper and lower
molars is the anterior primitive triangle or tri</<m, which has a cutting or
piercing function ; out of its three cones all ' secodont ' types of molars
are evolved. The part next developed is the talon, or heel, which has

1See also " Evolution of Mammalian Molars to and from the Tritubercular Type," Am.
NaL, December, 1888.

2Schlosser, "Die Difl'erenzierinig des Siiugethiergebisses, " Biologisches Centralblatt,
Juni. 189H.

3 Die Eociine Saugethier- Wtlt ron E</< i-L-in<i< », Zurich, 1891.

68 EVOLUTION OF MAMMALIAN MOLAR TEETH

a crushing or grinding function, and therefore plays a chief role in all
' bunodont ' types. The first diagram exhibits the relations of these two
portions of the crown in the upper and lower molars, and the six primary
and secondary cusps which typically develop upon each (Fig. 41).*

We will not enter here into the well-understood transformation of
this tuberculo-sectorial type into the sexitubercular bunodont type, seen
typically in the upper and lower molars of the Puerco Protogonia
[JBuprotogonia, Fig. 149, p. 169], which is the least specialized ancestral
bunodont form that has been discovered. We may lay emphasis upon
the fact that the parent form of ungulate molar has six tubercles both
above and below instead of six above and four below as formerly
supposed.

It is important to remember, as an exception to the law of sexitu-
bercular origin, that all the Amblypoda and all the Periptychidaet1 (among
the Condylarthra) developed their upper molars upon the trigonal basis,
out of the three tubercles of the tritubercular crown, and without
becoming sexitubercular, that is, without the addition of the hypocone
or talon.

Now how shall we study the molar teeth of the early Ungulates,
especially of the apparently similar primitive forms of Perissodactyls,
which are so difficult to distinguish ? The following steps must be
taken :

First. Locate each of the six primary and secondary cusps, as far as
they are present.

Second. Note the form of each, whether rounded (bunoid), crested
(lophoid), or crescentic (selenoid).

Third. Note the position of each upon the crown with relation to
the other cusps.

Fourth. Note the relative size or development of each.

Fifth. Note the relative development of the cinyidum, in different
parts of the contour.

Sixth. Note the presence of one or more peripheral secondary cusps,
which develop from the cingulum, or external borders of the crown.

Finally. If crests are formed or forming, note the points at which
the transverse crests unite with the external cusps (paracone and metacone,
paras tyle and mesostyle).

"[From the evidence furnished by the molars of the Jurassic and Cretaceous mammals,
of the Eocene trituberculate Creodonts, Insectivores, and Primates (e.g. Anaptomorphiis), it
seems probable that the talonid or heel in the lower molars appeared much earlier than the
so-called "talon" or hypocone in the upper molars. See Gidley's views on page 221, Fig. 208.
The hypocone of the upper molars (p. 59; Figs. 144, 147, 252), was developed part passu
with the degeneration of the paraconid, and fits in between the entoeonid of one lower molar
and the metaconid of the next lower molar. It is really a secondary crusher, analogous with
the protocone. — ED.]

1 There are considerable grounds for removing the Periptychidse from the Condylarthra
to Amblypoda.— O. f [See p. 164.]

TRITUBERCULY IN rxci'LATKS 69

These differential features, it will be observed, follow the progressive
order of evolution in the molar crowns, for in ' modernization ' we see,
first, a degeneration of one or more of the primary cusps, then a remoddl///'/
of the form of each cusp which may affect the twelve upper and lower
cusps very differently : for example, in such an ancient type as Mcni*-
cotherium we find one bunoid, two lophoid, and three selenoid cusps in
each of the upper molars. Third, the cusps begin to shift their positions
upon the crown. Fourth, they begin to develop unequally. Fifth, the
cingulum, which is primitively a complete peripheral band, begins to
disappear at certain points. Sixth, one or more peripheral cusps grow
up from the cingulum or upon the sides of the main cusps. Finally,
as the crests develop, the unequal development of the cusps causes the
transverse crests to unite at different points with the external crest.

We find that if such analysis be applied to the elements of the molar
teeth we derive an absolutely infallible means of distinguishing different
lines of descent, for the above are the main features of divergent evolution.

The primitive horse, tapir, rhinoceros and titanothere all stand apart
and cannot be confused ; each have their clear differentia. To check
the possibility of being misled by parallelism in molar form, we should
next observe the dental series as a whole, the proportionate development
of different members of the series — the metatropJiism ; this often furnishes
the final proof or disproof of relationship, so far at least as can be derived
from the dentition alone.

The above method of analysis is the outgrowth of an extremely
careful study and comparison of all the early Coudylarthra and Perisso-
dactyla, and it has been found necessary to exercise the closest scrutiny
to distinguish these early stages of divergence,*

Now to turn to the subject of nomenclature, the system of terms was
originally based upon the actual homologies of the primary elements of
the trigon and trigonid, but in extending it to the other parts of the
crown and to the secondary cusps it was found that we must apply
similar terms to some of the later elements in the upper and lower teeth,
which are merely analogous to each other (performing a similar function)
occupying a similar position, and developing at about the same period,
otherwise the terms soon multiply so as to become a burden rather than
a convenience. [See Figs. 43, 135.]

As far as possible, therefore, the same prefixes are retained for the
secondary parts of the molars as for the primary; thus the anterior
transverse crest of the upper molars is called the protoloph, as it is
invariably developed by the union of the protocone, protocouule and
paracone or parastyle, never from the metacone. The anterior transverse

[*It should be noted that the molars of the Lower Eocene ancestors of the horse
(Eohippus) and of the tapir (Syttemodon) are closely similar. — ED.]

70 KVOLUTION OF MAMMALIAN MOLAR TKKTH

crest of the lower molars is termed the metalophid because it is always
developed from the metaconid or metastyle, and protoconid, never from
the paraconid ; the posterior transverse crest of the lower molars is
termed the hypolophid, because it is mainly formed by the hypoconid and
entoconid, never from the metaconid or paraconid. The external crest
of the upper molars is composed of so many cusps that it requires a
distinct prefix, but is readily remembered as the ectoloph. So with the
peripheral cusps, one or more of which are developed in all Ungulates,
and are especially numerous in molars of the Equidse ; to these the
terminal -style is applied in lieu of the English term ' pillar ' proposed
by Huxley — we can readily locate the parastyle as the antero-external
buttress which is developed near the paracone, the mesostyle as developed
on the outer wall between the paracone and nietacone. Similarly, in the
lower molars, we find in several lines of Ungulates, but again most
conspicuously in the Equidae, that the metaconid and entoconid are
reinforced by little cusps which grow up behind them (a, a and l>, b,
Eiitimeyer) ; these may be termed respectively the metastylid and
entostylid, while the pillar arising secondarily in the primitive position
of the paraconid may be termed the parastylid.

The principles upon which this terminology is based are therefore
very simple.

1. The termination -cone is given to the main primary or central
cusps, and -conule to all intermediate cusps.

2. The termination -style is proposed for the peripheral cusps arising
mainly from the cingulum.

3. The termination -loph is applied to the crests.

4. The seven prefixes are based upon the succession and position
of the elements in the primitive evolution of the crown, viz. : proto-,para-,
meta-, hypo-, cnto-, ecto-, mcso-. The prefixes are first applied to the
cones ; then to the styles, according to their proximity to the cones ;
then to the crests, according to the cones which mainly compose them.

5. Homologous and analogous elements in the upper and lower jaws
are given similar terms, but distinguished arbitrarily by the terminal -id.

Upon the opposite page are given the terms formerly employed by
French, German, and English authors for the teeth of the Ungulates
before their common tritubercular origin had been discovered by Cope.
In his ' Enchainements du Monde Animal ' Professor Gaudry, as far back-
as 1878, worked out most clearly the homologies of the molar elements in
the Ungulates from the sexitubercular-quadritubercular stage onwards ;
the valuable earlier studies of Riitimeyer1 are well known. But now that
the ungulate molar has been found to converge to the uuguiculate molar
type, and both are found to contain the same elements, and to spring from

1 Beitrage zur Kenntniss der fossilen Pferde, Berlin, 1863.

i\ r.\<;rL.\TKs

71

W

%

&9

H

OJ

H

X

H
~

W

O
O
K

-
—

X
H

fa

?H

J

-:
O

05
W

H

1

p

s £_

.

^. -^

~<u ">,"-

•o

CO

/. L *

•e

2 |lo

•5"

C-

ft S ft

C -S

ft ft

= — s

—

ft £ s

"3
—

a

^ _ s

0

Terms based
\\}»<\i \tr\i\\\\\\'
homolngius

protocone

paracone
inetacone
protoconule

0 0
V -
— ~

± ^

protocoled
paraconid

metaconid
hypoconid
entoconid

3
o

§-_2 „§

"S ^rr-

•f ">>">:

•g "S w
i ° *

* 2 S

a^-r

« to -^
O flj A

-= £ S

itoconid, hypoc

IXLKY, Fl.OWKIl, LYl'KKKER,

lisli ami American authors.

^-internal, anterior pillar -

3-external, anterior crescent
•o- ,, posterior ,,
^-intermediate tubercle

; c

S

43
«.

O

"5
S

43

O O

i-fxlcrnal cusp -
internal, fifth cusp -

ft
M " r

0 — _

III

— ' X —
g CU .S

- ?

u

,0

Q

r- W

|1 =

13 o

11 "
o S

5 o.2

CU _ CO

1 ' '

CD"

CO
CD

43 Z r
"^ Ll

2 5-^

prominence, pillar
lication of metaconid
,, entoconid

•r Mul iirs. Talon (hypocone).
r ,, Talonid (hypoconid, ei

-> 50

tt

S- CD ~

CD O

•^

s

CU

C '<- -o

^ 7: , ,

~ .2 a

- ft

S5, ^

— ,~

CO

0 43 CU
-«-> i/3 43

— 43
CO CO

CD

- c »

43

CO

j— ' — >y^

-£ 1o ^

a — .^>

-*3 X tfj

~~.

*z. 5^ o

a

cos

0 0

rj

o

o

X = C

£ C '£

C 2 0

Q3

^

~

S ft rt

~^ . ^ _

-

ft

co c« a

rf ^— -^J

-* — ft

-

0 — '—

=

02

° "a ~

rS

• ' '"T?

rfl

i i

1 1 1

1

i - .

i i i

i i i

* 2 5

>

•"t?

_^

O

o

—

o-2

O

7. Z^*

m

•r— a

, i

^-C —

CO

o ,0^

^ £

O s

f-^l

^ ^

Q^

00 C« S

~ *'-O

i; £

t^ t-

r^

^^ "Q

2$ - -

•° "?

CO.&

4i ^"

CD

rW &

' '

1 , .

1

, i ,

i • ,

c5

, . ,

s I'l

._ 13 C.

Q

'— a

"ft

* ^ -r

i-g

C<H

< |-s

as" «

CO ^

1 I *r*

<D ff

i ,

1 1 1

1

s ' >

i r —

• i i

Q (» C5

« £
£ 5

Is s

S

,0^ 3

r^

^ _rj ~CJ

O

-? °

CO

g OQP4
5 ^i.

i -

£"1

, , -£

r r^-S

, ,

, 1 1

1

§ § .2,

^ .£, eg

o . .

. , .

H a>

73 a

S ^

GO

O " »•

o

:B

3

ce 't? 0

i^~ O r«

"^

a o :-

a3

js

, , s:

j

~ K° *

O c6 ^

CO

5 5

•"• o>

>• s

CO

. , ,

. . .

, . ,

. , ,

< jT

0

S P

o ~~~

S |

5^

2 o

05 2 "

CU

U

2 S .2

.2 '2 '

. , .

, , ,

1

^

" O CU

^

5 .2

s o
o o

t. — »-3^

'j~ ~*

oT

-T /

O '^

X

— —

!-* 3) *%

^

)_1 ^a yj

)^i yi

c

o £

I 0

ft

M §

5 r
o

,£ 5

ft t«

x s 6

Sft'

£} QJ

1 1 1

i i i

c

0 -

a '

o

3

CD

CD r-

CO

CU J

CD CU

r-

<CD <CD 'CD

<OJ <QJ

a

sf-s

Q 3

CD

0} "^

•" Ll
•" 0)

33 _g

" co S

'^S

cu o o

&H i-l '

O O

s'3

0 0

^ i->

-5

CD" -

aT ^

J ^

.5

X rH

.s

CU ^

CU .S

g

. ^2 CO

0 -0

^2 CU ,

o -a

, , ,

, , ,

'S )S

155

1-1 ^

— _o

o

£ -—

.2-r ,

1 . ,

, .

_§ r

43

— S

— t?

"5

M

CO o

CO o

JS

•S

, ftg

^ ^

»— tC i

. , ,

, , ,

^ *-

,

1

,

i CO

,

,

ig1^

"f.

tc

— , —

zfl •*

b

-<*-

1 i
O 8

sf

CC

^

»T o

<?*

e

•*.

1

5 3

^ "o

1

§r.

"o

o ^

"o

"g

•r •"

.„ "--

K

^^w

>-N *

**^

^

"^ c -•

||

' ~«

tf

O

n

"^ X

|j O-i

£ 5

V

^

> t. 5

S-gB

a s«

0

!§"

O K
1^ p |

S"

O

.- -•-

J-, fl
P^ g

72 EVOLUTION OF MAMMALIAN MOLAR TEETH

the same Mesozoic source, it is important to unify our methods of
description by adopting a set of terms which refer back to the primitive
form and position in place of those which were based upon the com-
paratively modern form and position.

o.

The above discussion of the homologies of the molar elements of
ungulates was followed in 1890 by a contribution which may be
entitled :

APPLICATION OF THE THEORY OF TRITUBEP.CULY TO THE

PERISSODACTYLA.

(Scott, W. B., and H. F. Osborn, "Preliminary Account of the Fossil Mammals
from the White River and Loup Fork Formations contained in the Museum of Comparative
Zoology. . . . The Perissodactyla," by Henry Fairfield Osborn. — Bull. Mus. Comp. Zool. XX.,
No. 3,1890, pp. 88-91.)

The Horse Molars.

The upper molars of Mesohippus [Fig. 161rfJ clearly show the first step
in the formation of the posterior pillar, pp., which is so conspicuous a
feature in Anchitherium, in the posterior valley. This can also be observed
in a still simpler stage in a specimen of Anckilophus, from the French
Phosphorites. Step by step with the development of this cusp appears
the posterior pillar, p, in the lower molars, behind the entoconid ; this
accessory cusp can be traced back to the teeth of Epiliippus. When it
finally unites with the entoconid, in Hipparion, it forms the posterior twin
cusp (b, b, Elitimeyer), which is analogous to the anterior pair formed by
the union of the metaconid and anterior pillar a (a, a, Eiitimeyer).

Thus the transition from the Mesohippus to the Anchitherium molars
is very gradual, as shown in the accompanying figures. By tracing back
the rise of the eleven elements which compose the upper Equus molar, we
find that six belong to the primitive sexitubercular bimodont crown. Two
elements of the ectoloph, the anterior pillar and median pillar, rise from
the simple primitive basal cingulum of the Hj/racotherium molar ; the
same mode of development, we have just seen, is true of the posterior
pillar. The eleventh element, the fold of the postero-external angle of
the crown, p, is not prominent until we reach Equus. The term
" posterior pillar " is taken from Lydekker ; the other terms, " median "
and " anterior," are applied to parts which have an analogous origin from
the basal cingulum. The remaining coronal cusps are readily identified
with their homologues in the primitive tritubercular molar.

TKITI:I:KI:< TLY ix rxci'LATKs

Tin' Rhinoceros Molar.

The peculiarities of the molars of A-phclop* [cf. Fig. 17-"., p. 181] will
be made more clear by a few observations upon the molars of the rhin-
oceroses in general. The three main crests of the lophodont crown
may now be distinguished in part by terms which express their
homologies with the elements of the sexitubercular superior and
quadritubercular inferior molars of the primitive ungulate, PTienacoduus.
In the upper molars, the outer crest is formed by the union of the
primitive paracone and metacone, to which is joined the anterior pillar
(see Mesohippus, p. 175); it may be called the cctoloph. As the anterior
crest is formed by the union of the protocone, protoconule, and paracone,
it may be termed the protolopli. The posterior crest, which unites the
primitive metacone, the metaconule, and the hypocone, may be termed the
metaloph. The outer surface of the ectoloph in the primitive molar of the
rhinoceros is marked by three vertical ridges corresponding to its three
primitive component elements, rn>', pa, ap [pas] ; one or all of these disappear
in the flattening of the surface. It will be observed that nothing corre-
sponding to the ' median pillar ' of the superior molar of the horse is
developed. In the lower molars (the paraconid disappearing), the union
of the metaconid and protocouid forms the anterior crest or metalophid,
while the hypoconid and entoconid unite to form the hypol<>i>lii<l.

The secondary enamel folds, which are developed from the three crests,
bear a most interesting analogy to those observed in the horse series,
beginning with Protohipjw.x [Euliippv*, Fig. 166]; they are outgrowths of the
same regions of the crown and subserve the same purpose. They are, more-
over, of like value in phylogeny. The useful descriptive terms introduced
by Busk, Flower, and Lydekker, should be adopted in part.1 These
secondary elements consist, first, of three folds projecting into the
median valley, one from the ectoloph the crista ; one from the proto-
loph, the crochet ; * one from the metaloph, the antecrochetJ Secondly,
the ectoloph unites with the posterior cingulum and metaloph. Thus
the anterior and posterior valleys may be cut off by the union of these
folds into from one to three ' fossettes,' precisely analogous to the ' lakes '
in the horse molar, except that they are not filled with cement.

1 The terms ' protoloph ' and ' metaloph ' are, however, substituted for ' anterior collis :
and 'posterior collis' of Lydekker. The term 'anterior pillar '=:' tirst costa,' and
' paracone ' = ' second costa.' The mode of evolution of the ' pillar ' must have been similar
to that in the horses, where Lydekker has proposed this term for the ' posterior pillar.' It
is very appropriate, because the pillars in their earliest development can be shown to rise
independently from the cingulum (see Mesohippus, p. 175), and not as folds of the main
elements of the crown, as we should infer from their fully developed stage.

:: [Should have read antecrochet.—T&D.]
I [Should have read crochet. — ED.]

CHAPTER V.

SECOND OUTLINE (1897) OF TRITUBERCULAR EVOLUTION IN
MAMMALIA. WITH DISCUSSION OF CRITICISMS.

[Reprinted from an article entitled " Trituberculy : A Review dedicated to the late
Professor Cope,'' The American Naturalist, December 1897, pp. 993-1016.]

THE morphology of the crowns of the mammalian teeth has sprung up
practically as a new branch of study since Edward D. Cope and other
palaeontologists have demonstrated the unity of derivation of all the com-
plex forms from the tritubercular type. The older works and ideas of
Cuvier, Owen, Huxley and others are of comparatively little service now,
for they treat the teeth of each order of mammals as of so many distinct
types, whereas they must now be treated as modifications of one type.
This new odontography of the mammalia may be dated from the time
when it was recognized that the crowns of the teeth of the Unguiculata
and Ungulata, in the comprehensive Linnsean sense, are based upon a
common type and are composed of homologous elements of similar origin,
as developed by Cope, Osborn, Scott, Schlosser and others. It dates also
from the new embryology of the teeth as studied by Leche, Kiikenthal,
Taeker, Rose, Woodward [M. R] and others, with the revelations as to
primitive form, number, and milk succession.

But to fully establish the morphological branch in its new era we
must first demonstrate the theory of a tritubercular archetype. This has
been opposed in one form or other by nearly all English morphologists,
namely : Lankester, Forsyth-Major, Newton-Parker, M. R Woodward, E.
S. Goodrich, Marion Tims. It has been accepted only by Flower and
Lydekker. In Germany it has been accepted by v. Zittel, Schlosser and
Riitimeyer ; Schlosser, especially, has made important contributions to the
evidence. The theory is accepted somewhat reservedly by the embryo-
logists Rose, Leche, Taeker and others, who have attacked rather the
homologies of the upper and lower cusps than the theory itself. In
France it appears to have made little headway. In America, Scott,
Allen, Wortman, Earle and many others are working upon the trituber-
cular theory and have made important additions to it. It is difficult

SKCOXl) OUTLINK OF TKITUHKItf T I.Y 75

for the writer to take the " primitive polybuny " hypothesis seriously,
although it is advocated more or less positively by such able inorpho-
logists'" as For syth- Major, f Lankester, Goodrich^ and Parker. The fact
that the Multituberculates and Monotrenies and certain Rodents
exhibiting this type are primitive is no evidence that the polybunic
type itself is primitive. We know nothing of the history of the
degenerate Monotreme teeth, but we know that the further we go
back among the ancestors of the Multituberculates and Eoclents the
less " polybunic " and more tritubercular they appear.

This demonstration once made, as a matter of convenience in
thought and description, we must revise the old systems of nomen-
clature which were based upon secondary forms rather than upon
primary homologies, and which, as a rule, differ in every type of
mammals and among odontologists of every land and establish a new
odontography or descriptive method. Finally, we must trace out all
the lines of divergence in both forms and determine the principles
which guide them. The importance of a uniform nomenclature is seen
at once in the accompanying table of terms used among the rhinoceroses
and horses alone. It could not have been anticipated that the diverse
molars of the horse and of the rhinoceros, for example, would be limited
in their variations, in a late geological period, by their unity of origin in
an extremely early^ geological period. Yet such is undoubtedly the case.
Compare the accompanying figures of Merychippus (Fig. 162) and of
Acc/'xf/nrium (Figs. 175, 176). Imagine that you see the simple
bunodont molar of such a form as Owen's Hyracotherium rt'lpiceps
[l&porinum] (Fig. 159), underlying these diverse crests and crescents.
Consult Taeker's Zxr Kenntiri^ ff/r Odontogenese lei Unt/n/tt/cn and you
will find that this sexitubercular archetype is not imaginary, but is a
constantly recurring fact of embryonic development — all the crests and
crescents being preceded in the embryo by simple cones. Then compare
carefully the variations in the two teeth as follows : The two " cement
lakes " of Merychippus with the two " fossettes " of Aceratherium, enclosed
in the former by crescentic spurs, and in the latter by the " antecrochet '
and " crochet " ; the posterior " lake " and " fossette " similarly enclosed
by an upgrowth of the posterior basal cingulum. Can any one question
the homologies between these secondary adaptations to a diet of grasses
when it is seen that they spring from the same primary cusp centres ?
In the lower Eocene the sexitubercular prototype passes directly back
into the tritubercular archetype. So throughout the whole mammalian
scale not only ungulates, but primates, carnivores, insectivores, rodents are

*[This list should have included especially Dr. Florentine Ameghino (see pp. 201-204 of
this work).— H.F.O.]' t [See p. '205.]

I [See p. 206.] §[That is, relatively early, i.e.. the Cretaceous period. — En.]

CD

_*

CC

rM

tf

*

<

ij

^1

0=0

JS-

cS

fes

,— CC'

^ ex:

S ^i

•/.

(S

>^

o

CD

HH

•s.

HH

3

K

-f.

»

<

_;

O

MS

i sj;-
•s » g

33 -H >^

3 S <U

0 w

p& H

f^:3 g

-^ S
^(jj 0

J

CM

a

« a1-

o

C O ' — '

S5
P

S ^ 'TC

l^s*

53 S csd

O

» CJ

.23oo

a5

i'>Sc

pa
<

o-Sd.

, i

o

^•3 ^,

3

i — CD ^

« K
O C

^ F*

iall
"3os

S

o
o

M§

_o^^^-, S

z

1

"o co to j2

•Illl

t4

CD

H

OLARS.

§g
hB

o
fe

I— (

P3

^3 w ^^' — '

» .a o c

is

ior pillar
ior crescen
ior cresce
ior pillar

Ant
Ant
Pos
Pos

Vo

Nac

eile

p interne clu pr. lub
e externe, pr. lobe
c externe, sec. lobe
e interne, sec. lobe

Dentic
Dentic
Dentic
Dentic

| W ^

^ «" ^
•*-. s- _ -T5

^P°P"cS

pis

I

ence

Anterior ridge
Middle' ridge
Posterior ridge
Posterior prom

Lake

Cem

'Z ru

- 58

O -t>
o vi
oj o
ZCO

Pericones, Randgipfeln

te externe Aussenwa
crete ou lohe anterieure - Vorjoch
ine crete ou lobe posterieure Nachjoch

cu

<cj — '

CD

5

t- -r^

O CD O

CD 'S

CD

p ~* O

S "CD

*S- ^

11 g „

3 -f=

^ 71

3 2

8 S

— 2

~ "o

CD -3

^ ^

ca o

M-s;
S'g

^n -3

"o c«
OiJ

Vallon o

Fossette

SSl

Hi

ud-3

!

+3 i

Q^

C

-^ O

1

S^ S ca

^ -r; nj -*j

§ o^.S

c i-i — S-

s

1

to«)o

, — .

S

flj CJ

&

2

5

ft

o
Q

•~ o3

^

,_

II

tiD bO

3

•IM

1 1

.

c s

P=

Cfi

a

§3

2 3

II

J

"o
o

^ls

S3

S S

S o

y.

O

0

.+3
CC

S

Posterior

.2 .2 c

o CD ;s

Anterior
Posteriol

.2 o

^ 1

Pw «J

eu

po

F°

Fo

Bo

ngulum,

CS
ft

o
o

•- "2

.2 S

_S M

VI

O
Pi

H-t

<! ^

o 3

at u CD "
cc ••-? _^ ^

cfi ,_ "5 <«
X! ^ _S a>

' '

* m

•

, , ,

£ s

SB S "^ « ^

C 2

W S
W <!

^S
§^

^S
o f-

^ -i -*3'

"2 ^ '" -~ ^
"1 g 0-^"3

g&lll

CD

£ ^ 2 S
° S 5 °
| 8 § §

j" Deuteroco
1 Tritocone
^Tetartocoi

<B « CD,2i
5-^5-g

Illl
S OJ CD >>

^SW^

III

-g "o -g

i-' S-4 i»

% M

^'o M

i =§

"o a-

— CO ^

•r -fJ
r^ ^j

.13 °

l§

5 ^^

a ?

O ^H

'SSo

, ^

O ""^ (V)

So

PH

S

jgS

CD

w

Praesinus
Postsinus

CD

^3

o

-t> O

^ o *

1 -s.1
o ^5

CD CD

gu

bo

3

SECOND OUTLINE OF Tltl'IT I'.KIM ULY 77

found playing similar variations upon the primitive tritubercular type.
There are surprisingly few distinct types, but an almost unlimited number
of sub-types, or variations of form. As we descend among the older rocks
and the various series begin to converge, it becomes increasingly difficult
to distinguish the different orders by their teeth alone. Thus it came
about that all the Eocene monkeys were at first referred to the ungulates,
or to transition groups, as expressed in M. Filhol's composite term Pacii u-
Umuriens.

Tritubercular Homologie*.

Embryological Evidence. — The progress which has been made in the
embryology of the teeth is largely in the matter of the succession of
double series, as indicated by vestiges of earlier and later sets of teeth,
the so-called milk and permanent sets. Embryogenesis, however, has also
led to a very minute study of the order of succession of the cones of the
molar teeth, and without entering into the matter in detail, it may be
briefly stated that all authors are unanimous in describing the cones of
the lower molar teeth in different groups as developing in the same order
in which they are supposed to have arisen in the past, according to the
tritubercular theory, namely : Protoconid, Paraconid, Metaconid, Hypo-
conid. In the upper teeth, on the other hand, embryogenesis has been
found to contradict the conclusions reached by the tritubercular theory of
palingenesis, for all authors have agreed that the order is Paracone, Metacone,
Protocone, instead of Protocone, Paracone, Metacone. When these facts were
first brought out by Taeker, Eose and others, the writer, with undiminished
confidence in the force of paheontological evidence, advanced as an explana-
tion the fact that the protocone had become secondarily reduced in the
upper molars, and that the embryogeny no longer recapitulated the order
of evolution. This explanation has received a measure of support in the
latest researches by Woodward [M. F.], in which it is shown that in those
Insectivora in which the protocone is still the most prominent cusp of the
x/il>cn'or molars, this cusp also appears first in embryogeny, the paracone and
metacone following. Woodward points out that this is not the case in
other Insectivora, for they agree with the Primates, Ungulates and other
types which have been carefully investigated, in the late appearance of
the protocone. Woodward infers from these conflicting facts that there
were two modes of cusp evolution within the order Insectivora, one in
which the protocone appeared first, and another in which the protocone
appeared third or last. Such a double genesis seems to the writer highly
improbable.*

It is, however, certainly important, as Woodward and many others
have observed, to strengthen the palasontological evidence for the trituber-

*[See, however, the opposing views on pages 123- 126, 227.]

78 INVOLUTION OF MAMMALIAN MOLAR TEETH

cular theory. The writer has recently made strenuous efforts to secure
additional evidence, which have not thus far been successful. In the
meantime too great emphasis cannot be laid upon the fact that all the
existing palcKontological evidence points in tl« xtt»ir <li ruction, namely, to the
presence of the chief cone upon the inner side of the upper molars, and
upon the outer side of the lower molars. An important oversight on the
part of those who are still unconvinced of the tritubercular theory, is the
necessity of a mechanical adaptation of the upper to the lower teeth
in every stage of development, which is perfectly met by the tritubercular
theory.* Given the universally acknowledged trigonid or triangular
arrangement of cusps in the lower teeth, no mechanical relations can be
imagined in an upper molar crown which originated with the external
cusps, paracone and metacone.

If the main object of palseontological research is to trace back various
lines of descent as far as possible, the very unity of primitive type makes
this apparently more difficult than before, but not really so. We were
working before upon a false basis, or no basis at all ; we can now advance
upon the certain basis of primitive form and the one requisite of progress
is to employ much more exact methods of description and analysis.

The Three Primary Forms.

So far as the molar teeth were concerned, there were, to our present
knowledge, but three great primary forms, which succeeded each other as
stages and also persisted. From one or other of these all the known
recent or fossil mammalian teeth have diverged, including probably
the Multituberculates. These types are illustrated in the accompanying
cut. First, the liaplodont crown, which links the mammals with the
reptiles ; second, the triconodont crown which was predominant in the
Lower Jurassic period ; third, the tritnl»'rcnlar crown which appeared in
the [Upper Jurassic or] Lower Cretaceous T and has been by far the
most productive. The transitions between these great types are found
among the Mesozoic mammalia and have already been worked out
with considerable care.t

From each of these great primary stages it would at first appear that
some of the mammalia directly derived their dental type, for both the
" haplodont " and " triconodont " crowns are seen to-day among the
Cetacea. Yet there is ground for uncertainty here, for as the progressive

* [See especially Cope, "On the Mechanical Causes of the Development of the Hard
Parts of the Mammalia," Jour. Morph., Vol. III., Sept. 1889, pp. 226-274. In opposition
see pp. 61*, 68*, 82*, Figs. 2U8, 209, of this volume.— ED.]

1 It now appears advisable that the so-called Como (Atlantosaurns) Beds of North
America and the Purbeck Beds of England should be placed in the base of the
Cretaceous instead of in the Upper Jurassic as formerly. [See p. 22*.]

t [See, however, p. 222.]

SKC(tXI) orTUXK OF TIIITrr.KIU UA 7'.*

stages are " haplodont," " triconodout," " tritubercular," so the retrogressive
stages reverse this order, passing from "tritubercular" back to " tricono-
dont " then into " haplodont." Another view therefore is that such primary
forms have been secondarily acquired. The apparently " triconodont '
lower molar of Thylaci/nux is, for example, an indirect retrogression from
a tritubercular ancestral form. Again among the aquatic carnivora, in
the series of molars of the Seals, the eared Seals and the Walruses, we see
the backward stages from the " triconodont " to the " haplodont " ; and it
is therefore probable that the " tritubercular " was the form of molar
possessed by the Pinnipedia when they diverged from the Fissipedia.
There is considerable evidence that a similar retrogression has simplified
the molar crowns of modern Edentates, for it is now certain that at least
the Gravigrada were descended from tritubercular ancestors, the Gano-
donta [Ta?niodonta], Again, among the Cetacea, all their oldest allies,
such as Zeuglodon, are triconodont, not haplodont. With both these
groups, therefore, there are the possibilities of direct or of retrogressive
origin of the " triconodont " molar.

This uncertainty hardly extends to the " triconodont " stage, which is
typically shown in the Lower Jurassic Ampliilestcs, Phascolof/n r//<m and
the later Trin<,i»<!n,i. It is a very significant fact that this type dies out
in the Upper Jurassic. It is true we find many more recent " tri-
conodont " teeth, the lower molar of Mcsonyx for example, which are
positively known to be of tritubercular origin. Richard Owen compared
the lower molars of Thylaciims with those of Triconodon, but we have
found that what appeared to him to be similar cusps are not really
homologous. Thus while it is possible that the ancestors of some of the
modern haplodent and triconodont mammals never reached the tri-
tubercular stage, it is by no means a settled fact. On the other hand,
excepting the isolated group of Multituberculates and the single genus
Dic/'iirt/iiiiifn/i Marsh, the- molars of crcri/ l-nou'n foxsil mamma/* from /In-
close of the Ln/'-rr Cretaceous until tin- clow »f Ilir E<H'<'H<- [>< '/•/<></ bear ///<•
trituh rcular *f«n>/>.

This would appear to support the generalization that all mammals
passed through the third primary or tritubercular stage, yet it must
be borne in mind that all our evidence is derived from inhabitants
of fresh water basins,t and that the persistent haplodont and triconodont
types may have been living contemporaneously in the seas.

But the Multituberculates and Monotremes, were they tritubercular
in origin ? The teeth of Ornithorhynchus are so degenerate and irregular

* [This statement applies chiefly to the orders Insectivora, Carnivora, Primates,
Ungulata, since the trituberculate derivation of the Monotremata, Multituberculata,
Edentata, Rodentia, Cetacea, remains to be proven. — ED.]

t [Or at least from epicontinental as opposed to marine deposits. — ED.]

80 EVOLUTION OF MAMMALIAN MOLAR TEETH

that many features of primitive form may be lost ; they may quite
as readily be interpreted as tritubercular as multitubercular, especially in
the embryonic stage as described by Poulton.

It is not difficult however to establish the principle that a true multi-
tubercular tooth may spring from a tritubercular tooth. As pointed out
elsewhere, my friend, Prof. J. A. Allen, directed my attention to the
" multituberculate " rodents. A comparison of Mas, Dipodomys and
Perognathus beautifully illustrates the stages between " trituberculy "
and " multituberculy " in living types. The three rows containing twelve
tubercles in the latter genus are derived respectively from the " external,"
" intermediate " and " internal " cusps of a sexitubercular bunodont type
similar to the Hyracothcrium molar on a small scale. The additional
cusps are successively added to each row. Thus the upper molar of
Perog natlius is closely analogous to that of the Mesozoic Multituberculata,
especially to such a type as Tritylodon. Passing also from the higher
Multituberculata to the lower and more ancient, we find fewer and fewer
cusps until we reach a " paucitubercular " parent form in the Upper
Triassic Microlcstes. Microlestes itself was not tritubercular ; it had
a basin-shaped crown surrounded by irregular tubercles ; this basin,
however, was not dissimilar to that in molars of the Eocene rodent
Plesiarctomys which is obviously of tritubercular origin.*

This evidence has been recently reinforced in a most striking manner
by the discoveries of Professor Seeley in the Karoo Beds of South Africa,
from which two principal conclusions may be derived : First, that Tritylodon,
formerly placed with the mammalia, contains a large number of reptilian
characters. Since the fossil is closely related on the other hand to
the remaining Multituberculata, it appears possible that we have in the
Gomplwdontia the group from which the Multituberculates sprang.
A study of the dentition of other Theriodonts in the Karoo Beds shows
that while Tritylodon and Trirachodon are typically Multituberculates,
others, such as Diademodon t have a trituberculate pattern, exactly such a
pattern as we find in certain Lower Eocene mammals. Altogether there
is certainly increasing support for the writer's hypothesis, that the
multituberculate tooth is of tritubercular origin.

The Early Stages of Sexituberculy.

The Tr if/on. Kespect for Cope's priority should not prevent our
ultimately adopting the late Professor Kutimeyer's term trigonodont for

* [The derivation of some multituberculate types from trituberculate types does Dot
prove that all multituberculate types have been derived from trituberculate types, and
reasons are presented on p. 105*" for thinking that the ancestral multituberculate molar
as represented in Mirrolestfis was not derived from a typical tuberculo-sectorial lower
molar. — ED.]

tfSee p. 92.]

SKCON'I) ori'LINK OF TKlTrilKIK'l'LY 81

the third stage, retaining- the term " tritubercular " as descriptive of
the whole transformation, and as peculiarly appropriate to certain types
of teeth, such as the superior molars of the lemurs. " Trigonodont '
is most appropriate because the first step in molar morphology is to
identify the "primitive triangle," and the term "tubercular" hardly
applies to a lofty pointed cutting crown. Our studies among the Mesozoic
mammals have left no doubt that the upper and lower triangles, or
" trigon " and " trigonid," were derived from the reptilian protocone by
the addition of lateral cusps. The mechanical perfection of this type
consisted in the fact that the lateral cusps were developed upon or
shifted to the outer side in the upper molars, and to the inner side
in the lower molars, thus producing an interlocking " shear." The
" trigon " was essentially a cutting apparatus, so perfect that many
mammals retained it without further evolution. Thus Chrysoddoris, the
little Insectivore of the Cape, presents a fine example of this type,
persistent in its molars* (Fig 42*).

The Talon. But in a great majority of tritnberculates the " talon "
was added as a crushing apparatus. It invariably appeared first in the
lower molars (where we may distinguish it as the " talonid ") and pressed
into the basin of the superior " trigon." At first it was a mere spur
(hypocone) as in Amphitherium or in the existing Calcochloris (allied to
Chrysochloris), but between the Jurassic and Upper Cretaceous periods the
talonid widened into a basin-like shelf supporting an outer cusp, the
•' hypoconid " ; an intermediate cusp, the " hypoconulid," and an inner
cusp, the " entoconid." Thus we find in the majority of the Upper
Cretaceous (Laramie) and Puerco or lowest Eocene mammals that the
lower molars bear six cusps ; the above-mentioned three on the talonid
and three on the trigonid (protoconid, paraconid, metaconid). With these
six cusps the equipment of the lower molar was complete, and it was
ready for transformation into the molar of a primate, ungulate or
carnivore, as the case might be.

But why notice such a detail as the posterior intermediate cusp or
hypoconulid ? Because, to give only two reasons, this cusp plays an
important role in the ungulates; it is invariably present, t except perhaps
in the Coryphodons, and forms the third lobe of the last lower molar,
which is thus proved to be a primitive character ; again, it is found
throughout all the Primates, and although seldom availed of, this cusp
constitutes an important and distinctive character as between the different
races of man. Its extreme antiquity is appreciated by few anthropologists,
and at the present time it is degenerating. (See Fig. 40, Nos. 11, 12.)

* [It should be borne in mind, however, that some authors (e.'/. Fovsyth Major) hold
that the heelless condition of the Chn/ii,,-/,/,,, <x molars is secondary. See also pp. 124,
•_'•_>.-> below.— ED.]

t[That is, in the more generalized forms. — Kn.]

F

EVOLUTION OF MAMMALIAN MOLAR TEETH

While these changes were taking place, the upper molars remained
comparatively stationary in the persistence of the simple trigon, up to the
close of the Cretaceous period, the main change being a depression of the
level of the trigon. All three cusps in some groups were depressed from
the high secodont to the low bunodont level. In the majority of the
carnivorous types we find that only the protocone was depressed and that
the pair of outer cusps, paracone and metacone, persisted in their high
primitive level ; the crown being thus prepared for the transformation
into the true " sectorial." But in the omnivorous and herbivorous types,
all three cusps are depressed and the upper molars always increased their
crushing area by the addition of a heel or " talon," exactly analogous to
that previously developed upon the lower molars.* As is well known, this
" hypocone " is an upgrowth from the cingulum and its typical mode of
development is well shown in the Primates (Figs. 128-132). While this
was going on the trigon was also supplementing its bunodont equipment
by the addition of the little intermediate cusps "protoconule" and "meta-
conule."t These always appeared where the " talonid " abuts against the
" trigon." Thus, finally, the upper molar, like the lower, was provided
with six cusps and both were ready to diverge into any ungulate form.

All these foregoing stages persist and may be readily studied and
verified among some of the living marsupials, insectivores, lemurs and
monkeys, and can be seen in any well-equipped osteological museum
almost as well as among the fossil series.

The Nomenclature of the Molar Cusps and Crests.

The system proposed by the writer some years ago has now been
adopted by many of the American, English and German writers who are
studying the fossil series. It is based upon simple principles :

1. The termination "-cone" is employed for all the primary central
cusps derived from the crown of the tooth, while the diminutive -conulc is
employed for the smaller " intermediates " or cuspules.

2. All peripheral cusps or elements developed mainly from the cingu-
lum or external borders of the crown are distinguished as -styles (" pillar "
or " buttress "). The only exception is the " hypocoue," which, while
arising from the cingulum, soon takes its place upon the crown.

* [The hypocone of the upper molars is analogous in position to the hypoconid of the
lower molars, but in function it is more analogous to the paraconid which it replaces, since
it fits into the space between the entoconid of one and the metaconid of the following lower
molar. A closer functional analogy is with the protocone of the upper molars, since the
protocone and hypocone of one upper molar fit into the talonid (hypoconid) and trigonid
of two successive lower molars. — Eu.]

t [In many, if not all, cases the protoconule and metaconule were developed long before
the hypocone. Compare the Jurassic and Cretaceous Trituberculates (pp. 96, 218, 220),
Pomtolambda (Fig. 140), and certain insectivores (p. 128).— ED.]

SKCOX1) OUTLINE OF TRITUIJKRCULY 83

3. The crests, transverse and longitudinal, are always composed of
two or more cusps and styles, and are distinguished by the termination
-lopli.

4. The prefixes "proto-" "para-" •' meta-," "hypo-" " ento-," etc.,
refer back to the primitive position or order of development in the tricon-
odont and tritubercular stages.

5. The suffix -id is employed arbitrarily to distinguish the elements of
the lower molars from those of the upper.

The use of the terms " trigon " and " talon " for the cutting and crush-
ing regions of the crown, respectively, is especially advantageous among
the upper Mesozoic and lower Cainozoic mammals, where it is necessary
to refer constantly to the relations of the upper and lower crowns in
apposition, as in the evolution of the sectorial and lophodout types. As
to the form of the cusps, we pass from simple pointed cusps to three well-
known modes of modification to which the adjective " bunoid," " lophoid,"
and " selenoid " may be applied. A combination of these terms gives us a
permanent system of distinguishing the complex forms of ungulate molars
from each other, by referring first to the form of the protocone ; second, to
that of the outer paracone and metacone. Thus in Palceosi/ops, as the
protocone is bunoid and the outer cusps are selenoid, the crown may be
distinguished as " buno-selenodont." In Pcdceotherium the protocone is
"lophoid," and it may be described as " lopho-selenodont." Rhinoceros is
truly " lophodont," since all its six cusps are " lophoid." These are pre-
ferable to the terms " tapirodont," " symborodont," " bathmodont," "loxolo-
phodont," etc., proposed by Cope, because the latter are associated with
generic types.

The Evolution of the Ungulate Molar.

The fact of derivation of all ungulate molars (excepting in the Ambly-
poda) from sexitubercular upper and lower crowns, leads us to look sharply
for traces of these six tubercles [as modified] from the primitive plan of
Euprotogonia. These six cusps are almost invariably found in the upper
molars of both perissodactyls and artiodactyls up to the middle of the
Eocene period, as typified in Hyracotherium and Homacodon or Dicholune.
In the lower molar the trigon loses the " paraconid " and the talon loses
the " hypoconulid," the latter persisting only in the last molar as the
" third lobe." This loss was accompanied by the complete transformation
of the lower molars from the " secodont " to the comparative " bunodont "
type, as effected in the lowering of the " trigonid " to the level of the
" talonid." This is exemplified in the steps between the first and third
molars of the creodont genus Miacis (Fig. 38, No. 6). In a side view of
all early ungulate molars, such as Hyracotherium, we see that the " trigonid "
is still the highest portion of the crown. In the ungulates, unlike the

84 EVOLUTION OF MAMMALIAN MOLAR TEETH

carnivores, all three molars were affected simultaneously. An exactly
similar levelling process can now be observed in a comparative series of
recent Lemurs and Monkeys. To summarize the five steps toward the
establishment of the ungulate primitive type : the addition of the lower
talonid, the lowering of the cusps of the upper trigon, the addition of the
upper talon and simultaneous lowering of the lower trigonid, the loss of
the paraconid and hypoconulid. By these changes the cutting was trans-
formed into the crushing type. The development of the talon necessitated
the loss of the " paraconid," for they both occupy the same space when
the jaws are closed ; the stages of this gain to the upper molar and loss
to the lower are well shown in the species of JZuprotogonia.

All these changes belonged to the constructive period and took place
presumably before the great divergence of the ungulate orders began ; or
it may have been partly due to parallelism or homoplasy, because we find
that the molars of Triyonolestes, the earliest known artiodactyl, are tritu-
bercular.* Some groups, such as those to which Coryphodon, Uintatherium
and Periptyclms belong, built up their whole molar structure upon the tri-
tubercular or trigonal basis.

From this point onward dated the period of " modernization." An
important legacy of the old triangular form was the obliqur arrniit/emnit of
the outer and inner cusps parallel with the sides of the primitive triangles.
Thus all the primitive crests developed upon these cusps were oblique and
not directly transverse. The main features of modernization upon which
we must now closely direct attention are :

1. The addition of one or more peripheral cusps or "styles" as up-
growths from the cingulum. These reached their most extreme develop-
ment in the EquicUe. (See Fig. 49.)

2. The persistence or degeneration of the cingulum at certain points,
for all primitive molars are completely invested by a broad cingulum.''"

3. The modelling of the cusps into the " bunoid," " lophoid " or
" selenoid " form.

4. The metatrophic or unequal growth of the cusps, especially as
affecting the external pair, protocone and metacone, in the upper molars.

5. The shifting of the cusps from their primitive position upon the
crowns.

6. The shifting point of union of these transverse crests with the
external crest.

The differential features of the development of ungulate molars all
group around these six heads. If we were examining an isolated molar
tooth from the lower Eocene, the first step would be to locate its primary
cusps and then note its divergence as tested by the above differentia.
We would then be in a position to make a conjecture as to the series in

* [See pages 171, 172. — ED.] t [More or less. — ED.]

SKCiiM) OUTLINK OF TUITriJKKCl'I.V 85

which this molar belonged — as no two series are modified similarly in
all these respects. Yet the prevailing method among many pal;i-<ui-
tologists is to pass lightly over most of the differentia and, for example,
group widely divergent forms under the Lophiodontidse as if in the con-
stitution of these dense enamelled tissues nature could lightly pass from
one to another.

A few words now upon the secondary "styles."' Their function is
evidently to increase and elaborate the crushing surface of the crown. In
Pht ii'irot/i/x the first to appear is the " mesostyle " between the paracone
and metacone, but this genus was on a side line of the Condylarthra. In
all true perissodactyls and artiodactyls, the first peripheral cusp to appear
is the antero-external buttress of the upper molars, which we call the
" parastyle," since it adjoins the paracone. The " mesostyle " appears
later, and only in those ungulates in which the paracone and metacone
are moulded into crescents. Thus the lower Eocene Hyracotherium does
not exhibit this cusp, but it appears as a distinctive feature of the middle
and upper Eocene Pachynolophus (Orohippus). The mesostyle was strongly
developed in all the selenodont, buno-selenodont and lopho-selenodont
types, such as the Artiodactyla and J/o//.srn/A(r/>?tt,, Chalicotherium,
Palceosyops, the palseotheres and horses. Look at an upper molar of
Merych-ippus and see what an important role these styles play (Fig. 162).
First, we observe the " parastyle " and " mesostyle," next most important
is the " hypostyle," which develops near the hypocone upon the posterior
cingulum of Mesohippus and Anchitherium and finally completes the
border of the ''anterior fossette " or cement lake. The horse molar, by
the way, furnishes the best illustration of the value of tracing back the
various portions of the crown to their birth-place in the primitive crown
of Hurocothi'iitim. Every turn in this labyrinth of folds is thus made
perfectly clear.1

A corresponding set of styles grows ap on the lower molars, and it is
very easy to locate them with reference to the reciprocal upper set if we
simply keep in mind the fact that throughout the whole course of develop-
ment the elements of each trigonid are placed just in front of those of the
corresponding trigon, that is, the protoconid and uietaconid fit iust in front
of the paracone and protocone, as shown in the diagram (Figs. 37, 396-).
Thus the inferior entostylid is developed near the entoconid, while the
superior hypostyle develops near the hypocone. The first of the inferior

'"' [The para* and metastyles as well as the proto- and metaconules are very ancient
elements of the molar crowns, since they appear in the Upper Jurassic and Upper
Cretaceous trituberculates. The termination "style" as used in this book is applied (1)
to all cusps originally external to the para- and metacones ; (2) to the " protostyle,"
"hypostyle," "entostylid," etc.— ED.]

1 Mr. Lydekker has courteously called attention to the fact that in the earlier study
of this subject the writer misinterpreted the descriptive terms employed by Huxley.

86 EVOLUTION OF MAMMALIAN MOLAR TEETH

styles to develop is the " metastyle," a reduplication of the metacone, the
well known " a-a " of Ptiitimeyer.

In all ungulates in which the " mesostyle " is developed the external
cusps remain of the same size. In the tapirs no " mesostyle " appears,
yet these cusps are symmetrical ; but in the rhinoceroses, which also lack
the mesostyle, the first fact to note is the asymmetrical growth of these
cusps ; the metacone is elongated while the paracone is reduced and
crowded up against the parastyle. This point was observed by Cope in
seeking for a definition of the Ehinocerotidas in 18*75. The rhinocerotine
molar, whether of Hyrachyus, Amynodon or Acemthcrium, has the further
distinction that it is the only type in which a complete ectoloph is formed,
and second, as Cope has already observed, the asymmetry of the external
cusps is emphasized by the flattened metacone and conic paracone.
Figure 175 illustrates also the three projections from the ectoloph,
protoloph and metaloph, namely, the "crista," "antecrochet" and "crochet."
These, with the three " fossettes " formed by them, were noted and named
by Cuvier, and, as shown by Falconer, Flower, Lydekker and others, are of
great specific value.1 We have already seen that Cuvier's term "fossette"
may be substituted for the " cement lakes " in the horse's molar. The
terms formerly adopted, or proposed, by Lydekker,2 after English usage,
and those in German and French usage, have already been given in the
Table.

There is another line of perissodactyls in which the metacone is
flattened but not elongated, and no complete ectoloph is formed. I refer
to the little Wasatch genus Heptodon (which Cope has erroneously placed
in the ancestry of Hyrachyus), also Hdaldes of the Bridger, an undoubted
successor of Heptodon, which Marsh was wrongly led to consider an
ancestor of the Tapirs. The molars, studied by our six differentia, are
found to differ from those of the rhinocerotine Hi/mclti/ns by the incom-
plete ectoloph, also by the shifting inwards of the metacone and con-
sequent shortening of the metaloph. In looking about for molars with
similar differentia, we find those of the true Lopliiodon of Europe,
L, isselensc, for example, stand nearest.

Now, how shall we distinguish the early Tapirs ? First, there is no
mesostyle ; second, the paracone and metacone (as observed by Cope) are
both conic and symmetrical ; third, a feature of great importance, appar-
ently unnoticed hitherto, is that the protoloph and metaloph spring from
the anterior bases of the paracone and metacone, and not from near the
apices of these external cusps as in all molars of rhinocerotine affinity.
We find, as a general law, that where the external cusps are symmetrical

1 As pointed out by Lydekker, the writer mistakenly transposed these terms " crochet "
and "antecrochet" in a former paper, Bull. Mus. Comp. Zool., 1890, p. 81.

2 " Siwalik Rhinocerotida," Pal. Indica.

SKCON'l) OUTLIXK OF TRITUBERCULY 87

as in Palreotheres, Horses and Tapirs, the transverse crests always arise in
front ; where they tend to asymmetry as in Helaletes, Lophiodon and
Rhinoceros, the crests tend to rise from or near the apices.

Enough has been said to make clear the new method of procedure in
the analysis and discrimination of early ungulate molars. Let us apply
this form of statement and description to the aberrant lower Wasatch
genus Meniscotherium as a resume :

Upper Molars, buno-seleiiodont ; paracone, metacone and protoconule
selenoid ; metaconule reduced, lophoid, united with hypocone ; a large
parastyle and mesostyle. Lower Molars, seleno-lophodont ; metaconid re-
duplicated by metastylid. We h'nd that a similar analysis may be given
of Chalicotherium, excepting only " protoconule reduced." It is thus
suggested that Me.n.ixcnfJt<'ri//in may be related to Chalicotherium*

This method may be summarized as follows : Look for traces of
/>/•///> i fire ancestral structure in the form and position of the cusps.
Second, determine the divergent form, position, proportions and relations
of the cusps. Third, determine the secondary cusps, crests and foldings,
their form and relations. Finally, let us turn to a wholly different molar
type and examine the complex and aberrant molars of Cori/pJwdon. Can
we establish any homologies between its elements and those of any of the
ungulates we have been considering ? Fortunately we are partly guided by
the molar of the Puerco genus Pantolambda Cope, which is even older than
the Coryphodons. This is our key to the ancestral or primitive form, and
by its aid Cope has, we think, rightly interpreted the homologies of the
Coryphodon molar elements. We first note that nature has here evolved
a lophodout crown from the tritubercular or trigonal basis, for there is no
distinct talon or hypocone except in the unique form Manteodon.
Pantolambda has no parastyle, t but a prominent mesostyle and a pair of
selenoid external cusps, also a seleuoid protocone with a spur leading
toward a protoconule and suggesting an incipient protoloph. The selenoid
external cusps of this type suggest a comparison with the lopho-selenodont
perissodactyls, and we are able to reach the following result.

In a large series of Corijphodon molars we see first that the protoloph
is formed of the protocone, protoconule and parastyle, exactly as in the
horses. Unlike the horse (Anchitherium), the ectoloph is more or less
detached from the protoloph, but the examination of a large series of
specimens in the American Museum and Cope's collection convince us that
it is composed of the same elements as in AnrJi/fJ/erium, namely, the
paracone, which has almost lost its crescentic form, the mesostyle, which
is much less prominent, and the metacone, which is still crescentic. This

* [See however p. 184, where Chalicotherium is held to be more probably au aberrant
Perissodactyl. — ED. ]

t [This is obviously a lapxus calami, the parastyle being especially prominent. See
Fig. 140. -ED.]

88

EVOLUTION OF MAMMALIAN MOLAR TEETH

FIG. 43. Epitome of the stages leading up to the typical Eutherian Molars, according to the
theory of trituberculy, chiefly illustrating the nomenclature.

SKCOXD OUTLIXK OK TKI'IT I'.KIiCULY Ml

enables us to describe this molar as follows : It is of buno-selenodont
origin and has a complete protoloph and ectoloph, 1ml no ///</<'/«/'//. Its
homologies with the elements of the Anchitherium molar are clearly shown
by a comparison of Fig. 144 and Fig. 1GO. This illustrates again the
necessity of starting upon the trigonal basis instead of upon the basis of
two lobes, as in the work of French paleontologists. In his h'ur/nt.ine-
ments dv Monde An'mml, Prof. Gauclry has admirably worked out the
upper molars of the perissodactyla and artiodactyla from the sexitubercul.-n
stage onwards. He divides the tooth into two lobes, a " premier lobe,"
including our protocone, protoconule and paracone and a " second

FIGURE 43.

No. 1. "Haplodont" or simple conical molars, the hypothetical starting point.

•2. Relations of haplodont upper and lower molars in the Dolphin. (The haplo-
donty of Dolphins may be secondary, see p. 190.)

3. '' Protodont " lower molar of Dromatherium (p. 1!)), showing- main cone and

accessory cuspules.

4. Protodont lower molar of Microconodon (p. 19), accessory cuspules better

developed and incipient heel or talonid.

5. Progressive stages of " triconodont" inferior molars. Three cusps in line with

"cingnlum" out-growths from the base (a) Amphilestes (inner face), (b)
Pkascolotherium (inner face), (<•) Triconodon (inner face), (d) Amphilestes
(outer face) (p. 24).

6. Triconodont inferior molars of a seal (Ogmorhinus). (Secondary, p. 143.)

7. " Tritubercular " inferior molars of Menacodon ; («) external, (b) internal view.

The small cusps are internal to the protoconicl (p. 32).

8. Tritubercular inferior and superior molars of Chrysochloris. In the upper ja\\

the apex of the triangle is internal, in the lower <'.ii>Tn<d (see also No. 10).

9. " Tuberculo-sectorial " inferior molar of Amphitheriurn showing pointed-cusped

"trigonid," and a low posterior, obtuse "talonid" or heel, an out-growth
chiefly from the base of the crown. The heel is " tubercular," functioning
as a pestle or crusher, the trigonid "sectorial," with cutting ridges and
piercing cusps (p. 27).

10. (a-e) Typical tuberculo-sectorial inferior and superior molars. The talonid
bears three cusps "hypoconid" (hyd postero-external), "entocouid" (en'1
postero -internal), " hypoconulid " (hld posterior). Somewhat incorrect (see
Fig. 39).

I I. Inferior molars of Miacis (a primitive Carnivore) showing in in.,, in., depression
of trigonid to level of talonid; all cusps now "bunoid" or f<>w and conical
(p. 84).

12. Inferior molars of Anaptomorphus (primitive Primate) showing (in m:;)

degeneration of paraconid, elongation of hypoconulid (pp. 50, 158).

13. A " quadritubercular " superior molar (Olbodotes, an Eocene Insectivore, or

Rodent) showing three main cusps (pr, pa, me) with a fourth ("hypocone,"
postero-internal) growing up from the cingnlum, changing the crown con-
tour from a triangular to a quadrate outline ; rudimentary "para," " meso,"
and "metastyles" are present on the external edge of the crown (Fig. 104).

14. Quadritubercular superior molar of an Eocene Primate Adapis (Fig. 132).

I"). " Sexitubercular " superior molar of Notharctus (an Eocene Pi'imate) showing
the four main cusps (pr, pa, me, hy) with the addition of two small inter-
mediate conules "pi," "ml," also the well developed external parastylr.
mesostyle, and metastyle (Fig. 77).

90 EVOLUTION OF MAMMALIAN MOLAR TEETH

lobe " including our hypocone, metaconule and metacone. All subsequent
authors in France follow this system, which indeed works well for one
group. But what we need now is a system which will apply not only to
all groups of ungulates, but to unguiculates as well, so that when we
reach the upper Cretaceous borderland between unguiculates and ungulates
we can employ the same set of terms and the same basis of description.

1 can only conclude by expressing the conviction that the tritubercular
theory of Cope rests upon such conclusive evidence* that its universal
adoption as the key to the interpretation of all molar teeth cannot be long
deferred. It is one of the chief anatomical generalizations of the present
century.

[The Bibliography given in the foregoing article is here omitted. See Biblio-
graphy at end of volume. — ED.]

* See pp. 2-26, 227.

CHAPTKE VI.
CHRONOLOGICAL OR GEOLOGICAL SUCCESSION OK MOLAR TYPES.

A REVIEW of the dental types as observed in the successive geological periods
presents us with the ideas of successive stages belonging to these periods,
of the enormously long era of time required for the transformation of the
teeth, and of the very significant fact that only two fundamentally distinct
dental types have thus far been discovered ; first, the (i) tubercular, and
its derivative, the nmltitubercular, the origin of which is still a matter of
hypothesis, and second the great (n) haplodont, protodont, triconodont,
trityhercular sequence, among the manifold offsprings of which we find
again arising secondarily the haplodont, triconodont, tubercular and
mulci tubercular.

1. EEPTILIAX ANCESTORS OF MAMMALS IN THE TRIAS.

It is now generally believed that the Theriodontia, an order of reptiles
found chiefly in South Africa, are, as the name indicates, not far from the

i. a.

FIG. 44. Lateral view of the Skull of (.'/inoijnuthiis rrnti >-onotiis, showing the five simple pre-
rnolars and triconodont (protodont) molars with grooved fangs. After Seeley.

actual ancestors of the mammalia. It is true they exhibit a very large
number of reptilian characters : but mingled with these are features of the

92

EVOLUTION OF MAMMALIAN MOLAR TKKTH

skeleton and characters of the teeth in which they closely resemble
primitive mammals. There is moreover the negative evidence that they
constitute the only known group from which the mammals could have
descended.

The teeth of Theriodonts exhibit four types as follows : The first
or haplodoiit type consists of simple, recurved reptilian -li ke crowns,
implanted with single fangs, as in the genus dSlurosaurus.

Second, of the protortani type, with crowns implanted by partially
divided or very slightly grooved fangs, consisting of a single main cone,
with lateral denticles somewhat irregular in character. The best example
of this type is Cynognatlius (Fig. 44).

The third, or tt<b<'rt:nlfttc, type is entirely different, and probably

Pie. 2.

Fig 11

FIG. 45. Dental and Cninial structures of the Theriodonts. Dimi, ,,«ul<i,i miix/na's, analogous
to a low-crowned tritubercuhite, like Arctoci/on. D. tetn/;/on/is, trituberculate structure more
obscure ; note the general resemblance to the molar crown of certain squirrels. 2). brnchiitiura,
showing the incipient division of the fang. The skull of Triti/loilon is greatly reduced. All
upper teeth except Figs. 4, 5. After Seeley.

belongs to omnivorous or herbivorous animals ;* it is typified by the genus
Gromphognathus (Fig. 45); the molar teeth are covered with irregular
tubercles (Diademodon mastacus), in which the writer has observed some
disposition towards a tritubercular arrangement, or triangle of three larger

*[0r more probably to carrion feeders (Broom).— ED. ]

GEOLOGICAL Sf< '< 'KSSK i.\ OK MOI.Ai; TYl'KS

93

cusps, or the crown is more or less basin-shaped, with an irregular,
raised border (D. br achy tiara) \ the latter teeth are analogous to thn-r
of .!//>/ Wr.sYf.s, of the Upper Triassic or Ehsetic of (iermany (Fig. 48).*
It will Vie observed that the opposite series of grinders converge towards
each other anteriorly ; these teeth are therefore multicuspidate but not,
strictly " multituberculate," because in the true Multituberculates the
opposite tooth rows are parallel and the jaw motion palinal or from in
front backward. These teeth, with irregularly disposed tubercles or
basin-shaped crowns, are most closely analogous to those of certain
squirrels.

The fourth type is the multituberculate, seen in the genus Tritylodon^
(Figs. 46, 48), iu which the opposite series of teeth arc parallel: tin-

ftlHX

Kn;. 4'i. Ti-il"iijilni>.

>,-,ix. anterior portion of the skull iiy,un the left face, two thirds natural
size. After Owen.

grinding teeth are covered with tubercles arranged in regular, parallel
rows, with grooves between, adapted to a fore-and-aft motion of the
jaw.

Certain of these TheriodontsJ present the same dental formula as in the
generalized dentition of the most primitive mammals. Their resemblances
to the Protodonta and Multituberculata in tooth structure are so striking
that since they belong practically to the same geological period the
probability of actual affinity is to be considered.

*[But these are upper not lower teeth; they are elongate transversely (as in upper
molars generally) not antero-posteriorly as in the lower molars of M'« -r<>l, xte*. The lower
molars (Fig. 45, Nos. 4, 5) were single-rooted, round teeth, with a depressed crown bearing
a low transverse median ridge (Broom). — ED.]

t [Broom has recently shown that Tritylodon is more probably a mammal. — Kr>.]

J [Certain Theriodonts (e.g. Trinu-ltodon) had more or less lophodont molars with a
low, irregular medium cross crest. Traces of such a crest are seen also in Ornithorynchus (1)
Kurfodoit, Diplocynodon, Dryolextex, Sriiirn*, Lcpu*. Similarly in the lower molars a
transverse crest connecting the protoconid and metaconid is seen in Diademodon, Trim-
clwiloii, Paurodon, Amblotherium, C<nt</>*, etc. This cross crest may possibly liave some
bearing on the origin of the tritubercular type. — ED.]

94 EVOLUTION OF MAMMALIAN MOLAR TEETH

2. THE TPJASSIC MAMMALS.

The supposed Triassic mammals again include the two grand divisions
of tuberculate or grinding teeth, and of pointed or piercing teeth. They
are distinguished as mammalian by the more or less complete division
of the root into two fangs.

Among the former is the genus Microlestes of the Upper Triassic
or Rhsetic of Germany, known from a single molar (Fig. 48, No. 1).
The so-called Microlestes teeth in the Rhretic of England (Fig. 48,
Nos. 2, 3) are rounded or basin-shaped, with irregular tubercules on the
sides ; they undoubtedly belong to a different species or even genus, as
they are broader and more basin-shaped, resembling in fact the hinder
molars of Plagiaulax.

The other division is believed to be represented by the Protodouta,
which we have already described in detail (pp. 18-21), with large single
cones, more or less regularly developed lateral denticles, and partially
divided fangs. Numerically, these molars have one fang and three cusps,
namely, protocone, paracone, metacone.

3. MAMMALS OF THE JURASSIC.

We now meet with the multituberculates, considerably specialized
into different types.

The other mammals present two great types of teeth : first, the
triconodont with large main cones and lateral denticles both in the upper
and lower molars, as exhibited in Ampliilcstcs (Fig. 5) and Pliascolotlicrinm.
(Fig. 6) ; these animals are believed to be Marsupials. The second
type is the tritubercular, or more strictly, tuberculosectorial, that is,
with tubercular or crushing heel and sectorial trigon, as exhibited in
Amphitherium (Figs. 15, 17). (See pp. 21-30.)

The Upper Jurassic* exhibits the surviving triconodont types, among
which are such teeth as those of Spalacothcrium (Fig. 11), which appear
to exhibit a transition between the triconodont and tritubercular. In the
same geological period the tritubercular types are diversified into two
distinct kinds, both of which have their parallels among later Tnsec-
tivora.t Numerically, these upper molars have three fangs and three
main cusps, namely, protocone, paracone, metacone, also a parastyle,
metaconule ; the lower molars have two fangs and from three to
four cusps, namely, protoconid, paraconid, metaconid and hypoconid.

* [As stated on p. 22, the Purbeck and Atlantosaurus Beds are by some regarded as of
Lower Cretaceous age. — ED.]

t[7. e. one type, represented by Peramus, Amphitherium, resembles the teeth of Microp-
ternodus (Fig. 71) : a second type represented by Stylodon parallels the high piercing type
of Chrysochloris. — ED. ]

GEOLOGICAL SUCCESSION OF .MnLAi: TYI'KS 95

Mr. Gidley's observations upon American Jurassic mammals are given
on pages 219-223.

4. UPPER CRETACEOUS MAMMALS.

(See also pp. 105, 115, 11<>.)

A very long geological interval, representing an enormous period
of time, separates these mammals from those of the Upper Jurassic. With
few exceptions our entire knowledge of the upper Cretaceous mammals is
derived from collections made in the Laramie formation of the Kocky
Mountain region. These collections again embrace the two great dental
types of multituberculates (Fig. 55) and trituberculates. All the multi-
tuberculate types of teeth and jaws are too far specialized to give origin
to any of the modern divisions of mammalia, either placental or marsupial.

The tf if /'!><' /'dilates (Fig. 47; see also pp. 115, 116) are represented
in the American Museum collection by a series of isolated upper and
lower molars, also by a number of fragmentary jaws ; and among the latter
are some in which the angle is apparently inflected, indicating Marsupial
relationship. It is possible that, besides Marsupials, we find here Insec-
tivores, primitive Carnivores, and the ancestors of ancient types of
Ungulates ; but it is obvious that the determination of relationships
from such isolated material is a very difficult and hazardous matter.

Nevertheless, the teeth of trituberculate type are of extraordinary
interest, since they lend strong support to the theory of the unity of origin
of the molar types of the higher mammals from a tritubercular stem, instead
of from a multitubercular, as has been suggested.

The upper molars so far as known, are of the simple, generally low-
crowned tritubercular type as distinguished from the high sharp cusped
crowns of the Jurassic molars. The main triangle of cones, or trigon,
is of symmetrical form. In the majority of specimens, the largest
cone is the main internal one, or protocone (Fig. 47, A, D), while the
external cones, paracone and metacone, are symmetrical, invariably of
smaller diameter, and as a rule less elevated. These proportions sustain
the 'palseontological' theory that the reptilian cone is internal* rather than
the embryological and premolar analogy theories, that the reptilian cone
is antero-external in the true molars as in the prernolars. The outer wall
of the crown is reinforced by a cingulum, on which secondary spurs or
cuspules are developed, sometimes of very large size, more or less homo-
logous with the 'styles' of the tertiary types of molars (e.g. Fig. 68).

Another feature recently discovered among Jurassic mammal teeth is
the presence of intermediate tubercles or conules on the trigon. A very

* [This may be a non sequifnr, because the great elevation and size of the paraconid in
HI, Al, Fig. 47, need not be interpreted as indicating its relative antiquity. — ED.]

96

KVOLUTION OF .MAMMALIAN MOLAR TEETH

/< " in

i y- \

/'" >' s ,..

p

r '«'

hi

talcmd

iff-. 7//

— Al

II

me* triaomd

Fii.. 47. Upper Cretiiceous (Zo/vniuV) Trituborculates. [All figuies three times natural size.]

ABBREVIATIONS.— /);-. protocone ; v>a. paracone ; me. inetacone ; pi. protoconule; ml. meta-
conule ; j>». parastyle ; nils, metastyle ; :n,i. entoconid; hi. hypoconulid.

GEOLOGICAL SUCCESSION OF MOLAR TYPES 97

primitive feature is the entire absence of an internal cingulum and con-
sequent absence of any cusp corresponding with the hypocone ; in fact,
all these teeth are still strictly triangular in form, although the outer wall
may be somewhat irregular. A prominent cutting metastyle, analogous
to that seen in primitive Creodonts, is developed.

In the lower molars we observe an elevated anterior triangle or
trigonid. As a rule the protoconid or reptilian cone is the most
prominent cusp, but in certain specimens (Fig. 47 HI} this cusp is some-
what depressed. A progressive character is the reduction of the antero-
internal cusp, or paraconid. But the most strikingly progressive feature,
as compared with Jurassic molars, is the broadening out of the heel of the
crown or talonid to support three cusps, the hypoconid, hypoconulid, and
entoconid respectively.*

Thus there are three main features, as compared with the Jurassic
molars : (1) general depression of the crown of the upper molars,
accompanied by a strong development of the external cint/t/lar cvxj>*
and of conules, but not by an internal cingular development ; therefore
there is (2) an entire absence of the postero-internal cusp or hypocone ; (3)
there is a striking development of the heel or talonid of the lower molars
with two or three cusps, and more or less depression of the trigonid. In
general, these progressive features relate these teeth very much more
closely to those of the Basal Eocene than to those of the Upper Jurassic.

Numerically, the most progressive of these upper molars have from
two to three fangs ; from three to five cusps, namely protocone, paracone,
metacone, protoconule and metaconule ; and two to three cingules,
parastyle, mesostyle, and metastyle ; the lower molars have two or more
fangs and from five to six cusps, protoconid, paraconid, metaconid,
hypoconid, hypoconulid and entoconid.

* [This broadening of the talonid was probably con-elated with the depression and
transverse broadening of the protocone. — ED.]

In Figs. A, B, D, the long sectorial spur from the outer wall is more probably the metastyle
than the parastyle, consequently this is probably the posterior side of the tooth ; although if the
tooth happened to be in'>, the elongate spur would be anterior, ;., . parastyle.

A, — Crown and side views of three superior molars, probably of the right side. Types of
Pi'OtolnmW.a liatclu.ri Osborn.

B. — Crown views of four superior molars of the left and right sides. Genus /)«?<'n/////.< Marsh.

C. — A superior molar, and an inferior molar of the right side. Type of Sijiifunmloii sexicuspi
( isliorn.

D. — Crown and anterior views of two superior molars of the left side. Genus not determined.

E. — Crown view of a right superior molar. bi<l< l/i/m/ix vorax Marsh, after Marsh.

F. — Crown views of two superior molars, probably of the left side. Type of Ertoconn,i:>,i
ti> t: rsoni Osborn.

O. — Crown and side views of a superior premolar or molar (ef. Jiri/nlixtis, Iciojix). Genus
not determined.

El. — An inferior molar of the left side. ? Di<i, Iphodon. Crown and inside views.
A!. — Two inferior molars of the left side. Genus undetermined. Crown and inside views.
HI. — Crown and inside views of a lower molar of the left side. Genus not determined.
II. — Crown and inside views of a lower molar of the left side. Genus not determined.

G

98 EVOLUTION OF MAMMALIAN MOLAR TEETH

5. BASAL EOCENE MAMMALS.

In the basal Eocene, Puerco, Torrejon, and Fort Union formations of
North America, we again find the mammals divided into multituberculates
and trituberculates.

Among the latter the law of trituberculy was first perceived by Cope ;
they include animals possibly related to the Primates and Rodentia, also
ancestral Edentata (?), Creodonta ancestral to the Garni vora, Condylarthra
collateral ancestors of the Perissodactyla and Artiodactyla, and Ambly-
poda, a very primitive order of Ungulates totally distinguished from all
others by the fact that the teeth evolve throughout on the triangular or
tritubercular basis.

The American Museum collection as revised by Dr. W. D. Matthew,
omitting the Multituberculates, contains 1200 identified specimens,
referred to 62 species, 33 genera, 6 orders. All flic upper molar* of these
animals without exception are either : (].} purely tritubercular or triangular,
or (2) transitional to the quadritubercular and quadrate form Tjy the addition
of a hypocone. This more extended knowledge of the basal Eocene not
only reinforces the evidence originally advanced by Cope, but affords an
almost overwhelming body of proof as to the fundamentally trigonal
tritubercular character of the upper molars of these orders.

The molar teeth exhibit a marked progression upon those of the Upper
Cretaceous, in three very important characters: (1) there is an internal
cingulum in a very large percentage of the upper molar teeth, (2) a
hypocone or postero-internal cusp is developing from this cingulum in a
somewhat smaller percentage of upper molars : (3) the paraconid or
antero-internal cusp of the lower molars is disappearing in some of the
lower molars.

Numerically, the most progressive upper molars present three fangs,
six cusps : namely, protocone, paracone, metacone, hypocone, protoconule,
metaconule, and two styles, parastyle, mesostyle, and metastyle. The
lower molars present also six cusps, protoconid, paracouid, metaconid,
hypoconid, hypoconulid and entoconid, the latter three arising from the
talonid or heel.

6. LOWER EOCENE MAMMALS.

In the lower Eocene, Suessonien of Europe, Wasatch of North America,
the majority of modern groups of mammals are represented, including the
following orders : Primates, llodentia, Insectivora, Creodonta, Carnivora,
Tillodontia, Edentata, Condylarthra, Amblypoda, Perissodactyla, Artio-
dactyla.

All of these animals are either tritubercular or in stages evolving from

GEOLOGICAL SUCCESSION OF MOLAR TYPES

99

trituberculy into higher forms of molar teeth as described in othci
chapters (see pp. 37, 45, 132, 151, 158, 170, 171, 175, 178).

RESUME OF GEOLOGICAL SUCCESSION OF TYPES.

It does not appear necessary to present any further proofs of the fact
that apart from the Multituberculates the geological succession of types of
molar teeth presents so few exceptions to the main law of trituberculy and
the principle of cusp addition that these exceptions may be designated as
aberrant. The advance of the teeth from type to type corresponds with

the advance of geological time as follows :

UPPER MOLAR TYPE.

MAXIMUM TOTAL OF CUSPS

IN ONE PAIR OF
UPPER AND LOWER MOLARS.

Triassic,

6

Haplodont, Protodont, Tri-

conodout, with single or

grooved fangs,
Jurassic, Triconodont, Tritubercular,

with 2 or 3 fangs, 8 cusps, 1-2 styles.

Cretaceous, Tritubercular, 11 cusps, 2 styles.

Basal Eocene, Tritubercular, Quadrituber-

cular, and higher stages, - 12 cusps, 3-4 styles.
Upper Eocene, Tritubercular and all the

derived types, 12 cusps, 6 styles.

CHAPTER VII.

THE ORIGINAL OR PRIMITIVE STRUCTURE OF THE MOLAR TEETH
IN THE DIFFERENT ORDERS OF MAMMALS.

MAJOR CLASSIFICATION OF THE MAMMALIA.

THE Table of Classification on pages 11-16 naturally introduces the brief
review of the various orders of mammals with reference to the structure
of their grinding teeth which forms the subject of this chapter.

The reptilian superorder Therapsida, including the order Cynodontia,
is the only one of the numerous orders of reptiles in which striking
resemblances to the mammals are observed in the skeleton as well as
in the teeth (see p. 91). These animals appear to stand nearest the
reptilian stock from which the mammals sprang, and to most nearly
fulfil the hypothetical groups, Hypotlieria, Promammalia, Sauromammalia
proposed respectively by Huxley, Haeckel, and Baur.

The two grand divisions of the Mammalia proper, the (1) Prototheria
or egg-laying mammals, and (2) Eutheria or viviparous mammals,
include fourteen extinct and eighteen living orders, or thirty-two orders
altogether. It will be shown that in thirty of these orders tritubercular
evolution can either be directly demonstrated or is rendered in a high
degree probable by analogy. In three orders only the Monotrernata,
Allotheria (Multituberculata), and Cetacea, does there remain considerable
doubt as to the mode of dental evolution.

In the two great divisions of the Eutheria, namely, the Didelphia
(Marsupialia) and Monodelphia (Placentalia), tritubercular evolution,
although proceeding independently, has arrived at closely similar adaptive
types, thus affording remarkable instances of parallel evolution.

The Placentalia are conveniently divided into four great groups,
which it is true have a descriptive rather than strictly taxonomic value,
namely :

1. The Unguiculata, including all the clawed animals of primarily
insectivorous (Insectivora, Cheiroptera), or carnivorous habits (Carnivora),
and secondarily of prevailing herbivorous habits (Rodentia, Edentata).

ORDINAL TYPES OF MOLARS : PROTODONTA, ETC, 101

In a general way these animals are conservative and exhibit some of
the earlier stages of trituberculy as well as the extreme specializations
for exclusively carnivorous and herbivorous diet.

2. The second group includes the Primates, which are primarily
frugivorous and omnivorous. In a general way also, these monkeys
and lemurs are transitional between an earlier insectivorous and a
frugivorous and herbivorous diet; thus the study of their teeth forms an
admirable introduction to the study of the teeth of the Ungulata.

3. The third great group includes all the hoofed animals or
Ungulata, of which the food is primarily herbivorous and secondarily
omnivorous in certain cases. In a general way again, the Ungulata
exhibit the extremes of complication of the crowns by cusp addition
and cusp modelling.

4. The fourth great group includes the Cetacea, in which the teeth
are specialized for various kinds of sea food, or entirely wanting.
They exhibit what is probably a retrogressive dentition from an earlier
type, such as that presented by certain of the Unguiculata.

PROTODONTA.

The molar teeth of the Protodonta have been described on page 18 as
transitional from the haplodont to the triconodout condition. Similar
changes are observed among the Theriodontia, as described on page
91 ; and analogous changes, that is, the addition of lateral cusps to a
central cusp, have occurred frequently among the Vertebrata.

It is possible that the Protodonta are not mammals ; but the in-
cipient division of the fang and the apparent absence of angular and
articular bones of the jaw render it probable that they are at least very
early types of mammals, probably of egg-laying habits, like the Mono-
tremes. A jaw somewhat resembling those of Dromatherium and
Microconodon has recently been described by Broom* in South Africa
from the Upper Beaufort Series (Karoo system), which are considered of
Upper Permian or Triassic age.

SPEC I A L R INFERENCES.

Osborn, H. F., "On the Structure and Classification of the Mesozoic Mammalia,"
Journ. Acad. Nat. Sci. Phila. (2) IX., 1888, p. 222, antea, pp. 31-35.

Osborn, H. F., " The Triassic Mammals, Ui-oniathtrium and Microconodon ; Proc.
Amer. I'lu'lns. Soc., Apr. 1887, <nit<'«, pp. 18-21.

ALLOTHEKIA 01; MULTITUBERCULATA.

Because the Multituberculata are among the oldest known mammals,
first occurring in the Knit-tic or Upper Triassic, it has been urged that the

*Geol. M«<i. Deo., IV., Vol. X., 190:?, p. 34r>.

102

EVOLUTION OF MAMMALIAN MOLAR TEETH

\ 7n> t

]•'](:. 4s. 1. M',1-1 -ill: .sV. ,s ,i,>>i'ii'nx (Stuttgart .Collection), a lower molar viewed from above ; lc,
posterior face; 16, external face, greatly enlarged. '2. .l/'V,Wr.\v. .-,• (FJlii<i/iin/ii.i- -) ntinn-'i. from
above, o. riii<ii,inlti.i- minor, the lower molars viewed from above; 3n, external face of same
enlarged Oj diameters. 4. Ptilodus trovessartianus, lower molars viewed from above; 4((, ex-
ternal face of same ; ifl, internal tubercles ; eel, external tubercles. Original.

KII:. 4'.i. Mi ,i/.vc«r.iM!x, Laramie or Upper Cretaceous of North America. Probably an upper
molar viewed upon the wearing surface, enlarged two diameters.

Fio. 50. «, Left lower jaw of Cteiini-oi/uii .f rr/iliin, Marsh, from the Upper Jurassic or Basal
Cretaceous of North America, inner view, three times natural size ; b, Hight upper jaw of
C. potcns, inner view x 4. c, The same seen from below, a, First premolar, li, fourth preuiolar
as interpreted by Professor Marsh. After Marsh.

Figs. 48-50. The molar tooth forms of the Multituberculate family Plagiauluci<ho.

ORDINAL TYPES OF MOLARS : MULTITUBERCULATA

103

Multitubercular pattern is a primitive one. Forsyth Major's argument on
this subject is discussed on pages 205-208, and the conclusion is reached
that there is quite as much evidence for considering the rnultituberculate
molar as secondary or evolved from a still earlier tri tubercular type.
There are several grounds for considering that such an evolution mav
have taken place.

1. The earliest multitubercular molars known, those of J//Vro/,.s/V.s '
(Fig. 48, Xos. 1, 2) are comparatively simple, consisting of a central
depression or basin indented on the edges by larger and smaller tubercles.
Thus the earliest rnultituberculate molars known have comparatively few
•cusps, in other words, are paucitubercular, whereas the latest rnultituber-
culate molars known have a very large number of tubercles, or are
multitubercular.

Fig. 51. The molar tooth forms of typical Multituberculates.

.'it, from the Rhsetic (Triassic) of Germany, an upper molar ; la, ditto, in side view,
natural size. 2, 'la, Ti'itylmlim, from the Triassic of South Africa, an upper molar, in2, wearing
surface and outside view, natural size. 3. Pt>liiin«st<«i<j,i, from the Basal Eocene of North America ;
the second upper molar, natural si/.e. 4. Boloil.oii, from the Upper Jurassic of England, the third
and fourth upper molars, enlarged about 6 diameters. 5. Shnoymtthus, from the Upper Jurassic
of England, a lower molar enlarged about 2J diameters. 6. Chirox, from the Basal Eocene of
North America, the upper molars enlarged 1 J diameters.

2. Basin-shaped crowns more or less similar to those of Microl?*1<-x
moorei are known to be secondary or of tritubercular origin in several
distinct families of mammals. Thus certain Eodents (p. 205), Fruit-bak
•(p. 129), the Kinkajou (CVra^/Vx, p. 142), Myrmccobius, (p. 112), exhibit
more or less distinctly basin-shaped crowns of which the edges are
•elevated, and yet retain more or less clearly the primitive tritubercular
pattern.

3. The depression of the centre of the crown to form a basin out of
what was primitively a projecting point or cone finds its analogies in the
•invagination of the incisor teeth of the horse, and in the basin-shaped

[J The evidence tending to connect Microlextes with the Multituberculates is briefly that :
(1) the M. moorei molar closely resembles that of Pkujinnlax minor (Fig. 48, Nos. 3, 3«); (2)
Plagiaulax minor presents many resemblances to Ctenacodon and Ptit.odus, typical Multi-
tuberculates ; (3) an enlarged grooved premolar, probably representing a less specialized
condition than that seen in the later Multituberculates has been recorded by Dawkins from
the Rhtetic beds in which the Microlestes moorei molar was found ; (4) Dr. Ameghino
brings evidence to connect the Multituberculates (including J/MVO/<:*/VX) with the Diproto-
•dont Marsupials, especially the South American forms described by him. — ED.]

104

EVOLUTION OF MAMMALIAN MOLAR TEETH

third incisors of the Eat Kangaroo (J-Jctto/if/ia). In Aninci/on (p. 133)
the crown is depressed but not at all basin-shaped. The molars which
gave origin to the Kinetic Microlfst.cs basin-shaped type may, therefore,
have been either simply conic or haplodont, or even possibly tritubercular.

FIG. 52. The outer surface of the right maxilla of So/mlon, from the Middle Purbeck Beds

(Upper Jurassic, England). XA

c,

53. Views of the maxilla of Allodon laticeps, Marsh, seen from below, from the Atlaiitosaurus,
Beds (Upper Jurassic, North America), x 4.

llllJJi ^3

• ,s

FIG. 54. Cldro.r phratuis, Cope, x §. a, viewed from below, palate with dentition, three
promolars and two molars in situ ; b, viewed from the outer side. From the Torrejon Formation,
Basal Eocene, Stage II. New Mexico. After Cope.

Figs. 52-54. The molar tooth forms of the Multituberculate families

BolodonticUe, Chirogidte.

It must be clearly stated, however, that the Microlestes antiquus molar
bears no close resemblance to any of these types and there is no clear
evidence in the Microlestes molar itself of derivation from the tritubercular
type.

ORDINAL TYPES OF MOLARS: MONOTREMATA 105

4. Some additional evidence on this point is derived from the study of
the dentition of the molar teeth of the Monotremes (p. 107).

5. Still another possibility is that the molar form of Microlestes
antiquus may be derived from that seen in Phascolotherium by the
upgrowth of heavy cusps from the internal ciuguluni (compare Fig. 48,
No. 1, and Fig. 6).

6. Again, this molar might represent a pattern allied to that of the
Amphitherium molar (Figs. 15, 17), that is the highest anterior (?) cusp
above the anterior (?) root (Fig. 48, Nos. la, Ib) may be a protoconid,
the first cusp on the opposite side of the tooth a metaconid, and all that
portion of the tooth above the second root may be the talonid. Or again,
the Microlestes molar might have been derived from that of the Amblo-
theriida- by the upgrowth of the cingulum cusps on the inner sides of
the crown and the loss of the transverse connecting ridge.*

SPECIAL REFERENCES.

Ameghino, F., "Los Diprodontes del Orden de los Plagiaulacicleos y el Origen de 1<»
Roedores y de los Poliniastodoutes," Anales del Mus. Nac. de Buenos Aires, Tom. IX.,
1903, pp. 81 a 192.

Osborn, H. F., " The Structure and Classification of the Mesozoic Mammalia,''
Jour. A cad. Nat. Sci. Phila., Vol. IX., No. 2, July, 1888.

Owen, Richard, " Monograph of the Fossil Mammalia of the Mesozoic Formations,"
Palceontogr. Soc. Lond., 1871.

Owen, Richard, "On the Skull and Dentition of a Triassic Mammal (Tritylodon
longcevus)" from South Africa ; Qti«r. .l<mr. Oeol. Soc., 1884.

Lemoiue, Victor, "Etude sur le Neoplagiaulax de la Faune Eocene inferieur d-s
Environs de Reims," Extr. Bull. Soc. Geol. de France, Feb. 1883.

Cope, E. D., "The Tertiary Marsupialia," Amer. Naturalist, 1884, p. 687.

Marsh, O. C., "The American Jurassic Mammals," Amer. Jour. Sci. and Arts.,
April, 1887.

Marsh, O. C., " Review of Osborn's Memoir on the Structure and Classification of
the Mesozoic Mammalia," Proc. Acud. Nat. Sci. Phila., June, 1887.

Falconer, H., "On the disputed Affinity of the Mammalian Genus Plagiaulax from
the Purbeck Beds," Collected Memoirs, Vol. II., pp. 430-451.

MONOTREMATA.

The true molars of OrnithorhyncJms when first described (Poulton,1
1888) were adduced by Cope as inultituberculate and as tending to
demonstrate the affinity of the Multituberculata to the Monotremata.
When critically examined, however, the molars of Ornitkorhynchus are
found to be very degenerate both in structure and in pattern, and it

* [Against all these speculations one might advance another speculation, that Microlexti s,
Tritylodon and the Multituberculates, appearing iu the Triassic, were not closely related
to trituberculate mammals (which are first known in the Upper Jurassic or Basal Cretaceous)
but were independent offshoots from the Tlieriodontia. — -En.]

1 Poulton, E. B., "The True Teeth and the Horny Plates of Ornif/ior/n/ii,-!,,!*." V'""'-
Jour. Microsc. Sci, Vol. XXIX., Aug. 1888.

106

EVOLUTION OF MAMMALIAN MOLAR TEETH

- in* -

">'

-/n'-

-m-

m-

FIG. 55. Dentition of North American Upper Cretaceous Plagiaulacidse (Ptilodus) and PoljT-

mastodontidse (Meniscoessus).

1. Ptilmlvs.— First and second inferior molars ; on left side in situ, 011 right side reversed.
These teeth belong to two individuals.

ORDINAL TYPES OF MOLARS: MONOTl: KM ATA

107

cannot truly be said that they actually resemble those of any Multituber-
culate in the strict sense, because all the higher Multituberculates exhibit
an extremely regular mechanical disposition of the cusps, whereas in this
living Monotreme the cusps are extremely irregular. Secondly, it does

External Internal
Inferior

xio

Internal External
Superior

Fig. to. Molar teeth of Oi~nithorhynchus /IC.-.K'O./'KS, before being shed and functionally re-
placed by the horny epithelial plates. After Oldfield Thomas. Xos. 1, 2. Superior and inferior
teeth and horny dental pads. X'o. 5. Crown view of a molar of " Microlestes"

not appear that the Oniithorliynclius molars can be cited as evidence either
for or against the tritubercular theory because of the evidently secondary
and largely degenerative changes which they have undergone ; they
bear evidence (Fig. 56) of descent from a more primitive regularly

'2. Ptiloilus. — First and second inferior molars, worn considerably. These teeth belong to two
individuals.

3. 1'tilvilim. — First superior molar of the left side.

4. Ptiloiltis. — Fourth superior prevuolar, first and second molars placed together and reversed
in outline to show probable relations. The three shaded teeth on the left side of drawing belong
to three individuals.

.">. ii. Ptilmli's. — External and superior views of two lower jaws, showing proportions of the
teeth.

7. -V. nixt-nfxsiix.— First and second inferior molars of two individuals placed tou,-tliri- ,'ind
reversed to exhibit the natural position.

8. A/< /uV(.i/.<xi!.v - i-iiiHiuisti's. — First and second superior molars of two individuals placed to-
gether and reversed to show the natural position.

9. Mi iiixruexgH.*. — Composition side view of upper and lower dentition as far as known. Teeth
and jaws combined from eight individuals. The superior premolars are not yet known with
certainty.

108 EVOLUTION OF MAMMALIAN MOLAR TEETH

cuspidate condition. As described and figured by Thomas,1 the posterior
superior molars exhibit two high conical internal cusps, from which
minute ridges run downward to the external borders of the crowns, the
edges of which are peculiarly crenulate rather than cuspidate, in the
ordinary sense of the word. In the posterior •////<•/•/>>/• molar there are two
high cusps on the external side connected by transverse ridges, with a
series of crenulations on the internal side.

It is especially noteworthy (1) that unlike the Multituberculates the
lower molars reverse the pattern of the upper molars (as in tritubercular
teeth generally) and (2) that the highest cusps are on the inner side of the
upper molars and on the outer side of the lower molars. So far as these
facts are of rulni' they would support fltc hypothesis that these are degenerate
tritubercular teeth.

MAESUPIALIA.

It was early perceived by Cope that the molars of Didelphys (Figs.
57, 58 c) are of the tritubercular type, very similar, in fact, to those of
the early Eocene Creodonts. We are indebted to Dr. B. Arthur Bensley-
for a careful study of the molar teeth of Marsupials in general and an
exposition of the very striking parallels which they present to the evolu-
tion of the molar teeth in Placentals. In fact, on the supposition that

FIG. 57. Skull and dentition of the common Opossum (Didilphys i-ii-i/iiiinint) ', a typical Poly-
protodont Marsupial, with the dental formula Inc. f , Can. i, Pms. f, Ms. f or 50 teeth in all.
x £. After Matthew.

the Marsupials separated at a very early period from the Placentals and
that subsequently the dental as well as the general evolution of the two
groups was entirely separate and independent, the teeth exhibit some of
the most convincing proofs of parallel or homoplastic evolution of which
we know.

1 Thomas, O., ''On the Dentition of Ornithorhynckus," Proc. Roy. Soc., Vol. XLVI. ,
1889, pp. 126-131, PI. 2.

- Amer. Naturalist, Vol. XXXV., 1001, p. 251; also Tran*.:-Linn. Sor.,- London, 2nd
Ser., "Zool.," Vol. IX., Pt. 3, Dec. 1903.

ORDINAL TYPES OF MOLARS: MARSUPIALIA 109

Primitive Tuberculo-sectorial Types.

Bensley shows that the Oligocene opossum (Pemtherium) molar may
be taken as the theoretical starting point from which development took
place along two main lines, namely (1) the carnivorous and (2) the omni-
vorous and herbivorous.

If we arrange the teeth of the Australian Marsupials according to the
analogy with the Placentals, we obtain such a result as is shown diagram-
matically in Figure 58, Nos. 1, 2. The primitive tritubercular, tuberculo-
sectorial type is here represented by the teeth of Daxyurus vircrrinnx
(58 r?), but would be still more strikingly illustrated by the teeth of one
of the purely insectivorous forms of the Dasyurida? (Sminthopsis, Ante-
chinomys, Phascologale}. All these Marsupials exhibit molar teeth which
may be readily compared in pattern cusp for cusp with those of
Didelphys.

Carnivorous Marsupials.

From this archetype, so to speak, the first line is the carnivorous line
to be compared among the Placentals with the specialized Creodonta
(Oxycena, etc.). Among Marsupials the carnivorous evolution is entirely
confined to the single family Dasyuridae, and it culminates in the teeth
of the Tasmanian wolf (Thylacynus cynocephalus). The teeth of this
animal are represented in Figure 58, g\ they show all the essential
characters of the teeth of Dasyurus except that in the lower teeth the
metaconid is absent (this cusp also disappears in the lower cutting-
teeth of the Felidse among Placentals). The progressive carnivorous
adaptation is effected in the upper molars by a conversion of the
paracone and metacone and metastyle into cutting blades and by a
degeneration of the other styles, which become reduced to very incon-
spicuous tubercles. In the lower molars the corresponding cutting blades
are formed by the elevation and lateral compression of the paraconid,
protoconid and hypoconid, while at the same time the cusps on the
other side of the crown, the metaconid, hypoconulid and entoconid are
either reduced or disappear (metaconid).

Omnivorous and Herbivorous Marsupials,

Turning to the second group, which leads to the omnivorous and
herbivorous adaptations of the teeth, we find that the first stage is
represented in the molars of the Bandicoots (Peramelidas) Fig. 58c, 59.
Analogy with the evolution of these teeth is to be sought among the
primitive insectivorous-omnivorous Placentals, c.<j. among the modern
Insectivora, in Tupaia and Myogal:.

The main trend is toward cusp addition ; thus the upper molars

110

INVOLUTION OF MAMMALIAN MOLAR TEETH

-Pa.

-Me

Upper Molars of the Left Side.

FIG. 58, No. 1. Adaptive Radiation of the Tritubercular Molar Type in the Marsupialia.
From Bensley. (Cf. Fig. 58, No. -I.)

a, b, c, Insectivorous type, re, Oligocene Opossum (Pi i-ut/n i-ium /".•/».'•); fr, Recent Opossum
(Diillpliys azai-if); c, l>nlt.l/>/i>/s ririiinianu. a, d, r/, Progressive carnivorous specialization.
</, Dnsimrus rii-ei-riitus; ft, 'J'lti/lacitnus ciinvreplinlus. e, Progressive insectivorous-omnivorous
type (Perameles imsi'tit) ; /, Omnivorous to herbivorous type (Petauroidex rolans). h, Incipient
lophodont, herbivorous type (Trirl«i*<n '"•< fi'l/iiculu). i, Perfected lophodont, herbivorous type
(Mucropus sp.). j, Selenodont herbivorous type (Phascolttrctos cineri ex).

of Perameles nasuta1 (Fig. 58, c) show all the essential characters of
those of Da*>/nriix, and they show the addition of an incipient hypo-
cone ; in the third molar the hypocone is not very pronounced and
the tooth is still triangular ; but in the second molar the hypocone
is well developed and the tooth is now quadrate. Thomas (1888,
p. 220) describes the triangular and quadrate modifications as character-
istic of this family (the Bandicoots). The lower molars of Perameles

1 The upper molars of Perameles macrura (Figs. 59, 60) closely resemble those of the
Insectivore Myoyale, as shown in unworn teeth ; thus they exhibit the many sharp cusps
and the double mesostyle characteristic of several insectivorous forms.

ORDINAL TYPES OF MOLARS: MARSUPIALIA

111

Pa'

\

V<|

\ -fc^*s

^m-^

End-

Lower Molars of the Left Side.

Flu. os, Xo. '2. Adaptive Radiation of the Tritubeivular Molar Type in the Marsupialia
(continued). From Bensley. (Cf. Fig. 58, X». 1.)

o, b, c, Insectivorous type. «, Oligocene Opossum (Peral/ni-inin .""""'); !l, Recent Opossum
(Diilelphii.i a :i.i i-ii1); c, DiiU-l/ilii/x cu-'juiinnu . n , t1, ;/, Progressive carnivorous specialization.
(', Daxyui-ux viverrinus \ <i, Thtilnri/,t,'.-: ,-n,tm; jilmlun. c, Progressive insectivorous-omnivorous
type (Pi rameles nasntn); f, Omnivorous to herbivorous type (I'ltmii-niiiix i-i>/n,ix). It, Incipient
lophodont, herbivorous type (Trichosurv.x I'ul/n'i'ulu). i, Perfected lophodout, herbivorous type
(Macropits sp.). j, Selenodont herbivorous tj-pe (P/iascolarctos cinert i> .-•).

nasi it a (Fig. 58, No. 2, c} persistently resemble still more closely those
of Das//t>i"vv, the omnivorous modification being only developed in the
posterior heel where the hypoconulid is reduced.

Perfected Omnivorous Marsupials.

The perfected omnivorous modification among Marsupials is met
with in the Phalangeridse. The two examples of the Phalangers
examined by Bensley were Tricliosurus rnlpecida, in which the molars
are highly specialized, and Petanroides volans, the teeth of the latter
being represented in Fig. 58, /. Although these teeth illustrate the
completed quadritubercular condition, they are only approximately transi-
tional ; as among the Placentals, the hypocone is completely formed
in the upper molars and the paraconid is correspondingly reduced

112 EVOLUTION OF MAMMALIAN MOLAR TEETH

in the lower molars. The anterior shelf in the lower molars (a.s.)
which is prominent in the Peramelida? and Dasyuridae and very con-
spicuous in the molars of the Macropodidae (Fig. 58, i.\ is here absent.
There are also no vestiges of the external styles in the upper teeth.

Crcscentic and Crested Herbivorous Marsupials.

The crescentic or selenoid modification of the cusps, which is so
characteristic of the Artiodactyl Ungulates among Placentals, appears
to be developed only in Phascolarctos (Fig. 5 8 j.) among the Marsupials.

The crested or lophoid modification of the cusps so common among
the perissodactyl ungulate Placentals is widely represented in the
Macropodidae (Fig. 58 i.}. The incipient stages leading up to this
crested condition are seen in the teeth of the specialized Phalanger
Trichosurus vulpecula (Fig. 58 k) and also in those of Hypsiprymnodon
moschatus among the Macropodidie, or Kangaroos. A further parallel
to the perissodactyl Ungulate placentals is witnessed in the develop-
ment of a hypselodont modification of the crested crowns among certain
Macropodidae.

The derivation of all the molar types in the Marsupials from a
tritubercular pattern is rendered still further probable by the existence
of annectant forms which tend to unite the specialized, quadri-to
sexitubercular Diprotodonts (Kangaroos, Phalangers, Wombats), with the
generalized tritubercular Polyprotodonts (Dasyures and Opossums).
Thus Ccenolestes, the only living American Diprotodont lacks the
characteristic Diprotodont reduction and syndactyly of the second and
third digits of the hind foot and is further allied to the Polyprotodonts
by its close external resemblance to the Dasyurid genus Phascologale.
On the other hand, the Polyprotodont Bandicoots (Peramelidas) exhibit
the syndactyly of the Diprotodonts. Among fossil forms the gap
between the two sub-orders is largely bridged over by the extraordinary
genus Wynyardia of Baldwin Spencer,* which presents a perfect
melange of characters seen elsewhere only in the Opossums and
Daysures (Polyprotodonts), and in the Phalangers and Kangaroos
(Diprotodonts).

Aberrant and Specialized Tyrjcs.

The apparently triconodont lower molars of Tki/laci/nus are thus
shown to be secondary.

The multicuspidate true molars of MyrmecoMus further support the
view that an elongate or basin-shaped and polybunous crown may have
arisen from a more tritubercular crown, because they still retain traces

* [" A Description of Wynyardia bassiana, a Fossil Marsupial from the Tertiary Beds of
Table Cape, Tasmania," Proc. Zoo/. Soc. Lond., Nov. 20, 1900, pp. 776-795, PI. XLIX.]

ORDINAL TYPES OF MOLARS: MARSUPIALIA

of the tritubercular crown. The secondary development of a number of
sharp piercing cuspules seems here as in the Insectivora (p. 117) to

Notoryctes, the Marsupial Mole

pas mt* rr?t*

be correlated with insectivorous diet.
(Figs. 61, 62) has tritubercular upper
and lower molars ; in the upper molars
the high pointed protocone is at the
apex of a V, formed by two ridges
diverging downward and outward to
the basal external portion of the
crown, ending in the para- and meta-
cone respectively ; on the internal or
lingual side, at the base of the crown,
the V is surrounded by a deep U-
like internal cingulum The lower
molars reverse this pattern except
that they lack the U-like cingulum.*
According to Gidley the principal
cusp (marked pi') is really the para-
cone + metacone, the protocone ap-
pears as an internal ledge (marked
cing), but there is no evidence for
this except analogy with the supposed
case in Insectivora (see pp. 124, 22V).
The molars of the Bandicoot (Pcra-
mcles, Figs. 59, 60) show a general

parallelism with those of Didclpli'us, G-aleopithecus, certain Moles and
Shrews, and certain Eocene Artiodactyls (Ccenotherium, Tapirulus) in

Fiu. 59. Molars of Perameles macri'j-n fivm
a specimen in the Yale Museum, xf. ^4, Crown
view left upper molars. B, Crown view left lower
molars. C, Upper and lower molars in opposition,
as seen obliquely from the outer side and below.

711

FIG. 60. Diagram showing spatial relations of the patterns of the upper and lower molars
when in cont&ct of Perameles maerura, (Cf. Fig. &.>). Compare the somewhat similar pattern of
the molars in the Mole, and in

several respects. First the mesostyle is doubled (ws1, ///*") permitting
the formation of two complete triangles from the paracone, metacone,
and styles of the upper molars. Second, the lower molar pattern consists

* [As in other mammals without a talonid in the lower molars, the protocone is much
elevated. The internal ledge (tiny.) appears to correspond with that in CJvrysochloris
(Fig. 68).— ED.]

H

114

EVOLUTION OF MAMMALIAN MOLAR TEETH

of two triangles, the hinder one being formed from the enlarged talonid,
the apices of these triangles being external and fitting in the spaces
between successive outer triangles of the upper molars.*

rne.

Fin. 61. Dentition of the Marsupial Mole, Notoryctes typlilops. XT. After E. C. Stirling.

Thus, in conclusion, the various Marsupials afford the most con-
vincing evidence of tritubercular origin. It is only necessary to

understand the intermediate stages and to
compare even the most highly specialized
teeth with analogous types among the Pla-
centals, to be convinced of the universal
tritubercular derivation. Howes (Sept. -Brit.
Assoc., Belfast, ]902) declared he could not
conceive the derivation of the Diprotodon
molar from the tritubercular type, but com-
parison of these teeth with those of Macropus
and the stages leading to the Macropus type,
FIG. t;-2. First upper molar of together with the still stronger but more

NotorycUs lHf>ld»p*, upper figure, m°

crown view -, lower figure, anterior indirect evidence anoi'ded by the Jrlacentals1

view. XT- After E. C. Stirling. • T 7-, • 7 * -\ -n ij.'

indicates that the Diprotodon molar will ulti-
mately be found to be of tritubercular origin.

SPECIAL REFERENCES.

Dentition of Marsupials (figures and descriptions).

Owen, Eichard, Odontography, 1840-1845.

Tomes, C. S., A Manual of Dental Anatomy, 1898.

Bronn, H. G., Klassen und Ordnungen des Thier-reichs, Bd. I., pp. 169-177.

Schlosser, M., Die Affen . . . Marsupialier . . . des Europaischen Tertitirs, etc.,
Wien, 1887-90.

De Blainville, H. M. D., Osteographie des Mammiferes, Tome III., Paris, 1839-1864.

Bensley, B. A., " Oh the Evolution of the Australian Marsupialia ; with Remarks
on the Relationships of the Marsupials in general." Trans. Linn. Soc., 28 Ser.,
Vol. IX., Pt. 3, Dec. 1903, pp. 83-217.

*[As in other mammals the protocone opposes the talonid of the preceding lower molar,
the hypocone opposes the trigonid of the corresponding lower molar. —ED.]

1 Compare the similar lophodont molars of the Tapir with the lophodont to quadrituber-
cular molars of Syxtemodon (Fig. 172).

ORDINAL TYPES OF MOLARS 115

MOLAR TEETH OF MULTITUBERCULATES, MARSI.TIALS (?), AND PLACENTALS
^1 IN THE UPPER CRETACEOTS.

Some years ago the interesting discovery was made of a great number
and variety of teeth in fragmentary jaws of mammals in beds of Upper
Cretaceous age in North America. The specimens are so fragmentary
and isolated that it is difficult to classify them ; some probably represent
Placentals, others probably represent Marsupials, whilst still others are
possibly Prototheria (Multituberculata).

As shown in Figs. 47 and 55, from Osborn, among these forms are two
prevailing types of teeth, namely, the highly specialized mxltitt/lerculate
teeth, which had reached almost their final extreme of evolution, and the
less specialized trituberculate teeth which are in a comparatively early
stage of evolution.

The Multituberculates (see p. 95 and Fig. 55) prove that regularly
succi'xxi r<- cn*i> ii'Iilifion had taken place since the earlier Jurassic stage, as
a result of which the first upper and lower molars exhibit as many as 23 and
12 tubercles respectively in Ptilodus, or 21 cusps and 9 cusps respectively
in Meniscoessn*. This law of successive cusp addition from the posterior
basal cingulum is entirely analogous to that which occurs in the complicated
molars of the Proboscidea. It demonstrates that cusp origin in the Multi-
tuberculata at least is not by concrescence but by cusp addition.

Trituberculates. — In this group (see also pp. 95-97, Fig. 47), all the
upper molars known are of the simple, generally low- crowned, tritubercular
type, that is, they consist of the trigon to which no trace of the hypocone
or postero-internal tubercle has been added. About ten distinct kinds of
upper molars have been found altogether. They are all much more recent
in type than the few upper molars known of Jurassic mammals, which are
invariably high-crowned or piercing ; on the other hand, they are some-
what older in type than the prevailing molar teeth of the basal Eocene
or Puerco, because they lack all traces of the hypocone, which is very
common in Puerco teeth. These Cretaceous up^cr molars exhibit, how-
ever, two other characters found also in the Jurassic, namely, (1) small
comdi's or intermediate tubercles; (2) external sft/fcs, or extensions of the
cingulum on the outer wall ; in some cases these styles are very
prominent, forming distinct cusps (Fig. 47).

The primitive features of the upper molars all point toward a tri-
tubercular ancestry. We observe first the large size of the internal
cusp or protocone, which certainly supports the theory that this cusp
was earlier in evolution than either of the outer cusps.* Second, we
observe the striking symmetry of the external cusps or paracone and
metacone, giving a symmetrical form to the trigon or main triangle

*[See criticism of the similar view expressed on p. 95. — ED.]

116 EVOLUTION OF MAMMALIAN MOLAR TEETH

of cusps,* which also lends strong support to the theory that the large
protocone was the original cusp of a trigon and that the para- and
metacones were added as lateral or external cusps of equal size. The
" premolar analogy " or " paracone theory," on the other hand, receives no
support whatever from the study of these superior molar teeth.

In general it may be said that the upper molars are intermediate
between the Jurassic and the basal Eocene stages of evolution. They
thus lend overwhelming proof, if any more were needed, of (1) the
progressive evolution of trituberculy among the mammals, (2) of the law
of adaptation of the crown by cusp addition. These principles are equally
well illustrated in the lower molars.

The lower molars (Fig. 47, Al, HI, etc.) are relatively more pro-
gressive, because (1) not only is the broad heel or talonid well developed,
but on this heel the three typical cusps, hypoconid, hypoconulid, and
entoconid, are in some cases present. (2) Another progressive character
is that the usually elevated trigonid has already been modified in one
type (Fig. 47, II) by the degeneration of the antero-internal cusp or
paraconid, although possibly the small size of the paraconid may be a
primitive character. (3) These lower teeth are of two types ; first, the
secodont or tuberculo-sectorial with elevated triangle or trigonid and a
lower heel or talonid, as illustrated in Fig. 47, HI, Al, El; second,
a slightly more bunodont type, in which the trigonid has become
secondarily depressed although still above the talonid. This secondary
depression is carried much further in certain Primates and Ungulates
of the basal Eocene. All the other lower molars present unusual
features. The one marked HI has an extremely high paraconid, a feature
noticeable, though in a less degree, in the two marked Al. On the other
hand in No. II. the paraconid is small and the metaconid is higher
than the protoconid.t In the upper molars, No. 7 may be regarded
either as a fourth premolar of the type seen in the Creodont, Pseudop-
terodon minutus, or as a molar of the type seen in the Oligocene
insectivore Ictops tJwmsoni (Fig. 66). A noticeable feature of all the
upper molars is the indentation in the middle of the outer border of
the crown, as in Peralcdes (Fig. 12), Kurtodon (Fig. 13), Dryolestes
(Fig. 14), and many modern trituberculate insectivores (Figs. 65-80).

SPECIAL REFERENCES.

Osborn, H. F., "Fossil Mammals of the Upper Cretaceous Beds," Bull. Amer.
Mm. Nat. Hist., Vol. V., 1893, pp. 311-330.

*[A marked advance upon the condition seen in the Upper Jurassic Dryolestes,
where the paracone was large and medio-external instead of at the anterior tip of the
triangle. — ED.]

t [The metaconid is almost directly internal to the protoconid and connected with it
by a transverse cutting crest, as in Amblothermm and other Upper Jurassic genera. — ED.]

ORDINAL TYPES OF MOLARS: INSECT1VORA 117

[NSECTIVORA.

The first statements to be made are : (1) while Insectivora show
a persistent and prevailing tritubercnly or triangular arrangement of the
upper and lower cusps, it is also true that (2) the mode or sequence
of origin of these upper cusps, and the homologies of different cusps
and of different portions of the upper molar crowns are uncertain.
According to the evidence presented by Mivart (Proc. Zool. Soc., 1868)
and Gidley (quoted below pp. 12:5-126) in the trigonal molars of
Centetes, ChrysoMoris and other Zalambdodonta, the main cusp is
homologous with the paracone or paracone + metacone of other animals,
the protocone being represented by the internal ledge marked hij in
our Figs. 68, 69, B, c.

Among extinct and modern Insectivora, owing apparently to the
persistence of very primitive feeding habits, we observe also the
persistence of very primitive stages of tritubercular evolution in
the molar teeth. The lower molars of certain Eocene and Oligocene

FIG. 63. Skull and dentition of the Hedgehog (Erinaceus europcvus), a typical dilambdodout

insectivore.

Insectivores, such as Apternodus and Micropternodus, are practically
identical in pattern with those of the Jurassic Amphitlierium, Amblo-
therium, and Dryolestes, presenting an extraordinary instance of persistence
of type.

Insectivores retaining these primitive tritubercular molars have
been distinguished by Gill as Zalambdodonta (with a single external
lambda or crescent), whereas the more specialized Insectivores with
quadrate or subquadrate molars, such as Erinaceus, have been termed
Dilambdodonta (with two external lambdas or crescents). There is
no doubt that the quadrate or dilambdodont condition is secondary
because we have, first, analogy with other groups in which the
quadrate is always found to succeed the tritubercular form ; second,
there is the direct evidence in the unworn molars of JSrinaceus, and
especially of G-ymnura, and some of their fossil predecessors, as shown

118

EVOLUTION OF MAMMALIAN MOLAR TEETH

FIG. 64. A, top view of facial portion of skull and A' upper teeth and palate of Proterix loomisi,
a primitive Erinaceid from the Upper Oreodon Beds, Lower Oligocene, showing molars of

tritubercular origin.

After Matthew.

77?'

' FIG. 65. Upper molars of a primitive recent
member of the Erinaceidje (Gymnura sp), showing
clear traces of the tritubercular pattern. Xote
the upgrowth of the hypocone. From a specimen
at Yale University, x ^.

FIG 66. Upper jj-l- mS of letups
thomsoni from the Titanotherium
Beds, Lower Oligocene. These teeth
somewhat resemble those of the
Mesozoic genus Dryolestes (Figs. 14,
207, No. 2). xf. After Matthew.

FIG. 67. Upper and lower teeth of Ictops acutiilens, from the Titanotherium Beds, Lower
Oligocene. A member of the family Leptictidas, which family, according to Matthew, might
"without serious straining of relationships be included as a primitive sub-family of Erinaceidw,
with which they agree well enough in skeleton and in most skull characters." In accordance
with their primitive condition they have tritubercular molars, with a rapidly developing
hypocone. The molars of the modern Giriiinv.nt are of the same general type (cf. Fig. 65).
x -2. After Matthew.

ORDINAL TVI'KS OF MOLARS: IXSKCTIVORA

19

ms. pa.
me. \

mis. pa-

nic.

mts. me. pa.

pr.

FIG. 68. Molars of tritubercular Insectivores : A, "Chrysochloris," ///-right; B. Potamormle, in'2
right (after Allman, Tmns. Zoo!. Hoc. , ISiitj, pp. 1 - t seq.) ; Centetes sp. (All enlarged.) According to
Gidley the outer cusps marked ps, jm, mt , i»tx, are stylar cusps, the cusp marked ///• is really the
paracone or paracone + metacone while the ledge-like internal projection marked //// in A and
indicated by the anterior basal upgrowth in C is the so-called protocone, the pusterior IKISM!
upgrowth being the hypocone ; in B the cusps marked pi, iiil are almost certainly the para- and
metacones respectively. See Addendum. IP. 225.

PS.

mts.

pa. ins.

pa.

ps.

pa. me.

>lts.

pi-.

pr.

ins1

mts.

nits.

ps.

nits.

Fie. ti'.i. Possible but uncertain homologies (according to the tritubercular theory) of the true
molar cusps in Zalambdodont (A-E) and Dilambdodont (/•'-//) Insectivora, and Cheiroptera (/).
A. Centetes; B. Ictops thoumfini (after Matthew). C. "Chrysochloris"; D. Solenodon (after Dnl^son),
lSsS'2. E. PotiniiO'jule (after Allman). F. .l/.i/"""'' ; G.Galeopithcct'x; H. Proscnlops (after Matthew).
/. Suctinomus brasiliemis, a bat (after H. Allen, IM'H). P'or another interpretation of the cusps in
A see the text. In E especially the homologies of the cusps are very uncertain.

Postscript.— December, 1006. An apparent solution of this puzzling problem is given on
page 225.

The abbreviations in Figs. 68, 69 are incorrect and should be disregarded. The
true interpretations are given in the text.

9

FIG. 69o. Cheek teeth of Zalambdodont Insectivores and certain bats. From Gidley.
{All figures except No. 9 three times natural sizi.)

No. 1. Potamor/alc— Left upper jaw (No. 124327, U.S.N.M.) ; habitat, Africa.

•2. Solenotlon— Left upper jaw (No. 2230, U.S.N.M.) ; habitat, Cuba.

3. Centetcs — Left upper jaw (No. 03316, U.S.N.M.) ; habitat, Madagascar.

4. Ericulus — Left upper jaw (No. 1224S-*, U.S.N.M.); habitat, Madagascar.

5. Hi niir, ni, i, .< — Left upper jaw (No. 63310, U.S.N.M.) ; habitat, Africa.
ii. Cli fi/sochloris — Left upper jaw (No. 61686, U.S.N.M.); habitat, Africa.

7. Vfspertilio fuscus— Left upper jaw (No. 62736. U.S.N.M.); habitat, Washington, D.C.

8. Scotophilus kiil/li — Left upper jaw (No. 113463, U.S.N.M.); habitat, Philippines.

9. Hai-jiirn-i'/ilidlux — Hight upper jaw. (Outline drawing taken from a plate prepared in 1880

by Wilhelm Peters for a monograph of the bats. This monograph was never
published.)

ORDINAL TYPES OF MOLARS: INSKCTIVORA

121

FIG. 70. Lower jaw and teeth of Apternodus medicevus. From the Oligocene of Montana. Pro-
visionally assigned to the Insectivora, but possibly allied to the Cheiroptera (Matthew). The
talonid is represented by the minute posterior basal spur. A, upper, At, external and A- internal
views. XT. After Matthew. The molars somewhat resemble those of certain Mesozoic trituber-
culates (e.g. Permnus, Fig. 18, Peraspalar, Fig. 22).

FIG. VI.

FIG. 73.

FIG. 71. Lower jaw and teeth of Micropternodus borealis. From the Middle Oligocene of
Montana. Probably allied to the Zalarabdodont Insectivores. x J. After Matthew.

FIG. 72. Upper molars of one of the Centetid;u (Ericulus setosus) viewed obliquely from the
rear,, showing resemblances to the teeth of the Upper Jurassic genus Kurtodon (Fig. 13). Compare
also Fig. 207, No. 2. x |.. From a specimen at Yale University (Peabody Museum).

Fio. 73. Inferior view of skull and teeth of Proscalops mioccenus, a primitive Mole from the
Upper Oligocene of Colorado, x r. After Matthew.

122

EVOLUTION OF MAMMALIAN MOLAR TEETH

in Figs. 64-67, that the hypocone is secondarily developed. Moreover,
even in the quadritubercular Hylomys, Necrogymnurus, and G-alerix,
the third upper molar is trituhercular.1

The remote ancestors of the Zalambdodonta or more primitive division,
with trigonal or triangular molars, have already been described under
the " Insectivora Primitiva" (pp. 22-30). Under the modern repre-
sentatives of this primitive type have usually been included such forms
as Ccntctcs, Soleuodon, and Chrysochloris (Figs. 68, 69). More recently,
on embryological grounds (pp. 209-213), this comparison has been
seriously questioned, and there are a number of authorities who believe
that the homologies proposed between the cusps in the recent Insectivores
and in the Jurassic Insectivores are unfounded. Our most recent
discoveries, however, seem to lend fresh support to the older view ;
these discoveries include a lower Oligocene Insectivore fauna from
Montana, fully described and figured by Matthew.2 Among these
animals we especially note Apternodus (Fig. 70), in which the
lower teeth strikingly resemble those of Amphitheriiim ; and Micro-
pternodus, in which the lower molars resemble those of Solcnodon.
Again the teeth of Ictops tJwmsoni (Fig. 66) remind us of the upper
teeth of Pkascolestes (Uryolesfcs, Fig. 14). A higher stage of evolution
is represented by the lower and upper teeth of Ictops acutidens

pr^/yc*

A

D

FIG. 74. Tuberculosectorial and cainassial lower molars in a Viverrid and Insectivores.
A. Euptires govdoti, a very primitive Viverrid, resembling an Insectivore (crown view of /«,).
x }. B. Cr,,t,tes ccaudaius (crown view of MJ), representing the ancestral form of C. xf.
(', D. Carnassial modification of m2 in an Insectivore, Hemiccntetes mculago.scariensis. C. Crown
view. XT. D. Inner side view. XT- Note the lateral compression of the tooth, enlargement
of pt-'i, pud, reduction of med, tuld, as in the carnassial of Carnivora (Fig. 94, D). All from Forsyth
Major, Philos. Trans., Vol. 185 (18<>4), B. p. 24.

(Fig. 67). Moreover, the teeth of Ericulus setosiis (Fig. 72) suggest
those of the Upper Jurassic (or Basal Cretaceous) genus Kvrtodovi
•(Fig. 13).

From another locality and from a more recent horizon in the
American Oligocene is the form Proscalops mioca-mis3 (Fig. 73), a
primitive mole with rather simple trigonal molars, in which, however,

1 Matthew, W. D., Bull. Amer. j)/«.v. Nat. Hist., Vol. XIX., 1903, p. 228.

2 " The Fauna of the Titanotherium Beds at Pipe stone Springs, Montana," Bull. Amer.
Mus. Nat, Hist., Vol. XIX., 1903.

3 Matthew, W. D., Mtm. Amer. Mm. Nat. Hist., Vol. I., Pt. VII., 1901.

ORDINAL TYPES OF MOLARS: INSECTIVORA 123

the three external styles, parastyle, mesostyle, and metastyle, arc
strongly developed, as in the molars of Paiitolamlda (Fig. 140).

A very interesting fact is that the molar dentition of tin-
('entetidie has in one case followed the same line of carnassial

^-ftas

'W:7 /\^fa"

Tne — -'

FIG. 75. Tritubercular and carnassial upper molars in a Yiverrid and two Insectivores. A.
Ettplcres gouiloti. Crown view, ;/<l. x:*.. B. C'.nti.tcs ti:u tulutiis. Crown view, /«!. x i.
C. Hcmicentcti'S madayascariensis, rn-. x J-. Note the long sectorial modification of the metastyle,
correlated with the blade-like character of the lower molars. In A and B, /»• probably represents
the fused para + metacone, the proto and hypocones being represented by the small basal cingular
outgrowths in B. All from Forsyth Major (op. cit. p. 23), but relettered. (Cf. Figs. (58, I'i'.i, i',:'./.)

specialization as in Oxycena (Fig. 92), and Patriqfelis among the
Creodonta, the upper and lower sectorials becoming highly shear-like
(Figs. 74, 75)*

Discussion of the G-idley- Woodward, Interpretation.

The following interpretation of Gidley supports that of Woodward,
which is presented on pp. 208-213 (see also pp. 226, 227):

' Woodward l found that in Centetes and Ericulus the main internal
cusp, usually termed the protocone, was first to develop, but he believed
this cusp to be the paracone, the whole tooth representing only
the antero-external triangle of such a form as Talpa, the protocone
and metacone not having been developed. This, as stated by Wood-
ward, is a modification of Mivart's view published in 1868," in which
he states his belief that in Centetes, Chrysocldoris 3 and like forms, the
main portion of the crown represents the union of the two external
prisms of Talpa and like forms. According to Mivart, the main
internal cusp of Centetes, Ericulus, Chrysochloris, etc. [Fig. 62 a, Nos. 3,
4, 6], was derived by the fusion of the paracone and metacone,
while the protocone and hypocone are wanting or rapidly diminishing
in size and importance. According to both Woodward and Mivart,

* [In Erinaceus also, p4 is a generalized sectorial with depressed antero-internal cusp
and oblique shear, and niL similarly has an enlarged oblique protoconid-paraconid shear
and reduced heel. — ED.]

1 Proc. Zool. Soc. London, 1896, 588-589.

-Journ. Anatomy and Physiol., Vol. II., 139, 1868.

a " The form figured by Mivart has since been removed to a distinct genus, Bematisc.us,
Cope, Am. Nat., XXVI., 1892, 127. The typical Chrysochloris upper molar has 110 trace
of a protocone."

124 EVOLUTION OF MAMMALIAN MOLAR TEKTH

therefore, in these forms, which have been considered typical trituber-
culates, the outer cusps are developments of the cingulum, while the
main internal cusp has been wrongly termed the protocone and is in
reality the paracone, according to Woodward, or combined paracone
and metacone, according to Mivart, while the inner cusp (protocone)
is greatly diminished in size or lias entirely disappeared. These two
authorities, therefore, are agreed on the two points of principal
importance regarding Centetes and Erwulus, viz. : (1) the location of the
paracone in the main internal cusp and (2) the ultimate loss of the
protocone. I strongly concur in these views, for in a series of upper
molars, including Potamogale, Solenodon, Centetes, Ericulus, Hemicentetes
and Chrysochloris (see Figs. 1-6, PL IV. [our Fig. 69 «]), the stages
suggesting the gradual diminishing and final disappearance of the
protocone are very complete, amounting almost to demonstration, and
there can be little doubt that the molars of the Centetes and Chrysochloris
type have been derived from forms similar to that of Potamogale,
involving the loss of the protocone. In consequence of this the
paracone, or combined paracone and metacone, conies to be the principal
inner cusp. In Potamogale [Fig. 69 a, No. 1] the protocone is quite
prominent and still typical in form, while in Solenodon [No. 2] it is
much reduced and is beginning to divide transversely, or more pro-
bably is beginning to separate from a likewise reducing hypocone.
This is in favour of the view held by Mivart that the simple inner
cusp in Potamogale and like forms is in reality the fused protocone
and hypocone. The reduction is carried still farther in Centetes [No. 3],
in which two inner cinguliim-like cusps appear, one on each side of
the enlarged paracone. In Chrysochloris [No. 6] and Hemicentetes the
inner cusp (protocone and hypocone) has entirely disappeared.

" Regarding Mivart' s ' fusion theory,' I am inclined to believe that
Woodward has not given due weight to the evidence cited by Mivart
and that there is considerable support for this theory to be found in
the modern bats and insectivores. Mivart considered the Potamogale
molar as an intermediate form between molars of the Talpa type,
having two external triangular prisms, and those of Centetes and
Ericulus [No. 4], having only one such prism. He pointed out that
in Potamogale there is ' a very interesting approximation of the
triangular prisms,' in which the paracone and metacone, although
still remaining distinct, are in very close juxtaposition. This view is
strongly supported by a series of bat molars to which Mr. G. S.
Miller has kindly called my attention. In this series, which includes
Vespertilio [No. 7], Seotophilus [No. 8] and Harpioc&phalus [No. 9], are
suggested the successive steps from Talpa to Potamogale in the insectivore
group. Vespertilio represents the normal or more generalized form,

ORDINAL TYPES OF MOLARS : INSECTIVORA 125

in which the protocone is large, the paracone and inetacone are
widely separated, and the external styles are nearly equal in size.
The mesostyle is much reduced in Scotophilus and is drawn inward,
the paracone and inetacone are more closely oppressed and the protocols
is somewhat shortened. In Harpiocephcdm * the mesostyle has dis-
appeared, the parastyle and metastyle have drawn closer together ami
compose the entire outer portion of the crown, while the paracone and
inetacone are closely approximated, forming the greater part of the
inner portion of the crown, the protocone being very much reduced.
Thus in Harpioceplialus a stage is reached nearly analogous to that of
Potamogale, the principal difference being that the inetacone is the
dominant cusp instead of the paracone, as in the latter genus.

' From these comparisons it seems reasonably clear that such forms
as Oentetes, Ericulus and Chrysochloris have attained a secondary or
pseudo-tritubercular form by passing through some such stages of
evolution as are suggested by the two series here selected. Other
examples of a fusing paracone and inetacone and reducing protocone
may be found in the molars of some of the creodonts and carnivorous
marsupials and in the sectorials of many of the carnivores.

' From the foregoing it now seems to be demonstrated beyond question
that the main inner cone of Centetes and Ericulus is not the protocone
as observed in normal groups, but, if not entirely made up of the
primary cusp (paracone), it at least involves that element, and
Woodward's contention that the evidence of embryology is in entire
harmony for the molars and preniolars is not controverted by these
seeming exceptions as supposed by Osborn."

Homologies. The reasons the cusps of the superior molar teeth
of Insectivora are difficult to homologize are: (1) that in certain types
the strong development of the styles on the external cingulimi tends
to confusion with the paracone and metacone, as also in the case of
certain Cretaceous mammals (e.g. Didelphops, Figs. 47, F, E], of certain
Amblypoda (e.g. Coryphodon, Fig. 141), and in Marsupials (Figs. 58, 59);
(j!) the strong development of the paracone in certain types tends
to confusion with the protocone; (3) in order to secure a greater
number of sharp piercing cusps the mesostyle divides t into two

1 "The skull of Harpiocephai «.$ from which this description was taken was obtained by
Mr. G. S. Miller through the kindness of Oldfielcl Thomas, of the British Museum.

Unfortunately it came too late to be photographed and figured uniformly with the
series. Its place is taken on Plate III., by an outline drawing from a figure for Wilhelm
Peters' Fledermause des Berlines Museums fiir Naturkunde (a projected monograph of
the bats)."

* [Gidley, J. W., "Evidence bearing on Tooth Cusp Development," Proc. Waxhhif/ton
Acad. Sci., Vol. VIII. , 1906, pp. 93-95.]

t [In certain Marsupials and Insectivores the presence of two distinct mesostyles may be
a primitive character. — ED.]

126 EVOLUTION OF MAMMALIAN MOLAR TEETH

distinct cusps in Myo;/ale (Fig. 69 F}, and the intermediate conules-
(pl, ml) are very prominent in Galcopithecus (Fig. 69 G). It is these
confusing elements which have led to the proposal of such diverse
homologies by different authors. Mr. Gidley interprets the cusp
homologies as follows: In Fig. 69 A-I the cusps marked ps, pa, ms,
ms", me, mts are all peripheral or stylar cusps, none of them
homologous with the paracone and metacone of other mammals. In
A, C, D the cusps marked pr are really paracone or paracone + metacone,
the protocone being represented by the internal ledge marked hy.
In E-I the ledge-protocone is fully established. In E the cusps
marked pi, ml are really the paracone and reduced metacone. On
this interpretation A, C, D, instead of being primitive, represent a
secondary simplification by fusion of the paracone and metacone.
This interpretation also harmonizes with the probability of an inward
displacement of the para- and metacones in F-I, as in many genera,
(Cf. Figs. 67, 75, 83, 85 (m3, m2), 90, 94, 116, etc.) Again, the molar
pattern of Ictops acutidcns tends to connect the tritubercular pattern of
Ictops thomsoni, with the quadritubercular pattern of Erinaceus and
higher types, but it does not justify the identification of the main
internal cusp as homologous in modern Zalambdodont and Dilambdodont
Insectivores.

The teeth of Insectivora, therefore, require thorough re-examin-
ation on the basis of the tritubercular theory to test these disputed
points. (See Addendum, page 225.)

For those to whom the original materials are not accessible, the
following references to works which contain numerous figures will be
especially valuable :

SPECIAL REFERENCES.

FIGURES OF TEETH OF INSECTIVORA.

Bronn, H. G., Klaus, u. Ord. d. Thierreich's, Bd. I., pp. 202-211, 233 (Galeopithecus).

Dobson, G. E., Monograph of the Insectivora, Pt. I. 4to. 1882.

Owen, E,., Odontography, 1840-45.

Tomes, C. S., A Manual of Dental Anatomy, 1898.

De Blainville, Osteographie des Mammiferes. (See esp. " Cladobates," Tupaia.)
Paris, 1839-64.

Schlo?ser, M., Die A /en. . . . Insectivoren . . . des Europaischen Tertiary etc.
Wien, 1887-90.

Leche, W., Zur Morphologic des Zahnsystems der Insectivoren. Anat. Anz., Bd.
XII., 1897, p. 513.

Leche, W., Zur Entwickelungsgeschichte des Zahnsystems der Saugetiere. Stuttgart,
1895.

Matthew, W. L)., "The Fauna of the Titan otherium Beds at Pipestone Springs,
Montana," Bull. Amer. J//w. Nat. Hist., Vol. XIX., 1903, pp. 197-226; "A Fossil
Hedgehog from the American Oligocene," Bull. Amer. Mus. Nat. Hist., Vol. XIX.,
1903", p. 228 ; " Fossil Mammals from the Tertiary of North-eastern Colorado,"
Mem. Amer. Mus. Nat. Hist., Vol. I., Pt, VII., pp. 375-376.

Fn:. 76. Upper and lower teeth of Palu'miiKi/m veterrima, an Iiisectivoi-e frota tlie \Vasati-h
Formation, Lower Eocene. Wyoming, the teeth showing resemblances to the teeth of the most
primitive Creodonts. x j . The upper molars have the internal cusp somewhat depressed, a
feature emphasized in the Creodonts (Figs. 84-90).

f.l.'m. \

s

Fin. 77. Lateral, inferior, and superior views of the skull of Ifyopsfxlug pui'l".*. fiMin the
Wasatch Formation, Lower Eocene Wyoming, x ^. This animal was long thought to belong
to the Eocene Primates, but Dr. Wortman (op. <•;/. p.in-.') has adduced sum'- evidence for its
removal to the Insectivora, and Dr. Matthew regards it as a very generalized Eutherian. The
teeth are of the omnivorous-frugivorous type and in a general way conform to the ideal
ancestral pattern of the molars of Primates and Ungulat. -~.

128

EVOLUTION OF MAMMALIAN MOLAR TEETH

FIG. 78. Upper teeth of Hyopsodv.s marshi from the Bridger Formation, Middle Eocene. Note
the quadrate contour of the molars, the well-developed intermediate conules and hypocone.
(The opposite dental series converge too much in the figure.) (Cf. Figs. 77, 79.)

X|.

FIG. 79. Lower teeth (crown view) of Hyopsodus Icmoinianus, from the Wasatch Formation,
Lower Eocene. Note the absence of the paraconid, the presence of a well-developed postero-
median cusp, the hypoconulid on mlt «z3. (Cf. Figs. 77, 78.) xi.

TTL 3

FIG. 80. Teeth of Diaeodexis (Hyopsodus) laticuneus, a genus now referred to Hyopsodv.s.
From the Wasatch Formation, Lower Eocene, xf.

ORDINAL TYPES OF MOLARS: CHKIROl'TKII A

129

CHEIROPTERA.

The following quotation may be made from the writings of an
author (Oldfield Thomas) who is in no way committed to the tri-
tubercular theory; it becomes especially clear by reference to the
preceding section on the Insectivora :

"The earliest Bats, or rul;i-ochimptera \vuuld liave been cuspi-
date-toothed and insectivuroiis like their ancestors the terrestrial
Insectivora. Among them there would presently have arisen a
form like Harpyia, fruit-eating but still with cuspidate teeth and

*••

FIG. 81. Skull, crown view of uppi.-r to-th and internal view of right upper canine of Pteralopex
i/t,-nta Thomas, a Fruit-Bat, which, instead of having cusp-less cheek teeth as in other Mega-
cheiroptera, has retained cuspidate teeth more or less suggestive of the tuberculo-sectorial type.

x L. After Oldfield Thomas.

no doubt markedly ' tuberculo-sectorial ' premolars and molars.
Then, while the modern J In /•/>// i" would have arisen in one direction
by the reduction of the incisors, in another there would have
followed some form like Ptcralopex, still retaining to a certain
extent cuspidate teeth. Then the cusps would have more and
more tended to disappear, the result being Pfi rn/n^ and its allied

i

130

EVOLUTION OF MAMMALIAN MOLAR TEETH

genus, of which some few species (e.;/. Ptcropus aneiteanus and
leucopterus, and Cynoptcrus) retain remnants of the ancient cuspi-
date structure, while others (e.g. Pteropus coronatus, . . .) have
lost all trace of molar cusps." Oldfield Thomas, Proc. Zool. Soc.,
1888, pp. 474-475.

In the treatises especially of Harrison Allen1 and of Matschie1
we have examples of the teeth of Cheiroptera studied respectively

FIG. 82. Superior view of the right half of the lower jaw of a Microcheiropter, ArtH* us
pcrspicillatus, showing the progressive change in nij, r».2 of a cuspidate into a "basin-shaped"
molar, x£. After H. Allen.

from the trituhercular standpoint and from the older standpoint in
which there was no especial attempt to establish tritubercular homo-
logies. The very interesting and important work of Matschie on
the Fruit Bats develops particularly the secondary formation from a

A

B

hy. pr.

Fid. 83. Progressive stages in the development of the hypocone in Microoheiroptera. A. Tri-
tubercular stage (Atalnpha cinerea). x i. B. Transitional stage (Promops perotis californicus).
X i. C. Quadritubercular stage with strongly-developed hypocone (Nyctinomus brasilicnsis). x |-.
After H. Allen. Possibly the series may be reversed, and we may be witnessing in the stages
C, B, A, the secondary loss of the hypocone, caused by the shortening of the skull and
tooth row.

tritubercular crown of a basin-shaped crown with a depressed centre
and a rim surmounted with irregular cusps, analogous to that which
we find among the most primitive Multituberculates. If all the

1 See titles under "Special References."

ORDINAL TYI'KS OF MOLARS: CAKNIVORA 131

stages leading to this secondary basin-shaped formation arc c;ird'ully
examined it will be found that they invariably point back to a more
symmetrical tritubercular arrangement. A similar case occurs in the
Microcheiroptera in the genus Arfil^ii* (Fig. 82), where the first and
second inferior molars ' have become basin-shaped and irregularly cuspi-
date remotely resembling the Microlestes molars (p. 102). The nearest
relatives of this genus are plainly tritubercular.

According to Mivart and Gidley certain bats show a secondary
simplification of the molar pattern by fusion of the para- and metacones
as in the supposed case of the Zalambdodont Insectivores.

In general, therefore, the Cheiroptera definitely support the tri-
tubercular theory, since the molar teeth are clearly derivable from
simple tritubercular types such as we find among the Insectivora.

SPECIAL REFERENCES.

Bronn, Klassen u. Ord. d. T/tierreichs, Bd. I., pp. 211-2:23.
Owen, R., Odontography, 1840-45.
Giebel, C. G., Odontographie, 18."..").

Schlosser, M., Die Affen . . . Clriropteren . . . dex Europaischen Tertiurs,
etc. Wieu, 1887-90.

Allen, H., "A Monograph of the bats of North America," Bull. U.S. Nat. Mus.,
No. 43. Washington, 1893, p. 198, 38 pll.

Matschie, P., Die Megacheiroptera des Berliner Museums fur Naturkunde. Berlin,
1899. (This monograph contains 14 fine lithographic plates on the skull and
body-form of the Fruit-Bats, Pteropodidse.)

CARNIVORA.

The carnivorous quadrupedal placenta! mammals may be divided
into three grand divisions, namely :

The Creodonta, which were primitive Carnivores of Cretaceous
and Lower Eocene age, the special peculiarity of which is that the
carnassial or specialized cutting teeth in the upper and lower jaws
are not the same as those in the true Carnivora. In the true
Carnivora invariably the fourth upper premolar and first lower
molar are transformed into carnassials ; whereas in the Creodonta
one or more of their first, second, and third upper and lower true
molars may transform into carnassials. The second great division is
the Fissipedia, including the modern terrestrial Carnivora, which
have been termed Carnassidentia by Wortman, in reference to the
distinctive possession of sectorials formed out of the fourth upper
premolar and first lower molar. The third great group is the
Pinnipedia, the water living forms, in which the teeth have been

JH. Allen, "A Monograph of the Bats of North America," Bull. U.S. Xat. Mus.,
No. 43, 1893, PI. V.

132

EVOLUTION OF MAMMALIAN MOLAR TEETH

secondarily modified for effective prehension rather than for the fine
cutting or mastication of the food, which consists principally of fishes.
Here we find the haplodont and triconodont forms secondarily
attained.

Creodonta. In general the earliest or basal Eocene Creodont
molars are very important because they present types of upper and
lower teeth which are transitional between those of the Insectivora
Primitiva, the trigonodont Insectivora, the upper Cretaceous mammals,

P.! f.J /.# m.i m.z »:.J

.j m.2 m.i p. 4 p. 3 p.2 p.i

KM;. S4. Typical trituberculatc molars in very primitive Creodonts. Left upper figure,
Tricuites eubtrigonus ', right upper figure, Chriucus bntdirini ', central figure, Chriacus truiico.tus ',
all Oxyclsenids from the Torrejou Formation, Basal Eocene, Stage II, New Mexico. Lower figure,
Triisodon keilprinianus, family Triisodontidrc ('? Mesonychidse), Puerco Formation. Basal Eocene,
Stage I, New Mexico. All xi.

and those of the higher or more specialized Creodonta and Fissipedia.
Many early types of molars referred to the Creodonta closely resemble
certain molars in the Upper Cretaceous, but the two distinctive
features are that they almost invariably present (1) an internal
cingulum (wanting in the Upper Cretaceous mammals) and (2) a
more or less well developed hypocone rising from this cingulum.

It was among these Lower Eocene Creodonts that Professor Cope
discovered his great generalization as to primitive trituberculy, the
upper molars being universally tritubercular, while the lower molars
are universally tuberculo-sectorial.

ORDINAL TYPES OF MOLARS: CARNIVOUA

133

Molar evolution in the Creodonts follows three general lines : h'rst,
development of carnassial teeth as an adaptation in the purely
carnivorous forms (Figs. 90, (.-)l); second,
the persistence of more hlunt trituber-
cular teeth in the omnivorous forms
(Figs. 84, 85) ; third, development of
irregularly low-crowned teeth in certain
very specialized omnivorous forms, as in
Arctocyon and Anacodon (Fig. 86). The

1 i i P ,1 ,1 • -i cusp from the basal cingulum.

low, irregular molars of the third type

are so specialized in Anacodon that the primitive tritubercular pattern

is vanishing; but in the less specialized Arctocyon and the still less

specialized Clcenodon, the tritubercular origin of these teeth is perfectly

apparent.

FIG. 85. Upper teeth of Deltnthi rium fun-
daminis, an < >xyel;enid Creodont from the
Torrojon Formation, Basal Eocene, Stage III.
x 1. The hypocone is developing as a small

B

FIG. 86. C, Crown view, p4.m3 of Anacodon ui-sii.li as, family Arctocyonidfe, of the Creodonta,
from the Wasatch Formation, Lower Eocene, x%. Note the secondary obscurement of the
tritubercular pattern by the upgrowth of the basal cingulum, especially in the region of the
hypocone, the flattening of the crown (compare the side view B), the wrinkling of the surface.
Analogous conditions are seen in the Gorilla and the Bears.

3

FIG. 87. Upper ami lower teeth of Dissacus s<i.ui-(i!h"ii/u'x. a Mesonychid Creodont, from the
Basal Eocene (Torrejon). The upper molars differ from those of the majority of the c< intern) n>r.-iry
Creodonts in the small size of the metueone, the ledge-like character of the protocone; the lower
molars are laterally compressed, subtrenchant, with greatly reduced metaconid. They have
bean cited by L»r. Wortman as favouring the premular analogy theory. (See j'a-e I'll''. )

FIG. 88. Inferior surface of the skull of Mcsonyx uintensis, family Mesouychidaj, of the
Creodonta, from the Uinta Formation, Upper Eocene, xl. Note the bluntly tritubercular
molars.

A

/r~ — ^i. ?~>w- -

& te

tn.2 m I p. 4 p.3 p.2 P '

FIG. 89. Upper and lower teeth of Sinopa o^ixtkutmna, family Hyaenodontidse, sub-order
•Creodonta, from the Wasatch Formation, Lower Eocene, Wyoming, showing the sectorial modi-
fication of ml, in- and 7H2, )»3. A, Superior view, upper teeth ; B, Outer side view, lower teeth ;
( ', Superior view, lower teeth, x ^. After Matthew.

ORDINAL TYI'KS OF MOLARS: CARNIVO1IA

135

FIG. 90. Skull (partial) and dentition of Palaxmictis occidentalis, a, short-faced Creodont from
the Wasatch Formation, Lower Eocene. 1. Side view (xp, showing enlargement and sectorial

modification of p*, ml (x£). 2. Front view (xi). 3. Palatal view, upper jaw (x-i), showing
tritubercular molars, sectorial modification of the posterior side of jj4, mi, reduction of //(-.
4. Internal view, upper and lower teeth, .showing sectorial modification of /<•», ml, and MI, m.,.
•i. Superior view of lower teeth, showing tuberculosectorial character uf „/,, m%.

Evolution of Carnassial Teeth in Creodonta and Fissipedia.

1. Lover Carnassials in Fix*ij>i-</i«.

In order to thoroughly grasp the principles of transformation of a
trituhercular into a sectorial type, let us first examine such a series
of the lower teeth as are represented in Fig. 95.1 In A and B
we see the external and internal aspects of the tuberculo-sectorial

1 From Osborn and Wortman, " Foss. Mamm. Wasatch and Wind River Beds/'
Hull. Amer. Mus. Nat. His/., 1892, p. 99.

136 E VOLUTION OF MAMMALIAN MOLAR TEETH

crown, in which the trigonid or triangular arrangement is still
perfectly apparent and the broad heel or talonid is still per-
sistent in these three characteristic cusps. In the types repre-
sented in C and I) the heel has begun to degenerate. In E the
talonid has become very simple, the inetaconid is degenerating, fthe
trigonal arrangement has been lost because the paraconid is rotating

FIG. 91. Upper and lower teeth of a Creodont, Hitienml.on. Carnassial teeth, — — (first and

second upper, and second and third lower molars). The perfected carnassials -^ are shaded.
After Matthew.

FIG. 91 Upper and lower teeth of a Creodont, O.ma MI. Carnassial teeth — . (First upper

IK*

and second lower molar). After Matthew.

outward and forward, — lending support, by the way, to the cusp
rotation theory (p. 32 (6)). In F (Dinidis) the paraconid is nearly
subequal with the protoconid, and the metaconid is vestigial, while the
talonid is still further reduced. In G (Fclis) the metaconid is altogether
wanting, the talonid is vestigial, the paraconid and protoconid are in the
same fore-and-aft line, of nearly equal size, and sharply lophoid or sectorial.
The i-irniassicd specialization in this case involved the development or

ORDINAL TYPKS OF MOLARS: CARNIVORA

137

transformation of two cusps (paraconid and protoconid) at the expense or
degeneration of four cusps (metaconid, hypoconid, hypoconulid, entoconid).
Carnassial specialization in general involves the great reduction
of a number of elements. It is thus diametrically the reverse of
certain types of herbivorous specialization, in which there is a
constant increase of the number of elements.

Fio 93. Upper and lower teeth of a Fissipede or true Carnivore, the Lion. Carnassial
teeth '*—, (fourth upper premolar and first lower molar). After Matthew.

pa.

me

pa., me.

D

FIG. 94. Evolution of the sectorial upper and lower molars in the Creodonts. A. Typu-.il
tritubercular upper and tuberculoscctorial lower molars in 7Y<V.,,, ./.,,/ heilprinianus, family
Mesonychidse, Puerco Formation. Stage I., Basal Eocene. B. S<'."v>" opigtJwtoma, family Hyit-no-
dontida'. Wasatch Formation Lower Eocene. Note: the shearing modification of the posterior
side of the upper molar and of the anteruinternal side of the lower molar, the small si/.e of the
taloiiid, the reduction of the metaconid. ('. O.fini.-im fordpata, familj- Oxy;i'nid:e, Wasatch
Formation. Note the further accentuation of the characters mentioned under K. D. Ptir<i<in,>
dasyuroides, family Hyasnodontida:, Upper Eocene and Oligocene, France. Xot<-: the con-
crescence of the metacone with the paraeone, the loss of the metaconid, and (almost) of the
talonid (hyl), the anterior shifting of the protocone, the lateral compnjssimi ,,f tin- ]..\vor to,.tli.
E. Hi/itiiini'1,1 Iwrridus, family Hysenodontidse, White River Formation, Oligocene. Completed
carnassial modification, resulting in long shearing blades, in the upper sectorial o imposed of the
paracone, vestigial metacone and enlargi-l m.-tastyle ; iu the lower sectorial of the enlarged para-
conid and protoconid. All the figures represent the first upper molar (./-1) and the second lower
molar (nt.2). From Scott and Osborn.

138

EVOLUTION OF MAMMALIAN MOLAR TEETH

- E

- — D

— C

FIG. 95. Inferior teeth of various Creodoiits (A-h) and true Carnivores (F, G), showing tlu-
homologies of the cusj'S of the specialized lower carnassial tooth. A. Pulaonictis gigantea, Lower
Eocene, France, outer view, showing tuberculo-sectorial molars, with a small low talonid. B.
Palcfonictis occidentcUis, Wasatch Formation. Lower Eocene, Wyoming, inner view. C. Ambloe-
tonv.s xinosus, Wasatch Formation. Worn teeth, inner view, showing reduced talonid (?)• D.
The same, outer view. E. Milk teeth of Pttti-io/ilis, family Oxysenida:, sub-order Creodonta,
(inner view). Bridger Formation, Middle Eocene, Wyoming. Note the lateral compression of
the teeth, the enlargement of the protocoiiid and paraconid, the reduction of the metaconid.
F. Dinictis felina, mie of the Machserodont Felida;, sub-order Carnivora Vera (Fissijiedia), White
River Formation, Oligocene. Inner view showing blade-like character of the tooth, enlargement
and separation of the paraconid, reduction of the metaconid and talonid. G. Fclis concolor
(Puma). Inner view. Note vestigial character of the talonid (me4), disappearance of the
metaconid.

i
i

2. Convergence of Upper Carnassials in Creodonta and Fissipedia.

The independent evolution of the carnassial teeth among Creodonta
and Fissipedia affords the most distinctive and interesting example of con-
vergent evolution, whereby similar adaptations arc reached from dissimilar

ORDINAL TYPES OF MOLAKS: CARXIYOKA

139

In i / innings, so that if we did not know the intermediate history \\«-
would be entirely misled. This results from the fact as noted above,
that, in the Creodonta, teeth becoming upper sectorials are chiefly the
molars, whereas in the Fissipeclia they are invariably the fourth upper
premolars ; although the initial pattern of the upper molars and of the
upper premolars is different, the crown being composed of cusps some
of which at least are not homologous with each other, the result of
adaptation is to make these two teeth appear to be entirely similar.
(Series I., II., Fig. 96.)

protocam liy/iocon?

protocone

e- - f^jf ^ V

*

protocone

FIG. 96. Convergent Evolution of G'aruassials in Creodonta and Fis.sipedia. Arranged by Dr.
W. D. Matthew from specimens in the American Museum of Natural History. These series
represent morphological but not direct evolutionary sequences.

I. (Upper row). Creodonta. First upper true molar of the right side evolving from the tri-
tubercular into the carnassial type in the Hyienodontidse.

A1. Deltutherium of the Basal Eocene. Generalized tritubercular molar with three primary
cusps (pf., pa., i/ii.), a rudimentary hypocone, and two external styles (/«'-*., mts.).

A'1. Sinopn op<stl<i>tn,,,a of the Lower Eocene. Forward shifting of the protocone (/»'.), back-
ward prolongation and cutting shape of the metastyle (nits.'), reduction of the parastyle (y<».<.).

A3. Sinopa trhitice of the Middle Eocene. The progressive changes described under A- more
strongly accentuated; also incipient reduction of the protocone (pr.), and its approach toward
the paracone (pu.), but especially the approximation of the paracoue to the metacone.

A*. Ptfi-oilon of the Lower Oligocene. Still further accentuation of the above tendencies,
namely, approximation of the metacone to the paracone (me., pa.), reduction of metacone (/,« .),
enlargement of metastyle (mtu.), reduction of parastyle (pas.), auteroversion of protocone (pr.)
and its approximation to the paracone.

A5, lliiif.niii,!,! of the Middle Oligocene, representing the final stage. Protocone (pr.) reduced
to a mi- re- ciiisulum, paracone and metacone (pa.., me.) completely confluent, metastyle (;„?.<.)
greatly elevated, laterally compressed, giving us a cutting blade laterally compressed, composed
mainly of two elements, namely, jiiii-in'une+iiietaeoit.i:, and metastyle.

II. (Lower row). Fissipedia. Superior fui'i-t/i premolnr evolving from tritubercular into
carnassial form in Palseouictidif , Miacidse, Canidse.

B1. Palaonicti*. A Creodont of the Lower Eocene. Generalized fourth premolar type, with
protocone, deuterocone, and cusps corresponding to parastyle and metastyle.

JS2. Did>/mictis of the Basal Eocene. We note the forward shifting of the deuterocone, the
reduction of the parastyle, the enlargement of the inutastyle.

£'•': Ditplictiiuis, a dog of the Lower Oligocene. Forward shifting of the deuterocone, still
further reduction of the parastyle, elevation of the protocone and of the metastyle.

B*. Cuiiis, a recent dog, showing the vestigial deuterocone and parastyle, and the subequal
metastyle and protocone.

140

EVOLUTION OF MAMMALIAN MOLAR TEETH

Na.-

P.m

Fin. 97. Skull and jaw of Diniclis syualidens, a primitive Machaerodont Felid from the Oligo-
cene of Colorado. Note the position of the carnassials ( /i4, ntj) in relation to the areas of insertion
of the muscles of mastication. The action of the lower teeth on the upper teeth is from behind
forward and upward, x y. After Matthew.

Fio. 98. Inferior view of the skull of Pl<l'« »•<">.' !• ucosh us, an ancestral Raccoon (Procyonid)
from the (Upper) White River Formation, Oligocene Colorado, showing the subsectorial modifi-
cation of j>4, the crushing character of -,/ji, ml, the absence of ntf. X {-. After Matthew.

ORDINAL TYPES OF MOLARS: CARNIV01IA

141

pa.

m.l

m.2

FIG. 90. Crown view of the worn upper cheek teeth of a gigantic Amphlcyonine Canid
Borophagus (ii'H HI, from the Upper Miocene of Texas, showing tritubercular molars (ml, m%), and
carnassial p4. the enlargement of the paraoonc (/>'(.) is apparently secondary, xj. After

Matthew.

Fie;. 100. Skull aud jaw of Mi'it'ln oinni'm, an extinct species of marten (family Mustelidw),
from the Upper Miocene of Colorado. -Note especially the carnassial teeth, which as in the
Felidas have been developed in the fourth upper premolar, and first lower molar. Note also the
reduction of the first upper molar and the absence of the other post-carnassial teeth, another
analogy with the Felidse. x 1. After Matthew.

p.3
\~-V-*

.f

FIG. 101. Inferior view of the skull of Buncelwwt l<i!i<>/>li"t/i'i<, a primitive Musteline, from the
Oligocene of Colorado. Note especially the carnassial modification of p\ the reduction of );ii, the
practical disappearance of m~. x|. After Matthew (1902).

142

EVOLUTION OF MAMMALIAN MOLAR TEETH

me.

pa.

m.2

m.l

pr.

de.

Fig. 102. Crown view of the right upper cheek teeth of the Binturong (Aretictis binturong),
showing the loss of the third molar, the great reduction of the second (»;-), the simplification of
the first (/»i) by the loss of the metacone (mi), and the reduction of the protocone (pr), the con-
vergence in form between i/(i and p4, the greater or less flattening down of the crowns of p^ — mz ;
all this apparently in adaptation to frugivorous diet. The nearest allies of the Binturong, the
Paradoxures, have more normal teeth, while the most primitive, more carnivorous members of
the family Viverridre («.</. <i> n< tta) have tritubercular upper and tuberculosectorial lower molars.

FlSSIPEDIA.

The tritubercular and tuberculo-sectorial teeth of primitive Procyon-
iclse, Mustelidte, Viverridte, Ursida?, Canida3, Felidre, and Hysenidte are
too well known to require special emphasis. In general the Procyonida?
exhibit the bunodont, tritubercular type. The Ursidae exhibit a
depressed and secondarily bunodont, quadri tubercular and tuberculo-
sectorial dentition ; as seen by comparison of the more sectorial molars
of the Polar Bear, Thalassarctos, with the more depressed molars of the
relatively omnivorous Black Bear. The teeth of the bear were originally
more elevated like those of the dog, then secondarily elongated, and
finally depressed and irregularly tubercular.

Degenerate Types.

Adaptation to different habits has given rise to a great variety of
secondary modifications. For example, (1) to the flat or even basin-
shaped crown of Gercolcptcs, (2) to the degenerate tuberculate teeth
of Aretictis (Fig. 102); the extreme similarity between p4 and m1 in
this animal, the enlargement of the paracone and the reduction of
the metacone, the ledge-like appearance of the protocone, make the
resemblance which has arisen between these teeth analogous to that
which has arisen in Mesony.r, — namely it is a case of convergence.1

In general answer to the analogy argument (p. 215) it has been
shown in the case of the sectorials of Creodonta and Fissipedia above,.
that (a) final similarity of form is no indication of derivation from
homologous parts ; (b) the exact similarity of p* and m1 in Aretictis
(Fig. 102) is another case of independent or convergent evolution, or the
production of analogous crowns from non-homologous cusps (see pp.
138, 139); (c) fruit-eating habits (Aretictis is said to be frugivorous)
frequently lead to the degeneration or aberrancy of the molar crowns

l[But see also the indications in Figs. 14, 66. 67, 69a, 76, 84, 85, 105, 116, 117,
118, 131, 139, etc., that the similarity between p4 and m1 is not wholly due to convergent
evolution. — ED. ]

ORDINAL TYPES OF MOLARS: PINNIPKDIA 143

(p. 10.')); 00 the enlargement of the paraeone and reduction of the
metacone occurs also in Dinocyon (Fig. 99), in the Mesonychid Creodonts
(Figs. 87, 88, and p. L'lO), and is probably secondary: (••) no fossil
ancestral viverrines are known with such aberrant molars.

M'KCIA L REFERENCES.
Figures of Molars of ( 'arnivora.

Bronu, H. G., Klass. u. Orel. <!. TiderreicKs, Bil. I., pp. 17!)-i'Oi'.
Owen, R., Odontograpky, 1840-1845. See especially for histology of the teeth.
Giebel, C. G., Odontograpfiie, 1855.

Schlosser, M., Die Aff<'/i . . . Carnivoren de* Enrnpiiischeii Ti-rfft'irx, etc. Wien,
1887-90.

PlNNIPEDIA.

The teetli of the aquatic Carnivora, or Pinnipedia, are so much
modih'ed secondarily that until we trace their ancestral history we
cannot feel any confidence in attempts either to homologize the cusps
or to trace these teeth back to a tritubercular or triconodont stage.
Weber l adduces much evidence in favour of their derivation from the
Bears. In that case the ultimate derivation of their molars from the
tritubercular type would be obvious.

As figured above, Phoca gichigensis (Fig. lOo A) exhibits a tooth
analogous to that of the Triconodonta among the primitive Marsupials
(Fig. 6), that is, with a main central and two lateral cusps. We
have seen that somewhat similar molars with several cusps in a fore-
and-aft line have evolved secondarily out of tuberculo-sectorial molars
in the case of the Marsupial Thylacymis (Fig. 58, II. g) and of the
Creodonts Mesonyx (Fig. 87, lower teeth) and Hycenodon (Fig. 91).

Phoca, vitulina exhibits a type more suggestive of Zcuglodoii
(Fig. 194, p. 191), the central cusp is less prominent, and the lateral
and posterior cusps more elevated and connate. According to Dr.
J. A. Allen,'2 within this single species, Phoca vitulina, there is a wide
range of individual and sexual variation in the cheek teeth, as
regards the number and position of the cusps, the size of the crown,
its position in the tooth row, etc., so that we may infer that at
least in this species a high rate of evolution or variation is even
now in progress, while in the past the evolution has obliterated the
primitive pattern of the teeth. In Phoca f/rcenlandica the lower molars
are suggestive of the tuberculo-sectorial type.

The third type, Zalophus calif or nianus, the Sea-lion (Fig. 103 0),
presents in the upper molars at least an example of the secondarily
haplodont croini, in which the main outer cusp is probably the paracone,

1 Die Sauyetiere, 8vo, 1904, p. 551.

'-Bull. Amer. Mus. Nat. ///*/•., Vol. XVI., 1900, pp. 468-470.

144

EVOLUTION OF MAMMALIAN MOLAR TEETH

which has been developed at the expense of the degenerating . inner
cusps, which included originally an elevated protocone.

The secondarily triconodont cheek teeth of the Leopard Seal
(Ogmorhinus leptonyx] are shown in Fig. 42*.

SPEC I A L REFERENCES.

Bronn, H. G., Klassen und Ordnungen des Thierreic/i's, B<1. I., pp. 105-108.

Owen, R., Odontography.

Tomes, C. S., A Manual of Dental Anatomy, 1898.

De Blainville, H. D., Osteogruphie des Mammi. feres.

•>>>'>

Fio. 103. Secondary evolution of triconodont and baplodont types in Pinnipedia. Internal
view. A. Phoc'a gichigensis (family Phocidae or Earless Seals). B. flioca ritulhia (Harbor Seal).
C. Zalophus California us (California Sea Lion, family Otariidse or Eared Seals). Allx i.

PtODENTIA.

We naturally look among the brachyodont, short-crowned types
of Rodents, such as the squirrels and mice, for the ancestral form
of Eodent teeth. Matthew 1 and Osborn 2 have hypothetically traced
the Kodents back to a lower Eocene ancestor in the family Mixodec-
tidae ; Osborn has gone so far as to call these animals Proglires,8
whereas Wortman 4 has revived the view expressed by Cope that

1('A Revision of the Puerco Fauna," Bull. Amer. Mas. Nat. Hist., Vol. IX., 1897,
pp. 259-323.

2 " American Eocene Primates and the Supposed LJodent Family Mixodectidfe," Bull.
Amer. Mus. \at. Hist., Vol. XVI., 1902, pp. 203-213.

3 Recent studies by Matthew indicate the nearer affinity of these animals to the
Insectivora, with possible relations to the Lemuroids.

4 " Studies of Eocene Mammalia in the Marsh Collection, Peabody Museum," Amer.
Journ. Sci., Vol. XVI., Nov. 1903, pp. 345-352.

ORDINAL TYPES OF Mnl.AKS: IIODKXTIA

145

these animals are not Rodents l»ut Lemurs related to the aberrant
order represented by the living Cheirom//* or Aye-Aye. The evidence
for either of these antagonistic hypotheses is by no means final. It'
these animals are truly primitive Rodents or I'roglires, they settle the
tritubercular question so far as Rodents are concerned, bemuse the
teeth are typically tritubercular above and tuberculo-sectorial below
(Fig. 104). Pending the positive discovery of the remote ancestor-
of the Rodents, it may be said that the most primitive existing and
fossil forms of brachyodont rodents, as the Eocene and Oligocene Ischy-
romyidse, Sciuridse, exhibit apparent traces of the tritubercular pattern
in the molar teeth.

FIG. 104. Tritubercular molars in the " Proglires," possibly related to the Rodents. Upper
figure, an upper molar of Olbodotfg copei from the Torrejon Formation, Stage II. Basal Eocene,
showing a primitive tritubercular crown, with a hypocone growing up from the basal congulum.
Lower figure, lower jaw and teeth of Mi.codectus pungcns also from the Torrejon Formation,
showing enlarged incisor and tuberculo-sectorial molars. See note 3, page 144. x ?.

SlMPLICIDEXTATA.

The animals included within the sub-order Simplicidentata (i.e.
with a single pair of incisors, as contrasted with the Duplicidentata,
or Rabbits and Hares) are traced back by Tullberg in his great
monograph,1 to an ancestral type in which the molars exhibit four
cusps. Schlosser2 also has figured a morphological series of upper and
lower molars, showing the probable stages of evolution from the bunodont
tritubercular (?) molars of Ardomys to the hypsodont complex molars of
Hystrix. Forsyth Major, on the contrary, who has made an exhaustive
study of the teeth of Rodents, regards the teeth of primitive squirrels,
which are apparently tritubercular, as secondarily derived from a
polybunous form by the loss of certain cusps. It is more consistent

1 Ueler das System der ^"ayethiere. Upsala, 1S99.

2 "Die Differenzierung des Saugetiergebisses," Biol. Ctntrti/1,/., Bd. X., Nr. 8, 1890,
S. 251.

K

146

EVOLUTION OF MAMMALIAN MOLAR TEETH

with the evolution of the mammalian molar teetli in general to
suppose that the primitive Oligocene Sciurid of Fig. 105 actually
exhibits a tritubercular crown above without the hypocone, and a
(juadritubercular crown below, in which the paraconid is vestigial.
(Compare Fig. 106.)

Among the rats and mice we observe a secondary cusp addition

O i/ j.

and evolution, which has obscured the primitive tritubercular pattern
it" such existed, closely parallel or analogous to that of the multi-
tubercnlates in the development of three parallel rows of tubercles
above and two parallel rows of tubercles below. (See pp. 102-104.)

FIG. 105. Left upper cheek teeth of a primitive Squirrel, Sciurus (Prosciurus) rctustus from
the Titanotherium Beds (Lower Oligocene) of Montana, the molars showing apparent derivation
from the tritubercular pattern, x * After Matthew.

Fio. 106. • Lower jaw and teeth of Gyiiin<>/ii.>icl<n.< minimus, a primitive Sciuromorph (?) from
the Titanotherium Beds of Montana ; the molars apparently indicate derivation from the
tuberculo-sectorial type. MI seems to show a vestige of the paraconid. x A. After Matthew.

FIG. 107. Upper teeth (crown view) of Mylarmulus monodon (Cf. Fig. 108), showing the small
size of in1, i/i'2, the great development of p$. In the rootless hypsodont teeth the enamel is
folded into longitudinal "lakes.'' All traces of a simpler crown pattern have been lost, as in
many other Hystricomorphs, Castoridse, etc. x jj. After Matthew.

ORDINAL TYJ'KS OF MOLARS: RODKXTIA

147

FIG. 108. Facial portion of skull, and lower jaw of M/ilngaulus monoi/on, from the Upper
Miocene of Colorado, a Sciuromorph Rodent (family Mylagaulidse), in which the true molars are
reduced and the fourth premolar greatly enlarged (Cf. Fig. 107). x J . From Matthew.

Mylagaulus,
Middle Miocene.

Meniscomys,
Upper Oligocene.

Lower Miocene
(After Sinclair).

Hdplodontia
(the Sewellel or Moun-
tain Beaver). Recent.

ml.

' pr.

FIG. 1011. Traces of trituberculy in the fourth upper premolars of the Rodent families
Mylagaulidte and Haplodontiidas. " Osborn's molar cusp nomenclature is used as a matter of
convenience, not necessarily implying homology with the cusps of the true molars." (Matthew.)
If the fourth upper premolar in these families has been derived from a tooth with the trituber-
eular/om, the first molar was probably tritubercular inform and in origin. After Matthew.

FIG. 110. Upper check teeth of E-utypomys tliomsoni, an Oligocene Castorid Rodent from
the Oreodon Beds, allied to the Beavers, but with less hypselodont and more primitive teeth,
which apparently retain traces of the protocone, hypocone, paracone, metacone, and of the
external styles (pas, >nx, mix), x-3.. After Matthew.

148

EVOLUTION OF MAMMALIAN MOLAR TEETH

Fin. 111. Side view of the skull and jaws of Eutypomys thomsoni (cf. Fig. 110) showing sub-

hypsodont cheek teeth.

FIG. 112. Skull of the Rat (Mus rattux), illustrating the type of dentition characteristic of the
Myomorph Rodents. The motion of the jaw is proa], i.e. from in front upward and backward, x j-.
After Matthew.

DUPLICIDENTATA.

In the Duplicidentate Koclents (or Eabbits, Hares, and Pikas),
otherwise known as Lagomorpha, the crown has undergone a complete

p. 2 3 4 m'.l 2 3

FIG. 113. Skull of Palaolagus intermedius, an ancestral Hare from the Oligocene of Colorado,
illustrating the upper dentition of the Duplicidentate Rodents, x y. After Matthew.

metamorphosis into hypselodoiit, columnar and rectangular form. These
specialized adult crowns show no trace of separate tubercules or of
anything but a strictly transverse pattern, yet where the permanent
teeth are unworn, especially among certain fossil species, we find

ORDINAL TYPKS OF MOLARS: RODENTIA

149

more or less evident traces of the primitive pattern. The milk
teeth also still retain vestiges of a triangular pattern with three
main tubercles, internal to which is a deep notch or internal folding.
This notch in the adult teeth extends entirely across the crown,
forming the double transverse ridge which so deeply divides the
crown that the anterior and posterior moieties have been supposed
by Marett Tims l to represent the fusion of two originally separate
elements. This notch, however, is entirely secondary.

The ancient pattern of the molars is bomologized by Forsyth
Major 2 himself with that of the molars of Pdycodus, a primitive and
strictly tritubercular primate ; it certainly suggests as the ancestral
condition a triangular to quadrate, low-cusped, brachyodont, three-

7376

D

pr.

FIG. 114. —Apparent traces of trituberculy in Lagomorpha (Duplicidentata). After Forsyth Major.

A. Upper molar of a primitive Oehotonid (Lagomyid) Tltanoiwis fontanitcsi from the Middle
Miocene of Em-ope. Note : (1) the beginning of the groove on the internal side, which in the
later Hares, etc., sinks inwards and divides the crown into two portions ; (2) apparent vestiges
of the tritubercular pattern.

B. Anterior view of the same tooth. Note the reduction of the external roots, the hyper-
trophy of the internal root, the spreading of the enamel upon the anterior and inferior portions
of the crown.

C. Lcpus cuniculus, milk molar 1, showing retention of much less specialized condition than
in the adult.

D. Second upper molar of a Flying Squirrel (Ptertmvys i»,l<iaotis) to illustrate the ancestral
type of molars from which the specialized type in the Duplicidentata molars proabably arose.
(Cf. Fig. 115.) All figures »..

rooted tooth, with the protocone centrally placed. This type of tooth
is associated with omnivorous or insectivorous habits and a chiefly
vertical motion of the mandible, as compared with the oblique to
fore-and-aft grinding motion and herbivorous diet to which the adult
teeth of the Duplicidentata are adapted. Thus the milk teeth of even
this highly specialized group of Rodents revert to a tritubercular
pattern.

Palseontological evidence leads to the same conclusion. If the
known fossil American ancestors of the Hare be arranged in chrono-
logical order (Fig. 115), it is seen that as we go back in time the
molars are less hypselodont, the broad band of enamel which is con-
fined to the inner side of the teeth in Lepus spreads out over the
whole tooth, and in the earliest stage the vestiges of two roots are

*Journ. Anat. a. Phyxiol., Vol. XXXVII., 1903, p. 144.

-"Oil Fossil and Recent Lagomorpha," Trans. Linn. Soc. Loud. ('2), VII 1S99
pp. 433-520.

150

EVOLUTION OF MAMMALIAN MOLAR TEETH

clearly seen on the buccal or external side of the tooth, which in
some specimens have large and well defined alveoli. Two morpho-
logically older but geologically more recent stages are figured by
Forsyth Major (Fig. 114). Taken in connection with the fact that
the crowns of these teeth show a vestigial triangular pattern and in
comparison with the molar of Sciuroptrrus xanthipcs (op. cit., PI. 39,
Fig. 15), and with the fact that the hypertrophy of the lingual
portion of the tooth is secondary, little doubt remains that we have
here a tooth with three originally subequal roots disposed in a
triangle, and a trigonal and probably tritubercnlar crown.

Thus both divisions of the Bodentia, when studied from the ana-
tomical or comparative zoological standpoint, or from the standpoint

A

C

FIG. 115. — Evolution of hypselodont rootless crowns in the upper molars of American ancestors
of the Hares (Leporidse), showing (1) the disappearance of the external roots of the tooth, (2) the
limiting of the enamel to a broad band on the inner side. From Matthew.

A. Paltvolayus brachyodun. Titanotherinm Beds, Lower Oligocene.

B. Palcuolagus turgidus (less worn). Oreodon Beds, Middle Oligocene.

C. Palceolagus turgidus (more worn). Oreodon Beds.

D. Palceolagus inter medius. Leptauchenia Beds, Upper Oligocene, the crown is now rootless,
the enamel is becoming confined to the anterior face.

E. Lepus americanus. Recent. The enamel is now confined to the anterior face. xf.

of the milk teeth, or from the standpoint of palaeontology, apparently
lead back to a tritubercular, trigonal pattern of the crown. As among
all the other divisions of the mammalia we confidently predict the
absolute demonstration by palaeontology of the derivation of the Eodentia
from trituberculate ancestors. On the other hand, Dr. Wortman l has
shown that in certain ancestral Eocene Sciuromorphs (Paramys) in
the conrse of molarization of the inferior premolars, the apex of the
original single-pointed premolar remains in the antero-internal cusp
and not the antero-external cusp of 2h — P±> whereas in the Ungulata,
Carnivora, Insectivora, Primates, and probably other orders, the primi-
tive tip or true protoconid is in the antero-external cusp. Hence, if
the true molar cusps are homologous with similarly placed cusps in
the premolars (see pp. 195-200) the so-called protoconid or antero-
external cusp in the molars of brachyodont Rodents is not homologous
with the similarly placed cusp of other orders. Dr. Wortman also

1 Amer. Jour. Sci., Vol. XV., June, 1903, pp. '216--218.

ORDINAL TYPES OF MOLARS: TI LLODOXTIA

151

points out that the hypsodont and folded molars of Stencojibcr, Palccocaxtm-
and Castor, may have been derived from the more brachyodont teeth
of Sciuranis, which in many respects is closely related to Paramys.
' In like manner Mysops and Sciu-mn/v afibrd the stem types from
\vliich both the Hystricomorphs and Myomorphs were in all probability
derived/' All of this, if proved, would show (1) the derivation of all
Rodent molars from the brachyodont Sciuromorph type, and (2) that
the nomenclature of Trituberculy could only lie applied to the molar
cusps of Rodentia as a matter of convenience not as indicating homologies
with similarly placed cusps in other orders.

SPECL I /> REFERENCES.
Owen, R., Odontography, 1840-45.
Giebel, C'. G., Odontographie, 1855.

Broim, H. G., Klas*. K. Ord. des Thierreidts, Bd. I., pp. 150-169.
Tullberg, T., System der Nngethiere. Upsala, 1899.

Schlosser, M., " Die Differeuzienmg des Saugetiergebisses," Biol. CentralbL,
Bd. X., Nr. 8 u. 9, 1890 (especially pp. 250-251).

TlLLODONTIA.

Esthonyx from the Wasatch or Lower Eocene almost certainly
represents an early stage of the Tillodontia.1 The pure trituberculy
of its molars brings this group also in line with the great majority
of early mammals. (Fig. 116.)

mTs

FIG. 11(3. Upper cheek teeth of Esthmu/.:- <i,'-i<tii(enfs, a Tillodont from the Wind River
Formation, Lower Eocene, representing the ancestral pattern from which the Tillotl,, ,-ii<-,,i
teeth must have been derived, x A.

EDENTATA AND TJGXIODONTA.

The teeth of all the known specialized Edentates are so highly
modified not only l»y the loss of enamel, but by the simplification of
the pattern of the crown to a conical or haplodont condition, to a
tubular condition, or to the compressed columns of the (Jravigrade
Sloths, that until recently no light whatever was thrown by comparative
zoology on the ancestral forms of the molar teeth. The first discovery

JSee Wortman, J. L., in Bull. Ame.r. Mm. AT«/. Hint., Vol. IX., 1897.

152

EVOLUTION OF MAMMALIAN MOLAR TEETH

pointing toward such ancestry was recorded by Thomas1 in the vestigial
milk teeth of Orycteropm, the posterior milk molar retaining a simple
crown with two roots.

ff

Fin. 117. A. Skull of Conoryctes comma, family Conoryctidse, order Tseniodonta, from the
Torrejon Formation, Stage II. Basal Eocene, New Mexico.

B. Superior view, lower teeth.

6*. Superior view, upper teeth. The upper teeth show apparent traces of tritubercular deriva-
tion, the lower of tuberculo-sectorial derivation, x j. (From Wortman, after Scott and Osborn.)

FIG. 118. Upper cheek teeth of Owichodectes tissonensis, family Conoryctidie, order Tsenio-
donta, from the Puerco Formation, Stage I. Basal Eocene, New Mexico.

A. Outer side view. B. Crown view. C. Inner or lingual side view. The molar teeth are
obviously tritubercular, suggesting those of the Oxyclrenid Creodonts (Fig. 84). x+. (From
Wortman, after Scott and Osborn.)

It has, of course, been generally assumed that in the ancestors of
the Edentates, the molar teeth were not only rooted and covered with
enamel, but that they possessed a more complicated pattern of the
crown.

1 "A Milk Dentition in Orycteropus," Proc. Roy. Soc., Vol. XLVIL, 1890, pp. 246-248,
PI. S.

ORDINAL TVl'KS OF MOLARS: T.KXJODO.XTA

153

In 18 90 Dr. J. L. Wortman1 of the American Museum expedition
discovered that the animals long1 described by Cope partly as Tseniodonta
and partly as Creodonta exhibited strong resemblances in the skeleton
to the Gravigrade Sloths, there being an especial similarity of structure

PIG. 119. Lower jaw and teeth of Ony<-l,c,:/,<-l(* t;.<« Hie.mi*, family Conoryctidje, order Tfenio-
donta, from the Puerco Formation, the most Creodont-like member of the order Ganodonta ;
showing lower molars of tuberculo-sectorial derivation, x ±. (From Wortman, after Scott and
Osborn.)

FIG. 120. Skull and dentition of Ilfmli/niius ot<mhl< its, family Stylinodontidaj, order T;«nio-
donta, from the Puerco Formation, Stage I , Basal Eocene, showing the enlarged gnawing
canines and other characters pointing toward Psittacotherium and Calamodon. xL (From
Wortman, after Scott and Osborn.)

between Psittacothcrium and Megalonyx. This discovery was a direct
confirmation of the prophetic remark of Dr. Max Schlosser,2 that
certain forms (Esfhonyx, Calamodon, Psittacothcrium), which, on the
one hand, are evidently (Onychodcctes and Hcniinmnis}, related to the

1 " Psittacotheriuni, a member of a new and primitive sub-order of the Edentata,"
Bui/. Amer. Mu«. Xat. Hint., Vol. VIII., 1896, pp. 259--2G2.

2 Quoted by Wortman from Schlosser's "Die Differenzierung des Siiugetier Gebisses,"
Biolofj. Gentralblatt, June, 1890, p. "2~i'2.

154

EVOLUTION OF MAMMALIAN MOLAR TEETH

Creodonts, can, on the other hand, be regarded as the ancestors of
a part, at least, of the Edentates ; to the extent that in such a line
the formation of prismatic teeth out of the tritubercular and tuberculo-
sectorial type can be traced.1

FIG. 121. Side and top views of the skill of PsMacoth<rii<m multifraguw, family Stylino-
dontidre, order Tteniodonta, from the Torrejon Formation, Stage II., Basal Eocene, showing the
enlarged gnawing canines, x ^. (From Wortman, after Scott and Osborn.)

To this group Wortman1 gave the name Ganodonta (equivalent to
Tamiodonta Cope) in reference to their enamelled teeth.

All the early members of this group of Ta-niodonta have tubercu-
late teeth, in which, however, the enamel is so delicate that it rapidly
wears off. In Psittacotherium, for example, the lower teeth are
([uadritubercular (Fig. 123), in Calamodon (Fig. 124) the unworn lower
teeth are crested and exhibit all the five cusps (proto-, meta-, hypo-,
entoconids, and hypoconulid) of the primitive crown, having completely
lost, however, the paraconid. In the remotely related Onycliodcch-x
(Figs. 118, 119) the upper molars are again strictly tritubercular, while

1 Wortman, J. L., " The Ganodonta and their Relationship to the Edentata," Bull. Amer.
J/H.S. Nat. Hist., Vol. IX., 1897, pp. 59-110.

Fui. 122. Front and side views of the jaw of Psittucotltfi-imn multifraiji/ui, showing enlarged
gnawing canines, and molars of tuberculo-sectorial derivation. xA. (From Wortrnan, after
Scott and Osborn.)

FIG. 123. Superior and side view of the second upper premolar and the first and second lower
molars of Psittaeotlnrium multif<-it<inii<. family Stylinodontidfe, order Tteniodonta, from the
Torrejon Formation, Stage II., Basal Eocene, New Mexico. The lower molars may have Ix-en
derived from the tuberculo-sectorial type by the loss of the paraconid. x^. (From Wortman,
after Scott and Osborn.)

P-l p. 2 p'-3 p-4

Flo. 124. Side and superior view of the lower jaw and teeth of Cain motion simples, family
Stylinodontidse, order Tseniodonta, from the Wasatch Formation, Lower Eocene. Note the
greatly enlarged rootless canine, the cylindrical molars with imperfect enamel, xi. (From
Wortman, after Scott and Osborn.)

156

INVOLUTION OF MAMMALIAN MOLAR TEETH

the lower molars are also obviously derived by depression of the
trituberctilo-sectorial pattern.

Opinions differ as to the value of the evidence that these forms
are true ancestors of the American Edentates, but the balance of
structural characters is certainly very strongly in favor of this theory.
If it shall be finally positively confirmed by future discovery, the
American Edentata will also be definitely ranged in the tritubercular
ranks. It should be stated, however, that Professor Scott seriously

we*

ni.2

en"

me''

m-3

Fie. 125. Crown view of unworn lower molars of Culnmodon simplex, family Stylinodontidse,
order Taniiodonta, from the Wasatch Formation, Lower Eocene, x -| . Molars apparently derived
from the tuberculo-sectorial type by the loss of the paraconid (compare analogous conditions
among Primates and Ungulates). (From Wortman, after Scott and Osborn.)

FIG. 12G. Single tooth and fragment of the lower jaw with teeth of Sti/linodon minis, family
Stylinodontidfe, order Tfeniodonta, from the Wind River Formation, Stage II., Lower Eocene,
representing a highly specialized Taeniodont with hypsodont rootless cheek teeth, which have
lost the enamel on the inner and outer sides, and all traces of tuberculo-sectorial derivation, x 4.
(From Wortman, after Marsh.)

questions the supposed ancestral relationship of the Tteniodonta to the
American Edentates, because he says the Tseniodonts are much less
like the Santa Cruz Miocene Ground Sloths than like the descendants
of the latter, the Pleistocene Ground Sloths ; and hence he interprets
the resemblances between Ganodonts and Edentates as an instance
of convergent evolution.

SPECIAL REFERENCES.
Owen, B., Odontography, 1840-45.
Giebel, C. G., Odontographie, 1855.
Bronn, H. G., Klassen u. Ordnungen des Tlrierreich's, Bd. I., pp. 147-150.

ORDINAL TYl'KS OF MOLARS: I'l; I MATKS

157

PRIMATES.

Cope's contention us to the tritubercular origin of the teeth of
Primates rested upon the strongest possible proofs both from com-
parative zoology and from pakeontology. The tri tubercular pattern
is still the prevailing one among the Lemuroidea, while the Antbropoidea

A I

B

C

c. p. 3 P-

c. P-3 p.4

V m:> m, 2 m-3

FIG. 1:27. Jaws of American Eocene Primates, etc., natural sixo.

A. Pelycodus tutus, family Notharctidae, order Primates.

13. Iliinii«i:l,ix /mulus, probably an Insectivore.

''. ATiaptomorphus » mulug, analogous to Tarsius.

D. Microsyops sp., family Mixodectidae, one of the so-called " Proglircs." xi.

radiate from trituberculy into quadrituberculy, and into crested forms.
Osborn's recent revision1 of the American Eocene Primates proves
that the molars exhibit a fundamentally triangular pattern in every
one of the twenty-two or more known species. The various types
exhibit a familiar succession of stages from a more triangular condition

r? <_/

with an extremely rudimentary hypocone, to a quadrate, sexitubercular

1 null. Amtr. Mus. Nat. Hint., Vol. XVI., 1902, pp. 169-214.

158

EVOLUTION OF MAMMALIAN MOLAR TEETH

condition, stages which have already been treated in the evolution of
the human molar teeth (pp. 50, 55).

The special characters of this evolution were brought out in the
same paper. An interesting feature of some of these American
Eocene monkeys 'is that some of them pass from a tritubercular into
a quatlri- and finally into a sexitubercular condition, with a prominent

ml:-

ml:-

Vic,. 128. Superior molars of (A) Adapis maynus, a lemuroid from the Eocene of France;
(B) Hi/opsodus uintrnsis, an Insectivore (?) from the Eocene of North America ; and (C) Notharctvs
sp., a Primate, from the Eocene of North America ; all three apparently derived from the same
tritubercular ground plan, by the upgrowth of the hypocone. A and B, x^.

hypocone and large intermediate tubercles, closely homoplastic with
the grinding teeth of primitive Ungulates (Figs. 132, 149). Similarly
the premolars progress by the addition of internal cusps.

The normal dentition of man is beautifully illustrated in Figs. 1 and
134, p. 161, taken from Selenka's admirable monograph.1 In modifying
his figure (Fig. 1) we have expressed the tritubercular homologies on
one side of the jaw and the embryonic order of evolution of the cusps
in numerals on the opposite side. In this connection reference may be

FIG. 129. Skull and dentition of a Lower Eocene (Wasatch) Primate Anaptomorphv*
homunculus, with tritubercular upper molars. Partially reconstructed. The premaxillary portion
of the skull is wanting, x^-.

made to the discussions of the relative value of embryological and
palaeontological evidence on pp. 49, 214.

Certain peculiar variations of the human molar teeth may be referred
to here which are often described by anthropologists as anomalies, but
which really are either homogenetic or homoplastic with cusps well
known in the lower mammalia.

Protostylc or tdbercvlus anomalus. On the anterior side of the
protocone in the upper molars we have observed in many of the lower
mammals, especially in the Periptychidse (Periptychus, Ectoconus,

Menschenaffen, Wiesbaden, I'.Mio.

ORDINAL TYl'KS OF MOLAKS : I'KIMATKS

159

Fig. 137), that a special cusp is developed, to which we have given
the name protostyle, from its proximity to the prolix-one.

From a recent paper by I*. Adloft'1 we learn that this was originally
designated by Carabelli as occasionally occurring in the human molars,

Fir,. 130. Upper (A) and lower (A1, A'-) teeth of Anaptomorphus homv.nculus. (Cf. Fig. 129.)
The upper teeth are tritubercular, in the lower teeth the paraconid is seen to be much
reduced, x£.

and hence named by him tnltarnlus anomal/'s. Batujeff regarded this
as a progressive structure, pointing out (1) that the tuberculus anomalus,
while most frequently found on the first upper molar, is also occasionally
found on the second and third molars ; (2) that it is more frequently

FIG. 131. Origin of the hypocone from the basal cingulum, as shown in l,nl , w«,i >,,<//«/-,x, a Basal
Eocene (Torrejon) Primate (?) of uncertain relationship. > y.

observed in higher races than in lower races, to which Adloff adds,
(3) that among recent and extinct anthropoid apes the tuberculus
anomalus is certainly not present.

Adloff' in this connection calls attention to the anomalous detach-
ment of cusp components of the human molar crown as evidence of

1 "Zur Frage nach der Entstehung der heutigen Saugethierzsahnformen," Ztitschrift ./'.
Morphologic u. Anthropologie, Bd. V., pp. 357-384.

160

EVOLUTION OF MAMMALIAN MOLAR TEETH

a process the reverse of concrescence. He concludes (p. 379), that
as each tooth has primitively sprung by concrescence of cusps derived
from successive dentitions, so in incipient retrogression these cusps
break apart again into their original components.

A

m.2 m.3

FIG. 132. Evolution of the upper molars in the Notharctidae, a family of American Eocene
Primates. A, Peli/iMlusj'sviiiroi-tis, Wasatch Formation (Lower Eocene), the teeth showing clear
traces of trituberculy. B, Notharctus nunienus, Wasatch Formation, p* and nii-m3 more quadrate.
C, Notharctus sp., Bridger Formation (Middle Eocene), p* quadrate, with two external cusps;
vnl-m;5 large, with well-developed mesostyle (ms). x £.

p.3 p. 4 m.l m.2

m.3

FIG. 133. Notharctus sp., Bridger Formation, Middle Eocene. Crown view, lower teeth,
showing the loss of the paraconid in ni^m*, enlargement of the postero-internal cusp or
entoconid. x ^.

Apart from this new evidence for the concrescence theory, to
which we do not feel able to attach much weight, the question of the
existence of the progressive evolution of the ' tuberculus anomalus ' or
protostyle in the human teeth is especially interesting, as another
instance of homoplasy , or the independent evolution of an apparently
homologous cusp in different orders (see p. 236).

ORDINAL TYPES OF MOLAKS: I'RLMATKS

K.I

M. HKMIIKXCKS.

Owen, R., Odontography, 1840-4").

Giebel, C. G., Odontographie, 1855.

Bronn, H. G., AV,,.w. u. '>,-,/. d>.>* Tl,n ,•,•<•;, -Us, Bd I., pp. 223-23*!.

Srhlosser, M., Die Affen Lemuren . . . dex Em-npiiischen 7V,v/V//-.<, etc. Wicii, lssT-90.

Selenka, E., Menschena/en, Wiesbaden, 1900.

Topinard, P., " De 1'Evolution des Molaires et Prcmolaires chez les Primates et <-u
particulier chez 1'Homme," L' Anthropologie, No. 6, Nov.-Dec., 1892.

Gaiuhy, A., "Sur la Similitude des Dents de I'Homme et de quelques Animaux,"
V Anthropologie, T. XII., 1901.

Osborn, H. F., "Revision of the American Eocene Primates," Bull. Amer. .!/".•>•.
Nat. Hist., Vol. XVII, 1902, pp. 159-184.

Wortman, J. L. " Studies of Eocene Mammals, ..." Part II. Primates, J //«<•/•.
Jour. Sc. [4], Vol. XV., Mar. 1903, May 1903, June 1903.

FIG. 134. Xormal, ideal, human (Caucasian) dentition, based upon a photograph of a specimen,
modified after several other specimens. From Selenka, after Rose. Traces of trituberculy are
evident in the upper molars ; the lower molars have lost the paraconid. The crowns of the
teeth are bluntly tuberculate, in adaptation to omnivorous diet. • ;.

162

EVOLUTION OF MAMMALIAN MOLAR TEETH

pa. mts,

P?. • me. •'

crista

A1

Jjosffosseffc ,, medifossette

c r'tsta.

pa*

pa*

•ochet

crochet.

pa.

nits.

c/ng.

hi''

m£<* m$<t

me1*

me^ fn^

FIG. 135. Diagram showing typical modifications of the tritubereular pattern in

Ungulates (cf. Fig. 43).

ORDINAL TYI'KS OF MOLARS: TXCl'LATA 163

Eii Ungulata.

S !'!•:< 7 A L REFEllEX<'l-:s.

Dealing with the teeth of Ungulates from the tritubi-ri'tifur *tn adjoint.

Schlosser, M., "Beitrage zur Keinitniss der Starnmesgeschichte der Hufthiere uiul
\Vrsuch einer Systematik der Paar- und Unpaarhufer," Mor/ilml. ././///•/>., 12, Taf. V.,
VI. (see especially the " Allegemeiner Theil. Das Gebiss der Perissodaetylen und
Artiodactylen und seine Beziehungen zu dem der Condylarthra." pp. 97-112).

"Die Differenzierung des Saugetiergebisses," BioL Centr<dl>L, Bd. X., Nr. 8, 1890
(especially pp. 247-250).

Osborn, H. F., Antea, pp.

FIGURE 135.

No. 1. " Sexitubercular " superior and " quinquetubercular " inferior molar of Hyra-
cotherium (Eocene ancestor of the Horse), showing in the upper molar six
main cusps, a parastyle developing from the cingulum, and an encircling
cingnlum, continuous in the lower molar with the hypoconulid. The para-
conid of the lower molar is vestigial or wanting. The crown is low
(" brachyodont ") with low, conic cusps (" bunodont ").

•2. Superior molars of Pantolambda (Eocene Aniblypod). The triangular pattern
of the crown has been retained, the protocone remaining central in posi-
tion ; the outer cusps have become crescentic (" selenoid "), the parastyle
(ps) is extremely large (Fig. 140).

3. Three stages in the degeneration of the paraconid seen in the lower molars of

Amblypoda (Fig. 144).

4. Superior and inferior molars of A. Systemodon (Eocene tapiroid) and B.

Tapirus. The parastyle is seen developing from the cingulum, the inter-
mediates (pi., ml.) are conspiring with the outer and inner cusps to form
the "ectoloph," "protoloph," and "metaloph" (cf. 8) of the modern Tapir
(B). In the lower molars the paraconid is vestigial or absent.

n. " Lophodont" type. Primitive Rhocerotoid molar (Hyrachyus) showing com-
pleted "protoloph," "metaloph," "ectoloph."

6. " Lophodont " type. Modern Rhinoceros molar showing accessory folds,

" antecrochet," " crista," " crochet."

7. " Bunoselenodont " type. Primitive Titanothere (Palizosyops) . Internal

cusps "bunoid," external "selenoid."

8. " Lophoselenodont" type. Primitive (Eocene) Horse, Eohippus. Traces of

the original triangular pattern of the crown are still discernible.

9. "Lophoselenodont type." Primitive Horse (Packynolophus). The "inter-

mediate connles" (pi., ml.) will become crenulate and with the ectoloph,
and the hypostyle, will produce the complex crown pattern of the modern
Horse.

10. J. Upper molar of Ai«_-li />/"•/•//'//*, B. Lower molar of Merychippus, Miocene

horses (cf. Fig. 160).

11. " Selenodont " Artiodactyl type (Protoceras). All cusps crescentic.

12. Ground plan of molars in various Ungulate sub-orders. A. Condylarthra (!)

(Meniscot/terium), B. Amblypoda (Periptychus), C. Perissodat-tyla (Hyraco-
therium), D. Condylarthra (Phenacodus).

164

EVOLUTION OF MAMMALIAN MOLAR TEETH

AMBLYPODA.

The indisputably triangular and tritubercular nature of the molars
of the primitve Amblypoda is demonstrated in the accompanying
figures (Figs 137-146) of several of the lower Eocene types.1 All
of these teeth bear a close eneral resemblance to the tritubercular

Fio. 136. Lower teeth of Ectoconus ditricionux, a Basal Eocene (Puerco), relative of Pcripti/chus.
Compare the lower molars of Euprotogonia (Fig. 149) and Protoyonodon (Fig. 148) among the
Condylarthra. j .

molars of Creodonts. The peculiar and distinctive feature of the
evolution of the upper molar teeth in the Amblypoda is that they
do not pass into a quadri tubercular 01 quadrangular stage by the
forward shifting of the protocone and upgrowth of the hypocone, like
all the other Ungulates, but develop special types of bunodont,
selenodont, and lophodont molars out of the primitive triangle.

FIG. 137. Upper and lower teeth of a primitive Amblypod, l'n-ijil<fchus rhabdodoa, from the
Puerco Formation. Stage I. Basal Eocene, x ^. The upper molars are clearly derived from
the tritubercular type, the lower from the tubei-culo-sectorial type. The upper molars develop
a protostyle (liji), the protocone remaining central in position. (Contrast Phenacodus, Fig. ISO).
The bluntly cuspidate crowns apparently indicate an omnivorous-herbivorous habit.

Fig.

136,

Iii some forms (e.g. Periptychus, Fig. 137, Ectocoi
Conacodon, Fig. 139) the cusps remain bunoid.

In other basal Eocene forms (Pantolambda, Fig. 140) the cusps
become crescentic or selenoid, reminding us strongly of the teeth of
some Insectivora (for example, Proscalops, Fig. 73) on a large scale.

1 [According to Matthew these primitive types belong to the group which gave rise to
the typical Condylarths and Amblypods, but are much more primitive than the typical
Amblypods, and cannot be ordinally separated from the Condylarthra. — ED.]

ORDINAL TYl'KS OF MOLARS: AMI5LYPODA

165

a i

Fig. 13S. Upper and lower teeth of Hcmithln us kowali '••</. lunus, from the Basal Eocene (Puerco).
Compare Figs. 137, 130. x -}-. After Matthew.

Fig. 130. Upper and lower teeth of Conacodon entoconus, a Basal Eocene (Puerco) relative of
I'ti-iptychv.s. x |.. Observe the reduction of the paraconid in the lower molars correlated with
the development of the hypocone from the basal cingulum of the upper molars. The upper
molars remain tritubercular, although the protocone is somewhat pushed forward. (Cf.
s). The cusp marked msd is the hypoconulid.

ints ,

mesostyle

c

D

FIG. 140. Primitive Amblypod, Pantolambda cavil-ictus, from the Torre j on Formation. Stage
II. Basal Eocene. A. Superior molars ; B. Diagram of same ; C. Inferior grinders ; D. Mandible.
Observe the central position of the protocone as in all the Amblypoda, the crescentic cutting
modification of the para- and metacones, the strong development of the para-, nieso-, and meta-
styles. In the lower molars, the paraconid is reduced and there is a tendency to form two sharp
ridges, the first from the protoconid and metaconid, the second from the hypoconid and ento-
conid. (Cf. Cor>i,,h«,l,,,', Fig. 141.)

166

INVOLUTION OF .MAMMALIAN MOLAR TEETH

Out of this three-crescent type the crested or lophoid type of
Cnrt/jihoi/an (Figs. 141, 143) and Uintatherium (Fig. 142) have evolved,
as partially discussed on page S7. As shown in Figs. 140, 141, 144,
145, 146 the loicer molars of Uintatherium are closely linked to those of
Coryphotlon and Pantolambda through the genus Bafhyopsis (Fig. 144-146),

ps.

me.

m

.2

m.3

FII;. 141. Left upper and right lower teeth of Corypliadon tmtis, a hornless Amblypod from
the Wasatch Formation, Lower Eocene. As compared with the molars especially m'A, of Panto-
/iniiiiilii (Pig. 140), observe the great developmentof the ridge from the protocone to the parastyle
(protoloph), the posterior displacement of the paracone, the reduction of the mesostyle, the
reduction of the posterior limb of the V formed by the metacone. In the lower molars observe
the heightening of the anterior and posterior ridges, x £.

Fi<;. 142. A. Second upper molar of Uintathrrnnn from the Bridger Formation, Middle
Eocene. £. Diagram of same. As compared with the molars of Con/phoilon (Fig. 141) observe
that the ectoloph has apparently been rotated posteriorly around the metacone as an axis, while
the metacone itself has approached the protocone, with the final result that the protoloph and
ectoloph diverge toward the external side of the tooth, x J.

which is strictly intermediate in its mandible and inferior molars, and
thus supports the view that the upper molars also of Uintatherium
have passed through stages represented in a general way by Pantolambda
and Coryphodon. The steps in this evolution are the most com-
plicated and difficult to understand, especially the rotation of the
ectoloph, a feature which is less positively demonstrated than the
other features of this exceptional evolution.

SPEC I A L R INFERENCES.

Osborn, H. F., "Evolution of the Amblypoda, Pt. I., Taligrada and Pantodonta."
Bull. Amer. Mus. Nat. Hist., Vol. X., 1898, pp. 169-218.

ORDINAL TYl'KS OF MOLARS: AMBLYPODA

167

paracone

parasty/e'-t

protocone

protoconid

B

metacone

hypocone

hypoconid

^^^^^^^^^^ metaconid ^-— -^ entoconid

FIG. 143. Scheme of upper and lower molars in Coryphodon. (Compare Fig. 141.)

'en*
Uintatlierium.

protoconid hypoconid

paraconid-

-entoconid

metaconid , Entoconid*

metasty/id

me'

Pantolambda.
FIG. 144. Evolution of the lower molars in the Amblypoda. Crown view.

168

EVOLUTION OF MAMMALIAN MOLAR TEETH

m.l

Batliyopsis.

Coryphodon.

m 3

m .3

Pantolainbda.
Firs. 145. Evolution of the lower molars of Arnblypoda. Internal view.

A

p. 2 p. 3 p. 4 m,/ m2

m, 3

FIG. 140. Lower teeth ( x \} and mandible ( x %) of Bi'tltyopsis tissidens, an Amblypod from the
Wind River Formation, Stage II., Lower Eocene. In the structure of its lower molars this animal
is more like Uintathcriwn than Coryphodon. (Of. Fig. 144.)

-CONDYLAKTHKA.

These Ungulates are contemporary with the Amblypoda, but
unlike them their molar teeth evolve from trituberculy into quadri-
tuberculy.

The proof of the tritubercular ancestry of these oldest and most
generalized hoofed animals is furnished by the oldest form Protogono-
don (Fig. 148) and by the diminutive predecessor of Phenacodus, the
species Euprotogonia minor (Fig, 151); in this species the crown
is still triangular, but the hypocone is seen developing on the second

c 2 $ It 1 3

Fro. 147. Upper and lower teeth of Mindu ,<i<.< ii'i-iiiiinx, A primitive Condylarth (?) (family
Miochenidaj), from the Torrejon Formation, Stage II., Basal Eocene. Note the loss of the para-
conid in the lower molars, the bluntly rounded character of the cusps of the upper molars, x | .
Possibly an Insectivore. (Cf. Hfio/tswlug, Fig. 78.) From Osborn and Earle.

ml

FIG. 148. Upper molars, lower jaw and superior view of lower teeth of Protoiionodo,' j,^,^^^.*,
a very primitive Condylarth possibly ancestral to the Phenacoilontidre, from the Puerco Formation,
Stage I., Basal Eocene. (Cf. Figs. 147-151). This animal was apparently intermediate between
the Creodonta and the Condylarthra. (Cf. Figs. S4, 14'.i.) x f . After Osborii and Earle.

FIG. 149. Upper and lower teeth of £uprotorionia pv.trcensis, from the Torrejou Formation,
Stage II., Basal Eocene, a Condylarth more primitive than P In ixirodus (Fig. 150). x1. After
Osborn and Earle.

170

INVOLUTION OF MAMMALIAN MOLAR TEETH

superior molar (A, m2), and in the lower molars (0, ml5 w2) the
paraconid still persists. The intermediate cusps or conules in the
upper molars are well developed.

me

FIG. 150. (Left hand figure). Upper and lower molars of Euprolofionin puercensis, from the
Basal Eocene (Torrejon) representing the ancestral molar patterns of Plunacodus primavv.s (right
hand figures) from the Lower Eocene (Wasatch).

Various species of Protogonodon, Eiiju-ofof/oiiift and Phniacodu* thus
exhibit beautifully the evolution into the sexitubercular, bunodont
superior molar crown, out of which the upper molars of all the
higher hoofed animals have evolved, the crown passing from the more

A

m-3

in.? fit.t dp. 4

FIG. 151. Upper and lower molars of Euprotofionin miiirn; a small Pheuacodout Coudylarth
from the Torrejon Formation, Stage II., Basal Eocene, showing traces of trituberculy combined
witli progressive upgrowth of the basal cingulum, intermediate cu.spules and hypocone in the
upper molars, and with the reduction of the paraconid in the lower molars, xi.

triangular into the more quadrate form. Similarly, the lower molars
lose the paraconid but retain the hypoconulid.

SPECIAL REFERENCES.

Schlosser, M., " Beitrage zur Kenntniss der Starumesgeschiehte der Hufthiere
und Versuch einer Systematik der Paar- und Unpaarhnfer," Morphol. Jakrb. 12,
Taf. V., VI. See especially pp. 97-112.

Matthew, W. D., " A Revision of the Puerco Fauna," Bull. Amer. Mus. Nat.
Hist., Vol. IX., 1897. Especially pp. 293-320

ORDINAL TYI'KS OF MOLAKS: ARTIODACTYLA

171

ARTIODACTYLA.

Of this even-toed group of hoofi.-d mammals the must ancient
representative is found in the lower Eocene or \Yasatdi lieds. It is

eft me p

en" me" pa"

hf /;/

A hf pr<

hlj en" me" Da" en" me"

Fi<;. 152. Lower teeth of 2Y,>/,!../(^..< a very primitive Artiodactyl from the \Vasatch Forma-
tion, Lower Eocene.

A. Triiivnolestes c/tc.censis, paraconid small but distinct. X-3-. B. TI-'«J<>, «//<.</, s metsiocus,
paraconid reduced. x|-. C. Ti'i'.iv<i<>l<.!ttis ttmt</irns, parauonid vestigial or absent, all cusps low
and rounded. xA.

Fn;. 1S3. A quudritubercular to bilophodoiit Peccary, P/<iti/.i<,ntix l>~,cul,-n ,;,1 ,<x. family Dicoty-
lida?, from the Blanco Beds, Pliocene, of Texas, x^. After Gidley. The incipient bilophodont
condition* is seen in Dicoiyti-x and even among Baboons (r//,.

the little species Trigonolestes c/iacciiftis (Fig. 152) of the Wasatch, which
is undouhtedly an Artiodactyl, as shown by the associated astragalus.

* [The more or less completely bilophodont type has been secondarily developed
independently in many other families. Among Artiodactyla we also have Lixtriodon
(SuicUe), Tapirnlus (Anoplotheriidae) ; among Perissodactyls, the Tapirida?, not to mention
the early Equida- and the RhinocerotidaB, Lophiodontidse in which the anterior and posterior
crests are connected by the high external cusps or ci-est. Among the Proboscidea we have
especially Uinotheriidtw and Palceomastodon ; among Pyrotheria, Pyrotlierium ; in certain
other Ungulates, e.r/. Uintatherium, Arsinoitherium, the bilophodonty is 7iot strictly typical.
Among Sirenia we have Manatux ; among Marsupials, the Kangaroo and Diprotodon. Even
certain Theriodont reptiles (Trirachodon) developed a transverse crest in each molar
and the whole series of cheek teeth was thus functionally analogous to a bilophodont
dentition. — ED.]

172

KVOIA'TIOX OF MAMMALIAN MOLAR TKKTH

As the name indicates, the upper molars are triangular ; the inferior
molars exhibit a well defined trigonid, with the paraconid reduced but
still distinct, in fact the teeth are almost tuberculo-sectorial. Comparison
of a series of lower molar teeth of species belonging to this genus
(T. metsiacus, T. ctsayicus) show a progressive degeneration of the
paraconid, depression of the trigonid, and elevation of the talonid :
in other words, the development of a bunodont crown (Fig. 152).

Fin. 154. Upper and lower cheek teeth of Bv.nomtry.f ile<i«.iix, from the Uinta Formation
(Upper Eocene), a small bunoselenodont Artiodactyl. Note the crescentic outer cusps, the large
crescentic intermediates (j>l. ml.), especially the very large mctaconule which is preoccupying the
position usually assumed by the hypocone, which here remains minute (/<//). x 1. After Wortmau.

}?l /

Fio. 155. Upper molars of Anthracotherium Icun.ise, a bunoselenodont Artiodactyl from the
Protoceras Beds, Upper Oligocene, showing the greatly enlarged metaconule, which takes the
place of a hypocone. (Cf. Bunomtryx, Fig. 154.) xi.

Fir;. 150. Principal types of molars among Artiodactyls.

A. Bunodont. All cusps conic. Examples Primitive Suilliiies (Elotherium, Peccaries, etc.).

B. Bunoselenodont. Outer cusps only crescentic, inner cusps conic, or subcrescentic. Example,
Aiithracotheres (A i < <•< » ' / ' s ).

C. Selfiwdoiit. All cusps crescentic. Examples, J\Iu->/coj>otamus, Deer, Antelopes, Camels.
In C'the cusp marked h;/ appears to be the enlarged metaconule. (Cf. Figs. 154, 155.)

In the Eocene Helohyus the molars are fully bunodont, that is,
with low, rounded cusps ; the paraconid is extremely minute. The
series also shows the progressive molarization of the premolars and
lends support to the theory that the molar pattern was originally
triangular, and sharply differentiated from the premolar pattern, and
that the premolars are becoming molariform by adding the cusps in
a different order from that of the molars (see pp. 194, 195).

OKD1XAL TYI'KS OF MOLAIIS: A I!TIO]>A< TV LA

173

g the Oligoeene Suilline or Pig-like Artiodactyls we tind in
Leptochcerus (Fig. 157) persistent tritubercular molars which \vere
mistakenly referred to the Primates.

Thus we have the most direct evidence of trituliercular ancestry
among the Artiodactyls, in which the bunoselenodont, and finally the
purely selenodont types were evolved."*

Fig. Io7. Upper check teeth of /,>/</«(•//(• ,•«.•.• iir-m-ili*, a very primitive Artiodaetyl retaining
tritubercular molars, from the Oreodon Beds, Middle Oligocene, xi. After Marsh.

Another line of evolution is by the formation of transverse crests
sometimes forming a bilophodont crown as in Platyyonus, a Pliocene
Peccary (Fig. 153).

In the bunodont type of Elotln'rimn (Fig. 156 A) the upper molars
show the protocone, paracone, metacone and hypocone, and the inter -

VK-. l.'iS. Upper and lower cheek teeth of _D/V/,.//n<,/' leporinum, a primitive Anoplotheriid
from the Upper Eocene of France, retaining conic cusps in the molars. (Cf. Figs. 147, 157).
About twice natural size. The premolars are laterally compressed as in Eocene Camelidas,
Oreodontidse Protoceratidse, etc. The enlarged metaconnle and the small hypocone (cf. Fig. 154)
is well shown. The upper teeth belong to the milk set, the lower to the permanent set.

mediate conules (-pi, ml}. The lower molars, however, have already lost
the paraconid. In all Artiodactyls the metaconule is very large, often
replacing completely the cingulum-hypocone (Figs. 154, 155).

SPECIA L REFERENCES.

Schlosser, M., " Beitrage znr Kenutniss der Stammesgeschichte der Huftliiere
und Versuch einer Systematik der Paar- and Unpaarhufer," Morphol. Jahrb. 1±
See especially pp. 97-.112.

" [It is an interesting fact that the main internal cusps (pr) of the upper molars and the
external cusps of the lower molars of the most ancient Artiodactyla were not perfectly round
or bunoid as commonly supposed, but subcrescentic, as in the most ancient Perissodactyla
(e.g. Lambdotherium, Eohippus), Condylarthra (Protogonodon), Amblypoda (Hemit/i/n-n-..
Gonacodon, Pantolambda), Primates (Indrodon), Creodonts (Oxyclcenids), Insectivores
(Ii-topx, Dryolestes}. The purely bunoid or round cusped condition is probably secondary,
like the perfected crescentic condition. — ED.]

174

EVOLUTION OF MAMMALIAN MOLAR TEETH

PERISSOBACTYLA.

The probably tritubercular origin of tbe molar teetb of the odd-toed
Ungulates or Perissodactyla has been discussed in the case of horses,
rhinoceroses, and tapirs on pp. 72, 75, 85-87. The key to the

parastyle -
paracone -

protoconule —
proiocone —

protocontd hypoconid

metaconid entoconid

FIG. 159. Upper and lower molar of Hyrocotlteriuin leporinum from the London Clay (Spar-
nacien) Lower Eocene, showing the fundamental ancestral pattern of the Perissodactyl molars,
i.e. Upper molars with four well rounded cusps and two small conules, the rudiments of the proto-
and metaloph, lower molars with four main cusps, lacking the paraconid. and with an incipient
hypoconulid. The hypocone and metaconule are apparently twin cusps, the hypocone not being
derived from the cingulum. xf. After Owen.

rnesostyle

in, Irifstijle

protolojih--\-

I'lttlixtllltll-

KII; 100. The Secondary cusps of an Ungulate Molar. A. Anchithc, :>',„. upper. B. Mery-
is, lower. The Primary cusps are indicated by abbreviations.

origin of the primitive sexitubercular superior molars, from which the
elaborate pattern of all these teeth were evolved, is to be found not
only among the Condylarthra (pp. 168-170), but also in the study of
the various types of teeth in the horses themselves. From this sexi-
tubercular type the crown evolves by a bunoid, lophoid, or selenoid

ORDINAL TVI'KS OF MOLARS: I'KKISSODAf 'TV I.A

175

modelling of the cusps. In the Titanotheres, fur example (Figs. 1G7-
171), the protocone and hvpocone remain Imnoid, the paracone and
inetacone have become selenoid. and the small ridges tunned by th<-

D

E

FIG. 161. Evolution of the upper molars in the Equida;. E. after Kowalevsky. The
series figured is not a phylogenetic one, since animals belonging to several different lines of
descent are represented ; but it is a morphological series and (with Figs. 1(32-1(54) shows thu
principal successive stages of molar evolution in the F.quidse. All figures natural size.

A. Hiiracothenum, Lower Eocene.

R. Pachynolophus, Middle Eocene. Note ps, ms, and crescentie />c, me,

C. Aitchilophus, Lower Oligocene. Note mts and hypostyle Its.

D. Hfi.«>hi{itit'x, Middle Oligocene.

E. Anchitherium, Lower Miocene.

pro to- and metacomiles have become vestigial, the crown thus consisting
of two outer crescents and two inner cones ; but even in this specialized
crown traces of the primitive triangular arrangement of three primary
cusps remain. In the little La mix kit her inm (Fig. 167) we have a

postfosaette = c-

crochet

antecrochet

Fiu. 102. Upper milk molar of M ,-i/r/,i/ilni.< sp. from the Upper Miocene, showing the
completed ground plan of the molar pattern of the modern Horse. (Cf. Figs. liiO, lid.) x |

perissodactyl (probably an ancestral Titauo there), in which the trituber-
cular derivation of the molars is indisputable. In these teeth among
the horses, the styles and intermediate conules d>1, ml) play an important
role (p. 85). The key to the evolution of the teeth of the horses as
compared with that of the Titanotheres is given in Figs. 135, 159,
160. The type attained is lopho-selenodont.

17G

EVOLUTION OF MAMMALIAN MOLAR TEETH

mesostyle

hypocone-—"^

—iantecrochet

D

i
i

FIG. 163. Cusps, Crests, Styles, Crochets, and Fossettes in the molar teeth of the Horse.
The specimens figured belong to a North American Pleistocene species, Equus complicatus.

A. Unworn crown.

B-D. Successive stages of wear. All figures natural size.

The Tapiridne, as shown in Figs. 172-174, develop a sub-lophodont
dentition, the protoloph and nietaloph being well developed, while the
ectoloph does not form a completely united crest.

In the Khinocerotida? we have the extreme lophoid evolution, in
which the crown consists of the completed proto-, meta-, and ectolophs
(Figs. 175-182).

ORDINAL TYPES OF MOLAKS : PKRISSODACTYLA

177

«* ,

®'

m't'ms" "'

Fie;. 1(J4. Evolution of the lower molars in the Equida. Morphological not phylogenetic.
(Cf. Fig. li \1.)

A. Hi/i-'ii-nt/i' i-ium, Lower Eocene.

B. Piichimolophus, Middle Eocene. Note appearance of metastylid (MI.V') by fissure of the
metaconid.

D. M' Koliiiipus, Middle Oligocene. Note appearance of entostylid. (Marked msd).

E. Hipparlon, Pliocene. Note great expansion and posterior fold of the metaconid
//<.<•'), protostylid (psfi), endostylid (i s'l). Worn tooth.

F. Equus. All figures nitural size. Worn.

FIG. 105. Premolar Terminnlogy, proposed by Scott and adoyited in this volume. Primitive Un-
gulate Types. Fourth upper premolar and first molar of A. Euprotoyoiiia, and B. Hyracotherium.

Through all these higher stages the homologies with the primitive
tritubercular crowns can readily be traced.

SPECIAL REFERENCES.

Hatcher, J. B., "Recent and Fossil Tapirs," Amer. Jour. Sci. (4), Vol. L, 1896,
pp. 161-180.

Wortman, J. L., and Earle, Charles, "Ancestors of the Tapir from the Lower
Miocene of Dakota," Bull. Amer. Mus. Nat. Hist., Vol. V., 1893, pp. 159-180.

"Species of Hyracotherium and allied Perissodactjls from the Wahsatcli and
Wind River Beds of North America," Bull. Amer. Mus. Nat. Hist., Vol. VIII., 1896,
pp. 82-110, pi. II.

Osborn. H. F., "The Extinct Rhinoceroses,'' Mem. Amer. J/»x. Nut. Hist., Vol. I.,
1898, pp. 75-164.

Schlosser, M., " Beitrage znr Kenntniss der Stammesgeschichte der Hufthiere mid
Versuch einer Systematik der Paar- und Uiipaarhufer," Morphol, •l<iln-l>. \~2. Especialljr
pp. 97-112.

M

178

EVOLUTION OF MAMMALIAN MOLAR TEETH

de pi '. pi Pr

de

mca

FIG. 166. Upper and lower cheek teeth of several Lower Eocene Equidse. All x 1.

First or upper jiiiun . Upper cheek teeth of Eoltippvs (Protorohippus) venticolus, Wind River
Formation, Lower Eocene, Stage II. Side view.

Second figure. Crown view of the same.

Third figure. Lower cheek teeth of Eohippus cristatus, Wasatch Formation, Lower Eocene,
Stage I.

Fourth (Join rfiiii'n ). Lower teeth of Ef.ihiiniv.it index, Wasatch Formation, Lower Eocene.

All figures natural size. After Wortman.

Observe: (1) The general similarity in the molar pattern to that of the Eocene Tapiridie, as
shown in the development of two transverse ridges in the upper and lower teeth, a feature
destined to be emphasized in the Tapiridre but highly modified in the Equidse. (Of. Figs. 100-
164, 172-174.) (2) The incipient character of the mesostyle (/<(.<.), and the appearance of the
hypostyle (As.) in »ii-ms. (3) The complication of the fourth premolar pi, by the development
of a protoloph from the protoconule and deuterocone, and of a small metaloph from the
metacouule, whereas in p:i, the protoloph is developed from the protoconule only, the meta-
loph from the deuterocone. (4) In the lower molars note (1) the molarization of p4, (2) the
development of an oblique crest running from the hypoconid (/<?/<*.) to the metaconid, (3) the
faintly incipient twinning of the metaconid ; all three features being progressively developed in
the later Horses.

Fio. 167. Upper cheek teeth of Lambdotherium popoagicum, from the Wind River Formation.
Lower Eocene, a primitive Titanothere with molars of tritubercular derivation, x \ . Observe
the subcrescentic character of the protocone,the continuity of the metaconule with the hypocone.

FIG. 168. Lower jaw and teeth of Lambdotherium popotifficuni (cf. Fig. 167). The inferior
molars retain a vestige of the paraconid, the fourth premolar is becoming molariform. ; i.

FIG. 169. Lower premolars of an undescribed species of the family TitanotheriidiB, Dinta Formation,
Upper Eocene, showing progressive molarization of 2h'P+ X j|.

ms

m.r

tip. i

.mts

"ml

Fio. 170. Upper milk molars <ij>1 - >'/<-!, and first permanent molar of a primitive Middle
Eocene Titanothere (Palceosyops major). The external cusps are becoming crescentic, producing a
bunoselenodont type ; the primitively triangular arrangement of the three main cusps is still
evident. The fourth milk molar is seen to be closely similar to the first permanent molar. The
cusps of the milk molar are lettered in accordance with Scott's premolar nomenclature (see pages
195-200), but it is probable that the cusp marked te in rlp3 corresponds to the postero-internal
cusp of dp2 and the aiitero -internal cusp (dc) of i!pl, as in the Equidte (cf. Fig. 166). Unlike the
Equidse the Titanotheres have greatly reduced the proto- and metaconules of the molars, x ^-.

FIG. 171. Upper dentition of a specialized Oligoceue Titanothere (Brontotherium tichoarus).
The molars are now quadrate, the outer cusps sharply V-shaped, x J.

FIG. 172. A. Upper and lower molar of a Lower Eocene (Wasatch) Tapiroid Systemodon

.*' mt/tiaiix (.^)

B. Upper molar of a recent Tapir Tiijiiri'* innericanv.s, showing the completed bilophodont
pattern, x ^-.

is

FIG. 173. Bunolophodont teeth of a primitive Tapiroid (Systcmodon priinn rug), from the
Wasatch Formation, Lower Eocene. Observe the general similarity in pattern to the teeth of
contemporary Equidfe (Fig. 166) combined with a stronger development of the protoloph
and metaloph, a greater obliquity of the tooth as a whole, and a somewhat more central position
and larger size of the paracone. x i . After Wnrtman.

FIG. 174. Evolution of the premolars in Protapims and Tnpirus. After Wortman and Earle.

A. Protapirus simplex. Oreodon Beds, Middle Oligfocene.

B. Protapirus obliquidena. Protoceras Beds, Upper Oligocene.

C. Tupii'us a/inericanus.

In A, premolars 1-4 are all comparatively simple and there is no metaloph.

In B, the internal cusp (deuterocone) of p3, JD-I has split into two and there is an imperfect
metaloph.

InfC, p2-p* are fully molariform. x^. After Wortman and Earle.

medifossetta

postfossette

prefossctti
entecrochet

crocfiet
FIG. 17-x Typical Rhinoceros molar, showing terminology of the crests and folds.

metacone-

metaconule- / — -f

hypocone-

-protoconule

protocone

Fro. 176. Topographic relations of the parts of the typical Rhinoceros molar pattern to the
sexitubercular ancestral ground plan.

faraslytf faraccne mitatone

parastylc paracone- mttacont

ftroColoptt m:tatoph

parastyle paracone ntttatone

protohpk mttalopk

prolobpk

mclaiofk

mctalophid hypolopkid ItypaanuliJ mclalophid hypoloplmt It?

nielaloptuii

liypolopkltt

Tapiroiil. Lophiodout. Rhinocerotoid.

(P rot «p! r >'.<). (L»ii/iic»lori). (Hi/rachyux).

FIG. 177. Typical lophodont molars of Perissodactyls.

In the Tapirnid type the paracone and metacone are subequal or symmetrical, strongly convex
externally. In the Rhinocerotoid type, the paracone is more or less convex, the metacone
elongate, externally flattened to concave. The Lophiodont type is intermediate. It is noteworthy
that in the Middle Eocene or Bridsjer Stage, the molars of the various Perissodactyls are
rather similar. Thus some molars of the Tapirid S;i.<i: i,,,,,l,,,i approach the /'/••<'"<•»/,//>/)/'.•.• type
among Equidas, some molars of the Lophiodont H<.lnl' tea approach the lIiii-<K'l,nii.< type ; while tin-
latest representatives of Hin-nclii/us show some approach toward the Hi/i-i"-" '»<• type among
Rhinoceroses.

182

EVOLUTION OF MAMMALIAN MOLAR TEETH

FIG. 178. Superior molar pattern of a primitive Hyracodont Rhinoceros Hvrachyus agrarius
from the Bridger Formation, Middle Eocene, showing a general resemblance to the Lophiodont
type, from which this differs in the smaller paracone and flatter metacone. The crista as in
lihinoceroses generally is merely a portion of the internal face of the paracone (cf. Fig. 175).

frtnloph

taloplt

protolopli

arlocone

jirot

D

FIG. 179. Molarization of the fourth upper premolar in Ccnwpvs occidcntalis, a Rhinoceros
from the Oreodon Beds, Middle Oligocene as shown in four specimens from successive geological
levels. In A, the protoloph is imperfect, the metaloph irregular, in D the protoloph is complete,
the metaloph fairly well developed. 6 and C are more or less intermediate.

protocone

protocone

tritocone

tetartocone

deuierocone'

tetactocone

FIG. 180. Progressive molarization of the preruulars from pi - />4 in Hyrttchyu^agrarius, family
Hyracodoutidre, a Rhiuocerotoid from the Bridger Formation, Middle Eocene

A

Flo. 181. Four stages in the evolution of the molars in the Rhinoceroses. After Gaudry.

A. Aco-ntherium Umaneiise, Upper Oligocene, Europe.

B. Rhinoceros pachygnathus, Lower Pliocene, Europe.

C. Rhinoceros antiquitatis, Pleistocene, Europe.

D. Elaxinotfierium sibericum, Pleistocene, Europe. After Gaudry and Boule.

We note (1) the progression from brachyodonty to hypselodonty, ('2) from roots to open-pulps,
<3) the increasing verticality of the slopes of the crests, (4) the development of the cement, (5) the
increasing plication of the enamel. Stage A parallels the Stegodonts among Proboscidea
(pp. 180-188). Stage D parallels the elephants, and also many of the South American Ungulates,
and even the rootless grinders of many Rodents.

protocone

-ff metalopMd

h)lpolof)h i ri

FIG. 182. Extreme specialization of the Rhinoceros molar type, in Elasmotherium sibericum,
from the Pleistocene of Siberia. The plicatiun of the enamel has mere IM-.I the rutting surface,
and the efficiency of the tooth for grinding hard dry grasses or shrubs. (Compare the enanu-l
lakes in the Horse molar, Fig. 103). x +. After Gaudry and Boule.

184

EVOLUTION OF MAMMALIAN . MOLAR TEETH

CHALICOTHEKOIDEA OR ANCYLOPODA.

The Ancylopoda are apparently a group of aberrant Perissodactyla
in which the nails or generalized hoofs have become secondarily
modified into large claws. The types of this order are the genera
Macrotherium, Ckalicotherium, and Anci/lotherium, in which the superior
molar teeth exhibit a strong general resemblance to those of

FIG. 1S3. Bunoselenodont upper cheek teeth of Meniscotherium tcrrcerubrce, a Condylarth (?),
from the Wasatch Formation, Lower Eocene. Note the large selenodont protoconule, the oblique
metaconule-hypocone ; in the lower molars, the twinning of the metaconid (me<t), a feature also
independently developed in the Equidas, Fig. 166. x£.

ml-

,-ps.

,pa.

pl.

by-

pr.

Fio. 184. Buno-lopho-selenodont molar (ufl) of Schizotheriwn modicum, an Ancylopod from the
Eocene of France. In the drawing the height and size of the protocone and hypocone is consider-
ably foreshortened.

Titanotheres among Perissodactyla : that is, they are buno-selenodont
(Fig. 184), or, strictly speaking, buno-lopho-selenodont, because they
combine bunoid, lophoid, and selenoid modelling of the cusps. These
molars have evidently evolved from a sexitubercular ancestral type,
and we reason by analogy that they were of tritubercular origin.

In Meniscotherium we have a genus of much greater geological age,
which exhibits a somewhat similar type of grinding tooth (Fig. 183).
It has been considered by Osborn for this and other reasons as a
possible ancestor of the Ancylopoda, though still a member of the
order Condylarthra. However recent observations tend to show that
these resemblances are not indicative of genetic relationship but that the

ORDINAL TYPES OF MOLARS: HYRACOIDKA

LSD

Chalicotheres have more probably been derived from Lower Eocene
Titanotheres.

SPEC! A L R HFERENCE.

Osborn, H. F., "The Ancylopoda, Chalicotherium, and Artionyx," A IDC/: X«tin-<ili*t,
Feb., 1898, pp. 118-133. [The discussion about Art/on//. <• is based upon a font n<>\\
known to appertain to the Oreodont Artiodactyl Agriochcerus.]

HYKACOIDEA.

The Hyraces, including many existing African and West Asiatic
species, also the fossil forms Pliohyrax Osborn, Saghatherium Andrews,
M<'</«lohyrax Andrews, exhibit strong evidences of derivation from a buno-

Fig. 185. Milk dentition of a Hyracoid (Scnjhatln rium antiqaum}), from the Upper Eocene
of Egypt, with buno-lopho-selenodont cheek teeth, xl.

pr.

Fig. 186. Upi>< i- fiijuri , Hiim.i- x/ir incus, upper clieek teeth, external view. x-g-. Lowa- X1"'1 ' •
A, Hiii-n.i- capensis, third upper molar, unworn, x ". B, II. si/riacus, third lower molar, worn.

lopho-selenodont ancestral type. The superior molar teeth (Figs. 1S.~>.
18G), consist of two short, transverse crests, the proto- and metalophs, and
an elongate external crest, the ectoloph. The latter has been eomp;ireil
with that of the rhinoceros, but actually resembles that of tbe

186

EVOLUTION OF MAMMALIAN MOLAR TEETH

horses more closely, because it consists of two halves, a paracone and a
metacone, divided by a faint vertical ridge, the mesostyle ; this ridge
is much more strongly marked in the extinct genera, mentioned above.
It proves that the ancestral types of teeth among the Hyracoidea were
lopho-selenodont, like those of Palceoiherium perhaps, the more remote
ancestors being sexitubercular and tritubercular. The lower molars
exhibit the double crested pattern, similar to that seen in so many
Perissodactyla.

It is an interesting fact that the fossil Hyracoids retain the primitive
double rooted canine, a tooth which undoubtedly (see p. 194) once
belonged to the premolar series.

SPECIAL REFERENCES.

Osborn, H. F., "On Pliohyrax Krupii Osbovn, A Fossil Hyracoid from Sarnos,
Lower Pliocene, in the Stuttgart Collection," Proc. Fourth International Congress
of Zoology, Cambridge, 1898, pp. 173-174, PI. 2.

Forsyth Major, C. J., "Pliohyrax grsecus from Samos,:) Geol. Mag., N.S., Dec.,
IV., Vol. XL, pp. 547-553, Dec. 1899.

Andrews, C. W., "Notes on an Expedition to the FayAm, Egypt, with Descriptions
of some New MammaJs," Geol. Mag., Dec., IV., Vol. X., No. 470, Aug. 1903, pp. 339.

PROBOSCIDEA.

The highly complex, plated tooth of Elcphas is so far remote from
the molar crown of the general type seen in Protogonodon (Fig. 148) or

ft or oc.

a/ fit.

FIG. 187. Side view of the skull of Pclaoumxtodon Icadnelli, from the Upper Eocene of Egypt.
After Andrews. About ^ natural size.

even in Hcmithlceus (Fig. 138) that it might seem hopeless to attempt
to show derivation of the former from the latter ; but the palaeontology
of the Proboscidea renders such a connection absolutely certain.

LSI

ORDINAL TYPES OK MOLARS: l'i;< >!',< )S( 'I I >KA

The gradations leading- back from the plated tt-cth of
through Sta/odon, with its numerous crests and short crowns, to Ma*ft>dnn,
were long ago followed in Falconer and Cuutlry's Fauna A/i//<//">
Sivalensis and in Gaudrv's Enchaincmcnts du Monde Animal/' : but. the*

nmz runs fcm+ mj. m.»- — =gj

* -. -. -t-~ ^r*-~ -r— -»• — ;'"
^/

Fio. 188. Upper cheek teeth and basal view of skull of Puheomastoilon beadnelli, (cf. Fig. IN;>.

After Andrews. Scale, about J_.

recent researches of Andrews have carried the evolution back through
Palceomastodon (Figs. 187, 188) to the quadritubercular molars of
Mceritkerium, a middle Eocene stage representing the African atavus
of the Proboscidea. By analogy with all the other groups we have

ex.oc.

Fio. ISO. Side view of the skull of M<> , itln /-//'//i Im

[Scale, about 1.

o

xi. (Cf. Fig. 100.) After Andrews.

been considering, there is no question that the quadritubercular
molars of Mceritherium (Figs. 180, 190) sprang from an ultimately
tritubercular type.

The quadritubercular molar of the Mceritkerium type by (1)
transverse connection of its pairs of cones, produced a bilophodont
crown similar to the lower molars of the Tapirs; (2) by continuous

188

EVOLUTION OF MAMMALIAN MOLAR TEETH

upgrowth of successive cingula (talons and talonids) it transformed a
bilophodont into a trilophodont crown, a trilophodont into a tetralo-
phodont, etc. Thus, the plates of the teeth of Elqihas owe their
origin to upgrowths of the posterior basal cingulum.

We have many analogies among other hoofed animals with Probos-
cidean molar evolution. Among Suina, Perchcerus has plainly quadri-
tubercular upper molars, and shows the origin of a double trefoil

FKJ. UK). Inferior view of the skull and teeth of Mcerithenum lyonsi, from the Middle Eocene
of Egypt. After Andrews. Scale, about i.

analogous to that seen in Mastodon. The Hippopotamus also shows
a double trefoil. Listriodon of the Middle Eocene of Europe exhibits
lophodont teeth remarkably similar to those of Dinothcrium. The
hindermost molar of the wart-hog (Phacochcerus) parallels the molars of
the Proboscidea in the development of complex many-columned teeth
from the constant upgrowth of the talon posteriori)'.

SPECIAL REFERENCES.

Falconer, H., and Cautlev, P. T., Fauna Antigua Sivalensis, Part I., pp. 1-64,
4to, 1846.

Andrews, C. H., "On the Evolution of the Proboscidea," Fhilos. Trans. Roy. Soc.
Ser. B, Vol. 196, pp. 99-118. London, 1903.

Gau dry, A., Les Enchamements du monde animal dans les temps geologiques
Mammiferes tertiaires, 8vo, 1878, pp. 172-191.

SlKENIA.

The bilophodont molars of the Sirenia (Fig. 191) do not offer any
difficulty to the theory of trituberculy, because of the many other
cases in which bilophodonty has evolved from sexi-, quadri-, and
trituberculy.

Considerable evidence has been adduced for the belief that the

ORDINAL TYPES OK MOLARS: SIRENIA, KT< '.

1 89

•me-

Sirenia are an aquatic offshoot from the Ungulata. I>e lUainville

first made the bold suggestion that they were related to the Proboscidea,

and in 1002 Andrews again directed attention

to the numerous anatomical similarities of these

two groups, reinforcing them by the new evidence

offered by the ancestral Mcerithcriitui, which

closely resembles Prorastoma not only in the

teeth but in the humerus. Lydekker, on the

other hand, in 1892, pointed out the likenesses

between the third and fourth upper milk molars

of Prorastoma vero/n'ti*' and those of the seleno-

dont Artiodactyl Merycopotamus dissimilis.

Of the affinities of the Sirenia and Ungulata in general, as
expressed in the table on p. 75, there can be little doubt, but for
their more specific relationships we must wait for additional evidence.

Fie;. 191. Unworn molar of
American Manatee (Ti-ichechus
inmiiih'x), showing "masked
selenodonty." x^.

SPECIAL REFERENCES.

Andrews, C. W., "On the Evolution of the Proboscidea," Phil. Trails. Roy.
Lond., Ser. B, Vol. 196, 1903, pp. 99-118.

Lydekker, K., "On a remarkable Sirenian Jaw from the Oligocene of Italy and its
bearings on the Evolution of the Sirenia," Proc. Zool. Soc. Lond., Febr. 2, 1892,
pp. 77-83.

SOUTH AMERICAN UNGULATES.

The Ungulates of South America include the highly specialized
Typotheria, Homalodotheria, Toxodontia, and Astrapotheria, in all of
which the teeth present the extreme of lophodont modification and of
elongation or hypselodontism. Close analogies are found among the

ni.3

Fio. 192. Upper cheek teeth of Prottrothi rium sp., a primitive Litoptern from the Santa Crux
Formation, Middle (?) Miocene, Patagonia. Compare the somewhat similar buno-lppho-selenodont
molars of Palaotherium, Protorohippus (Fig. 100), Sairftu.tliei imn (Fig. l.v'O, Mi nim-ntln r'nun (Fig.
183), SchizotJicrium (Fig. 184) and C'cenotherium, familv Anplotheriidse, order Artiodactyla.

teeth of the Equido_j and Ehinocerotidse, and especially of the Amyno-
dontidfe. Naturally there is little trace left of the archaic and simpler
constitution of the teeth in these highly specialized crowns.

It is very suggestive, however, that the most primitive of the
Litopterna, the fifth of these South American orders, retain unmis-
takable indications of a primarily triangular crown pattern. In

190 EVOLUTION OF MAMMALIAN MOLAR TEETH

Protcrothcrium (Fig. 192), for example, we see the trigonal disposition
of three main cusps. The still more ancient Ungulates from the
Notostylops beds of Patagonia, exhibit molar teeth of the type seen
among the Amblypoda and Condylarthra, namely the tritubercular,
bunodont type, and lend the strongest of all the recent evidence which
has come forward in support of the tritubercular theory.

Pyrotherium, believed by Ameghino to be ancestral to the Pro-
boscidea, has simple, bilophodont molars.

Doctor Ameghino, although rejecting Osborn's homologies of the
molar cusps, and holding widely divergent views as to the ultimate
« irigin of the molars, yet brings forward a great deal of evidence l to
show the derivation of all the inferior molar types of South American
orders from a Proteodidelphys (Fig. 202) type which has a typically
tuberculosectorial lower dentition more primitive even than that of
Diilclpliys (see pp. 202, 204).

SPECIAL REFERENCES.

Ameghino, F., Contribution cd Conocimiento de los Mamniiferos Fosiles de la
Republica Argentina. Buenos Aiies, 1889. Numerous contributions to Anales del
Museo National de Buenos Aires, Boletin del Instituto Geograjico Argentina, Boletin
Acad. Nac. Ciencias Cordoba, Anales Soc. Cientifica, Argentina, etc. ; especially
" Recherches de Morphologie Phylogunetique sur les Molaires Superieures des
Ongules," An. d. Mus. Nac. des Buenos Aires, Tom. IX., 1904.

Lydekker, R., Palaeontologia Argentina II. La Plata, 1893.

Owen, R., The Zoology of H.M.S. Beagle, Pt. L, "Fossil Mammalia," London, 1840.

CETACEA.

Aquatic adaptation has gone to such an extreme in the teeth
of the Cetacea that all traces of tritubercular ancestry, if such ever
existed, have been entirely obliterated. It has been suggested that
the Cetacea are so ancient that they branched oft' before the
haplodont reptilian crown had begun the series of modifications
leading to trituberculy (Fig. 43), but the presence of accessory cuspules
in the posterior molars of certain recent and Miocene Platanistidse
and of vestigial low cusped, two rooted teeth in embryos of Whale-
bone Whales, the analogy with the secondary haplodont molar teeth
of certain Pinnipedia (Fig. 103), and the fact that the placenta tion
and reproductive organs of the Cetacea are of a very high Eutheriaii
type, are all more in accordance with the hypothesis that the ancestors
of the Cetacea possessed more complicated tooth crowns, and that the
existing haplodont types are all secondary.

1 " On the Primitive Type of the Plexodont Molars of Mammals," Proc. Zoo!. Soc.
Lond., May -2, 1899, pp. 555-571.

ORDINAL TYPES OF MOLARS: CETACKA

191

ZEUGLODONTIA.

The derivation of the serrated cutting molars of Zeuylodon from a
more normal type retaining vestiges of the ancient inner portion of the
crown, is demonstrated by recent discoveries in Egypt (Figs. 193, 194)

FIG. 193. Inferior surface of the skull of Prolocttus atavus, from the Lower Middle Eocene of
Mokattam, near Cairo, Egypt ; a very primitive Zeuglodont, thought by Prof. Fraas to represent
an aquatic offshoot of the Creodonta. The cheek teeth while elongate anteroposteriorly retain
vestiges of the internal protocone and of its root. Aquatic adaptation is also indicated in the
elongation of the snout, the prehensile modification of the anterior teeth, the secondary bridging
over of the posterior nares, the enlargement of the auditory bullaj. About i. After Fraas.

•

FIG. 194. Restoration of tha skull of Ziuglodon osiris, from the Middle Eocene of Egypt,
showing a further advance of the specialization noted under Fig. 193. Much reduced. After
Stromer von Reichenbach.

192 EVOLUTION OF MAMMALIAN MOLAR TEETH

by which the Zeuglodontia are thought by Professor Fraas to have
been connected with the Creodonta and especially the Hyfenodonti<l;r.
If in turn the Zeuglodonts should be shown to be related to the
Odontocete Whales, through Squalodon and the Miocene Platanistidse,
the ultimate derivation of the Cetacean cheek teeth from the trituber-
cular type would no longer be in doubt.

CHAPTER VI 1 1.
EVOLUTION OF THE PREMOLARS.

IN the early geological periods the preinolar teeth of mammals were
very simple in structure, as shown in the following description and
figures by Osborn] of the premolars of the Triassic and Jurassic
mammalia :

1. PREMOLAES IN PRIMITIVE MAMMALS.

"The premolars of DromatJierium [Fig. 195^4] are very unique. They
are tall and styloid and single fanged ; the last premolar has a vertical
groove upon the posterior face. In Microconodon [Fig. 195#], which

A B 15 4 3 7 // /2 A5

FIG. 195. The premolar tooth forms of the mammals of the Mesozoic order Pantotheria
or Tritnberculata. The premolar represented is invariably the most posterior of the series ;
the anterior face is to the left. A. Dfunmtlt: i-ium, B. Microconodon. 15. Ki'i'tixlii,!. 4. Ti'ico-
nodon. 9. Peraspalax. 1. SpalcLcotheriwrn. 11. Phascolestes. V2. Dryolestes. > 3. Achyrodon. Nos.
A, 4, 9, 11, 12 are seen upon the inner surface, the remainder upon the outer surface.

belongs to a somewhat more recent type, the premolars have a faint
posterior heel, and the last shows the trace of a double fang. In all
the Jurassic genera the premolars, where fully functional, are bifanged,
and possess a convex anterior face and concave posterior slope terminating
frequently in a heel. As in the molars, the cingulum plays an important
part in connection with the basal cusps. It is present upon the internal
face of the premolars of all the Jurassic genera except Kurtodon
[Fig. 195 15], and is observed upon the outer surface in Diplocynodon
[Fig. '20]. It thus in many cases enables us to draw the line
between premolars and molars, as in both the Peralestid;u J'< fn/rstes,
Per<(xj>r//ii.<' [Figs. 12, 22, 195 9], and in the genera of the Insectivorous
Sub-group the inner faces of the molars are smooth.

10sborn, H. F., "The Structure and Classification of the Mesozoic Mammalia,"
Acad. Nat. ,SV«., Philadelphia, Vol. IX., No. '2, July, 1888, pp. iZo-^-'G and pp. •_':;'.i--_)4it.

N

194 EVOLUTION OF MAMMALIAN MOLAR TEETH

The cingulum generally embraces the entire inner face of the crown,
forming anterior and posterior cingulum cusps or cingules, which are
characteristic of the insectivorous forms, while in the supposed car-
nivorous and omnivorous forms, distinct basal cusps rise posteriorly and
sometimes anteriorly [Triconodon, Figs. 8, 195 4] above the cingulum. As
in the latter genera the cingulum is present with the basal cusps, it
probably precedes them in evolution, but there is no direct evidence
(in the Triconodonta) of the conversion of ciiinulcx into true basal cusps,
such as we find in the molars.

A review of the premolars of all the genera shows that they are
sharply distinguished from the incisors and from the molars, and less
distinctly from the canines in many instances.1 In several genera they
have undergone considerable specialization, as in the production into lofty
cones of pm^ of Achyrodon [Fig. 195 1J], or the apparently incipient
assumption of the molar pattern in Kurt od on [Figs. 195 15]"

2. ADAPTATION OF PREMOLARS.

The premolar teeth in general are quite as adaptive and independent
in evolution as the molars. While in many families of mammals the
first, second, and third, and more rarely the fourth promolars retain more
or less of this simple ancestral structure ; in other families the premolars
become greatly complicated. They either (a) enter upon an especial
adaptive evolution of their own, as for example in the upper sectorials
of the Cats (Felicia-1), or the elaborate fourth premolars of the Plagiau-
lacida? (pp. 102, 106), or (b) by a serial analogous development they more
or less closely mimic the structure and supplement the exact functions
and uses of the molar teeth : this mimicry reaches its highest extreme
among the Perissodactyl or odd-toed Ungulates, such as the horses, where
the premolars gradually metamorphose into the molar pattern and even
become superior to the molars in size and complication.

3. VARIOUS UPPER PREMOLAR TYPES.

1. Persistence of a simple conical form throughout. Realized in the
inferior premolars of Dromatlurium (Fig. 3).

2. The posterior or 4th superior premolar becomes sectorial, the
anterior premolars remaining more or less simple and conical, e.g. most
Carnivora.

3. The 4th, 3rd, and 2nd superior premolars become more or less
uniformly bicuspid, <\y. most Primates.

1 There is strong support among the Jurassic mammals and the recent Insectivora for
the opinion first expressed by E. Ray Lankester, that the canine is originally a bifanged
tooth and represents a modified anterior premolar.

EVOLUTION OF THE PREMOLARS 195

4. The 4th superior premolar more or less completely transforms
into the molar pattern, the third, second, and first remaining simpler
(e.g. some Artiodactyla, as Agriochcerus).

5. The 4th and 3rd superior premolars transform into the molar
pattern, the first and second adopting an entirely different order of
evolution (e.g. Galeopithecus).

6. The 4th, 3rd, and 2nd premolars successively partially transform
in the molar pattern. Example, many Perissodactyla, such as the
Titanotheres and lihinocerotidas.

7. The 4th, 3rd, and 2nd premolars completely transform into the
molar pattern. Example, some Perissodactyla (Equiclee).

I This premolar metamorphosis into the molar pattern observed in types
4, 5, 6, 7, above is a very gradual process, requiring hundreds of
thousands if not millions of years, and from the biological standpoint
most interesting as illustration of convergence, because form exactly
similar to that of the molars is finally attained from somewhat dis-
similar beginnings.

4. CUSP ADDITION IN THE PREMOLARS.

The first broad and systematic treatment of the subject of premolar
evolution was that by Professor W. B. Scott.1 It is somewhat too special
to be cited in full here. We accept Scott's interpretation in full as
regards the upper premolars, but have adopted a different interpretation
of the evolution of the lower premolars.

SUPERIOR PREMOLARS.

It is important to note here that all of the following description is
based on the older Cope-Osborn theory that the superior premolars
have followed a different order of cusp addition from the molars ; this
is now met by the newer theory (see Chapter IX.) that the premolars
follow practically the same order of cusp addition as that originally
followed by the molars. Pending the solution of this question the
comparisons which follow are made on the basis of the older theory.

We may take the progressive complication of the fourth superior
premolar as a standard ; the order of succession of the cusps in this
tooth is rather constant, while in the more anterior premolars there are
more various modes of complication.

First stage. As early as the Basal Eocene period the fourth upper
premolar, in every known genus in which the premolars tend to imitate

1 " The Evolution of the Premolar Teeth in Mammals," Proc. Acad. N«t. Sci. Phila.,
1892, pp. 40.1-444.

196 EVOLUTION OF MAMMALIAN MOLAR TEETH

the molars, is complicated by the addition on the inner or lingual side
of the protocone of a second cusp, which has appropriately been called
the deuterocone [Sevrepo?, second ; KWI/O?, cone] by Scott. This bicuspid
stage, which is retained in the ' bicuspids ' of man and other primates, is
the starting point, and brings out clearly the important initial fact
that in the premolars the protoconc remains upon the outer or buccal side

pr

FIG. 196. Fourth upper premolar of Creodonts and Carnivores in various stages of evolution.
(Cf. Figs. 84-92.) After Scott,

1. DeltatkeTium fundaminis, family Oxycteuid*.

2. Sinopa tahitue, family Hyaenodontidae. (Cf. Fig. 89.)

3. Cynodictis gracilis, family Canidae.

4. Felis concolor, family Felidse.

of. the crown, while in the molars (as 'we have tried to show above,
pp. 35, 217) it shifts to the inner or lingual side. From this it follows
that the deuterocone of the premolars has no exact serial homologue
in any of the cusps of the molars, although it becomes functionally
analogous to the protocone ; it follows, moreover, that all the cusps
which are subsequently added to the premolars are analogous, but not
serially homologous, to those in the molars (Fig. 196).

The Second xtaae of premolar complication usually consists in the
addition of a second outer cusp, posterior to the protocone, which, as the

third in the series, Scott has called the
fi'itocone [T^O/TO?, third] ; this in turn is
analogous in position and function with
the inetacone of the molars. This tri-

FIG. 197. Second upper premolar of tubercular fn'raspid Or trigOllOCloilt pre-
P>-otornh,pt>,<s n-,<licolns, a lower Eocene ,-,-,,,1.-,,. crarro imifafpa VPW nlriQplv flip
Equid(cf. Fig. 160), showing the division HlOiai Stage imitates \61} C1OS61}

°4ritVconI^inX7scTr''prot°coliei" tritubercular or trigonodont molar tooth,

and frequently the tritubercular types of

premolars display intermediate conules corresponding in position with
the proto- and meta-conules of the molars, but obviously not homologous
with them. The gradual development of this tritubercular premolar
stage is beautifully shown in series of teeth of JEuprotogonia and
Phenacodus, in which the tritocone may be seen in all stages of
development from its incipiency, and in which the more anterior
premolars are seen taking up the stages which have already been

EVOLUTION OF THE PREMOLARS

197

passed through by the fourth premolar, until it reaches the size of
a protocone (Figs. 197, 198).

pr tr

FIG. 198. Complication of the superior premolars in Coudylarthra. 1. PS, p4 of Phenaeodus
prinuevus, showing a complicated j>4. 2. P3, pi of Phenaeodus ioortmn,ii with simpler p*. 3. P± of
ilri'ta. After Scott.

This stage is very widely exhibited in the Middle and Upper Eocene
Ungulates and Creodonts and persists till the present time, with some
modifications in the sectorial or carnassial fourth premolar of the
Carnivora, in many Insectivora, and in some forms of the Artiodactyla.

The third and final stayc in the metamorphosis of the premolar into
the molar pattern is reached by the addition of a fourth main element,
which Scott has called the tetartocone [Teraprof, fourth] ; it corresponds

FIG. 199. Fourth upper and lower premolars of various Artiodactyla. After Scott.

1. Dicotyles torquatus, family Dicotylida?. Molarization nearly complete.

2. Tliinohyus lentus, family Dicotylida;. Molarization much less complete.

3. Perclin i-iix /1,-iiin's, family Dicotylidse.

4. TfiijfinoU ft> .-• iii-iH-lii/xtomus, family Trigonolestidre, showing simple upper and lower premolar.

in position and is analogous with the hypocone of the molars. As
observed by Scott, the tetartocone usually arises in the same manner
as the hypocone of the molars, namely, by the addition of a cusp at
the postero-internal angle of the crown immediately behind the
deuterocone (Fig. 199, No. 1).

In the normal or usual evolution of the tetartocone behind the
deuterocone variations occur characteristic of different families of
mammals. Sometimes (in some Perissodactyla, <•.//. Titanotheres) a
ridge extends back from the deuterocone, which splits in two, forming
the deuterocone anteriorly, the tetartocone posteriorly.

This brief survey of the steps of premolar evolution proves that these
teeth follow an order of differentiation quite at variance with that

198 EVOLUTION OF MAMMALIAN MOLAR TEETH

attributed to the molars, even when the final results are the same. It
is another remarkable instance of convergence in evolution, and proves
that similarity in form and in position affords an unsafe guide to serial
homology (see, however, Chapter XL).

INFERIOR PREMOLARS.

As shown in the foregoing comparison of the Jurassic mammals,
and in other primitive mammals, the primitive form of the lower
premolar is a simple, more or less recurved cone, obviously corresponding
to the protocone of the molars, and implanted by a single fang.

The metamorphosis of the inferior premolars also begins with the
fourth premolar, and extends anteriorly. While the order of cusp
development is less regular and constant than in the upper teeth, we
find that the lower premolars may be more closely compared with the
lower molars, that is, there is more evidence of serial homology between
the cusps of the lower premolars and lower molars. This fact fortunately
enables us to use the same terminology for the premolars as for the
molars. Professor Scott was led to take a different view, and proposed
several new homologues in the lower premolars, such as the paraconid,
deuteroconid, metaconid, tetartoconid, which may be eliminated by a
somewhat different interpretation of the development.

Comparison of the lower premolars should be made, not with the
molars of the Triconodonta, but with the molars of Trituberculates.
Such comparison shows that the premolar evolution may be interpreted
as substantially similar to the molar evolution.

The initial stage is the simple, single-fanged, conical cusp, the
protoconid of Dromatherium (Figs. 3, 195 A}.

Even in the Jurassic mammals the second stage appears in the two-
fanged crown with more or less developed posterior basal cusp, which
corresponds in position and function with the talonid or hypoconid
of the tritubercular molars of such a type as Amphithcrium (Fig. 15).
In the subsequent evolution of the crown this hypoconid remains on
the outer or buccal side of the crown as in the molars ; it is therefore
practically homologous serially with the hypoconid of the tritubercular
molars rather than with the metaconid of the triconodont molars.
Most of the existing Unguiculates as well as some recent, and many
extinct Ungulates, retain more or fewer premolar teeth which depart
but little from this type.

As a third *t«</<- an anterior basal cusp comparable to the paraconid
both of the tritubercular and triconodont molars is added. This cusp,
in fact, corresponds in the subsequent evolution of the teeth with the

KVOU'TIOX OF THK PREMOLARS

199

paraconid of the true molars; and may, therefore, be considered as
more or less serially homologous with that molar element.

As 'A fourth stayc (or in some cases as the third stage of development
the order of succession not being- constant) there appears a cuspule on
the inner side of the crown of the protoconid which corresponds in
position with the metaconid of the true molars of the tritubercular

j 4 ,s

FIG. 200. Fourth lower prcnaolars of Creodonts in various stages of complication. After Scott.
1. Tiicentes sultrigoaus. family Oxyolcenidse.
•_'. i 'In nodon protogonoides, family Arctocyonidfe.

3. Ckriacus stenops, family Oxyclsenidse.

4. Chriacus sc/ilijxaei-iniius.

5. Deltatherium fundaminis, family Hyjenodontidje, tuberculo-sectorial pattern nearly com-
pleted, with a high trigonid and low talonid.

Ani2>/tit/i<'riittH type, and subsequently develops into an exactly analogous
form. (This is the deuteroconid of Scott's terminology.)

As a fifth stage a cusp is sometimes added to the premolar crown
on the internal or lingual side of the hypoconid, occupying- the position
held by the entoconid in the true molars. (To this Scott gave the
name of ' tetartoconid,' regarding it as serially comparable with the
tetartocone of the upper premolars.)

Kn;. -201. Premolar complication in C'nndylarthra (Phenacodontid;e). Fourth lower premolar
nf right side, internal aspect. 1 and ~1. Pi-<itiii/oii<nlin> /" ,>f<"'tts, showing incipient metaconid and
t:il. mid. o. Ev.protorjoni.n /ilicif' /•«, showing better developed metaconid and talonid, incipient
parat-onid. The tuberculo-sectorial pattern is thus nearly attained. After Scott.

On the completion of these four stages, the premolar readies a condition
analogous to the ' tuberculo-sectorial ' stage of the molars, namely, with
an elevated fri</onid composed of the protoconid, paraconid and metaconid,
and with a depressed fn/u/m/ composed of the hypoconid and entoconid.

The subsequent transformation of the fourth inferior premolar in
various type is shown in the preceding and accompanying figures
(Figs. 200, 201).

200 EVOLUTION OF MAMMALIAN MOLAR TEETH

In comparatively few mammals, however, are these four stages
completed. With many exceptions (such as the sectorial p^ opposed
by the sectorial m^ in the Carnivora), the evolution of the lower
premolars is normally correlated with that of the upper premolars,
which agree roughly with the seven upper premolar types enumerated on
pages 194, 195. For example, in the Equidae, three of the corresponding-
upper and lower premolars assume the molar form, but in Gfaleopithecus,
p^-'-^ become molariforrn, while the independently adaptive p-^-^- become
elongate and denticulate.

CHAPTER IX.
OBJECTIONS AND DIFFICULTIES AND OTHER THEOEIES.

Two classes of criticism have developed :

I. That the tritubercular type is not primitive.

II. That the Cope-Osborn theory of the origin of the superior
molars is incorrect.

Let us consider these in order.

I. That the Tritubercular Type is not Primitive.

1. THE TLEXODOXT' OK PROGRESSIVE SIMPLIFICATION THEORY OF

AMEGHINO.

In 1884,11896,- and 1899 3 Dr. Florentine Ameghino fully stated
an original theory of the origin of the grinding teeth, in which there
are four main propositions :

Firxt (op. cit., 1889, p. 556), that aside from the haplodont
reptilian type, the most primitive type of inferior molar is not
the protodont, triconodont, or tritubercular, but the pli'.roiJout. The
original plexodont upper or lower molar is supposed to have been
quadrangular, quadrituberculate, quadriradiculate (op. cit., 1896, pp.
18, 19, 25, 61, 64). The oldest known form of complete plexodont
molar, that of Protn><H<t<'li>li>/x (Fig. 202 and op. cit., 90, p. 556)
has an anterior and a posterior lobe, each of them carrying three
cusps, which are designated in the following terms, the names in
parenthesis being those of Osborn's nomenclature (op. cit., p. 557).

ma, median-anterior (paraconid). pe, postero-external (hypoconid).

ae, autero-external (protoconid). />/, postero-internal (entoconidj.

ai, antero-internal (metaconid). ///_/>, median -posterior (hypoconulid).

1 Filof/enia, Svo, 1884.

2"Sur 1'Evolution des Dents des Mammiferes," Hal. Aca</. X"<-. ,1, Cienc., T. XIV..
1896, pp. 381-517.

a"0n the 1'riniitivc Type of the Plexodont Molars of Mammals," /Vor. Zoo/. S<,r.
Lond., May 2, 1899.

-202

EVOLUTION OF MAMMALIAN MOLAR TEETH

Second, that the plexodont molar made its appearance suddenly
(p. 571, bottom), the only theory which can explain it in a satis-
factory manner being that it arose by the fusion or the concrescence
of " the dental germs or embryos of several simple teeth " (op. cit.,
'99, p. 555).

Third (p. 556, § 1), that the tritubercular, triconodont and pro-
todont stages are secondary simplifications of the plexodont stage.

Fourth, that the plexodont type is the starting point of all the
post canine cheek teeth including the premolars. (The teeth of Mouo-
tremes, Edentates, Cetaceans are supposed to have been independently
derived from the haplodont type.)

The ' plexodont ' type of Ameghino as illustrated in Protcodi-
dclphys (Fig. 202) is identical with the six-cusped ' tuberculo-sectorial '

FIG. 202. Lower jaw and teeth of Proteodidetphys prcecursor, a small Polyprotodont Marsupial
from the Chubut Formation. (? Upper Cretaceous) of Patagonia, xi. The true molars (" 5?«,"
"6m," "7m") are in a very primitive stage of the tuberculo-sectorial modification. After
Ameghino.

type of Cope ; briefly stated, therefore, the plexodont theory is that
this type, instead of being an intermediate stage, as in the Cope-
Osborn theory, is a primordial stage from which, on the one hand,
the tritubercular, triconodont and protodont types have been derived
by retrogressive simplification or loss of cusps, and, on the other hand,
the higher omnivorous (bunodont) and herbivorous (hypselodont) stages
have been derived by progressive modification of the relative size
of the different parts without the addition of new cusps (p. 565,
op. cit.}. In the 1899 paper Dr. Ameghino demonstrates this unity
of type in all the South American mammals, namely the Marsupialia,
Toxodontia, Litopterna, Primates, etc.

We give below several of this author's ingenious arguments (op. cit.
1896).

P. 19. Teeth with a single root (e.g. incisors) and simple crown
are primitive. Molars with two, three or four roots generally repre-
sent the fusion of two, three or four teeth. But roots may coalesce
as well as crowns, so that the number of original teeth may be
greater in many cases than that of the roots. Coalesced roots often

OBJECTIONS AND DIFFICULTIES AND OTHER THEOK1KS 203

retain an external groove at the line of juncture, and distinct
nutrient canals, one for each component (/.//. ox teeth). Superior
molars of man, ruminants, horses and other Cumulates, represent the
fusion of four simple primitive teeth.

Thus Ameghino seems to exclude as improbable any secondary
division of roots. He also regards such specialized forms as Cetaceans
and Edentates as primitive in their teeth.

P. 20. Probably in all superior molars which are quadrangular,
but with three roots, the internal root represents a fusion of two
roots. Continuation of this process may result in a single root. Two
rooted canines = 2 fused teeth, in normal canines the coalescence has
extended to the roots. [Dr. Ameghino does not bring forward any
direct evidence to prove the alleged coalescence.]

P. 29. "We have shown (Filouniln, pp. 88 to 112) that after
the fusion which produced the plexodont teeth [quadrangular, quadri-
tubercular, quadriradiculate] they soon became either still further com-
plicated by the formation of new tubercles, or more simply by the
gradual atrophy of the cuspids, tubercles, etc. We have always
particularly emphasized these changes, for . . . having examined an
enormous number of teeth in almost all groups of mammals, we are
convinced of the great facility with which these organs change in form,
by the addition or suppression of cuspids, tubercles, enamel folds, creux
rentrants valleys, etc., of the crown.

" We do not share in the opinion that multituberculate molars must
be the result of the fusion of as many simple teeth as there are cuspids
on their crowns ; the number of fused teeth could only be four or five at
most for each tooth. The presence of this kind of molars in the Trias
supports the theory of fusion, because if these teeth were the result of
gradual complication not only would it be necessary to push back the
origin of the mammals to an excessively distant epoch, but also we
ought to find in pre-Triassic beds numerous forms intermediate between
the conic and the multicuspidate. However, as we do not find any
such forms, we are brought to the belief that plexodont teeth are
established by a rapid process such as fusion would be." (Translation.)
[Another equally rapid process would be the direct modelling up of a
cylindrical or conic crown into a basin-shaped crown with cuspidate
edges.]

P. 25. "It is incontestable that the triangular type dominates in
the ancient epochs, but this is quite natural since being a modification
of the quadrangular type it should of necessity be more abundant: in
consequence of the principle formulated elsewhere ' that modified or
collateral forms should always be infinitely more numerous than the
main stem from which lateral branches spring.'' (Translation.)

204 EVOLUTION OF MAMMALIAN MOLAR TEETH

In support of the simplification hypothesis as applied to the pre-
molars, the author endeavours to show (op. cit., 1899):

(1) The true molars of all these orders can be reduced to this type
(pp. 558-565).

(2) The so-called premolars of Proteodidelphys (pp. 556, 557) are
more complex than those of its descendants, Eodidelphys, Microbio-
f/irrium, Didclphys. [To the present author the premolars of
Proteodidelphys seem less complex and more primitive than those
of DidelphysJ] When complex lower premolars do occur in different
orders they are to lie interpreted as reversions to the ancestral
plexodont type (p. 566, §2).

(3) The less complex condition of lower premolars generally is
secondary (p. 568), due to the want of space for the complete develop-
ment of these teeth, which resulted from a fore-and-aft shortening of
the whole tooth row : for (a) modern placental cheek teeth represent
two series (p. 569), that is, the first series = true molars + milk molars,
the second series = replacing teeth or premolars; (b) formerly (p. 569)
(in South American forms) the first series was all in use at one time
and ctliilnted the same form from one end of the series to the other;
(c) on account of the gradual intercalation of the two series the space
left free for the incoming premolars was shortened and that for the
milk molars was increased so that the premolars progressively simpli-
fied and the milk molars became more complicated. Similar reasoning
applies to the upper premolars (p. 569, footnote).

He concludes (p. 571): " The clear result of all these facts is that the
famous theory of the gradual complication, of triconodonty and trituberculy,
is an untenable hypothesis. Nowhere do we meet with the stages leading
from haplodonty to plexodonty ; all those which have been mentioned
are, on the contrary, as I believe I have demonstrated, but the result of
simplification of molars which were formerly more complicated."

In answer to these ingenious arguments for the simplification theory
may be adduced the facts newly stated in Chapter VII., pp. 100-192.

2. OBJECTIONS BY FLEISCHMANN AND MAHX ANSWERED BY SCOTT.

In 1891 A. Fleischmann * and R. Malm2 rejected both the nomen-
clature and the homologies of Trituberculy for reasons which were
discussed in great detail and apparently shown to be insufficient, by
Prof. W. B. Scott in 1892.3

1 " Die Grundform der Backziihne bei Sangethieren," Sitzungsberichte d. kun. pn-uss.
Ahul. ,!. IVis*. zu Berlin, 1891, p. 891.

2 " Ban uncl Entwicklung d. Molaren bei Mus und Arvicola," Jlorphol. Jahrb., Bd. XVI.,
1>. 6.V2.

3 " The Evolution of the Premolar Teetli in the Mammals,'' Proc. A<iad. A'«/. Sd. Phi/a.,
1892, pp. 409 41'J.

OBJECTIONS AND DIFFICULTIES AND OTHER THEORIES 205

3. THE PRIMITIVE POLYP-I-NY THEORY.

In a very interesting and exhaustive paper, On Souu' Miw< //>
Squirrels, with Jtfina,-/.-* on iln- Dentition a ml Classification oj the Sciuri/m
(p. 179), Dr. C. J. Forsyth Major, after a very careful consideration of
the simpler dental types in Rodents, concludes (pp. 196-215) with a
full discussion entitled "On the Primitive Type of Sciurine Molar and
of the Eutherian Molar in General."

' Trituberculism," he observes, "or, as we rather ought to call it,
the reptilian-cone theory, is no more a theory, but has become a dogma.
I am a heretic, and may say that I opposed the theory already in
1873, viz. before it was invented;1 since that time I have kept silent
for various reasons. ... It would appear that the Allotheria, the Multi-
tuberculata (p. 202), ought to have been a stumbling-block for the theory.
But this is not the case ; they have been simply pushed aside on account
of being an aberrant order. Nevertheless, I shall refer to them later on.
. . . The adherents of trituberculism assert that they have proved the
Mammalian molar to be traced back to a more and more simple form.
I have tried to show that they have failed to do so, and in my turn
assert that the molar of Placentalia can be traced to a polybunous form,
and that the real tritubercular pattern is a more specialized secondary
stage. So that, as a matter of course, the cardinal point to be estab-
lished is to show, that the more complex forms, which in the Lower
Eocene as well as in the recent period are found side by side with
the simpler forms, trituberculate or otherwise, are indeed the primitive.
the more generalized type."

This point is supported by a detailed argument, of which we can
present merely a brief summary. The starting point is that (1) brachy-
odont teeth are more primitive than rootless or hy pselodont teeth ; ( 2 )
that the more brachyodont a molar is the more polybunous it is. The
latter statement is illustrated by comparison of Eocene and recent
squirrels, in which the author observes that the most bracliyodnnt molars
exhibit flat, elongate crowns covered with small cusps which tend toward
a longitudinal arrangement. An example of this (Type 1) is Sciuro*
i adieu*. Type 2, Sciurus m It/a fix, which he believes represents a less
brachyodont successive stage, shows four more or less transverse ridges
with three intervening valleys. Type 3, AV/v.s getulus, a sub-hypsodont
squirrel, exhibits a more distinctly lophodont or crested crown, approach-
ing that of the more specialized Kodents. Again (p. 213), among the
New Guinea mice (Chiruromys), we find multicuspidate teeth "of a

1 Forsyth Major, " Nageriiberreste aus Bohnerzen Siiddeutschlands uncl der Schweiz.
Nebst Beitragen zu einer vergleichenden Odontographie von Ungulaten uiul Unguiculaten."
1873, Palceontographica, XXII.

206 EVOLUTION OF MAMMALIAN MOLAR TEETH

very generalized type, precisely such as we anticipate to meet with in
a refuge for old and little modified forms."

The first conclusion drawn is that in the most lirachyodont Sciurine
molars the crowns are quadrate, and the cusps tend toward longitudinal
arrangement with two entire outer and inner ' marginal series ' of
cusps in the lower molars, and with two ' marginal ' and a more or less
complete ' intermediate series ' in the upper molars. From this it is
inferred (p. 205) that the nmlticuspidate or polybunous condition is the
most primitive, and that the triangular or tritubercular condition repre-
sents an extreme of specialization.

Analogous arguments are applied by this author to other mammals.
The polybunous molar of ^-Elurus is regarded as one of the most primitive
among the Carnivora, the polybunous Ardocyon is regarded as the most
primitive among the Oreoclonts. By similar reasoning we should consider
the polybunous tuberculate molars of the bear and of the orang as
respectively more primitive than the high pointedly cusped molars of
the typical Viverrines (cf. Fig. 102), or of the lower Primates
(see pp. 158-160). The author concedes (p. 185), that in certain bats
(see p. 129), and in Cheiromi/s the basin-shaped molar is retrogressive
on account of the secondary assumption of fruit-eating habits ; but regards
(p. 213) the basin-shaped polybunous molar of J/iV/Wr.sA'.s as primitive
and ancestral both to the multituberculate and trituberculate types.

By similar reasoning he finally reaches the conclusion that in the
primitive molars of placentals the cusps were arranged in longitudinal
rows, three rows with two intermediate grooves in the upper teeth, and
two rows with one intermediate groove in the lower teeth ; in other
words, that the ancestral type of Eutherian molar was multituberculate.

Dr. A. Smith Woodward (Vert. Pal, p. 269) feels the force of this
argument as well as that of the embryologists, and concludes. " Hence
this — at first sight — brilliant generalization [of primitive trituberculy]
can only be accepted at present as a convenient working hypothesis
which remains on its trial."

Mr. E. S. Goodrich1 in his discussion of the Lower Jurassic mammalia,
after enumerating several gaps in the argument for the tritubercular
theory, concludes: "The common ancestor of the tritubercular sectorial
mammals and of the Multituberculates probably had teeth of an indefinite
multituberculate pattern, which gave rise, on the one hand, to elaborate
multituberculate teeth, and on the other to the tritubercular sectorial.
Thus the development of two longitudinal rows of three cusps would give
rise to the type of lower molar common amongst the Multituberculata ;
the fusion of the two anterior of these cusps or the loss of one would

1 " On the Fossil Mammalia of the Stonesfield Slate," Quar. Jour. Mirr. Sci., Vol.
XXXV., 1894, pp. 407-432.

OBJECTIONS AND DIFFICULTIES AND OTHKl! THEORIES 207

yield the tri tubercular sectorial tooth common among the Marsupials
and Placentals : while the loss of the inner cusps would result in the
formation of a triconodont molar. The conclusion reached is, therefore,
that the primitive mammalian molar bore a crown with several cusps."

This seems to be a clear statement of the polybuny theory, although
the reader must consult the authors themselves for the detailed argu-
ments by which it is supported. According to the author last quoted.
not onlv the tritubercular but the triconodont molar is of multitubercular

origin.

4. SUMMARY OF OBJECTIONS TO THE POLYBUNY THEORY.

Our own contrary view may be advanced with equal show of
evidence that even the most prilnitive Multituberculates known had
already passed tli rough previous fewer-cusped or even tritubercular
stages (see p. 80). A third and perhaps preferable alternative is
that the simple " multituberculate " lower molar of the Triassic Micro-
lestes was derived from the triconodont type by the transverse broadening
of the base of the tooth and the upgrowth of the internal basal border.
Other possible modes of derivation are discussed in pages 103-105.

The general answer to this line of reasoning is that which applies
equally to many other generalizations which are founded chiefly upon
anatomical and zoological comparison, namely : that there is a danger
of inverting a series, of placing the most specialized teeth at the bottom
of a theoretical scale of evolution, while the most primitive are placed
at the top. This criticism certainly can be demonstrated as correct so
far as the argument applies to Arctocyon (see p. 133) and blunts (see
p. 142), as we feel reasonably certain that the presumed ancestors or
oldest representatives of each of these types have elevated and distinct
cusps, and in each case the low-crowned, tuberculate condition is
secondary. Among the Rodents we have, it is true, not yet traced
the phyletic succession as fully as in other series, but we have strong
grounds for considering trituberculate upper molars such as those of
Plcsiarctomys as the most primitive (see pp. 145-151).

The chief difficulties in the ' polybuny theory ' may be summarized
as follows : ( 1 ) The examples cited as primitive are tubercular low-
cusped crushing teeth ; all zoological and palaeontological evidence goes to
show that such crowns are secondary as compared with more pointedly
cusped, piercing-cutting-crushing teeth such as those seen in Polyproto-
dont Marsupials (p. 109), Insectivores (p. 117), Creodonts (p. 132). (2)
Palaeontology offers strong negative evidence against it, for Microles1<-*
very probably1 leads up through Plagiaulax and Ctenacodon into the

1 Osborn, H. F., "Structure and Classification of the Mesozoic Mammals," Jew. Acad.
Nat. So'., Fhila., 1888, pp. 214-216.

208 EVOLUTION OF MAMMALIAN MOLAR TEETH

later Multituberculates which are the only known group of polybunous
mammals in the Mesozoic period, and all of these have sharply defined
and modeled cusps, elongate crowns, parallel grinding series, enlarged
chisel like incisors, correlated with chiefly horizontal jaw motions and
gnawing habits — conditions far too specialized to have given rise to
any of the later fewer-cusped or tritubercular types. (3) Embryological
evidence is against the polybuny theory, there being no trace of previous
polybuny, but on the contrary quite generally in the true molars a
triangular disposition of the cusps first developed. (4) According to
the polybuny theory one cusp is as old as another, but if either the
' tritubercular ' or the ' embryological ' or the ' premolar-analogy ' theories
be correct one of the cusps is older than the others.

II. That the Cope-Osborn Theory of the Origin of the
Superior Molars is Incorrect.

This concerns the superior molars only. Strong evidence is brought
forward from embryogeny, from the analogy of premolars, from
palaeontology, that in the superior molars the protoconc (of Osborn) is
not homologous with the primitive or original reptilian cusp of the
crown.

In previous pages of this volume we have stated and discussed the
evidence against the Cope-Osborn theory of the origin of the upper
molars.

Let us now review this evidence.

1. CUSP HOMOLOGIES FOUNDED ON EMBRYOGENY.

The most thorough presentation of the development of tooth cusps in
its bearings upon the homologies of the upper cusps is that by M. F.
Woodward,1 published in 1896.

In course of this important article he refers to the deficiency of
the palaeontological evidence among trituberculates, so far as the upper
teeth are concerned, to establish the homology of the upper and lower
protocones beyond question, concluding : " We consequently have no
palseontological evidence to support the assumption that a tritubercular
stage is passed through by the mammalian upper molar in its evolution
from a protodont or possibly a triconodont tooth. ... If the triconodont
tooth be a stage in the evolution of the mammalian molar, then I
should believe that the anterior cone disappeared, the main cone
becoming enlarged as the paracone and the posterior as the
metacone. ... At this stage the upper teeth overhang and bite out-

1 "Contributions to the Study of Mammalian Dentition, Pt. II., On the Teeth of certain
Insectivora," Proc. Zool. Soc. Land., 1896, pp. 557-594.

OBJECTIONS AND DIFFICULT! KS AM) OTIIKI! THKnlilKS

209

side the lower molars, and the future antero-internal cone (protocone)
was developed as an internal shelf acting as a mortar for the cusps of
the lower teeth and at a much later period developed a cusp. The
hypocone arose in a similar way with the elongation of the teeth. . . .
In the Centetidce and Peralestes, the upper molars could not have over-
hung the lower ones to the same extent, consequently no internal lobe
bearing the protocone was developed and the external cingulum was
very largely developed." Thus Woodward's position agrees in the main
with the views recently put forward by Gidley mi pal;eontological
grounds (p. 219).

The author therefore accepts the tri tubercular theory from the Eocene
period onwards, but endeavours to establish another theory as to the

».p/r-s.«a
-'-'•;:•&'*

Fio. 203. Transverse section of the germ of the first upper molar of a young Mole
"showing the primitive dentine germ giving rise to the paracone (5)," the protocone appearing
as an internal extension or talon from the base of the paracone, which is hence supposed by
Woodward to be the most ancient portion of the crown. The dental lamina dl on the lingual side
of the developing true molar may represent the vestiges of a post-penn-ment //i1. ,< ° •>-. After
M. F. Woodward.

origin of the superior tritubercular type; he excludes the triconodont
and protodont stages, unless we suppose that the anterior cusp
disappeared.

As regards embryological testimony, he first refers to the papers of
Rose on the human teeth. Tucker on the teeth of Ungulates, and Leche
on the teeth of Marsupials, as concurring in the demonstration that ///>
paracone (of Osborn) ?.s tin: jir»f <-nsp to </rrrlop in 1h< vpj.>i ,' nmlur fn-fii.
He continues ((yy /•/'/.. p. .~>S,~>) from his own researches upon the relatively
primitive Insectivora, " If the protocone represents the summit of the
original protodont tooth of the ancestor of the Mammalia it must lie the
direct continuation of the primitive dentinal germ, and as such should be
found to develop in a line with the axis of that structure. That this is
not the case is well seen in Fig. 32, PI. XXVI. [Fig. 203], where the
paracone is found to be identical with the primitive dentinal germ, and

o

210 EVOLUTION OP" MAMMALIAN MOLAR TEETH

the protocol) e appears as a mere internal ledge growing out from the
base of this structure, the metacone and subsequently the hypocone
being similarly derived from a backward extension of the base of the
primitive dentinal germ." Incidentally he adds, " I have failed to find
any support for the concrescence theory, neither do I consider that
any of the evidence put forward by Rose and Kukenthal is at all
conclusive in its favor/'

Order <>/' Embryonic Cusp- Development, according to WoodvarJ.

GROUP I. GROUP II.

(4 genera: Eriiiaceiis, Gymnura, Sorex, Talpd). (-2 genera: Centetes, Ericulus).

With quadri- or quinque- With tritubercular upper
tubercular upper molars. molars.

1. Paracone. 1. 'Protocone' ( = pa + me. Seep.24">).

•2. Metacone. 2. 'Paracone' \ , , .,

, ,T , , f< together.

3. Protocone. 3. ' Metacone J

4. Hypocone.
(5. Metaconule.)

1. Protoconid. 1. Protocoled.

2. Metaconid. 2 or 3. Paraconid \ , . ,,

rr< . . , ». , • -i i ( together.

o TJ i f Entoconid. 3 or 2. MetacomdJ

3. Heel TT - ,

IHypocomd.

4. Paraconid. 4. Hypoconid.

As regards the adult structure of the teeth and the embryonic order
of development of the molar cusps, Woodward divides the Insectivora
into two groups, as above.

GROUP I. Quadritubercular Molars. The first of the quadritubercular
types investigated is the Malayan Gymnnra, the adult molar teeth of which,
according to Woodward, "resemble those of the hedgehog (Ennaccus) in
pattern;1 like that genus, they exhibit five cusps which are strongly
developed, and in the upper jaw there is a well-marked cingulum, with a
small anterior and posterior cusp present in addition : in the lower jaw
the paraconid is less developed than in Erin ace us." In the developing
foetus (p. 567) the second superior molar "was less developed, and here
the para- and metacones were the most strongly developed, while the
protocone was present in the form of a large antero-internal shelf [italics
ours], but hardly as yet developed into a distinct cusp, though the
hypocone and metaconule had done so." In the Shrews of the genus
Sorcx (op. cit., p. 570) the author also finds in a young stage that the
plan of the dental germ is " roughly triangular, the main and only cone
being situated at the anterior extremity and slightly nearer the external
border." From the position of this cone, and from a comparison of the

1 [With the important exception at least in some specimens (Fig. 65, p. 118) that the
hypocone is less developed, and the whole tooth strongly suggests that of the tritubercular
to quadritubercular Eocene genus Leptictis (Figs. 60, 67). — ED.]

OBJECTIONS AND DIFFICULTIES AND OTHER THF.niMKs I'll

cusp ontogeny as seen in the molar of the Talp», with which it is
identical in pattern, I think one may conclude thai ////'s N/////A r//.s// /* tin
paracone [italics ours], the posterior extension representing the metacone,
while the internal shelf indicates the position of the future proto- and
hypocone." In the moles (genus Ta/jiti, p. 579) the adult upper molars
are " mainly tritubercular, but a very small hypocone is present ; the
protocone is small, whereas the paracone and metacone, especially the
latter, are very large, and show a tendency to become crescentic or
V-shaped, . . ." Even in the earliest foetal stages examined two slight
prominences were already visible, corresponding to the para- and meta-
cones ; these cusps were alone conspicuous in the younger stages, the
antero-external or paracone being the largest, though in the adult Mole
this is a smaller cusp than the metacone.

This, the author thinks, shows that the paracone is the first to
develop; " tin- i/ifu-nnl protocone appears late, as a low and inward
extension of the base of the paracone [Fig. 203], and cannot possibly Ic
regarded as the original axis of the tooth"

The embryogenic succession of the upper and lower cusps is as
follows :

UPPER MOLARS. LOWER MOLARS.

1. Paracone. 1. Protocol) id.

•2. Metacon^. 2. Metaconid.

3. Protocone. 3. Hypoconid.

4. Parastyle. 4. Entoconid.

5. Hypocone. 5. Paraconid.

In all these three genera the order of cusp development in the loir<'r
molars corresponds substantially with that given by the palreontological
theory.

Thus Woodward confirms Bose and Taeker's. results, and says
(p. 584) that "the order of cusp ontogeny is in entire accord with the
supposed order of cusp phylogeny as advanced by the supporters of the
Cope-Osborn tritubercular theory."

GROUP II. Tritubercular Molars. The most important form examined
by Woodward is the Tenrec of Madagascar, the genus Cciitete*. In the
adult the molars have usually been regarded as of columnar and typical
trituberculate type, consisting of an elevated internal cusp (protocone)
and more depressed external cusps (para- and meta-cone) ; as we shall
see, Woodward interprets the homologies of these cusps in an entirely
different manner. He observes (p. 573), in foetal development, that the
first superior molar is composed of a prominent main cone slightly inclined
inwards, this is undoubtedly the cusp determined as protocone of the
adult tooth, according to the Cope-Osborn theory, while growing outwards
low down from this main dental germ are two smaller ones, a slightly
more pronounced anterior cone and a less developed postero-external

212 EVOLUTION OF MAMMALIAN MOLAR TEETH

cone, corresponding respective!}7 to the para- and meta-cone as homologized
by Cope and Osborn. The embryogenic sequence is as follows :

1. Protocone. 1. Protoconid.

2. Paracone. 2. Metaconid.

3. Metacone. 3. Paraconid.

4. Hypoconid.

Similarly in the young jaws of Ericuhis (op. tit., p. 574) Woodward
finds that in the foetal first superior molar the protocone (as homologized
by Cope and Osborn) " forms the main mass of the tooth, while the para-
and meta-cones (as homologized by Cope and Osborn) form two rounded
external shelves, not at present conical " ; in the second superior molar
" the protocone and the small antero-external paracone are alone visible."

Summary of Woodward's Conclusions.

Woodward (1) entirely confirms the conclusions of previous authors
that in the lower teeth the evidence of embryology and palaeontology
is identical. (2) In (I.) the quadrituberculate Insectivora he strongly
reinforces the previous testimony of Rose, Kukenthal, Leche and Taeker,
namely, that the paracone develops first and the protocone is a secondary
shelf. (3) In (II.) the trituberculate Insectivora, on the contrary, he
brings forward facts which seem to support the paheontological theory
and honiologies ; it is in these animals, according to the Cope-Osborn
theory, that the protocone is still the highest and chief cusp of the
crown and thus should appear first in ontogeny, while in the quadri-
tubercular types the protocone being more depressed and the para- and
meta-cone being more elevated, the earlier appearance of the latter
might be interpreted as due to adaptive acceleration or coenogenesis.
In his discussion of these facts, however, Woodward does not admit
this interpretation of adaptive acceleration or cienogenesis, but maintains
that the Cope-Osborn honiologies are incorrect, and that the main
internal cusp in the trituberculate group is homologous with the antero-
external cusp in the quadrituberculate group, as shown in the accom-
panying diagram (Fig. 204). More in detail, he observes: "With
regard to the tritubercular upper molars of the C<'nt<ti<l(c, I should
conclude that the main cone of this type of tooth, usually termed
the protocone, was really the paracone : the whole tooth representing
only the antero-external triangle of such a form as Talpa, i.e. the
crescentic paracone with its two external cingulum cusps, the last
named being commonly but incorrectly described as para- and meta-cone
in Ccntdcs: that in the Cf/tf<'/iilcc no marked indications of the protocone
or metacone are as yet visible, while in Chrysochloris (Fig. 36-7) the
first indication of the protocone has appeared, viz., the internal shelf."

Thus he concludes as to honiologies that the reptilian cone of the
upper cheek teeth of the ancestral mammal is homologous with the

OBJECTIONS AND DIFFICULTIES AND OTHER THEORIES

213

protocone of the premolars, with the paracone of most molars, and with
the falsely called protocone of the molars of trituberculate Insectivores
and of the Jurassic genus jFVmA'.s/V.s (Fig. 204). Stated in another way
his conclusions are that, (1) the antero-extcrnal cone, or paracone above
and protoconid below, is the reptilian cone both in the molars and premolars,
a conclusion supported by the Danish observer Winge;1 (2) the falsely
called protocone of the upper molars is borne on an internal ledge or
extension of the base of the crown of secondary origin; (3) the meta-
cone is a similar backward development of the paracone which rises very
early in ontogeny long before the protocone; (4) the evidence advanced

ps. pa. me. mis. pg. ms. mis.

ps.

ruts

FIG. 204. Shading illustrates Woodward's hypothesis of the homological relations of the
tritubercular, zalambdodont molar types of Centetes (A) and Chrysochloris (B) to the sexituber-
cular dilambdodont molars of Tnl/ia (C). The whole of the Ctntetc* molar is supposed to correspond
to the anterior outer portion of the Talpa molar, while the " hypocone " of the Chrysochloris molar
is homologized with the antero-internal portion of the crown of Talpa. The posterior half of the
Talpa molar is supposed to be a neomorph or secondary upgrowth of the posterior side of the
tooth. But according to Gidley's interpretation the single V in Centetes and Chrysochloris is
secondary and represents the fusion of the two V's (pa, me) seen in Talpa.

The abbreviations are incorrect. For true interpretation see p. 225, Addendum.

in support of the tritubercular theory is insufficient to prove that the
upper molars primarily evolved on the lines of that theory; (5) owing
to want of material trituberculists have been led to assume that the
upper molars of the early mammalia passed- through similar stages to
those which they have determined for the lower teeth, and consequently
they have in most cases incorrectly identified the primary cone: (6)
that as regards the actual primary cone (paracone above, protoconid
below), its ontogeny recapitulates its phylogeny.

Ontogenetic Order according to Marett Tims.

Dr. F. W. Marett Tims '2 has reached partly identical conclusions
chiefly from the embryology of the teeth of the dog, namely, that the
cheek teeth of the dog are composed of

(1) A Primary Cone = paracone of ms, protocone of pms — protoconid
of lower ms and pms.

ll'0m Pattedyrenes Tandskifte isaer med Hensyn til Taendernes Former." Vul<n-l.-.
Meddel. fra den Xaturh. Fortn. i. Kjoheiiharn, 1SS'2, p. 15.

2 "On the Tooth-genesis of the Canidiv," Jour. Linn. Soc., Vol. XXV., Zoo!., No. 164,
1806, pp. 445-480.

214 EVOLUTION OF MAMMALIAN MOLAR TEETH

(2) A Secondary Cone = metacone of ms, tritocone of pms = hypoconid
of lower ms and pms.

(3) Three cinguluin cusps

(a) anterior = deuterocone, minute in all except lower sectorial

where it = the paraconid.

(b) posterior = shear (metastyle) of sectorial.

(c) centro-internal =• protocone of <lpm4 = heel (internal cingulum

of m1) = metaconid of lower teeth.

This theory agrees with the " premolar analogy" theory and with
the theory of Gidley in identifying the paracone of the molars, the
protocone of the premolars, and the protoconid of the lower premolars
and molars as the primary cusp.

Summary of Objections to the Embryological Theory.

In reply to this strong presentation of the facts of embryogeny
the following points may be made: 1. It is possible that the superior
tritubercular molar arises independently of previous protodont and
triconodont stages ; but the evidence afforded by the lower molar
teeth of Spalacotherium, which are incipiently tritubercular, tends to
connect these stages (Figs. 11, 12, and p. 32 />). 2. We do not posi-
tively know the structure of the upper molars in Spalacotherium ; but the
presumption that they were somewhat similar to the lower molars is
strongly supported by the fact that in the closely related genus Tri-
conodon the upper molars are exactly similar.* 3. The supposed upper
molars of Spalacotherium, named Peralcstcs by Owen, although of some-
what irregular form, present a pattern reversing that in the lower
molars of Spalacotherium, namely, with a large internal cusp, with a
closely connected antero-external crest surmounted by a cuspule, and
with a freer and more detached postero-external cone. Figure 12 in
Chapter I., not heretofore published, was directly taken from camera-
lucida outlines of the upper molars of Peralcstcs shaded with pencil.
It shows that the earlier figures of Osborn and Woodward were only
partially correct, and it tends to demonstrate, apparently beyond a doubt,
the existence of a large, prominent internal protocone, and of an antero-
external paracone shelf as originally homologized by Osborn, a postero-
external conical metacone, and an external cingulum embracing the
outer side of the crown at the base corresponding with the internal
cingulum of the lower molars of Spalacotherium.^ 4. The comparison
of a series of zalambdodont (tritubercular) and dilambdodont (quadri-

* [The presumed relationship between Triconodon and Spalacotherium is denied by
some authors. — ED.]

t [The supposed "protocone" of Peralestes might possibly be homologous with the
pai'acone of later mammals. — ED.]

OI'.JKCTIONS AND DIFFICULTIES AND OTHER THEOKIKS -J 1 ">

to sexti-tubercular) Insectivora favors the view that the high internal
cusp (protocone) in the former is homologous with the low antero-
internal cus}) of the latter (Figs. 04-80).* 5. Since all the orders in
which the paracone appears first (Dilambdodont Ensectivores, Carnivores,
Primates, Perissodactyls and other Ungulates) are derived from Eocene
forms in which the inner side of the crown is depressed, there has been
l»li'nf;i of time for embryogeuy to have become adapted to this condition
and for the main axis of the developing tooth germ to have become
shifted to the outer side (p. 54). G. The high protocone of trituberculate
Insectivores and the depressed ledge-like protocone of quadrituberculate
Insectivores alike fit into the talonid of the lower molars, and thus
both function like the protocone of mammals in general (Gregory).t

The Cope-Osborn theory, however, has to meet three further sets
of objections :

First, those derived from the comparison of the premolars and molars
clearly set forth by Wortman (see pp. 195, 142, 210).

Second, those derived from a restudy of the teeth of Jurassic
mammals themselves, developed by J. W. Gidley (p. 219).

Third, further comparison of the teeth of Insectivores, Chiroptera,
and other orders, developed by Gidley (see p. 124).

2. THE PREMULAI; ANALOGY THEOIIV.

The theory that the superior molars originally acquired trituberculy
in a manner similar to that which can be traced in tin: premolar meta-
morphosis has been designated in the introduction as the ' premolar
analogy theory ' (pp. 6, 7).

Premolar evolution, as the key to molar evolution, was suggested in
1880 by Huxley, in the following passage:1 "The exact correspond-
ence in plan of these teeth [of Otocyon~] is the more interesting, since, in
Gentries, it is easy to trace the successive changes by which the simple
and primitive character of the Mammalian cheek-tooth exhibited by the
most anterior pnumolar passes into the complex structure of the crowns
of the posterior teeth" (Huxley, Collected Papers, Vol. IV., p. 450).

It was also advocated by Schlosser,2 but upon different grounds,
namely : that in primitive jaws the upper overhangs the lower, and the
upper teeth fit not simply between but also slightly outside of the lower
•ones; accordingly it is more probable that the true protocone must be

1 " Review of the Cranial and Dental Characters of the Canidie,'' Coll tc ted Memoirs,
Vol. IV., p. 450.

2 " Die Entwickelung der verschiedenen Saugethierzahnformen im Laut'e der geologisclien
Perioden," Sonder-Alxli: au* den Verh. d. odontoloy. Get* //*<•/!., lid. 3, Heft 2 u. 3, 1S91. p. (.i

*[But see pp. 126, 2-2-3 Addendum. — ED.]
t [Erroneous. See p. 225. — ED,]

216

EVOLUTION OF MAMMALIAN MOLAR TEETH

sought in one of the outer cusps, as it is in the premolars, which, indeed,
begin their complication on the inner side. In so far as this observa-
tion applies to the Creodonta, Condylarthra, Ungulata, there is indeed
much force in it.

Scott has also expressed the belief that premolar and molar cusps are
in the main serially homologous ; in other words, that the molars origin-
ally evolved as the premolars did subsequently.

More recently Wortman l supports the premolar analogy theory on
pakeontological grounds, and advises the total abandonment of the theory
of trituberculy, asserting emphatically that the cusps in the molars were
added in exactly the same manner and in precisely the same order as
in the premolars (Fig. 205).

FIG. 205. Upper figure, Upper cheek teeth of Dissacus saurognathus from the Torrejon Forma-
tion, Stage II, Basal Eocene. Loicer figure, Upper cheek teeth of Mesonyx obtusidens from
the Bridger Formation, Middle Eocene. Dr. Wortman regards the Dissacus teeth as representing
the ancestral pattern of the Me&onyx teeth and believes the internal cusp or "protocone" of the
molars to be a secondary upgrowth of the basal cingtilum like the corresponding cusps of the
premolars.'2 After Wortman.

The two great facts which apparently support this theory are : first,
that we can actually follow the premolars passing by cusp addition from
the haplodont into the sexitubercular condition exactly like the molars ;
second, that according to the Cope-Osborn theory the protocones or
reptilian cones are on the outer side of the upper premolars and on
the inner side of the upper molars, certainly an anatomical paradox
which has never yet been explained away ; third, when it is further
considered that embryogeny or ontogeny supports this inference, that
in the embryos of the majority of mammals the antero-external cusp
(Osborn's paracone) of the true upper molars is the first to develop,
do we not secure a concurrence of testimony which seems irresistible ?

'["Origin of the Tritubercnlar Molar,"] Amur. Jour. Sri., Vol. XIII.. Jan. 1902,
pp. 94-95; Vol. XVI., Nov 1903, pp. 365-368.

2 [Dr. Matthew's recent investigations on the Meponychiclffi do not substantiate the
supposed facts on which this argument is based. — EL>.]

OBJECTIONS AND DIFFICULTIES AND OTHER THEORIKs

•JIT

Osborn (1904) presents New Palcvontological Difficulties in the
Prcmolar Analogy Theory.

The crucial test of the premolar analogy theory is the hoinology of
the protocone of the superior molar.

If the protocone is the antero-external cusp in the upper pre-
molars and the antero-internal cusp in the upper molars, the premolar
analogy theory falls to the ground, and the Cope-Osborn theory remains
thoroughly substantiated. From Osborn's latest contribution (1904) to
the tritubercular theory 1 we make the following abstract :

In addition to the evidence afforded by the upper teeth of Tri-
conodon, in which the main cone is central, and that seen in the upper

External or -maxillary

/

i.of.

A

Painted cr internal view ,

Fio. 20ii. Superior molars of Dryolextes Marsh. Upper Jurassic, Wyoming. A. Series of the
; side external and crown views. B. Series of the right side, external, crown and internal

i. o. f. infraorbital foramen. Other

left

views. Yale Museum, c, c, c, external and internal cingula.

abbreviations as in Fig. 12. (Cf. Fig. '207 (2) and p. 220.)

teeth of Peralestes (in the British Museum), in which the main cone is
internal (Fig. 12), and the upper teeth of Kurtodon (British Museum), in
which the main cone is internal (Fig. 13), the original theory was sup-
ported by Professor Marsh's statement that in the upper molars of
Dnjolcstcs (Yale Museum) the main cone was internal ; in each case
the main cone is believed to be the protocone or reptilian cone.

The two specimens here referred to in the Yale Museum exhibit
perfectly the structure of both crowns and fangs, of seven superior
molar teeth, bringing out the following important points (Fig. 206):

(1) The molars are sharply distinguished from the premolars, which
are bifanged teeth with simple, laterally compressed crowns.

Osborn, H. F., " Palteontological Evidence for the Original Tritubercular Theory.'"
Amer. Jour. &•/., Vol. XVII., Apr. 1904. pp. 321-3-23.

EVOLUTION OF MAMMALIAN MOLAR TEETH

/

Ga

2a

4a

5a

51)

>

la

'

6c

Fi CURE 207

OBJECTIONS AND DIFFICULTIES AND OTHER THEORIES I'll)

(2) The molar crowns are broadly transverse or triangular, and upon
the internal side of each is a large, conical, pointed cusp, j>r, supported by
a large stout fang, Fig. 206 A, m-6, m7 ; around the inner side of each
of these cusps is a delicate cingulum, Fig. 206 A, c.

(o) The I'.i-ti'i'iinl portion of the broadly triangular crown is supported
on two smaller fangs, Fig. 206 A, m6, m7.

(4) The external portion of the crown is depressed, and bears one
large antero-external cusp ? pa and one smaller postero-external cusp ?me
which is either partly worn away or less pronounced in development.

(5) Outside of this external wall there is also a faint basal cingulum,
c, c, c.

(6) Connecting these low external cusps with the elevated internal
cusp are two transverse ridges ; the anterior transverse ridge is higher
and stronger than the posterior.

This palseontological evidence appears to lend no support to the
evidence of embryogeny that the paracone (_£>«) or antero-external cusp
is the oldest cusp. On the contrary, it appears to prove that the
pointed, conical, internally placed cusp supported upon a single stout
fang is the main cusp or protocone, serially homologous with the main
cusp of the simple preceding premolar teeth.

Gidley s Rest it dy (1906) of Jurassic Mammals Supports Embryogeny
and the Premolar Analogy Theory.

It will be seen that embryogeny and premolar analogy concur in
evidence that the reptilian cone is represented in the antero-external cusps
of the superior molars. Mr. J. W. Gidley * now brings palaeontology to
the support of this position.

His conclusions are based on a fresh study of unworn superior
molars of Jurassic mammals in the U.S. National Museum, which were
not accessible to Osborn. These teeth are represented in Fig. 207,
which may be compared with the worn teeth of Fig. 206.

1 Gidley, J. W., "Evidence bearing on Tooth-Cusp Development," Proc. Washington
Acad. ScL, Vol. VIII., 1906, p. 106.

FIGURE 207. FROM GIDLEV.

Xos. 1 and 1«. Tricnnotlon? bimlcus Marsh (Atlantosaurus beds), left upper molars, m- and m3 ;
crown and external views. Six times natural size (Xo. -ii'.'s, I'.S.N.M.).

2, ~2/i, and 26. Dryohstes sp. (Atlantosaurus beds), left upper molars ; crown, external, and

posterior views. Seven times natural size (No. 2s4">, U.S.X.M.).

3. Druolc&t'S, first right upper molar, in' ; crown view. Eight times natural size (Xu. •>::'.',

U.S.X.M.).

4 and 4a. Dicrocynodon sp. (Atlantosaurus beds), left upper molars ; crown and external

views. Six times natural size (Xo. 2715, U.S.X.M.).
5, 5a, 56, and 5c. Paurodon sp. (Atlantosaurus beds), right lower molar, m._. ; rr,i\vn, external,

internal, and posterior views. Kight times natural size (Xo. 273M, L'.S.X.M.).
0, (3«, 06, and Gc. ? Pediomi/s sp. (Laramie beds), left upper molar ; ero\vn, external, posterior,

and anterior views. Bight times natural size (Xo. 5Uti2, U.S.X.M.).

7, 7a, and 76. Gen. et. sp. indt. (Laramie beds), left upper molar ; crown, external, and
anterior views. Eight times natural size (Xo. 507G, U.S.X.M.).

220 EVOLUTION OF MAMMALIAN MOLAR TEETH

lief erring to Osborn's figures of Dryolestes (Fig. 206), Gidley * says:
' But there are two important cusps not noted by Osborn, one an
external cusp placed anterior to the main external cusp, the other
a small but well-defined intermediate cusp appearing on the posterior
transverse ridge. Thus there are five distinct cusps instead of three,
as stated by Osborn, and these do not form a trigon in the sense that
this term has been used, for the main external cusp is in the middle
of the base of the triangle instead of forming one of its angles.

o O

" Considering the outer portion of the Dryolestes molar as
homologous to the three cones and two fangs of Triconodon, the
derivation of this type of tooth is much simplified, it being not so far
removed from the primitive reptilian condition, and though diverging
on different lines, is no more specialized, as a whole, than the Triconodon
type of tooth, the differentiation being carried on more rapidly in the
latter in the special development of the anterior and posterior lateral
cones and their accessory cusps, while in Dryolestes the specialization
has apparently been centralized in the development of the high,
narrow, heel-like cusp and its supporting fang on the inner side of
the molar.

' This view is strongly supported by the evidence obtained from
still another characteristic Atlantosaurus-beds type of molar represented
by Dicrocynodon. In this form, PL V,t fig. 4, the same primitive
arrangement of three cusps and two fangs is preserved in the outer
portion of the tooth, while on the internal side a large secondary
cusp has been developed differing widely in character from that of
Dryolcstes. This cusp is a laterally compressed cone supported by two
rudimentary fangs and is joined to the outer portion of the tooth by
a high, wedge-shaped ridge. The base of the inner cone is greatly
expanded antero-posteriorly, curving gently outward toward the
external portion of the tooth. Thus the crown, as a whole, is greatly
constricted medially with the inner and outer portions superficially
resembling each other.

" From these observations two important conclusions may be drawn :
First, that, leaving out of consideration the multituberculates, there
are among the mammals of the Atlantosaurus beds at least three
distinct forms of upper molars representing three primitive types of
about equal specialization apparently leading off' in entirely independent
lines. Probably only one of these, Dryolcstes, represents an ancestral
type from which the Upper Cretaceous and later forms possessing
trigonodont molars may have been derived. Second, that the evidence

*" Evidence bearing on Tooth-Cusp Development," Proc. Washington Acad. Sci.,
Vol. VIII., 190(i, p. 96.

t[Fig. 207.]

OBJECTIONS AND DIFFICULTIES AND OTHER THEORIES

UlM

derived from the Atlantosaurus-beds mammals entirely supports the
evidence of embryology and agrees in general with the ' pre-molar

000

a

ooo

3

E 4

Fio. 208. [Gidley's] Suggested Plujletic History of Two Types or Complex: Molars. [As in
Osborn's diagram [see Fig. 41, p. 61], the solid black dots represent the cusps of the upper
molars, the circles, those of the lower molars.] 1 to 6, Phyletic history of the "Trituber-
cular " type ; a to d, Phyletic history of the " Triconodont " type ; < , /, From the brachyodont
Triconodont stage to the bilobed hypsodont type of molar.

A, B, C, E, and G compare witli A, B, /', K, and G in Osborn's diagram [see Fig. 41, p. 01] ;
4, Dryolestes type, Atlantosaurus beds ('? Upper Jurassic) ; 5 and 0, Protolambda or Pediomvs type,
Laramie beds (Upper Cretaceous); d, Tricododon type, Atlantosaurus beds (? Upper Jurassic);
/, Palwolagus type, White River beds (Oligocene). From Gidley.

analogy ' theory. Thus, the evidence from all sources points over-
whelmingly to the conclusion that the primary cone is to be found
on the outer side in the upper molars of primitive trituberculate forms

EVOLUTION OF MAMMALIAN MOLAR TEETH

and in all forms derived from a tri tubercular type of tooth as well,
except where the main inner cone (protocone) has been reduced
secondarily. The opposite view held by the tri tubercular theory now
apparently stands on very insufficient evidence, and the proposition that
the protocone, of Osborn, represents the primary cusp is entirely without
support.

' The lower molars of the Atlantosaurus beds mammals furnish
abundant additional evidence along the line of conclusions regarding
the shifting of three cusps from a straight line to form the primitive
triangle. In such forms as Dri/olcstes and Paurodon \\Q have trituber-
culate molars in the primitive or forming stage, and, what is most
significant, the cusps resemble very closely, both in position and relative
proportions, those of the premolars of later types in their early stages
of transition to the molariform pattern. In the lower molars of
Paurodon the crown consists of a high, pointed cusp (protoconid),
centrally placed, a low posterior heel, a small antero-internal cusp
(paraconid), and a very small median internal cusp (metaconid). The
last two form the base of the trigonid. In Dryolestes both the trigonid
and the pimitive heel are somewhat more advanced in development.
In still other forms, such as Mcnacodon and Tinodon, the two internal
cusps are relatively large and the trigonid is fully developed, while the
heel, or talonid, is very small or entirely wanting. In all the paraconid
and metaconid are entirely on the internal side of the crown, and in
these and all the material examined there is not the slightest evidence

O

of any shifting of the cusps, but they seem to have arisen in the
positions they now occupy.1 In Paurodon the heel is apparently as
much or more developed than either of the internal cusps and seems
to have made its appearance even in advance of the metaconid. Also
the metaconid is still very rudimentary and is just budding off near
the base of the protoconid, but little posterior to its apex and midway
of the entire length of the crown, while the place of origin assigned
to it by the tritubercular hypothesis is already occupied by the
comparatively large heel.

' From these observations it seems apparent that the trigonid of
the lower molars is not the reverse of the trigon of the upper molars,
as held by advocates of the tritubercular theory, and the homologues
of the elements of the upper and lower -molars, as proposed by this
theory, are far from being apparent. (This also accords with the
conclusions of Winge.)

' The lower molars of Triconodon differ from any of the forms just

1 This is in accord with the general conclusions on tooth cusp development reached
by Herluf Winge as early as 1882. Vidensk Meddelelscr fra den naturhist. Forening
i Kjobenhavn, 1882, p. 18.

OBJECTIONS AND DIFFICULTIES AND OTHER THKniMKS

described. They are composed of three nearly equal cone-like cusp-
arranged like those in the upper molars of this genus in an antero-
posterior line. There is no cusp corresponding with the metacuni'l
in Dryolestes. There is a continuous basal cingulum on the inner face
of the crown, and the posterior cusp is in no way homologous, except
in position, to the heel in the lower molars of Piun-mlon and
Dryolestes.

' The mammals from the upper Cretaceous Laramie beds show a
great advance in development. The molars of the trituberculate forms
of this horizon have passed into a second well-defined stage of special-
ization which, though varying greatly in detail in the various types,
conforms in general to a distinctive pattern which may readily have
been derived from some Atlantosaurus-beds form, such as Dryolestes.
An upper molar of Pcdiomi/s Marsh, a typical example of the Laramie
tritubercular molar, compared with the corresponding tooth of Dryolestes,
presents the following differences and indicates the principal lines of
progression :

"(1) The main internal cusp (protoconc) is much broadened
antero-posteriorly ; (2) a second small V-shaped intermediate cusp
( protocomde) has been added; (3) the postero-external cusp (metacone)
has greatly increased, nearly equaling, both in size and importance, the
median external, or primary, cone (paracone), while the antero-external
cusp (parastyle) has remained small and undeveloped. A correspond-
ingly progressive development marks the trigonid and heel of the lower
molars."

Summary ami Conclusions [G

" Summing up the evidence derived from this preliminary study, the
following conclusions are suggested :

' 1. That the evidence obtained from the Mesozoic mammal teeth
furnishes no support to the tritubercular theory in so far as it involves
the position of the protocone and the derivation of the trigonodont tooth
from the triconodont stage through the shifting of the lateral cones
outward in the upper molars and inward in the lower molars.

" '2. That it supports entirely the embryological evidence that the
primary cone is the main antero-external cusp, or paraconc, having
retained its position on the o/'/s/W/ in most upper molars (see exceptions
above, p. 124).

"3. That it agrees in the main with Huxley's ' premolar analogy
theory, as supported by Scott.

" 4. That the molars of the Multituberculates, Triconod'on, Dryolestes,
and Dicrocynodon, were apparently derived independently from the simple
reptilian cone : hence the supposition follows that the trituberculate type

224 EVOLUTION OF MAMMALIAN MOLAR TEETH

represents but one of several ways in which the complex molars of
different groups may have been derived.1

" 5. That in the forms derived from the trituberculate type
of molar the order of succession of the cusps is not the same in all
groups, and apparently homologous elements are sometimes developed
from different sources. Hence it follows that no theory involving an
absolute uniformity of succession in the development of complex molars will
hold true for all groups of mammals.

' In the foregoing pages I have restricted the use of Osborn's
tooth-cusp nomenclature for the reason that, in this particular discussion,
there are some cases in which it is not strictly applicable and might lead
to confusion.

" On similar grounds Dr. Wortman2 has expressed the opinion that all
attempts to establish a tooth-cusp nomenclature founded on supposed
homologies are ' foredoomed to failure ' and should be entirely abandoned
as 'useless and confusing.' I agree with the general sentiment expressed
(op. cit., p. 366) that, owing to the adoption of different plans in different
groups of mammals for increasing the complexity of their molars, no
terminology founded on the basis of cusp homologies can be made strictly
applicable to all the mammalia. I do not, however, consider this
sufficient ground for abandoning absolutely so convenient a system of
nomenclature as that proposed by Osborn. Granting that many of the
terms proposed are founded on mistaken homologies, it does not
necessarily follow that they need be in the least confusing, as suggested
by Wortman. For in any system used, in order to make that system of
greatest convenience and highest utility, the names once adopted should
be permanent and not subject to transfer or substitution on any ground
of changed conceptions of homologies or history, for the same reason that
generic and specific names are retained regardless of the fact that they
may have been given to denote some supposed affinity or characteristic
which may later have proved entirely erroneous.

" Viewed from the nomenclature standpoint, therefore, the convenient
names proposed by Osborn have come to assume an individuality which
conveys a far more definite meaning than any purely descriptive terms,
be they of relative position or supposed homologies. Moreover, they
have the valuable advantages of clearness and brevity in description.
On these grounds, in the opinion of the present writer, and for the added
reason that great confusion would inevitably result from any change in a
terminology that has found its way into so many publications, Osborn's
nomenclature should be retained as originally proposed. Thus the term

Somewhat similar conclusions have been reached from different reasoning by E. S.
Goodrich, M. Tims and others.

*Amer. Journ. Science (4), Vol. 16, 1903, 265-368.

OIUKC'I'IOXS AM) DIKKHTLTIKS AND OTHKII THKOlII KS 225

'protocone' always means the main anterointernal cusp <>f u normal
upper iiiitliirifonn tooth, whether that element is regarded as the original
primary cusp or otherwise.

'The objection that the terms are not universally applicable is
scarcely worthy of consideration since they are widely applicable to the
p-eat majority of mammalian molar types, without in the least interfering
with the use of terms descriptive of 'relative position only,' which may
be used in any cases where Osborn's terms do not apply.''

Ailili ad ii in, December, 1000 (7 //v//n /•//;.

A careful study of the excellent series of Insectivore skulls in the National Museum,
while fully confirming Gidley's interpretation of the homologies of the molar cusps
(p. 124), yet does not sustain his broader conclusion that the paracone in normal
tritubercular molars is older than the protocone.

(1) The upper molar pattern of I'utainoyale (Figs. 69, 69 «), is fundamentally similar
to that of Mt/ogale (Fig. 69 F). Its high central cusp is evidently composed of an
enlarged anterior and much reduced posterior cusp, and these cusps have the same
spatial and functional relations with the cusps of the lower teeth as the para- and
metacones of Myogale, while the low internal cusp functions like the protocone.
Sotf, /»</*,,/ also has a low internal platform like that in JJyogale, but much reduced,
bearing two small cusps which function like the proto- and hypocones of Myogale.
The main high pointed cusp in Solenodon is surely equivalent to the large para- and
closely appressed reduced meta-cone of Potamogale. This high main cusp (pa + me)
is evidently less specialized in Potamogale than in Solenodon, Centetes, etc., and
further, the small basal, proto- and hypocones are in a more primitive condition in
Solenodon than in Centetes, where they are vestigial and now functionally replaced
by the enlarged high para- + metacone. In Hemicentetes the true protocone has
vanished and the teeth parallel the carnassials of Carnivora (p. 137), where the
protocone is likewise secondarily reduced.

(2) Thus, contrary to the more usual view, the " trituberculy " of Centetes and
clii-ysochloris is a secondary acquirement or pseudotritubercufy, and the fact that
the high main cusp (pa + me) in Centetes develops in ontogeny more rapidly than the
vestigial internal ledge, or protocone, does not prove that in Potamogale, Myogale,
and in normal tritubercular molars the outer cusp (paracone) appeared phylogeneti-
cally earlier than the internal cusp (protocone).

226

EVOLUTION OF MAMMALIAN MOLAR TEETH

FIG. '20f». Internal view of the right upper and lower cheek teeth of the Zalambdodont
Insectivore Cfntitm (No. 1), the Marsupial Diddphys (No. "1), the Dilambdodont Insectivore
Erinaceus (No. 3) and the Titanotheroid Ungulate Telmatherium (Xo. 4). The main internal cu«p

small talonid (Olikc a normal protocone, but is somewhat external to it, and occupies the position

of an inwardly grown paracone.

(See Addendum, p. 225).— ED.]
p = protocone.
h =hypocone.
m = metaconid.

t = talonid.

OBJECTIONS AND DIFFICULTIES AXI) OTHER THEORIES 227

Conclusion (Od><>,-// ).

It must not lie understood by the reader that the author of this
volume is doggedly maintaining a theory of the origin of the upper
molars, in which he has had a part, simply from personal reasons.
On the contrary he believes the question to be still .s/'fr jiulice and
will be the first to acknowledge his error — if error it prove to be.

The author moreover feels the full force of the very strong evidence
arrayed against the Cope-Osborn view. The evolution of the upper
molars is certainly not so simple as it at first appeared. It was
never maintained by Osborn, as is proven in his full series of
papers here reproduced,1 that the molar of Triconodon (Fig. 11 a),
gave rise to that of Dryolestcs (Figs. 206, 207, Nos. 2, 3), or even
that all mammals passed fully into a triconodont stage, such as that
of Ainphilestes (Fig. 5). The starting point of the superior molars
was supposed to be an extremely primitive triconodont type either
with the cusps in line or with the pattern of molars of the
Spalacotherium (Fig. 11) type inverted as shown in Peralestes (Fig. 12).

pp. 8, 33.

CHAPTER X.

RECTIGRADATIONS AS A RESULT OF LATENT OR POTENTIAL
HOMOLOGIES IN THE TEETH.

BY reotigradations I refer to the origin of new cusps or cuspules which
appear determinately, definitely, orthogenetically in both the upper
and lower teeth — quite independently in different orders of mammals,
and separated perhaps by vast intervals of time.

There is some law of predisposition operating here. If it were
not for this law the cusps of the teeth of mammals would present
an infinite variety of origin, whereas they actually present a singular
uniformity of origin except in the multituberculates and other possible
exceptions noted in the preceding chapters. It is the modelling of
the cusps after they appear which gives the infinite variety.

We do not know what conditions this " law of predisposition " :
we only see evidence of the influence of community of origin or
hereditary kinship.

In 1902 I supposed this law was what E. Hay Lankester meant by
Homoplasy, as shown below, but it appears from his letter cited below
(p. 239) that I was mistaken — Lankester's homoplasy is equivalent to
analogous evolution, to parallelism, or convergence.

1. HOMOPLASY AS A LAW OF LATENT oit POTENTIAL HOMOLOGY.

(Reprinted under the title given above from The American Naturalist, Vol. XXXVL,

April, 1902, No. 424, pp. 259-271.)

My study of teeth in a great many phyla of Mammalia in past times has con-
vinced me that there are fundamental predispositions to vary in certain directions ;
that the evolution of the teeth is marked out beforehand by hereditary influences
which extend back hundreds of thousands of years. These predispositions are
arovised under certain exciting causes and the progress of tooth development takes
a certain form converting into actuality what has hitherto been potentiality.

Science, N.S., Vol. VI., No. 146 (Oct. 15, 1897), pp. 583-587.

Ill previous communications, as shown in the above quotation, I
have spoken of the " potential of similar variation," as covering cases

CUSP RECTIGRADATIONS

229

• i

the independent evolution of identical structures in the teeth of
different families of mammals, especially in relation to the homologous
" antecrochet " and "crochet' folds in the teeth of horses, rhinoceroses,
and we may now add, of titanotheres (Oshorn, '94, p. i^OS . In tin-
present communication I propose to treat somewhat more fully of
the same phenomenon, as a special form of hoinology which has heen
clearly defined hy Lankester in 1874 as homoplasy, hut into which
paleontology has brought the idea of " potential."

TlFK P.IJOAD SlGXlFK'ANCK OF ANAF.QCY.

We are familiar with the classic distinction of analogous organs
as having a similarity of function: ancdoyy (Owen, '43, p. 374), "a
part or organ in one animal which has the same function as another
part or organ in a different animal";
Lankester ('70): "Any two organs
having the same function are analogous,
whether closely resembling each other
in their structure and relation to
other parts or not : and it is well
to retain the word in that wide
sense." Analogous organs may or
may not be homologous. " Analogy '
is therefore an extremely broad and
comprehensive term, and it appears
that we must include under it all
cases of the similar evolution of organs
either of common or of different origin
due to similarity of function. For
exam} tie, the " analogous variation "
of Darwin, the " homoplasy ' of Flf, .m Fourth upper premolar and first

1 flnlcpsjfpv ill rnvf- it lpa«t flia "nnn molar of primitive ungulates. A, Ev.protogo;>i<i. ;
111 part at least, me B> Hln.a<.otheriUM. y0t believed to be geneti-

vpro-piw " of fipvniflii wvitprQ rlip call-v related, J'et exhibiting independent i.r

W1 homoplastic evolution of homologous cusps.

" homomorphy ' of Furbringer, the

" heterology," "parallels," and "parallelism" of Hyatt, of Cope ('68,
also Origin of the Fittest, p. 96), of Scott, and of most American
writers, are all illustrations of analogy and may Vie very misleading
as to hoinology.

As Scott observed in 1896, "Parallelism1 and

convergence

of

JThe term "parallelism'' was employed by Cope in his essay of 1868 on the "Origin
of Genera" (reprinted in the Origin of the Fittest) in two quite different senses: first,
in relation to recapitulation in ontogeny, — "Those which accomplish less [stages] are
parallel with the young of those which accomplish more [stages] " ; second, quite in the
modern sense (op. cit., pp. 96-104) of independently acquired resemblances in different

O

o

230 EVOLUTION or MAMMALIAN MOLAR TEETH

development are much more general and important modes of evolution
than is commonly supposed. By parallelism is meant the independent
acquisition of similar structure in forms [i.e., animals] which are
them selves nearly related, and by convergence such acquisition in
forms [i.e., animals] which are not closely related, and thus in one or
more respects come to be more nearly alike than were their
ancestors."

The term " homoplasy " (Lankester) has been long used by the
writer and others in a somewhat similar sense, but it is not equivalent
either to " parallelism " or " convergence." As will be seen below,
the fundamental idea is different, because homoplasy always involves
homology, while parallelism and convergence may or may not involve
homology.

ANALOGY IN EVOLUTION.
Analogous Variation (Darwin). Similar congenital variation in more or less

distantly related animals and plants.
Parallelism. Independent similar development of related animals, plants,

and organs.
Convergence. Independent similar development of unrelated animals, bringing

them apparently closer together.

Homoplasy (Lankester) (1 Homomorphy, Fiirbringer). Independent similar
development of homologous organs or regions giving rise to similar
new parts.

In brief, analogy embraces similar changes due to similar adaptation
in function both in homologous and in non-homologous organs, both in
related and in unrelated animals.

THE LIMITED SIGNIFICANCE OF HOMOLOGY.

Owen ('43, p. 379), Lankester ('70), and Fiirbringer have especially
defined and elaborated the very ancient conception of homology, as
employed by Oken, Geoffroy St. Hilaire, and Vicq d'Azyr : homology
(Owen, '43), "the same organ in different animals under every variety
of form and function"; homogeny (Lankester, '70): "Structures which
are genetically related, in so far as they have a single representative
in a common ancestor, may be called homogenous." E. B. Wilson ('95,
pp. 101-124) has shown that the comparative anatomical test of
homology is more reliable than the embryological. Gegenbaur ('98,
pp. 23-25) has given a full presentation of the distinctions as the
basis of comparative anatomy; in his recent great work ('98, p. 23)

groups. As employed by Scott in his essay "On the Mode of Evolution in the Mam-
malia" ('91, pp. 363-867), "parallelism" is used in a very broad sense as affecting the
skeleton and teeth, on the principle "that identical modifications of structure, con-
stituting evolution of types, have supervened on distinct lines of descent." as embracing
not only single characters but whole series of them.

crsi- I;K<TI<;I;.\I>ATIOXS

231

he presents the matter in terms which may lie briefly analyzed with
the usages of other authors, as follows:

I. HOMOLOGY, GENERAL : as of vertebrae and limits.

1. IJOMOTYPY : as of opposite limbs, eyes, kidneys, etc.

2. HOMODYNAMV : (in part the "general," in part the "serial," homology of

Owen ; the " meristic " homology of Bateson). Corresponding limbs,
parts, segments (e.g., the humerns and femur) on the same side of the
body.

?,. HMMOXOMY : parts which are in the same transverse axis of the body, or
on only one section of the longitudinal axis ; e.g., the rays of the fins
of fishes, the single fingers and toes of the higher vertebrates are
homouomous organs.

II. HOMOLOGY, SPECIAL: (the "homogeny" of Lankester).

1. COMPLETE HOMOLOGY of elements which have retained their relations un-
changed, as of single bones from the Amphibia to the Mammalia.
•2. INCOMPLETE HOMOLOGY, as of organs which have either gained new parts

or lost certain of their parts.
«. (li'fective, as in comparison of fins of teleosts and of selachians.

b. augmentative, as in the heart of cyclostomes and of the higher vertebrates.

c. imitative, as where difFereut vertebrae connect with the ilium and become

sacral.

III. HOMOMORPHY (Fu'rbringer) : from these homologies certain structures are to
be distinguished as homomorphic which are more or less similar to each
other but stand in no phylogenetic connection.1

Homomorphy comes nearest, as we understand it, to the " homoplasy " of
Lankester, but the latter term has the priority of definition.

metacone-

DlSTINCTION BETWEEN HOMOGENOUS AND HOMOPLASTIC ORGANS.

Iii the strictest sense, special or genetic homology, the " homogeny
of Lankester, is the only absolute homology. For example, in all
four-limbed vertebrates, or Tetra-
poda (Credner), the first and
second phalanges of the tibial
digit or hallux are homogenous;
the earliest tetrapods had such
phalanges, so far as we can judge
from both paleontology and em- metaconule.
bryology, and all others are deri-
vatives.

But suppose we should dis-
cover that these tWO phalanges Km. ill. Ideal embryonic ground plan of rhinoceros

' molar, showing relation of primitive t-usps to the folds

had originated independently in and crests,
several different classes of verte-
brates, and were not derivatives; should they then be considered analogous
or homologous? "Again," says Lankester ('70), "it may perhaps be

1 Literally translated from Gegenbaur.

hypocone-

-protoconule
protocone

232

EVOLUTION OF MAMMALIAN MOLAR TEETH

admitted that the common ancestors of the Osseous Fishes and Mammalia
had a skull of decidedly undifferentiated character, with a much less
amount of differentiation than is observed in the skulls of either of
these groups. It is only in so far as they have parts represented
in the common ancestor that we can trace homog<:ny in these groups ;
and yet the homology of a vast number of bones in the skull of the
two is discussed and pointed out." Suppose, accordingly, that in the

formation of dermal rooting

paracone

metacone

crochet
-postfossette

prefossette -|

antecrochet

protocone-

FIG. 212. Molar tooth of an Upper Miocene rhinoceros
( T< /' ocerns), showing origin of secondary folds.

bones in different orders of
tishes a pair of bones corre-
sponding in position to the
parietals should arise inde-
pendently, or that in the
evolution of the teeth cusps
should arise independently
having the same form and posi-
tion,— what criterion should
be applied ? All such struc-
tures are habitually regarded
as homologous, yet it is apparent that they are not derivatives of each
other and therefore not homogenous or homologous in the strictest
sense.

Such cases of independent evolution of apparently homologous organs
I recently proposed J to signify as potential, or latent homology, borrowing
the term " latent " from Galtoii as indicative of a germinal rather
than of a patent or adult character, and the physical term " potential "
as expressing the innate power or capacity to develop a certain organ.
But my colleague, Prof. Edmund B. Wilson, pointed out to me that
such cases were almost exactly covered by the original definition of
the word " homoplasy " by Lankester ("70, p. 42), as shown in the
subjoined quotations from his essay :

When identical or nearly similar forces, or environments, act on two or more
parts of an organism which are exactly or nearly alike, the resulting modifications"
of the various parts will be exactly or nearly alike. Further, if, instead of
similar parts in the same organism, we suppose the same forces to act on parts
in two organisms, which parts are exactly or nearly alike and sometimes homo-
genetic, the resulting correspondences called forth in the several parts in the
two organisms will be nearly or exactly alike. I propose to call this kind of
agreement homoplastic or homoplasy.3 . . . What is put forward here is this : that
under the term "homology," belonging to another philosophy, evolutionists have

1 In a communication before the National Academy of Science, Nov. 13, 1901.

2 Italics are mine.

:i At this time Lankester accepted Herbert Spencer's Lamarckian views. Subsequently
he abandoned the mechanical inheritance theory for the pure natural selection theory.

rrsp KK(TH:i:.\i)ATK»xs

described and do describe two kinds of agreement, — the one, now proposed to
be called "homogeny," depending simply on the inheritance of a common part;
the other, proposed to be called " homoplasy," depending on a common action of
evoking causes or moulding environment on such homogenous parts, or on parts
which for other reasons offer a likeness of material to begin with.

Homologv thus includes -

[ Homogeny.

It follows that subsequent writers, including myself, have misused
the term "homoplasy," confusing it with " parallelism " and "con-
vergence," which, as we have seen, may affect absolutely non-homologous
structures. Homoplasy should be confined to structures in which there
is an element of homology.

Independently of Lankester (that is, not familiar with his paper)
I had therefore reached a similar conclusion thiough years of observation
in paleontology. I would now like to expand an idea which he also
lightly suggested in 1870 in the words, "or on parts which for other
reasons show a likeness of material to begin with!'

THE LAW OF HOMOPLASY AS ix PART IDENTICAL WITH DEFINITE

OK DETERMINATE VARIATION.

As observed in the evolution of the teeth especially, homoplasy
appears to be of very great importance, not on the technical grounds
of uniformity in nomenclature, but because it seems to coincide with
the principle of definite or determinate evolution, a principle which

me.

or.

me.

pa.

pr.

Fin. -J13. Superior molars of primates, Anaptomorphus to Homo, showing independent or
homoplastic origin of the hypocone, hv, from the cinguliiin.

may be of wider application.1 From the time of the " < higin of
Species" it has been admitted that evolution, so far as it depends
upon variation, is not in every possible direction, but is limited to
certain changes, the expression of certain hereditary or constitutional
causes which we do not in the least understand. The evolution of
the teeth of mammals enabled me in 1889 to give many concrete

1 See especially the correspondence of Darwin and Asa Gray ; also Osborn, The
Palfeontological Evidence for the Transmission of Acquired Characters, Xafiire, Jan. 9,
1890; the Orthogenesis and Orthoplasy of Kimer, Lloyd Morgan and Baldwin; Baldwin's
Dictionary of Philosophy and Pxyclioloyy, Vol. I., p. '243.

234

EVOLUTION OF MAMMALIAN MOLAR TEETH

illustrations of this principle and to show that variation is hardly
the proper term to apply to rudiments which do not arise in a
variable but in a h'xed manner.

It appears that von Waagen suggested the term " mutation " for
immeasurable variations somewhat similar to these. Scott in 1891
('91, p. 388) pursued the idea further in the following striking passage:
" These facts at least suggest the possibility that individual variations
are not incipient species, but that the causes of transformation lie
deeper, and act with more or less uniformity upon large numbers of
individuals. It may, perhaps, be the outcome of future investigations,
that while variations are generally due to the union of changing hereditary
tendencies, mutations are the effect of dynamical agencies operating

Fie. '214. Superior molars of primates. A, Adapis; B, //yo/woc/us ; C, Xot/iarctus. Showing

homoplastic cusps, hi/, ml, ps, ins, mtx.

long in a uniform way, and the results controlled by natural selection.
While this may be true, a great many facts must be gathered in its
support, before it can be regarded as more than a suggestion." Scott
subsequently, in his article " Variations and Mutations," expanded this
idea : " Bateson's results, as compared with those of paleontology, confirm
this distinction in many significant ways and emphasize strongly the
difference between variation and that steady advance along definite
lines which Waagen called mutation." This paper in turn is said to
have influenced de Yries's recent work, Die Mutationstheorie.

It is a singular coincidence that the human teeth were selected
by both Empedocles and Aristotle to test the "survival of the fittest'
versus the purposive or teleological theory of evolution. I pointed
out in the papers above referred to (Osborn, '89, pp. 561-566; '90)
the significant fact that new cusps of the molar teeth do not appear
at random, but at certain definite points; that they are at first so
minute that they can barely be perceived, so that it is difficult to
theoretically assign them a survival value in the struggle for existence ;
that the mechanical or Lamarckian explanation is the only one which
can be offered:1 I laid the chief stress, however, not upon the
mechanical explanation, but upon definite or determinate origin, and
this has been confirmed by the subsequent study of thousands of

1 Ryder and Cope confidently advanced the mechanical explanation ; it is not
without grave difficulties, owing to the lack of an heredity theory.

<rsi> RKCTIGRADATIONS

235

teeth in different families of iii;mnn;ils. The still more significant
fact that this definite and determinate evolution was proceeding
independently in a great many different families of mammals did
not at the time impress itself so strongly upon my mind.

If molar teeth are found independently evolving in exactly similar
ways in such remote parts of the world as Switxerland, Wyoming,
and Patagonia, it is obvious that the process is not go\enied by
chance hut represents the operation of some similar or uniform law
deduced from the four following considerations :

Firstly, the teeth differ from all the other tissues and organs of
the body in being preformed, beneath the gum.1 Unlike all other
organs they are not modilied, improved, or rendered more adaptive
by use ; on the contrary, after the first stage of wear, the longer
they are used the more useless and less adaptive they become.

Fii.. -2 10. Superior molar of JI/e'/v/c/(//>/"'---. showing styles ps, ms, ruts, and conules pi, ml,
homoplastic with those of the wholly unrelated primate molar, Fig. u, C.

Thus, new structures in the teeth do not first appear as modifications
(as distinguished from congenital variations) in course of life, as is
so often if not invariably the case with new structures in the
skeleton. New cusps, folds, crests, and styles are invariably con-
genital. Thus, of all organs of the body the teeth most exclusively
and purely represent the current of stirp, germinal, or constitutional
evolution.

Secondly, the teeth are, nevertheless, among the most progressive
organs in the body. Whereas the adaptation of the skeleton, among
the mammals at least, is by a constant loss or numerical reduction
of parts, the adaptation of the teeth -is by a constant addition and
modelling of parts (Osborn, '88, pp. 1067-1079).

Thin////, according to the present paleontological evidence many
of the different families and orders of mammals diverged from each
other at a time when they possessed three cusps on the upper molar
teeth and from three to five cusps on the lower molar teeth. This
being the case, only the cusps comparable in different orders of

1 The importance of this fact was first pointed out to me l>y Prof. 10. B. Poulton
of Oxford.

236 EVOLUTION OF MAMMALIAN MOLAR TKK'I'H

mammals with these original three upper and five lower cusps are
derivatives or homogenous.

Fourthly, it follows that the new cusps of the teeth furnish an
example of homoplasy independent of the individual modification.

Thus, we may say that in the teeth at least homoplasy involves
a law of latent or potential homoloyy, without professing to understand
what is its significance.

We should, a priori, expect that if additional cusps were added
independently in different families and orders of mammals in different
parts of the world, under highly different conditions, the teeth of the
higher Mammalia would present very great diversity. As a matter
of fact, the new cusps in different families are absolutely uniform up
to a certain limit.1 In the twenty-three orders of placentals and in
the seven marsupial families, many of which are adaptively equivalent
to orders, the independently developed fourth to eleventh cusps of
the upper molars, if so many are developed, are uniform and may
be termed homologous ; the eight cusps and folds succeeding the
original homogenous three arising, if at all, at similar points and
presenting a latent homology or homoplasy. The record in the upper

molar teeth stands thus :

HOMOLOGY.

HOMOGENY HOMOPLASY

Primitive three cusps common to all Cusps or folds which are or may be

mammals. independently developed in different

orders.

Protocone Hypocone

Paracone Metaconule

Metacone Protocol! ule

Parastyle
Mesostyle
Metastyle
Protostyle
Hypostyle

This expresses the comparison of mammals as a whole. Within
many of the orders, such as the Perissodactyla, which arise from six
cusped ancestors, the homology is different.

HOMOLOGY.

HOMOGENY HOMOPLASY

Protocone Parastyle

Paracone Mesostyle

Metacone Crista

Hypocone Crochet

Protoconule Antecrochet, etc.
Metaconule

1 The excess of this limit is in multitubercnlism, or polybunodouty, where cuspules
are indefinitely multiplied.

CUSP KK<TH!KAI>ATin\S 237

The elements to which these terms are applied are lu-st exemplified
in the molar teeth of some of the primitive horses ! Fig. LMTtj.

The teeth are by no means the only structure* which evolve
under this principle, the skull, vertebral column, and limbs also
evolving under it more or less completely: but the teeth ;iltord a
singularly beautiful illustration of it because they exclude individual
modification.

The chief object of this communication is to enforce the recognition
of homoplasy as something which must be accounted for. These
homoplastic cusps do not arise from selection out of fortuitous varia-
tions, because they develop directly and are not picked from a number
of alternates. Xeither does it appear that the mechanical- inheritance
theory, if granted, would produce such a remarkable uniformity of
result. We are forced to the conclusion that in the original trituber-
cular constitution of the teeth there is some principle which unifies
the subsequent variation and evolution up to a certain point. Herein
lies the appropriateness of Lankester's phrase, "a likeness of material
to begin with."

Philosophically, predeterminate variation and evolution brings us
upon dangerous ground. If all that is evolved in the Tertiary molar
tooth is included in a latent or potential form, in the Cretaceous molar
tooth we are Hearing the emboitement hypothesis of Bonnet or the
archetype of Oken and Owen. Embryologists have recently gotten
into the same dilemma, and my colleague, Wilson, has proposed to
drop the idea " homology ' altogether and substitute the idea
"equivalent." In the present case, however, I think we have to deal
with homology, or, more strictly, with a principle intenin(!in/< I, firm/
limnology and analogy.

In a paper recently read before the American Morphological Society
(December, 1901), this author has urged the necessity of adhering
as closely as possible to the historical standard in the embryological
study of homology, and of avoiding the use of the term " homology '
when this standard is not available. He therefore suggests for
descriptive purposes the use of the non-committal terms "equivalent"
and " homoblastic," the former being applied to embryonic structures
of like fate (i.e., giving rise to homologous parts), the latter to those
of like embryonic origin. The only decisive test of the homology is
historic community of derivation (i.e., homogeny).

238 INVOLUTION OF MAMMALIAN MOLAR TEETH

BIBLIOGRAPHY.

'68 COPE, E. D., u On the Origin of Genera," Proc. Acad. Nat. Sci. Phila., October,

1868.

'87 COPE, E. D., The Origin of the Fittest, 8vo, New Yurk.

'98 GEGENBAUR, CARL, Vergleichende Anatomie Jer Wirlelthiere, 8vo, Leipzig, 1898.
'70 LANKESTER, E. R., " On the Use of the Term ' Homology ' in Modern Zoology

and tlie Distinction between Homogenetic and Homoplastic Agreements,"

Ann. Mag. Nat. Hist., Ser. 4, Vol. IV., pp. 34-43.

'89 NEUMAYR, M., Die Stiimme des Thierreiches, 4to, pp. 60-61, Vienna, 1889.
'88 OSBORN, HENRY F., '' The Evolution of Mammalian Molars to and from the

Tritubercular Type," Amer. Nat., Vol. XXII. (December, 1888), pp. 1067-1079.
'89 OSBORN, HENRY F., "The Pahvontological Evidence for the Transmission of

Acquired Characters," Amer. Nat., Vol. XXIII. (1889), pp. 561-566.
'90 OSBORN, HENRY F., " Evolution and Heredity," Biological Lectures, 1890, Marine

Biological Laboratory, Woods Hole.
;94 OSBORN, HENRY F., " The Rise of the Mammalia in North America," Proc.

Amer. Assoc. Adv. Sci., Vol. XLII. (1893), pp. 187-227.
'43 OWEN, RICHARD, Lectures on the Comparative Anatomy and Physiology of the

Invertebrate Animals, delivered at the Royal College of Surgeons in 1843, 8vo,

London, 1843.
'91 SCOTT, W. B., "On Some of the Factors in the Evolution of the Mammalia.

On the Osteology of Mesohippus and Leptomeryx, with Observations on the

Modes and Factors of Evolution in the Mammalia," Journ. of Morph., Vol.

V., No. 3 (1891), pp. 379-402.
'94 SCOTT, W. B., "On Variations and Mutations," Amer. Journ. Sci., Vol. XLVIIL

(November, 1894), pp. 355-374.
'96 SCOTT, W. B., " Palaeontology as a Morphological Discipline," Biological Lectures,

1895, Marine Biological Laboratory, Woods Hole, pp. 43-61, Boston, 1896.
'01 DE VRIES, HUGO, Die Mutationstheorie. Yersuche und Beobachtungen iiber die

Entstehung von Arten im Pflanzenreich, 8vo, Leipzig, 1902.
'95 WILSON, E. B., " The Embryological Criterion of Homology," Biological Lectures,

1895, Marine Biological Laboratory, Woods Hole, pp. 101-124, Boston, 1895.

'2. LANKESTER'S REPLY TO THE PRECEDING ARTICLE.

Dr. E. Kay Lankester demurred from Osborn's interpretation of his
definition of " homoplasy " as shown in abstract in the following letter :

" . . . . Homoplasy does not demand an element of homology. I
expressly say ' homogenous parts or parts which for other reasons offer
a likeness of material to begin with.' That alternative entirely
destroys your contention. I recognized (as hitherto combined under
one term ' homology ') only homogeny — or hereditary quality — and
homoplasy or moulded non-heredity quality. The ' likeness ' due
' to other reasons ' than homogeny — above spoken of — cannot be homo-
geny. The ' likeness ' which clearly enough is included and pointed
to in the whole paragraph — as favouring the action of homoplasy — is
either a likeness of true homogeny (that is of form and relation
inherited), or a likeness of similarity in material, in position, or in
initial form — not due to close homogeny — but possibly a likeness of

CUSP RECTN; i!. \DATIO\S 239

such a general character as the 'likeness of material' (noi of elaborated
form and parts) in two epidermal surfaces. Thus the beak nt' tin1
liinl and of the turtle might be developed homoplastically from a
common ancestor's snout which had no beak — but in both the bird-
line and the turtle-line horn-producing epidermis is tin- ehief material
bi-ouglit into greater development.

"So in the bi-chambered hearts of bird and mammal the -unit, rial
is muscular tissue and the peculiar muscular tissue of the heart.

" I should like to know what cases of ' convergence ' you can cite
in which there is not some remote homogeny as to the tissues at
work, which is a totally different thing from inherited community of
the special form considered. You would not call a four-chambered
seed-vessel and a four-chambered heart cases of convergence nor of
homoplasy ! (except in the very remotest degree which if entertained
makes all shape and structure in a minimal degree homoplastic with
all other). Can you name cases of convergence or parallelism which
are not covered by the definition I gave of homoplasy ? What organs
are parallel in any two animals and yet have no likeness at all — even
the most general — in their material.

' Yours sincerely,

" E. HAY LANKESTKK.

" P.S. — Every living thing owes its properties to homogeny, that
is to say, its fundamental properties to a homogeny common to it
with all other living things, and as you run down group within
group there must be a more and more specialized homogeny affecting
members of smaller and smaller groups.

" But this factor can be detached in our consideration of structure
from the homogeny of actual completed form and mechanism. The
one is much more remote and less specific than the other ; and they
need to be given each its due rank and place ; not to be confused."

o. CONCLUSION (OSBORN), 1900.

The newly and independently arising cusps, described above as
' homoplastic,' should be described as rcctigradations*

*The term ' rectigradation ' was defined in Science, N.S., Vol. XXI., June 23, 1905,
p. 961, as follows: "Fourth, rectigradation, a new term with which I propose to charac-
terize what in the year 1889 I described as ' definite variations ' ; it embraces changes which
many writers have described as ' orthogenetic,' under the supposed law of direct change,
usually in an adaptive direction, which is described as Orthogenesis ; these probably are
the 'mutations' of Waagen.''

BIBLIOGEAPHY.

ADLOFF, P. "Zur Kenntniss des Zahnsystems von Hyrax," Zeitsch. f. Morph. n.

Anthrop., Ed. V., Heft I., S. 181-200, Stuttgart, 1902.

"Zur Frage nach der Entstehung tier heutigen Saugethierzahnformen," Zeitsi-hr.
f. Morph. u. Anthrop., Bd. V., Heft II., S. 357-382, Stuttgart, 1902.

AMEGHINO, FL. Flloi/cnia, 8vo. 1884.

"Sur 1'Evolution des Dents des Mammiferes," Boll. A cad. Nac. Ctencios e/>

Cordoba, T. XIV., 1896, pp. 381-517.

" Eecherches de Morphologie phylogenetiqne snr les inolaires superieures des
Ongules," Aii. d. Mus. Na<\ de Buenos Aires, T. IX. (Ser. 3^, T. III.), pp. 1-541,
631, fig. 9.
"On the Primitive Type of the Plexodont Molars of Mammals," Proc. Zool. !*<>,-.

London, May 2, 1899.
BATESON, W. "On Numerical Variation in Teeth with a Discussion of the Conception

of Homology," Proc. Zool. Soc. Lond., 1892, pp. 102-11 H.
M«ti?ri«t for the Stiuli/ of Variation, 8vo. London, 1894.
(Great Variability of Mammalian Teeth.)

BAUME. Odontologische Forschungen. Versuch einer Entuncklungsgeschichte des Gebisses.

1882.

BEATI: KHARD. "Sur les Deux Dentitions des Mammiferes," Bull. So<\ de Biologie,

1888, Nos. 9, 10.

COPE, E. D. "On the Homologies and Origin of the Types of Molar Teeth of
Mammalia Educabilia," Jour. Acad. Nut. >'<•/. /'////</., n.s., VIII., Pt. I., pp. 71-89,

1874.

[On the Origin of the Sectorial Tooth of the. Carnivora.]
•'Systematic Catalogue of the Vertebrata of the Eocene of New Mexico. . . ."

Rept. Geogr. EJ-/I/. «n<l Surveys W. of the 100^/t Meridian (Wheeler), 1875, pp. 5-7.
" On the Mechanical Genesis of Tooth Forms," Proc. Acad. Nat. Sci. P/^7«.,1878,p.45.
"The Origin of the Specialized Teeth of the Carnivora," Amer. Naturalist, XIII.,

March, 1879, pp. 171-173.
"Note on the Trituberculate Type of Superior Molar and the Origin of the

Quadrituberculate," Amer. Naturalist, XVII., 1883, pp. 407-408.
" On the Trituberculate Type of Molar Tooth in the Mammalia," Pal. Bull., No. 37,

/'roc. Amer. Phil. Soc., XXI., Dec. 7, 1883, pp. 324-326.
"On the Evolution of the Vertebrata, Progressive and Retrogressive," Amer.

Natwralist, XIX., Mar., April, 1885, pp. 347, 350.
" On Lemurine Reversion in Human Dentition," Amer. Naturalist, XX., 1886,

pp. 941-947.

241

COPE, E. D.

T/K' Origin of the Fitt<>*t, 8vo, L887. (See especially pp. 215. iMi;, 259, 347, :'.",'.), :S'il

"On tin- Tritubereular Molar in Human Dentition," Ji»/,: .\/»,-/>/«i/., July, 1888,

p. 7.
"Synopsis of tin' Vertebrate Fauna of the Puerco Series," '/'/•</ /<.-•. Amer. I'lnlos.

Soc., XVI.. Aug. 1888, p. L".I:>.
"The Mechanical Causes of the Development uf the Hard Parts of the .Mammalia,''

Jour. MorphoJ., Vol. IFF., 1889, pp. i:>,7-:27].
DEWOLETSKY, It "Neuere Forsclmn^en iiber das Gebiss der Sa'uger," S<>/>. aus

,l,ni .l.iln-i'xlierlrhti' '/'« /•./•. Staats-Obergymnasiums in Czernowitz, ]M)4-95,

Czernowitz.
DYHOWSKI. " Stiulien iiber Saugetierzahne," Verli. <1. /•./•. Z<n>L, Iii>t<ri>. <!<'*. in }\'i,-,i.

8vo, 1889.

FLEISCHMANN, A. :' Die Grundform der ISack/.iihne bei Saugethieren mid die Homo-
lo^-ie der enzelnen Hocker," Sitzungsber. <>. /'. /'/Y'//.V.<. .I/--/-/. <1. II '/.«., "Berlin,
1891.

" Mittheilung iiber die Zahnentwicklung von Hyrax," Ahld. Xntm-. l<!*t. ^A-.W/.<,-//.
Nuremburg, 1894.

FLOWKU, \V. H. "Remarks on the Homoln-ies and Notation of the Teeth of the
Mammalia." Jour. .I//"/. Physiol, III., pp. -262-278, 1869.

GAUDRY, A. K, ,<•},, /in, -nn-nt* tin Muml' .(nii/ml <ltnx fes Temps gfologiques j Mmiiiiiiferes

Ti'i-tiitws, Paris, 1878. (Su])ports concrescence theory.)
" Sur la Similitude des Dents de I'homme et de qnelques Animaax," L'Anthropologie,

Janv.-Avr., 1901.
GERVAIS, P. /f;.<t<>i/->' naturelle d.-s Mammiftres, Paris, 1854. (Suggests concrescence

idea.)
GIDLEY, J. W. "Evidence bearing on Tooth-Cusp Development," Proc. Wml/i nijtni,

Acad. Sci., Vol. VIII., 1906, pp. 91-110, pll. iv.-v.

GOODRICH, E. S. "On tlie Fossil Mammalia of the Stonesfield Slate,"' (jmtr. Jour.
Jfin: Sci., XXXV., 1894, pp. 407-432. (Supports "polybuny" theory.)

HOFFMANX. " Ueber die Entwicklung des Kronencementes an den Backenzahnen der

Wiederkauer mit Beriicksichtiguiig der Zahnentwicklung in Allgemeinen,"

Zeittchr. fiir wiss. Zool., Bd. LVIIL, 1894, j.p. r.66-617.
HUXLEY, T. H. "On the Cranial and Dental Characters of the Canidae," Proc. Zoo!.

Soc. LnirL, 1880, pp. 283-284.

(Regards molar of Centetes as primitive for Carnivora. Interprets molar

cusps by the " premolar analogy " theory.)
KLEVER, E. " Zur Kemitniss der Morphogenese des Equidengebisses," Morpltu).

Jahrb., Bd. XV., 1889, Leipzig.
MAHN, R, " Bau und Entwicklung. d. Molaren, Ijei Mus und Arvicola," 3for/>hol.

Jahrb., Bd. XVI., p. 652. (Criticises Tritubercnly.)
MAJOR, C. J. FORSYTH. " Nageriiberreste aus Bohnerzen Siiddeutschlands und der

Schweiz. Nebst Beitragen zu einer vergleichenden Odontographie von

Ungulaten und Unguiculaten," Palaeontographica, 1873, XXII.
(Opposes three-cusped origin of mammalian molars.)

242 EVOLUTION OF MAMMALIAN MOLAR TEETH

MAJOR, C. J. FORSYTH.

" On some Miocene Squirrels, with Remarks on the Dentition and Classification

of the Sciuriiue," Proc. Zoo!. Soc., Feb. 28, 1893, pp. 17D-215. (See pp. 196-214

for polybuny theory.)
" On Meg al adapts madagascariensis, an extin ;t gigantic Lemuroid from Madagascar.

. . ." Philos. Trans. Roy. Soc., Vol. 185 (1894) B, pp. 15-38, especially

pp. -22-24. (Discusses the so-called tritubercular type of molar in Viverridte,

Centetid*, Lemuridse.)
"On the Malagasy Rodent Genus Brachyuromys \ and on the Mutual Relations of

some Groups of the Muridae (Hespi'roiHiiince, Microtiuce, Murince, and Spalaciil")

with each other and with the Malagasy Nesomyime," Proc. Zool. 8m:, June 1,

1897, pp. 713 et seq. ("Polybuny " theory.)
"On Fossil and Recent Lagomorpha," Trans. Linn. Soi: Loud. (2), VII.,

pp. 433-520.
"On the Change and Form of the Teeth in the Centetidte a Malagasy Family of

Insectivora," Ann. Mag. Nat. Hist., ser. 6, Vol. XX., Dec. 1897, pp. 525-544.

MIVART, ST. G. "On the Osteology of the Insectivora," Jour. Anat. and PhysioL,
Vol. I. and IT., 1863, p. 117 et seq. (Adopts idea of "concrescence of cusps" in
Chrysochloris.)

KUKENTHAL, W. " Das Gebiss von Didelpki/s,n Ant. Am., Bd. VI., 1891, pp. 658-666.
"Einige Bemerkungen iiber die Saugethierbezahnung," Anat. An:., 1891, Nr. 13.
"Ueber den UrspiTing und die Entwickelung der Saugethierzahne," Jenuim-ln-

Zcitschriftfiir Natiinn'ssenschaft, Bd. XXVIII., 1893.
" Kntwickelungsgeschtliche Untersuchungen am Pinnipediergebisse," Jenaisclie

Zeitschr.f. naturw., Bd. XXVIII., 1893, pp. 76-118.
"Die Bezahnnng der Zahnwale," Denkschr. d. med.-natwrw. Gesellsch., Bd. III., 1893,

pp. 387-448.

KOWALEVSKY. " Monographic der Gattung Anthracotherium, Cuv.," Palceontographica,
N.F. II., 3 (XXI F.), 1873, 1874, pp. 210, 263, 264.

LATASTE, F. "Snr le Systeme dentaire du Genre Daman et FEtnde de la Dent canine,''

Zool. Am., 1887, p. 268.

"Considerations sur les Deux Dentitions des Mammiferes," Jour, de PAnat. et de la
I'/it/siol., 1888, p. 200.

LKCHE, W. " Studien iiber die Entwicklung des Zahnsystems bei den Saugethieren,"
l. Jahrb., Bd. XIX., 1892, p. 502, XX., p. 113.

OSBORN, H. F. "The Upper Triassic Mammals Dromatherium and Microconodon"

extracted from Proc. Acad. Nat. Sci. P/tila., Dec., 1886, and Proc. Amcr. Phil.

Soc., April, 1887, antea, pp. 18-21.
"The Structure and Classification of the Mesozoic Mammalia," Jew. Acad. Nat.

Sci. Phila., IX., No. 2, July, 1888, antea, pp. 31-35.
"The Nomenclature of the Mammalian Molar Cusps," Amer. Naturalist, XXVI.,

Oct. 1888, pp. 436, 437, antea, p. 66.
" Additional Observations upon the Structure and Classification of the Mesozoic

Mammalia," 7'w. Acad. Nat. Sci. Phila., Nov., 1888, pp. 292-301.
"The Evolution of Mammalian Molars to and from the Tritubercular Type," Amer,

Naturalist, XXII., Dec., 1888, pp. 1067-1079, antea, pp. 36-47.

BIBLIOGRAPHY 243

OSBOR\, H. F., and \YOKTMAN, J. L. "Fossil Mammals of tin- \Yasatch and Wind
River Beds," Collection of, 1891, P»iJL Amer. .!///,<. .\at. ///.>•/., Yol. IV., No. 1,
Oct., 1892, pp. s !-'.»<>, -mtea, pp. 60-72.

OSBORN, H. F. " Odontogenesis in the Ungulates,'' Amer. \<it>t,ralist, XXVI., ISDi'.

pp. 621-623.
" The History and Homologies of the Human Molar < 'usps,'' .1 nut. .1 //:.. V 1 1., Jena,

1892, Nr. 23, 24, aitt<>a, pp. 48-55.

"The Rise of the Mammalia in North America," Proc. Ann-r. .!.«'»•. .!<//•. ,sv/. XLII.,
Aug., 1893, pp. 189-227, Jme/-. .A-/'/-. ,SW. (3), XLVI., p]». 379-392, 448-466.

"The History of the Cusps of the Human Molar Teeth," International Dental
Journal, July, 1895, a/iti'a, pp. 55-65.

"Tritnberculy : A Review Dedicated to the late Professor Cope," Amer. Nat unit 1st.
XXXI., Dec., 1897, pp. 993-1016, antea, pp. 74-90.

NAWROTH. •' Znr Ontogenese bei Schweinemolaren," Inaugural Dissertation, Berlin,
1898.

OWEX, R. ''Monograph of the Fossil Mammalia of the Mesozoic Formations," Mono-
graph Paloeontograph. Soc., XXIV. (for 1870), 4 to, 1871.

POUCHET et CHABRY. "Contribution a 1'Odontologie des Mammiferes," Jour, de
C Anat. el de la Physiol., 1884.

POULTON, E. " The True Teeth and the Horny Plates of Omithofhynchus" (Juar. ./<>>//•.

Mn-r. M. XXIX., 1889, pp. 9-48.
ROSE, C. " Ueber die Entwickeliing des Menschlichen Gebisses," Verh. d. Deutsch.

<>,l<mt. Ges., Bd. 3, 1891.
" Ueber die Zahnentwickelung des Menschen," Schweiz. Viertel-jakrssch.rift fiir Za/u>-

heilkunde, Bd. II., 1892.
"Ueber die Zahnentwickelung der Beuteltiere," Anat. An:., Bd. VII., 1892,

pp. 639-650, 693-707.
"Ueber die Eiitstehung und Formabanderungen der Menschlichen Molaren,"

Anat. Anz., Bd. VII., 1892, Nr. 13, 14, pp. 392-421.
" Zur Phylogenie des Saugetiergebisses," Biolog. Centralbl, Bd. XII.. Nr. 20, -2 1 .

pp. 624-638, 1892.
"Ueber die Zahnentwickelung der Bentelthiere," Anat. An.;., Bd. VIII., 1893,

pp. 566-577.
"Ueber die erste Anlage der Zahnleiste beim Menschen," Anat. Anz., Bd. VII.,

1893, Nr. 1, pp. 29-32.

" Das Zahnsystem der Wirbelthiere," Erueb/t d. Anat. it. Entw. Merkel u. Bonnet, IV..

1894, pp. 542-591.

ROST. Versuch einer Phylogenie des Saugethiergelnsses, Jena, 1884.

RUTIMEYER, W. " Beitrage zur Kenntniss der fossilen Pferde und zur vergleichenden
Odontographie der Hufthiere iiberhaupt," Verh. d. naturf. Ges. in Basel, III.,
1863, pp. 558.

RYDER, J. A. "On the Mechanical Genesis of Tooth Forms," Proc. Nat. Sci. Phil a.
Acad., 1878, pp. 45-80.

SCHLOSSER, M. "Beitrage zur Kenntniss der Stammesgeschichte der Hufthiere und
Versuch einer Systematik der Paar- und Unpaarhufer," Morpholog. .la

XII., 1887, pp. 1-136.

Q2

244 EVOLUTION OF MAMMALIAN MOLAR TEETH

SCHLOSSER, M.

" Ueber Beziehungen cler Ausgestorbenen Siiugethierfaunen und ihr Verhaltniss

zur Saugethierfauna der Gegemvart," Biol. Centralbl. Bd., VIII., No. 19,

Dec., 1888, pp. 582-631.
"Die Affen, Lemuren, Chiropteren, Insectivoren, Marsupialier, Creodonten und

Carnivoren des europaischen Tertiars und deren Beziehungen zu ihren lebenden

und fossilen aussereuropaischen Verwandten, I-VIIL, Wien, 1887-1890 (Beitraye

Palceontologie Oesterreich- Ungarns).

"Die Differenzierung des Saugetiergebisses," Biol. Centralbl, Bd. X., 1890.
"Die Entwickelung der verschiedenen Saugethierzahnformen im Laufe der

geologischen Perioden," Verk. deutsch. odontol. Ges., Bd. 3, Heft 2, 3, 1891.
" Ueber die systematise-he Stelhing der Gattungen, Plesiadapis, Protoadapis

Pleuraspidotherium und Orthaspidotherium" Neues Jahrb. f. Mineralogie, Geologic,

und Paliiontologie, Jahrgang, 1892, Bd. II., pp. 239, 240.
SCHWALBE, G. "Ueber Theorien der Dentition," Verh. d. Anat. Gesell. Strassbwrg,

13-16, Mai, 1894.
SCOTT, W. B. "The Evolution of the Premolar Teeth in the Mammals," Proa. A cad.

Nat. Sci. Phila., 1892, pp. 405-444.
SEELEV, H. G. "On the Nature and Limits of Eeptilian Characters in Mammalian

Teeth," Proc. Roy. Soc., Vol. 44, 1888, pp. 129-141.
STEHLIN, H. G. "Ueber die Geschichte des Suidengebisses," Abh. d. schiveiz.

paliiontolog. Ges., Vol. XXVI., 1899, Vol. XXVII., 1900, Zurich, 1899-1900.
TAEKKR, .1. " Zur Kenntniss der Odontogenese bei Ungulaten," Inaugural Dissertation,

Dorpat, 1892.
THOMAS, O. "On the Homologies and Succession of the Teeth in the Dasyuridpp, with

an Attempt to trace the History of the Evolution of Mammalian Teeth in

General," Philos. Trans., Vol. 1887, B., pp. 443-462.

" A Milk Dentition in Orycteropm? Proc. Roy. Soc., Vol. 47, 1890, pp. 246-248.
"On the Dentition of Ornithorkynchus," Proc. Rot/. Soc., Vol. 46, 1889, pp. 126-131.
TIMS, H. W. MARETT. "Notes on the Dentition of the Dog," Anat. Anz., Bd. XI.,

1896, pp. 537-546.
"On the Tooth-genesis in the Canidte," Jour. (Zool.) Linn. Soc. Lond. XXV.,

1896, pp. 445-480, XXVIII., 1901, pp. 261-290.
" On the Succession and Homologies of the Molar and Premolar Teeth in the

Mammalia," Jour. Anat. Phys., XXXVI., 1902, pp. 321-343.
"The Evolution of the Teeth in the Mammalia/' Jour. Anat. and Physiol., Vol.

XXXVII., 1903, pp. 131-149.

TOMES, C. S. A Manual of Dental Anatomy, Human and Comparative, 8vo, Phila-
delphia, 1898.

TOPINARD, P. "De 1'Evolution des Molaires et Pi'emolaires chez les Primates et en
particulier chez FHomme," L'Anthropologie, No. 6, Nov., Dec., 1892.

WEBER, M. Die Saugetiere, Jena, 1904.

WINGE, H. "Om Pattedyrenes Tandskifte, isrer med Hensyn til Ttendernes Former."

Vidensk. Meddel. Naturh. Foren i Kjobenhavn, 1882.

" Jordf undue og nulevende Pungdyr (Marsupialia) fra Lagoa Santa, Minas Geraes,
Brasilien," Med. Udsigt over Pungdyrenes Slcegtskab. E Museo Lundii, Kjoben-
havn, 1893.

BIBLIOGRAPHY 245

WOODWARD, A. SMITH. "Outlines of Vertebrate Pah«outol«>.u\ t-.i Students of

Zoology," University Press, < 'ambridge, 1898.

WOODWARD, M. F. "Contributions to the Study of Mammalian Dentition, Part I.,
On the Development of the Teetli of the Macrojxulidie," 1'ror. %,,•>[. Soc. Lond.,
May 2, 1893, p. 449-473.
"On the Succession and Genesis of the Mammalian Teeth," Science Progress, I.,

1894, pp. 438-453.
"On the Milk Dentition of the Rodentia, with a Description of a vestigial Milk

Incisor in the Mouse (J7V.< musci/ltis)," An«/. Anz., Bd. LX., Nr. 19, 20, 1894.
"Contributions to the Study of Mammalian Dentition, Pt. II., On the Teeth of

certain Insectivora," Proa. Zool. Soc. Loud., 1896, pp. 557-594.
"On the Teeth of the Marsupialia, with especial Reference to the PremilU

Dentition," Anat. -I/':. XII., Bd. XII., Nr. 12, 13, 1896.

WORTMAN, J. L. The Cornell i'n fir,- A,t<it«my of the Teeth of the Vertebra t<i, Svo, 1886.
[Origin of the Tritubercular Molar] "Studies of Eocene Mammalia in the Marsh
Collection. . . ." Parti., '• Carnivora," Amer. Jour. Sci., Vol. XIV., Jan. 1902,
pp. 93-98 ; Part II., Amer. Jour. Sci., Vol. XVI., Nov. 1903, pp. 365-368.

ZUCKERKANDL. Die Backenziihiie des Menschen.

Congress der dentschen und osterrischen anthropologischen Gesellschaft,
Wien, 1889.

INDEX.

Aceratherium, 75.
Achyrodon, 24, 29.
Adapis, 158.
Allodon, 104.
Allotheria, 11, 101.
Amblotherium, 24, 29.
Amblypocla, 14, 164.
Ameghino, 4, 57, 201.
Amphitherium, 23, 24, 27-
Anacodon, 133.
Analogy, 229.
Anaptomorphus, 89, 158.
Ancylopoda, 15, 184.
Anisognathisni (overhang), 43*, 215.
Antecrochet, 73.
Anthropoidea, 14, 157.
Apternodus, 121.
Arctocyon, 133.
Artiodactyla, 14, 84, 171.
Astrapotheria, 16.

Bathyopsis, 168.

Basin-shaped molars, 103, 130, 131, 142,

206.

Bilophodont, 171*.
Bolodon, 24, 103, 104.
Brachyodont, 163.
Broom, 101.
Bunodont, 163.
Bunoid (defined), 89.
Bunoselenodont, 163.

Ctenogenetic modifications, 54.
Cpenolestes, 112.
Calcochloris, 81.
Canine (bi-fanged), 1941.
Carnassial teeth, 122, 123, 135.
Carnivora, 12, 131.
Carnivora Fissipedia, 131, 142.
Cement lakes, 73, 86.

Centetes, 117, 11!)- 126, ±26, 227.

Cetacea, 37, 101, 19(1.

Chalicotherium, 87, 184.

Cheiroptera, 12, 119 (bat), 121, 129.

Chirox, 103, 104.

Chrysochloris, 64, 81, 119-126, 227.

Cingulum, 33, 89.

Classification, Mammalia, 11.

Como Beds, 781.

Conacodon, 1<>">.

Condylarthra, 14, 168.

Concrescence theory, 57-59, 202.

-cone, 82.

Convergent evolution, 139.

Cope, 1, 2, 4, 5, 38 l.

Coryphodon, 166.

Creodonta, 12, 132-139.

Cretaceous (Laramie) mammals, 95,

219.

Crochet, 73, 229.
Ctenacodon, 102.
Cusp development, 42.
Cusp differentiation theory, 56.
Cusps, nomenclature, 41.
Cynognathus, 91.

Dasyurid;v, 109.

Dermoptera, 12.

Determinate variation, 233.

Diademodon, 92.

Dicrocynodon (Diplocynodon), 219.

Didelphia, 100.

Didelphys, 108.

Dilambdodout Insectivores, 117, 2K>.

Diplocynodon, 28, 219.

Diprotodon, 1 14.

Diprotodontia, 12.

Dissacus, 133, 216.

Docodon, 28.

Dog, 213-

248

INDKX

Dromatherium, 18. 193.
Dryolestes, '23, 27, .SO, 218, 220.
Duplicidentata, 148.

Ectoloph, 73.
Edentata, 13, 151.
Elephas, 188.

Embryologioal order of cusp develop-
ment, 54, <>4, 77.
Embryology of molars, 77, 208.
Embryological theory, 6, 54, 214.
Enamel folds, 183.
Enneodon, 28.
Equidae, 175.
Entoconid, 41, 89.
Eocene Mammals, 98.
Ericulus, 120, 121, 123.
Erinaceus, 117, 118, 123*.
Eskimo, teeth, 55, 62.
Esthonyx, 151.
Eupleres, 122, 123.
Euprotogonia, 83, 169.
Eutheria, 11, 100.

Fissipedia, 12.

Fleischmann, 204.

Forsyth Major, 4, 145, 149, 205.

Fossettes in molars, 73.

Fruit-bats, 129.

Galeopithecus, lli».

Ganodonta (Tjeniodonta), 152, 154.

Gidley, 123-126, 219.

Gomphognathus, 92.

Goodrich, E. S., 2llli.

Grinding teeth, 9.

Gymnura, 118, 21(1.

Haplodont, 89.

Haplodont type, 39.

Hares, 148.

Heel or talon, 67.

Hemicentetes, 120, 122, 123.

Homalodotheria, 16.

Homodont teeth, 432.

Homogeny, 231.

Homologies, disputed, of cusps, 123-126,

209-213, 219, 224, 225.
Homoplasy, 228, 232.
Horses, 72, 174-178.

Human molar teeth, 54, 55, 56-59, 62-65.
Hyopsodus, 12S.
Hypocone, 68*, 89.

Hypoconid, 89.

Hypoconulid, 41, 81, 89.

Hypolophid, 73.

Hypsodonty (hypselodonty), 183.

Hyracoidea, 15, 185.

Hyracotherium, 83.

Ictops, 118.

Insectivora, 1, 12, 117, 223.
Insectivora primitiva, 12.
Isognathism, 43*.

Jaws, anisoguathous, 215 ; isogJiathous,

43.
Jurassic mammals, 94, 219.

Kangaroos (Macropus), 112, 114.
Kukenthal, 6, 57.
Kurtodon, 26.

Laodon, 29.
Lambdotherium, 175.
Lankester, 232, 238.
Laramie mammals, 96, 115.
Lemuroidea, 14, 157.
Leptocladus, 24, 29.
Listriodon, 188.
Litopterna, 16.

-lo[jh (proto-, meta-, ecto-), 163.
Lophiodonts, 86.
Lophodont, 163, 181.
Lophoselenodont, 163.

Maim, 204.

Major, Forsyth, 4, 145, 149, 205.

Mammalia classification, 11.

Mammals, Eocene, 37.

Mammals, Jurassic, 94, 219.

Mammals, Laramie, 96.

Mammals, Mesozoic, 18-35, 37-44.

Mammals, reptilian ancestors, 91, 100

Man, 6, 54-59, 62-65, 158.

Marsupialia, 11, 94, 108.

Mechanics of molars, 43, 44, 60, 61*,

82*, 113, 221.
Menacodon, 25.
Meniscoessus, 102, 106.
Meniscotherium, 87, 184.
Merychippus, 75.
Mesonyx, 79, 134, 142, 216.
Mesostyle, 85, 89.
Metacone defined, 41.
Metaconid defined, 41.

IXDK.X

249

Metaconule, 82 1, 89 ('ml ').
Metaconule, in Artiodaetyla. 46 , 17:!.
Metalopb, 73.
Metalophid, 73.
Metastyle, 80, 86, 89.
Microconodon, 18, 193.
Microlestes, 80, 93, 102, 103.
Mioclsenus, 109.
Mivart, 123, 124.
Mixodectes, 14.").
Mceritheriuiu, 1S7.
Molar cusps —

< >rder of appearance in Ontogeny,
4S-.11 ; terminology, 41, 63, US,
71, 76, 82, 224.
Mole, 209.

Monotremata, 11, 10.").
Multitubereulata, 11, 80, 101, 115.
Mutation (von Waagen), 234.
Mylagaulus, 147.
Myogale, 119, 227.
Myrmecobius, 112.
Mystacoceti, l(i.

Nomenclature of cusps, 41, 03, 08, 71,
76, 82, 224 ; of premolars, 1110 ; of
trituberculy, 89,

Notharctus, 89, 160.

Notoryctes, 113, 114.

Odontoceti, 16.

Olbodotes, 89.

Ontogeny of the teeth, 48, 49, 51-55,

63-65, 208-214.
OrnithorhynchuSj 105, 107.
Orthogenesis, 228, 239.
Osborn, 1-9.

Pachylemuriens, 77.

Palftolagus, 148, 150.

Palseomastodon, 187.

Paheontological theory (of molar develop-
ment), 5.

Palaeotherium, 63.

Pantolambda, 87, 165.

Pantotheria, 12, 22.

Paracone as oldest cusp, 8, 208-216, 221 ;
defined, 41.

Paraconid defined, 41 ; reduction of, 103.

Parallel evolution, 171", 137, 139.

Parallelism, 2291.

Parastyle, 89.

Paurodon, 29, 219.

Peralestes, 26.

IVrameles, 1 13.

Peramus, 24, 27. 2S.

IVraspalax, 29.

Peratherium, lo'.i.

Periptychus, Hi I.

Perissodactyla, 1"), 72, 174; cusp noincii-

elature, 76.
Phalangerida', 111.
1'hascolestes, 24. 29, .SO.
Phascologale, 1 12.
Phenacodus, 17".
1'holidota, 13.
Piiinipedia, 13, 131, 143.
Placentalia, 100.
Plagiaulacidae, loo.
Plagiaulax, 24.
Plexodoiit theory, 201.
Polybuny theory, 4, 2U5.
Poly mastodon, 103.
Polyprotodontia, 111.
Potamogale, 119-126.
Potential homology, 236.
Premolar analogy theory, 7, 215, 223.
Premolars, 193 ; evolution, 172.
Primates, 14, 48, 56, 101, 157.
Proboscidea, 15, 1S6.
Proglires, 13, 144, 14.').
Proscalops, 119, 121.
Proteodidelphys, 202.
Protocone, as oldest cusp, 49, 67, 217:

as originally selenoid, 173 ; defined,

41.

Protoconid defined, 41.
Protoconule, 41, 82f, 89; defined, 41 ;

(in Dryolestes), 223.
Protodont, 39, 89.
Protodonta, 11, 101.
Protogonodon, 169.
Protoloph, 69.
Prototheria, 11, ion.
Psittacotheriuin, 154.
Pteralopex, 129.
Ptilodus, 106.
Puerco mammals, 2.
Pyrotherium, 16, 190.

Quad ritubercular, 89.
Quinquetubercular, 1 63.

Rectigradations, 228, 239.
Rhinoceroses, 73, 76, 86, 176, 181-183.
Kodentia, 13, 144.

250

INDEX

Rose, 4, 49, 51, 57, 63.

Rotation of pa., me., 7, 8, 31, 52 ; of pad.,

med., 222.
Ryder, 43.

Schlosser, 145.

Sciurus, 146.

Scott, 234.

Sectorial (defined), 89.

Selenoid, 112, 163.

Selenodont, 163.

Sexituberculy, 68, 80, 89, 163.

Shrew, 210.

Simplicidentata, 145.

Sinopa, 134.

Sirenia, 15, 188.

South American Ungulates, 189.

Spalacotherium, 22, 24, 25, 26 ; rotation

of pa'1, and med. , 7.
Squirrels, 146, 205.
Stereognathus, 24, 103.
Styles, 82, 84.
Stylodon, 24, 29.

Taeker on Odontogenesis, 48, 51.

Taeniodonta, 151.

Talon, 67, 68*, 81.

Talonid, 53, 89.

Talpa, 209.

Tapirs, 86, 176, 180.

Tertiary formations, diagram, 17.

Therapsida, 100.

Theriodontia, 91.

Thylacynus, 109.

Tillodontia, 13, 151.

Tims, Marett, 149, 213.

Titanotheres, 175.

Toxodontia, 16, 189.

Triassic mammals, 18, 94, 101, 105*.

Triconodon, 25, 218, 222.

Triconodont, 89, 227.

Triconodonta, 11,21.

Triconodont molars, 21, 33, 40, 60, 79,
88, 89, 144.

Tricon odonty, secondary, 64.

Trigon, 67, 80.

Trigonid, 89.

Trigonolestes, 171.

Tritubercular nomenclature, 41. 54, 55,
66-73 ; teeth, stages in evolution of,
39 ; type (as ancestral), 2-4, 40, 89.

Trituberculata, 12, 22.

Trituberculates, 95, 115.

Trituberculy, objections to, 201.

Tritylodon, 93, 103.

Tubercular, 89.

Tuberculo-sectorial type, 89.

Tuberculus anomalus, 158.

Tubulidentata, 13.

Tupaia, 109.

Typotheria, 16, 189.

Uintatherinm, 166.

Unguiculata, 100.

Ungulata, 101, 163; nomenclature of

molar cusps, 67, 76, 83 ; South

American, 189.

Variation, determinate, 233.
Vulpavus, 89.

Wart-hog, 188.
Whales, 16, 190.
Wilson, E. B., 237.
Woodward, A. S., 2oii.
Woodward, M. F., 6, 77, 123.
Wortman, 150, 224.
Wynyardia, 112.

Zalambdodont Insectivores, 117-126, 210-

213, 226, 227.
Zeuglodon, 16, 191.

GI.ASfiOW : HKIXTED AT THE UNIVERSITY PRESS BY ROBERT MACLEHOSE AND CO. LTD.

loL