A OTJIDE

TO THE

1 ELEPHANTS

5 (RECENT AND FOSSIL)

EXHIBITED IN THE DEPARTMENT OF

GEOLOGY AND PALAEONTOLOGY

IN THE

BRITISH MUSEUM (NATURAL HISTORY),

CROMWELL ROAD, LONDON, S.W.

ILLUSTRATED BY 3! TEXT-FIGURES.

LONDON :

PRINTED BY ORDER OF THE TRUSTEES
OF THE BRITISH MUSEUM.

1908.

THORP.

Price Sixpence.

EARTH

NO

A GUIDE

TO THE

ELEPHANTS

(RECENT AND FOSSIL)

EXHIBITED IN, THE DEPARTMENT OF

GEOLOGY AND PALEONTOLOGY

IN THE

BRITISH MUSEUM (NATURAL HISTORY),

CROMWELL ROAD, LONDON, S.W.

ILLUSTRATED BY 31 TEXT-FIGURES.

LONDON :

PRINTED BY ORDER OF THE TRUSTEES
OF THE BRITISH MUSEUM. t

1908.

ALERE V FLAMMAM.

PRINTED BY TAYLOK AND FRANCIS,
KED LION COURT, FLEET STREET.

EARTH

SCIENCES

LIBRARY

EARTH

SCIENCES

LIBRARY

QMS 7
PREFACE. '

THE fossil remains of elephants in the Department of Geology
are so numerous, arid the ancestral history revealed by them is
of so much general interest, that they appear worthy of a
special Guide- Book. The present account of the collection has
therefore been prepared by Dr. Charles W. Andrews, who has
already published technical descriptions of some of the more
important specimens in the Philosophical Transactions of the
Royal Society (1903 and 1908), and in "A Descriptive
Catalogue of the Tertiary Vertebrata of the Fayum, Egypt"
(issued by the Trustees of the British Museum in 1906).

The immediate ancestors of the modern elephants have long
been known through the discoveries of Falconer and Cautley
in the Pliocene formations of the Siwalik Hills in India ; but
the earliest members of the group have only been found during
recent years in the Eocene deposits of Egypt. The whole
of the Cautley Collection is exhibited in the Museum
besides other fine specimens from the Siwalik Hills described
by Falconer. The Egyptian discoveries also form an extensive
collection, though many of the original specimens are in the
Geological Museum, Cairo, and only represented ,here by
plaster casts.

Among the illustrations, figs. 1 and 6 are reproduced from
the " Encyclopaedia Britannica " by permission of The Times.

A. SMITH WOODWARD.

DEPARTMENT OF GEOLOGY.
February, 1908.

W789861

GUIDE

TO

ELEPHANTS (RECENT AND FOSSIL).

IN THE GALLERY OF FOSSIL MAMMALIA.

AT the present day the different main sub-divisions of the
Mammalia are, as a rule, very distinctly marked off from one

Pig. 1

Upper and lower dentition of pig, showing the presence of the full set of
44 teeth and the low-crowned bunodont cheek-teeth adapted for a soft,
mainly vegetable diet, c., canine; t., incisors; m., molars; pm., pre-
molars.

another. For instance, the Carnivora (flesh- eaters, e. g., tiger,
bear) are now widely different from the Ungulata (hoofed-
animals, e. g.j horse and ox) , each of these groups being fitted
for some special manner of life, and particularly for living on

B

2 ELEPHANTS.

some special kind o£ food. Thus, the principal character of
most of the Carnivora is the possession of sharp claws and
teeth for killing and devouring other animals ; while, on the
other hand, the vegetarian Ungulata have teeth fitted for
grinding vegetable matter and feet adapted solely for moving
from place to place in search of pasturage. If, however, we
trace back through the earlier periods of the Earth's Geological

Fig. 2.

Skull and mandible of Striped Hyaena, showing the sharp cutting cheek-
teeth adapted for a flesh diet. Lettering as in fig. 1.

History, the extinct animals from which these quite distinct
modern groups are descended, we find that in nearly all cases
in which these earlier fossil forms are well known, there is a
tendency for them to become more and more alike. In the
early Eocene, indeed (see table, p. 6), the Carnivora and
Ungulata are not always to be distinguished from one another
with certainty, so that the animals from which they and some
other mammals have descended, may be placed in a single
group. Nearly all these early mammals have certain characters
in common : thus in most there are five toes on each foot and
forty-four comparatively simply constructed teeth. Characters
such as these are called "primitive/' and when as time goes
on they become gradually changed in different ways and
adapted for particular purposes, they are said to be more

SPECIALISATION.

3

"specialised." Thus in the horse the foot (fig. 3) is extremely
specialised, in that it has only a single complete toe instead of
the primitive number, five : its limbs being specially fitted for
swift movement over hard ground. To take another instance,
the teeth of the tiger are said to be highly " specialised/'
because there are only thirty of the original forty-four, and
these have become specially adapted for seizing living prey and

Diagram showing the gradual loss of toes on the fore foot (a) and increase
of complexity in the grinding teeth (b) of successive horse-like Ungulata
from Europe, namely Hyracotherium (Eocene), 4, Anchitherium (Mio-
cene), 3, Hipparion (Pliocene), 2, Equus (Pleistocene and Recent), 1.

cutting and tearing its flesh. It must be added that all the
characters of a group of animals do not necessarily become
specialised, but that some may remain in the primitive condi-
tion. Thus man, in some ways the most highly specialised of
mammals, still retains the primitive number of five digits on
both limbs.

In order to trace back a modern specialised group of mammals
to its early primitive ancestors, a long series of fossil remains
from the successive geological periods is necessary. Unfor-
tunately in many cases these fossils have yet to be found, but
every year further discoveries are made and gaps of more or
less importance are filled up. The series of changes undergone
by a group of mammals is perhaps best known in the horse

B2

4 ELEPHANTS.

family *, but recent discoveries of remains of early forms of the
elephant-group in the Eocene beds of the Libyan Desert in
Egypt, have made it possible to trace the history of the
elephants also with considerable completeness, and in the
present guide a short account of that history is given.

Among living mammals the elephants are perhaps the most
remarkable. Not only do they exceed in size all other living
land-animals, but they are further distinguished by the possession
of a mobile trunk or proboscis, which is at once a sensitive
organ of touch and a most efficient means of grasping objects,
both large and small. Furthermore, the structure of their teeth
reaches a degree of complication not to be found in any other
animals. At the same time, though in many respects so peculiar,
in others they retain primitive characters that have been lost in
most of the other Ungulata, with which they are usually classed.
The most notable of these primitive characters is the presence
of the original five toes on each foot, while in most hoofed-
animals the feet have become " specialised" by the loss of one
or more of the digits.

It is now proposed to describe some of the principal stages
by which the elephants gradually came to be what they are at
the present day, and to show that the earliest known forms
were much like other primitive hoofed-animals, a condition
to which the pigs and tapirs among living mammals perhaps
most nearly approach. It will be shown that the earliest
known animal belonging to the Proboscidea or elephants was,
in fact, not unlike a large pig (see fig. 8), though in some
respects an even more primitive creature. From this beginning
we can trace a gradual increase in size in the later forms, a
gradual development of the trunk or proboscis, first as the
upper part of a long snout supported by the elongated lower
jaw, afterwards as the familiar movable organ so characteristic
of the modern elephants. We can also observe the gradual
increase in the size and degree of complication of the griiiding-
teeth, accompanied by the complete loss of many of the teeth

* Casts of specimens showing the gradual " specialisation " of the teeth
and feet in the horses are shown in a case in the North Hall. See also
pier-case 10 and table-case 5 in Gallery of Fossil Mammals.

Recent
Pleistocene
Upper Pliocene

ANCESTRY OF ELEPHANTS.

Fig. 4.

ELEPHA5
(short chin)

Lower Pliocene TETRABELODON

[LONGIROSTRJS STAGE]
Ufijier Miocene (shortening chin)

Middle Miocene TETRABELODON

[ANGUSTIDENS STAGE]
Lower Mio cene ( long chin)

Upper OUgocene fron, Africa

into Europe -Asia

Lower Oligoce ne ?\

\ PAL AEO MASTODON
UpnerEocene \ (^^enwg chin)

Middle Eocene

Lower Eocene

MOERITHERIUM
(short chin)

Diagram showing some stages in the gradual increase in size, and alteration
in form of the skull and mandible, occurring in the Proboscidea from
the Eocene to the present day.

6 ELEPHANTS.

found in the earlier forms. Finally, we have materials for
discussing the probable relationship of the elephants to some
other groups of animals.

Fortunately for the present purpose the British Museum
possesses the most extensive series of Proboscidean remains
to be found anywhere, so that except in very few instances
readers can see in the S.E. Gallery of Geology the actual
specimens, or, at any rate, casts of the specimens, upon which
the following descriptions are based, and can to some extent
check the accuracy of the various statements for themselves.

Before proceeding to the description of the animals, it may
be advisable to refer to the geological horizons or periods of
the Earth's history during which they existed and in the rocks
of which their fossil remains are found ; for it is necessary to
know the order in which the different forms appeared on the
earth, just as in tracing the pedigree of a human family,
the dates of the documents upon which it is founded must
be known. It will be seen from the following table that the
history of the life of the Earth falls into several great periods

TABLE 1.

f QUATERNABY .............. f RECENT.

Ji PLEISTOCENE.

-

TEBT.ABY ... ...... ;^v... MIOCENE.

j OLIGOCENE,
I EOCENE. *~
f CRETACEOUS.
MESOZOIC or SECONDARY .............. J JURASSIC.

[ TBIASSIC.
f PERMIAN.

CARBONIFEROUS.
PALAEOZOIC or PRIMARY .......... ...... j DEVONIAN.

j SlLUBIAN.

I ORDOVICIAN.
^ CAMBRIAN.

to which names have been given by geologists. The earliest of
these is called the Primary or Palaeozoic Period, and during

GEOLOGICAL SUCCESSION. 7

it the only backboned animals were fishes, amphibians (repre-
sented at the present time by newts, frogs, &c.), and, towards
the end, some reptiles^ In the next great period, the Secondary
or Mesozoic, the reptiles were of the greatest importance : they
were very numerous and some were of gigantic size. They
became fitted for various modes of life, some inhabiting the
land, others the sea ; some living on a vegetable diet, others
on flesh. During this period also the birds began to come into
existence, and remains of the most remarkable of the early
forms of birds, Archaeopteryx, are shown in table-case 13. At
the same time the first of the warm-blooded mammals arose,
though they were as yet insignificant in size and numbers.
The third period, the Caenozoic, is that with which we are
chiefly concerned. During it the reptiles became of little
importance, while, on the other hand, the mammals took
their place, becoming extremely numerous, many of them
of great size, and adapted to every kind of life and food. The
latter part of this period, sometimes called the Quaternary,
extends till the present day, and during it the mammals
still continue to be the prominent backboned animals, but
one of them, man, has become by far the most important
inhabitant of the world, and instead of merely being slowly
fitted to new conditions of life, now to a large extent controls
the conditions and changes them to suit his own convenience.

The second table (p. 8) shows what are the chief forms
of Proboscideans living at the different periods and their
distribution over the world. It will be seen that the earliest
mammal which can be definitely called a Proboscidean is
Moeritherium, a small tapir-like creature from the Middle
Eocene beds of the Fayum district of Egypt. This species
existed also in the Upper Eocene of the same region, but
was then accompanied by a larger and much more elephant-
like animal, Palaomastodon. At this time Africa was cut off
from the rest of the world to the north by a broad and deep
sea which extended from the Atlantic to the Pacific by way of
Northern India and Southern China, and the separation of
Africa prevented these early forms of elephant from wandering
into other regions till after the Eocene. After Palceomas-

ELEPHANTS.

GEOLOGICAL SUCCESSION. V

todon there is a large gap in the series, no fossil Proboscidea
having yet been found in the Oligocene, though no doubt
their remains will be discovered somewhere in the freshwater
deposits of that age in Egypt. Hitherto no Proboscidean bones
and teeth have been met with again till the Lower Miocene,,
but in rocks of that period they are found abundantly, not only
in Egypt, but in Europe, Asia, and North America. It is there-
fore clear that during the long lapse of time after the Eocene,
the deep sea above referred to, must to some extent have been
replaced by land, over which the early elephants could spread
outwards from their home in Africa. The vast changes in the
distribution of land arid water that took place in this region,
will be apparent when it is understood that rocks crowded with
the shells that lived at the bottom of this ancient sea are to-day
found thousands of feet above the sea-level in India and else-
where.

From the Lower Miocene period onwards we meet with
elephant-like animals in great variety all over the Northern
Hemisphere, wherever suitable deposits for the preservation of
their remains occur. At the end of the Pliocene period the
group also spread into South America, but at the present day
it is totally wanting in the whole Western Hemisphere.

During the later Miocene and Pliocene periods the head-
quarters of these animals seem to have been in India, for it is
there that we meet with the greatest number and variety of
forms, showing all grades of structure between the Miocene
types above referred to and elephants almost like those now
existing. From the end of the Pliocene to the beginning of
the Quaternary Period may be regarded as the time during
which the elephants reached their most flourishing condition,
both in the number of kinds that existed and in the wide
range over which they were spread. After this a gradual
decline in the group took place, till, at the present day, it is
represented by two species only, the African elephant confined to
Tropical Africa, and the Indian elephant found in India, Ceylon,
Burma, the Malay Peninsula, and some of the neighbouring
islands. But for restrictions placed upon their slaughter
even these last remnants of one of the oldest, and in many

10 ELEPHANTS.

ways the most remarkable, groups of mammals would soon
disappear, just as has happened for instance, in the case of
the great ground-sloths of South America, the giant lemurs
of Madagascar, and the giant marsupials such as Diprotodon
in Australia.

A more detailed account of the changes that the Proboscidea
have passed through will now be given, the following animals
being selected for description as representing six of the most
important of the successive stages at present known.

1. Moeritherium lyonsi, Middle and Upper Eocene.

2. Palaomastodon beadnelli, Upper Eocene.

3. Tetrabelodon angustidens, Miocene.

4. ,, lonyirostris, Lower Pliocene.

5. Stegodon insignis, Pliocene.

6. Elephas, Pliocene. Pleistocene, and Recent.

Some reference will also be made to other types, such as
Dinotherium and Mastodon.

Wall- Moeritherium (figs. 5, 7, 8) was an animal about the size of a
case 43. large Newfoundland dog, or of the tapir, which it must have

Table- much resembled in general appearance. It was common in the
region that is now known as the Fayum in Lower Egypt,
where its fossil remains occur in considerable quantities in
the Middle Eocene beds, intermingled with bones of toothed
whales (Zeugloduri), sea-cows (Eosireri), marine turtles (Pse-
phophorus and Thalassochelys), and snakes (Pterosphenus), as
well as skeletons of fishes. From this mixture of land and
aquatic animals it may be concluded that Moeritherium lived
near the shore, probably in swamps at the mouth of a great
river, where the remains of both marine and of drowned land-
animals would be mingled and entombed together in the muds
and clays, which accumulated in the estuary, and now make up
much of the strata found in this locality. In the Upper
Eocene beds, overlying those just described, remains of
Moeritherium are also found; here, however, there is no inter-
mingling of marine animals, but instead we find remains of
many remarkable land-mammals, crocodiles, and immense

MOERITHERIUM. 11

quantities of trunks of fossil trees, embedded in the sands and
gravels of a great river. Probably both the animals and the
tree-trunks were swept away by floods, their remains piled up
in shallows and places where the current was slack, and buried
in the mud and sand carried down by the stream.

The skull of the Moeritherium (see fig. 5) differs in no

Skull and lower jaw of Moeritherium from the Middle Eocene of the Fayum,
Egypt. \ nat. size.

ant.orb.j antorbital foramen ; c., canine ; e.v.oc., exoccipital ; fr.,
frontal; *. 1-3 incisors; jit., jugal; m. 1-3, molars; mx., maxilla ;
fi., nasal; p.a., parietal; par., paroccipital ; pm. 2-4, premolars ;
p.nLT., premaxilla; pt., post-tympanic process of squamosal ; s.oc., supra-
occipital ; sq. squamosal.

very marked manner from that of other primitive hoofed-
animals, and shows scarcely any trace of the peculiarities of
the skulls of the later Proboscidea. The most important
feature is the large nasal opening not quite at the end of the
snoutj the nasal bones being short ; this indicates that probably
there was already a short proboscis, something like that of the
tapir. Another interesting point is that some of the bones at
the back of the skull are thickened and contain air-chambers ;
in the later elephants this development of air-cells is carried to
such an extent that the whole form of the skull, particularly the
posterior portion, is entirely altered by it (see the broken
skull of the Indian elephant, stand G in Gallery). The reason
for this swelling of the bones is, that as the head becomes
heavier, owing in great part to the development of the trunk

32 ELEPHANTS.

and tusks, a larger surface for the attachment of the muscles
which support the head is necessary, and even in the small
Eocene Moeritherium change in this direction had begun.

In the fossil mammals it is the teeth that are of the greatest
importance in settling the relationships of different species to
one another, and in forming an opinion as to their food and
probable manner of life.

The reason for the importance of fossil teeth is, that while in
many ways they are very conservative, long retaining traces of
the earlier forms from which they have originated, nevertheless
they readily undergo change in accordance with the kind of
food they are called upon to seize and masticate. Moreover,
from their hardness they are more frequently preserved than
most other parts of the skeleton. It will be well, therefore,
before considering the teeth of Moeritherium and the other
Proboscidea to give a short account of mammalian teeth in
general, so that the descriptions later may be understood.
Central The tooth of a mammal (see Mammalia in Index collection
T?alj in Central Hall) consists of a root or roots embedded in
a socket in the jaw, and a crown which is exposed and is
adapted for the work it has to perform. In a section of a tooth
(fig. 6) it is seen that the greater part is composed of a hard
material called dentine, and that the crown of the tooth is usually
covered with a still harder substance known as enamel, while
in some cases the root, and in others (usually the more com-
plicated forms) both root and crown, may be coated with a softer
bony matter, called cement. The form of the crown differs
enormously in different animals and in different parts of the
mouth : it may be a simple cone like the canine or dog-tooth,
or it may form a large complicated grinding surface as in the
back teeth of the horse. The teeth are usually divided into
different series according to their particular position in the
mouth and the duties they have to perform (see figs. 1 & 2).
The front teeth implanted in the anterior part of the jaw are
mainly concerned in grasping and biting off the food : these are
called incisors. In mammals which, like the pig (fig. 1), possess
the primitive number of teeth, there are three of these on either
side in both upper and lower jaws. Behind these come the

CHARACTERS OF TEETH.

13

canines, one on each side above and below ; these are generally
more or less pointed teeth, serving chiefly for fighting or defence.

Central
Hall,
Bay I.

Diagrammatic section of various teeth.

1. Section of tusk of elephant, a permanently growing tooth. II. Section
of a young human incisor still growing, the root not yet fully formed.

III. Section of human incisor fully formed, the root being complete.

IV. Section of human molar showing the low-crowned (brachyodont)
condition, the cusps being rounded tubercles (bunodont). V. Section of
the molar of an ox, showing the high (hypsodont) complexly folded
crown. In the figures the enamel is black, the pulp white, the
dentine represented by horizontal lines, the cement by dots.

Behind these again are the cheek-teeth which have to do mainly
with the breaking up of the food before it is swallowed : in the

14 ELEPHANTS.

complete dentition there are seven of these on either side in both
jaws arid they are divided into two groups, (1) four premolars
in front, and (2) three molars behind: the premolars, or at
least those posteriorly, replace the milk-molars of the young
animal, while the molars have no predecessors. It is in the
premolars and molars that the greatest variety of structure is
found, as might of course be expected, because it is these teeth
that are most affected by the nature of the food. Teeth suitable
for cutting up flesh would be quite un fitted for grinding hard
vegetable matter, and consequently in animals feeding on soft
material the teeth differ widely from those in which the food is
hard and requires much mastication. In the former the crowns
of the teeth are low and their cusps or tubercles are either
sharp and cutting in the case of flesh-feeders (e. g., lions and
tigers) or rounded (bunodont) in the case of animals feeding on
a soft vegetable or mixed diet (e. g.y pigs and bears). Teeth
of this sort are called brachyodont (fig. 6, IV). In the case of
animals whose food is hard and requires much grinding, the
wear of the teeth is so great that simply-constructed low crowns
would be quickly worn out, and it must be remembered that
the length of an animal's life is largely dependent #n the time
during which its teeth remain in good working order. To meet
the increased wear the crown of the tooth becomes higher and
moves up in the gum as wear takes place, sometimes through-
out the animal's life or only for a time; teeth of this sort
are called hypsodont (fig. 6, V). With this increase in height
of the crown there is generally greater complication resulting
from the infolding of the enamel in various ways, and the
development of cement on the crown as well as on the roots.
One example of this gradual increase of height and complication
is about to be described in the case of the elephants ; another
extremely good instance is found in the gradual evolution of
the teeth in the horses, as excellently illustrated in the case in
the North Hall and in table-case 8 (see fig. 3).

From the above account it will be seen that when the full
number of teeth is present there are three incisors, one canine,
four premolars, and three molars on each side in the upper jaw,
and the same in the lower jaw. This is usually expressed shortly

MOERITHERIUM.

15

by a formula, thus: — I. |, C. \, Pm. *-, M| = }*, the letters
denoting the kind of teeth, the upper numbers the number
of each kind on one side in the upper jaw, the lower
numbers those in the lower jaw, so that in the example above
given there would be eleven teeth on each side in both the
upper and lower jaws, or forty-four in all. As has already been
mentioned, a greater or less number of these teeth may be
wanting in different animals and the formula will differ
accordingly : thus in man it is I. %, C. ^ Pm. ^ M. 5 = -8 or
thirty-two in all. In the cat it is I. ®, C. J, Pm. |, M. f = f-
or thirty in all.

With these preliminary remarks we may proceed to describe

Fig. 7.

Upper and lower teeth of Moeritherium.

A. Upper teeth. B. Premaxilla, large tusk-like second incisor. C. Man-
dible from outer side. c.} socket of canine ; i. 1-3, incisors ; m. 1-3,
molars ; pm. 2-4, premolars. £ nat. size.

the teeth of Moeritherium (figs. 5 & 7). In this animal the
dental formula is I.-®, C. -J, Pm. |, M. | = £ or thirty-six in all.

16 ELEPHANTS.

From the formula it will be seen that in the upper jaw only one
premolar is wanting to complete the primitive number, while
in the lower jaw an incisor and the canine are missing on
each side in addition to the premolar.

Of the upper incisors the second pair (i. 2) are greatly
enlarged and form strong downwardly directed tusks,, the
beginning of the great tusks so characteristic of the later
elephants. The canine (c.) seems to have been quite small and
unimportant, being on the way to disappearance. The pre-
molars (pm.) are separated from the canine by a short interval,
and as already mentioned the anterior one of the full denti-
tion is wanting. The remaining premolars are all simpler in
structure than the molars behind them, and consist of three main
cusps only^ the two front cusps being arranged in a transverse
line in the third and fourth premolars. These teeth are preceded
by milk-teeth which they displace from above as in the ordinary
mammals ; in the later elephants we shall see that this usual
replacement of milk-teeth by premolars is gradually lost.

The molars (fig. 7 A,m.) are the most interesting and important
of the fceeth, because it is in them that the most nearly complete
series of gradually more and more complicated forms can be
traced. In Moeritherium the crown of each upper molar is
composed of two transversely arranged pairs of knobs, giving
rise to a pair of transverse crests ; there are also in many cases
small posterior knobs — the first trace of the tendency to increase
the number of transverse crests by additions to the back of
the tooth, which is characteristic of the whole group.

In the lower jaw (figs. 5 & 7 C) the middle incisors (i 1) are
small, the second pair (i 2) large and tusk-like ; both are
directed forward and their upper surface continues forward the
surface of the spout-like anterior portion of the jaw. The third
incisors, the canine, and the first premolar of the full dentition
are wanting. The remaining three premolars (pm.}3 which re-
place milk-molars, are simpler than the molars, and only in the
third and fourth is there any arrangement of the anterior
cusps to form a transverse ridge. The first and second molars
(fig. 70), like those of the upper jaw, consist of two transversely-
arranged pairs of knobs and a posterior knob which is larger

MOERITHERIUM.

17

than in the upper teeth. The last lower molar (fig. 10 A) has Wall- ;
third ridge forming a sort of heel or talon, as it is called ; this case ^*

Table-
case 24.

tooth, though much smaller, is remarkably similar to the molars
of some of the earlier forms of Mastodon.

The skeleton of Moeritherium is not well known, but some of
the more important points in its structure are : — (1) the neck
is proportionately longer than in the later elephants in which it

18 ELEPHANTS.

Wall- is much shortened, (2) the hip-bones are narrow, while in the

°' later forms, owing to the great increase in size and weight, they

case LM are mucn expanded. The hnmerus also is rather different,

particularly at its lower end, the greater supinator ridge, so

characteristic of the larger Proboscidea, being scarcely developed.

On the other hand, the thigh-bone is extremely like that of a

very small elephant.

To sum up the primitive characters of Moeritherium : —

1. The skull differs little from that of an ordinary ungulate.

2. There is a full set of three pairs of incisors in the upper,

and only one pair wanting in the lower jaw.

3. There is a canine on each side in the upper jaw.

4. There are three premolars in the upper and lower jaws

replacing milk-molars.

5. The molars have only two transverse ridges and a small

hind lobe ; each transverse ridge is composed of two
distinct cusps ; the third lower molar has also a heel
or talon.

Some of the more important of the Proboscidean characters
are : —

1. The large size of the nasal opening, its somewhat

backward situation, and the small size of the nasal
bones.

2. The commencement of the development of air-cells in

the bones of the back of the skull.

3. The enlargement of the second incisors in both jaws to

form tusks.

4. The transversely ridged character of the molars.

5. The spout-like anterior portion of the lower jaw.

All these characters become much more marked in the next
stage, which is represented by Palaomastodon from the Upper
Eocene of Egypt.

The genus Palaeomastodon (figs. 9-11, 13) is represented
by several species varying in size from an animal little
larger than Moeritherium to one nearly as large as a small
elephant, so that in the matter of size alone there is a great
advance in the direction of the modern elephants. In the

PAL^EOMASTODON. 19

structure of the skull and teeth, as well as in the rest of the Wall-
skeleton, so far as known, the advance is likewise very striking. case 43-
In the skull (see fig. 9) the opening of the nostril (nar.) has ,^fc
shifted far back from near the end of the snout, though it is
still in front of the orbit of the eye. The nasal bones are still

Fig. 9.

Skull arid lower jaw of Palceomastodon, showing the elongated chin with
a pair of terminal incisors (/.«.), from the Upper Eocene (? Lower
Oligocene) of the Fayum, Egypt. -^ nat. size.

nar., position of opening of nose ; u.i., upper second incisor or tusk.

shorter and smaller than in Moeritherium. At the back of the
skull the development of air-cells in some of the bones has
enormously increased, but has not yet invaded the roof of the
skull, so that the sides are only separated by a sharp median
crest. The posterior surface of the skull slopes forward, and
there is a deep pit in the middle line for the attachment of the
muscles necessary to support the increasingly heavy head.

Of the incisor teeth only the second pair now remains, and
these have been still further enlarged, forming downwardly
directed curved tusks with a band of enamel along their outer
sides only. The canine has disappeared, but there are still
three premolars replacing milk-molars. The anterior premolar
is a simple cone, while the crown of the posterior one consists
of two transverse ridges, these teeth being subjected to two
distinct influences, namely, the tendency to the reduction in the
front of the series, and the tendency to become more like the

ELEPHANTS.

Wall-
case 43.

Table-
case 24.

molars at the back. The molars themselves show a distinct
advance, the crown of each consisting of three transverse ridges,
each ridge composed primarily of two main cusps which may,
however, show small traces of sub-division into secondary cusps.
The mandible (figs. 9 & 13) differs from that of Moeritherium in
the much greater prolongation of the spout-like anterior portion;
this now projects a considerable distance in front of the skull
and is prolonged still further forward by the single remaining
(second) pair of incisor teeth, which meet in the middle line and
form a sort of shovel-shaped extension ; the edges are worn both

Fig. 10.

Posterior lower molars (m2, m3) of (A) Moeritherium and (B) Palaomastodon,.
showing the increase in the number of ridges of Palaomastodon. f nat_
size.

on the upper and lower surface, so that these teeth were pro-
bably used for grubbing about in the ground to procure food,
and the upper surface must have been further worn by working
against the lower surface of the trunk or elongated upper lip.

PAL^OMASTODON. 21

There are only two premolars in the lower jaw, the anterior Wall-
of the three milk-molars (figs. 9 & 13) falling out without Ctise4^
being replaced from below ; the molars are three in number, ca^e 24
the two anterior with three transverse ridges, the third some-
times having in addition a small heel (fig. 10). It should be
noticed that in the full-grown animal all the molars and pre-
molars are in position and use at the same time ; it will be seen
that in the later forms of elephant-ancestors this is not so.

The skeleton, so far as known, is almost exactly like that of
a small elephant, the only important differences being that the
neck is longer and the limbs most likely less massive. The animal
as a whole (fig. 11) must have been very like a small elephant,
but would be distinguishable by the longer head and neck, and
by the fact that instead of possessing a flexible trunk it had a
long snout, the lower portion consisting of the elongated lower
jaw, the upper without bony support and probably extending
beyond the lower; the projecting portion most likely was more
or less flexible and capable of seizing objects, and was the
beginning of the prehensile trunk.

The chief steps taken by Pal&omastodon in advance ot
Moeritherium towards greater likeness to the later elephants
are: —

1. Considerable increase in size.

2. Lengthening of the snout, as shown by the mandible.

3. Loss of canines and all the incisors except the second

pair in both jaws.

4. Three-ridged molars.

5. Greater development of air-cells at back of skull.

6. Shifting further back of the nose-opening and smaller

size of the nasal bones.

7. Greater similarity of the bones of the skeleton to those

of ordinary elephants.

The next stage is found in Tetrabelodon angustidens (figs. 12— Pier-
14), from the Lower Miocene of Northern Africa, Europe, and 41^42
probably Asia. This animal is as large as a medium-sized Table-
elephant, and its teeth and skull are much more elephant-like case 23.
than in Pal&omastodon. Thus the nostrils have shifted still
further back, and the great development of air-cells in the bones

22

ELEPHANTS.

TETRABELODON. 23

at the back of the skull has led to the disappearance of the Pier-
ridge along the middle of the roof, which is now flat. Further, 4™8^'
the tusks, which were quite small and flattened in Palaomastodon, faW
are now large and round, and differ from those of a modern case 23.
elephant only in curving downward instead of upward, and in
having a band of enamel along the outer side relatively much
narrower than in Pateomastodon, in which nearly the whole

Fig. 12.

Skull and mandible of Tetrabelodon anyustidens, showing the greatly elongated
chin with a pair of terminal cutting incisors (/./.), from the Lower and
Middle Miocene, France. ^V nat* s*ze'

war., position of opening of nose ; u.i., upper incisors.

outer face of the tooth is enamel-clad. As the tusks increase in
size, the dentine of which they are composed acquires a peculiar
structure, which is shown in transverse sections of the tusks (see
table-case 24) and appears as a series of crossing lines curving
out from the middle of the tooth and giving a pattern like the
engine-turning on the case of a watch. This is quite peculiar
to elephant-tusks, and by it even small pieces of ivory can be
at once distinguished. This structure depends on the frequent
bending of the tubules which make up the dentine, and one result
of it is, that ivory is one of the most perfectly elastic of sub-
stances, and is therefor especially suitable for making billiard balls.

Pier-
cases
41, 42.

Table-
cases
23, 24.

-44 ELEPHANTS.

This form of ivory is not found in either the upper or lower tusks
of Moeritherium and Palceomastodon or in the lower tusks of
Dinotherium, and appears for the first time in the large tipper
tusks of Tetrabelodon angustidens. In some of the American
Tetrabelodouts in which the lower tusks are very large, this
structure is also found in them.

In the lower teeth also of Tetrabelodon angustidens (fig. 16 A)

Fig. 13.

A

m.7

mm.* pin 3.

flI7l4

Lower milk-dentition of (A) Pal(so?nastodon, (B) Tetrabelodon angustidens.
Showing some of the milk-teeth in situ, with the germs of replacing
premolars. The germs of the 2nd molars are shown at the back of the
jaw. About I nat. size.

i.f incisor; m. 1-2, permanent molars ; m.m. 3-4, third and fourth
milk-molars ; m.m., socket of second milk-molar ; pm. 3-4, premolars ;
sym., symphysis of mandible.

great changes have taken place. The milk-molars are still
replaced by premolars, but these are quickly dropped out
(fig. 13 B). This is chiefly the result of the great increase in
size of the true molars, particularly of the third, which is not
accompanied by a corresponding increase in the length of the
tooth-bearing portion of the jaws, so that there is not room for

TETRABELODON.

25

the premolars and molars to remain in position at the same
time. The consequence of this is, that as the posterior molars
are cut they move forward to take up their position in the jaw>
thrusting out the teeth in front of them so that in the adult
only the two large back molars remain on each side in both
jaws, and in old individuals perhaps only the last is left. This
mode of replacement is shown in fig. 13 B. In this species,
while the first and second molars still have only three ridges^
as in Palaomastodon, the last may have five or six, and, as
already mentioned, all are proportionately very large.

Fig. 14.

Pier-

41, 42.

Table-
cases
23, 24.

Restoration of Teirabelodon angustidens.

In the mandible (fig. 12) the anterior portion is extremely long
and projects much further beyond the skull than it does in Palao-
mastodon • and indeed it is in this species that we have the
greatest degree of lengthening of the lower jaw. The two lower

26 ELEPHANTS.

Pier- incisors, as in Pal&omastodon, help to add to the length, and

41^42 were no doubt used for grubbing in the earth. The remarks

Tabie- made about the upper cheek-teeth are equally true of the

case 23. lower. The neck seems to have been a little longer and more

Fig. 15.

Skull and mandible of Tetrabelodon lonyirostris, from the Lower

Pliocene, Eppelsheim, Hesse-Darmstadt.
i.f lower incisor ; m. 2-3, second and third molars. About -J^. na^ size.

flexible than in the modern elephant, but the limbs and other
parts were much the same. This animal when living (fig. 14)
must have been still more like an elephant than Palao-
mastodon, and the most noticeable difference would be that
here also, instead of the flexible trunk there was a long stiff

TETRABELODOX.

27

snout which was supported by the elongated front of the lower
jaw. Probably the end of the upper lip and nose was free and
moveable, and may even have been able to grasp objects to some

Fig. 16.

A

Second and third lower molars of (A) Tetrabelodou angustidens, and
(B) Tetrabelodon longirostris. ± nat. size.

extent, but the whole arrangement seems to be rather clumsy.
In most groups of animals as size increases the length of
the neck becomes greater in proportion, so that the animal can
still reach the ground ; but in these early elephants, in spite of

28 ELEPHANTS.

Pier- the great increase in size, the neck actually shortened, and it was
4jas^ only this extraordinary lengthening of the snont that enabled
Table- tne animals to reach to feed or drink.

case 28. The next stage in this strange history is found in Tetra-
belodon longirostris (fig. 15), an elephant of which the remains
are common in the Lower Pliocene of Eppelsheim in Germany
and other localities. In this animal the skull, so far as known,
does not differ to any great extent from that of Tetrabelodon
angustidens. The teeth, however, have advanced considerably
in size and complication. The first and second molars may
have four or five transverse ridges, while in the last there may
be as many as six ridges (fig. 16B). Only one of the milk-
molars is now replaced by a premolar, and both this and the
other milk-molars are early pushed out by the forward growth
of the large molars, only two of which at most on each side
remain in position in old animals. It is in the lower jaw,
however, that the chief changes have taken place. Here the
elongated anterior part (fig. 15), so striking in the last type, has
become shortened till it projects but little in advance of the
skull, and although its upper surface is still deeply grooved and
spout-like as in the earlier forms, the lower incisors no longer
meet in the middle line and prolong the spout, but are rounded
and separated from one another. In this animal it is clear that
the lower jaw was shortening up and could no longer reach the
ground, but doubtless the fleshy upper lip and nose, now freed
from their bony support for at least part of their length,
became flexible and better adapted for grasping the animars
food. In fact, this species must have looked much the same as
a modern elephant, except that it had a longer chin bearing a
pair of small downwardly directed tusks.

In some of the American Tetrabelodons of about the same
age as T. longirostris, the lower tusks, instead of undergoing
reduction, seem to have become greatly enlarged, and at the
same time the symphysial portion of the mandible is slightly
deflected, so that the mandible with its tusks is to some degree
similar to that of Dinotherium. An example of this form of
jaw is seen in the case of the mandible of a Tetrabelodon, from
the Loup Fork Beds (Upper Miocene) of Kansas, exhibited in
pier-case 42.

DINOTHERIUM.

29

Skull of Dinotherium giganteum from the Lower Pliocene of Eppelsheim,
Hesse-Darmstadt. ^U nat. size.

L O

found in the same deposits as the earliest known remains of
Tetrabelodon angustidens, and, as in the case of that species
its ancestors probably lived in Africa, though up to the pre-
sent no traces of them have been discovered. In the later
Miocene beds occur a number of species, some of enormous size
(e. g. D. gigantissimum from Roumania). The genus finally
disappears in Lower Pliocene times. The chief peculiarity
of these animals is, that the front part of their lower jaw is
turned sharply downward and bears two large tusks (fig. 17>

This species has, of course, no near relationship with Dino- Wall
therium (fig. 17), which forms a side branch of the Probo-
scidea, and is widely different from all the other members of
the group. The earliest known member of the genus is
Dinotherium cuvieri, a comparatively small animal^which is

Fig. 17.

30 ELEPHANTS.

Wall- caseC.) ; it is not certain whether or not there were any tusks
ease4.>. jfl ^ne Upper jaw. The teeth are of a much simpler character
'" than those found in the Mastodon-elephant line. In the upper
jaw of the adult there are two premolars, both simpler than
the molars ; of these latter the anterior one consists of three
transverse crests, the other two of two crests only,, a notable
peculiarity, since in the other Proboscidea it is the hindermost
molar which is the most complex. The posterior milk-molar
also has three crests. In the lower jaw in addition to the
down-turned tusks there are two prernolars and three molars.
As in the upper jaw, the premolars are simpler than the molars :
the anterior molar has three crests, the second and third two
only, though in the last there may be a trace of a third. In
the young there are three milk-teeth, the hinder one having
three crests, like the anterior true molar. All the molars and
premolars remain in use throughout the animal's life, a condition
already lost in the earliest Tetrabelodous, as has been already
described. It has been suggested that Dinotherium was a
more aquatic animal, but there is nothing in the structure of
the limbs to give any support to this idea, though the lowness
of the crowns of the teeth probably indicates that it lived on
soft herbaceous vegetation, such as may have grown in swampy
places.

In Tetrabelodon longirostris the main characteristics of the
modern elephants are already established, and the later changes
of importance include, (1) the still further reduction of the
mandibular symphysis and the loss of the lower incisors, and
(2) the great increase in the size and complication of the
cheek-teeth. It will be convenient to consider these changes
separately.

In the shortening of the chin the next stage is found in
Pliocene forms like Mastodon atticus from the Lower Pliocene
of Pikermi. In this the symphysis,*though much shortened,
is still more or less spout-like,, and in very young indivi-
duals the incisors may be present, though they are soon shed ;
the absence of these teeth in the adult is the chief character
distinguishing the genus Mastodon (fig. 18) from Tetrabelodon.
One of the best known species belonging to this stage of

MASTODON. 31

development is Mastodon arvernensis, which is found in Europe Pier-
in Pliocene deposits ; in it the lower incisors seem to be entirely 37^40
Avanting, and the anterior molars have four transverse ridges.

Fig. 18.

Mandible of a young individual of Mastodon americanus showing the small
remnants of the lower incisors. In the adult these are lost. From a
Pleistocene Deposit, North America. About 1 nat. size.

A nearly allied species, M. sivalensis (fig. 20), from the Pliocene
of India, is notable as showing a tendency to acquire five ridges
to its molars.

The region in which the passage from the Mastodons to the
true elephants occurred, seems to have been Southern Asia,

32 ELEPHANTS.

Pier- where, in a succession of Pliocene and Pleistocene deposits,
39^49 there is a complete series of forms passing from the Mastodon
up to the recent Indian Elephant. How far these changes
may have gone on in the rest of the Northern Hemisphere is
not known, but the history of the Mastodons in America is
rather different from that of the Old World forms. The first
members of the group to appear in North America are found in
the Loup Fork (Upper Miocene) beds of Montana, Nebraska,
and Kansas, and include Tetrabelodonts and Mastodons,
with such species as T. campestris, T. euhyphodon, T. pro-
ductus, &c. Some of these species have simpler teeth than
T. longirostris, and in this respect approach T. angustidens.
The presence of these animals must be attributed to immigra-
tions probably from Asia, since no Proboscidean remains
occur in the earlier Tertiary beds of North America. The
great increase in the size of the lower teeth of T. (?) precursor,
just described, seems to show that some of these American
species at least had turned away from the main line of advance
and become changed in particular directions, so that it is uncer-
tain whether the later types, in which the lower tusks were

Fig. 19.

Vertical longitudinal section of a molar tooth of a Mastodon, showing the
low crown, the open valleys between the cross-ridges and the thick
enamel (b). c., dentine. | nat. size.

reduced and finally lost, are descendants of these or represent
further invasions from outside. These later forms spread into
South America, where several peculiar species are found which
Stand B. persisted with little change till the Pleistocene. Probably the
reason why these Mastodons, as well as M. americanus of North

MASTODON. 33

America,, continued in a comparatively primitive condition, is pier-
to be found in their considerable isolation and freedom from *Qa^S
competition. One South American species, M. andium, is par-
ticularly interesting on account of its variability, especially in
the length of the chin and in the presence or absence of the
lower tusks. Some of these differences are due to sex, and
perhaps age, but the variations are no doubt mainly due
to the fact that the symphysis and lower tusks were undergoing
reduction and that the latter were about to disappear altogether.
Neither in North nor in South America does it appear that the
Mastodons gave rise to more advanced types, and the presence in
the Pleistocene of North America of true elephants (E. columbi
and E. primiyenius) is due to immigration from Asia. No
species of Elephas reached South America.

To return to the series of stages of development found in Pier-
Southern Asia, the first species that need be considered is ^^le
Mastodon cautleyi, which in the character of its teeth is nearly cases
related to Tetrabelodon longirostris, but the ridges of the molars 23> 24*
are comparatively higher. There is, however, as yet no cement
in the valleys (fig. 19), which are more or less obstructed by
small tubercles, and the inner cusps wear into a trefoil pattern,

Fig. 20.

Grinding surface of a lower molar tooth of Mastodon sivulensis.
From the Lower Pliocene of the Siwalik Hills, India, f nat. size.

as in Tetrabelodon angustidens and T. lonyirostris. While,
however, the anterior molars are almost identical with those of
T. longirostris, the posterior lower molar is very similar to that

34 ELEPHANTS.

of M. latidens, which in its turn approaches Elephas (Stegodon)
clifti (fig. 21) very closely. In M. latidens there are as a rule
five transverse ridges in the second upper molar and six in
the last.

Pier- The next stage is represented by Elephas (Stegodon) clifti

35 3£ (fig. 21). With this species we reach the true elephants, though
Table- *ne molars have much lower crowns and fewer transverse ridges
case '24. than in the modern species of Elephas ; and, in fact, these
Stand J. earijer forms are sometimes separated into another genus called
Stegodon. Since, however, no real line can be drawn between
them and the later types it is perhaps best to call all Elephas, but
distinguish these earlier forms by adding the name Stegodon in
brackets as above. In all this group the lower incisors have-
entirely disappeared, the anterior elongation of the chin at

Fig. 21.

Grinding surface of an upper molar tooth of Elephas (Sfagodon) clifti. From
Lower Pliocene, Siwalik HillSj India. Showing six transverse ridges.
£ nat. size.

the same time being reduced to a mere peg-like process
(figs. 23 & 30), and a greater or less amount of cement (see
above, p. 12) fills the transverse valleys in the crowns of the
molars (fig. 22). In Elephas (Stegodon) clifti the number of
ridges is greater than in Mastodon cautleyi. In order to express
briefly the number of ridges in the molars of this and other
species, a formula is used thus : — In E. clifti the formula

ELEPHAS (STEGODON).

35

M 1 ?=1 M 2 v M 3 £=£, means that in the first true molar (M 1)
there are in the upper jaw 6-7 ridges, while in this case the
number in the lower is not known. In the second molar (M 2)
there are 6 in the upper, the lower being uncertain, while in
the last molar (M 3) there are 7-8 ridges in both the upper and
lower jaw. It will be seen, therefore, that the numerator of the
fractions represents the number of ridges in the upper teeth,,
the denominator the number in the lower. When the two
numbers are given thus, 6-8, it means that the number of
ridges varies between them. This formula will be employed
below in describing the molars, and a table showing the gradual
increase in the number of ridges will be given at the end.

In E. clifti premolars are still developed, but must have
been pushed out almost as soon as the milk-molars they replace.
The crowns of the molars are quite low, and there is only a
small quantity of cement in the valleys between the ridges.

In Elephas (Stegodori) bombifrons and E. insignis (fig. 22),
which represent the next stage, the transverse ridges are some-
what more numerous and at the same time are higher, and the

Fig. 22.

Vertical longitudinal section of molar tooth of Elephas (Steyodori) insigms.
From Lower Pliocene, Siwalik Hills, India. Showing the wide valleys
between the cross-ridges filled with cement («), the dark band marked
b being the enamel, and beneath that the dentine (c). § nat. size.

valleys are filled with cement to a greater degree. Nevertheless
these differences are not very marked, and in the case of indi-
vidual teeth it is often difficult to be sure to which of these

D2

Pier-
cases
35, 36.

Table-
case 24.
Stand J.

36 ELEPHANTS.

Stand J. species they belong* In some of the Stegodont elephants the
tusks attain an enormous size ; for instance, in a skull of

Fig. 23.

Skull and mandible of Elephas ganesa, showing the immense upper tusks
and the shortened chin. From Lower Pliocene, Siwalik Hills, India.
About Jg nat. size.

Elephas (Stegodon) ganesa (fig. 23), exhibited in the gallery
(stand J), the tusks project for a distance of 9 ft. 9 in. beyond
the sockets.

Pier- In the next stage we pass from the low-crowned Stegodont
case 34.

Fig. 24.

Grinding surface of an incomplete upper molar of Elephas planifrons.
From Lower Pliocene, Siwalik Hills, India. § nat. size.

group to animals in which the ridges are considerably higher
and the valleys completely filled with cement ; this is called
the Loxodont group. The most primitive member is Elephas

ELEPHAS.

37

planifrons (figs. 24 & 25), m which the posterior molars may have
as many as twelve ridges. This is the last of the elephants in
which premolars have been observed ; these teeth are small and

Fig. 25.

Vertical longitudinal section of molar tooth of Elephas planifrons, showing
the deeper and narrower valleys completely filled with cement (a). The
enamel layer is marked b, the dentine c. f nat. size.

closely crowded up beneath the milk-molars, so that when those
teeth are greatly worn the premolars are exposed to view, and
are no doubt shed as soon as, or even before, the posterior part
of the milk-molars they should replace.

The species E. meridionalis (fig. 26), of which remains are

Fig. 26.

Grinding surface of upper molar of Elephas meridionalis. Upper Pliocene,
Tuscany. About £ nat. size.

found in the Pliocene of Middle and South Europe, seems to be
closely related to E. planijrons, and is in about the same stage of

Pier-
case 34.

Table-
case 24.

38 ELEPHANTS.

Pier- evolution, or perhaps a little more advanced. The last molars
33t may have 13-14 ridges, and in some cases approach the condition

ca^e 20 seen m -^ hysudricus. E. meridionalis attained enormous dimen-
sions, some individuals probably standing about 15 feet in height
at the shoulder. Remains of this species occur in the Forest
Bed of Norfolk (see table-case 20) and in the Upper Pliocene
deposits of the Val d'Arno and the Auvergne. According to
Leith Adams (' British Fossil Elephants/ p. 232) the molars are
distinguished by the following characters : — " Crowns very broad ;
columns short as compared with E. antiquus, and generally as
compared with E. primigenius ; the enamel of the discs thick
and rarely crimped, but usually uneven, looped or channelled ;
plates wide apart, with thick wedges of cement." Numerous
teeth of this species are shown in table-case 20.

Pier- Returning to the Indian series, the next stage may be taken

as represented by Elephas hysudricus, in which there is a con-
siderable increase in the height of the teeth and in the number
of the ridges. The skull is in many respects similar to that
of E. maximw, the modern Indian elephant, which may have
been its direct descendant. E. hysudricus, like E. meridionalis,
sometimes attained a very large size. This species occurs at
the end of the Pliocene and perhaps in the early Pleistocene

Fig. 27.

Grinding surface of second lower molar of Elephas antiqum from the
Pleistocene of Grays, Essex. £ nat. size.

Pier- beds of the Narbada. Contemporary with it was E. antiquus
caseSa (fi^ 27), a large elephant found in the late Pliocene and early

cases" Pleistocene of Europe, and appearing in beds of the latter

19, 19 A. horizon of India, where it has been called. E. namadicus. In

this species the skull is peculiar from the development of a

ELEPHAS.

39

sort of overfolded ridge on its frontal portion, forming an Pier-
overhanging fold on the forehead (see pier-case 34). The
presence of this peculiarity makes it very unlikely that any
of the later forms in which it is absent were descended
from this species, which is of particular interest because
it was very widely distributed. In the changes that took
place in the distribution of land and water at the end of the
Pliocene and the beginning of the Pleistocene, poriions of
the regions inhabited by this form and its varieties became
isolated as islands, and in these restricted habitats the species
became dwarfed and the dwarf Forms in the different islands
at the same time became specifically distinct from one another.
Instances of these small forms are Elephas melitensis, E. mnai- Table-
driensis, E. Cypriotes, and E. creticus. E. melitensis and ^^/A
E. mnaidriensis are found in Malta. Of the first-named species
a small form sometimes called E. falcuneri did not stand more
than about three feet high at the shoulder. The ridge-formula
of the molar teeth is :— M 1 1=|, M 2 $, M 3 J£=J|. A large col-
lection of remains of these species, obtained by Admiral Spratt
and Professor Leith Adams, is shown in table-cases 21 & 21 A.
E. mnaidriensis is also found in Sicily ; probably this species,
which is larger than E. melitensis, represents the intermediate
stage between it and E. antiquus. E. Cypriotes from Cyprus,
and E. creticus from Crete, were both discovered and described
by Miss D. M. A. Bate, who collected the specimens shown
in table-case 17 A.

Turning again to the main line, we find that Elephas hysu- Table-
dricus probably passed into some such species as E. armeniacus,
which in many respects is intermediate between the Mammoth
(E. primigenius] and the living Indian Elephant (E. maximus).
The Mammoth seems to represent the highest pitch of evolu-
tion attained in the Elephantidse, being in some respects in
advance even of the Indian Elephant. It is here that we meet
with the greatest number of ridges in the molars (fig. 28), the
formula being M 1 ££, M 2 JJlJJ, M 3 J^JJ. These teeth
represent the culmination of the long series of changes above
described, all tending to increase the efficiency of the molars

40

ELEPHANTS.

Pier- as grinding organs. The great size, and especially height, of
30^32. ^ie crown giyes tnem a prolonged period of wear, while the
Table- numerous alternating plates of enamel, dentine, and cement,

cases of different degrees of hardness, ensure that the grinding
1 7—19.

Fig. 28.

Grinding surface of molar tooth of the Mammoth (Elephas primigenius),
showing some still unworn posterior plates. About f nat. size.

surface will remain sufficiently rough for its purpose through-
out the period during which the tooth remains in use. The
Mammoth was a very widely-distributed form, being found all
over Northern Europe, Asia, and America, and it seems to
have been particularly abundant in Siberia and the islands to the
north, where remains occur in great abundance, and whence
the tusks are actually exported for commercial purposes. The
extinction of the Mammoth appears to have been a compa-
ratively recent event, and in Siberia portions of carcases with
the skin and flesh in good preservation are found in the frozen
tundras. An instance of this kind is illustrated by drawings
and photographs on the pillar between pier-cases 31—32. In
this instance the animal seems to have fallen into a hole and to
have died in its efforts to scramble out. The mouth was found
still filled with the grass on which the animal was browsing at
the time when it met with the accident. This individual,
restored and mounted in the, attitude in which it was found,
is now exhibited in the Imperial Academy of Sciences at
St. Petersburg. This specimen, with many others, shows that
the Mammoth was covered with a reddish-brown wool and long
dark hair, while the tail ended in a large tassel of hair. A

ELEPHAS.

41

piece of the skin with its woolly covering and some of the long Pier-
hair are shown in pier-case 31. Further confirmation of this ('ase31-
peculiarity of the Mammoth is found in the rude sketches
scratched on ivory by early Man, who was evidently quite
familiar with the animal. A reproduction of one of these early
drawings is shown near wall-case 1.

The finest Mammoth skull hitherto collected in England Caae K.
is shown in case K : this specimen was obtained from the brick-
earth at Ilford in Essex. Most of the skeleton seems to have
been found with it, but the bones were unfortunately destroyed
before their interest was recognised. The tusks in this skull
are 10 feet 6 inches in length beyond the sockets.

Fig. 29.

A.

Grinding surface of upper molars of (A) the Asiatic, and (B) the
African Elephant. About ^ nat.size.

The Indian elephant, E. maximus (fig. 29 A), one of the two
surviving species of the suborder, is found in India, Ceylon,
and the Malay Peninsula to Sumatra. The chief peculiarities
of the species,, distinguishing it from the African elephant, are

Stande
D,E.

42 ELEPHANTS.

Stands the flatness of the forehead, the comparatively small ears, the
D, E. presence of a single finger-like process at the front of the end
of the trunk (fig. 31 A), and the presence of four or five nails
on the hind feet. As might be expected from the wide range
of this species, different local forms can be distinguished, and
in some cases these have even been regarded as specifically
distinct, as, for instance, the elephant of Sumatra, which has
been called E. sumatranus. It seems, however, that all are
merely geographical races of the same animal. The Ceylon
form is said to be as a rule tuskless, and although tusk-
bearing forms do occur in the island, they may be either animals
imported from the mainland of India or the result of former
interbreeding with such. In India, also, some individuals, called
Muchnas, are tuskless or have very small tusks (see mounted
skin, stand E). The Sumatran type differs in being rather
more slightly built, and in possessing a rather longer trunk and
more expanded end to its tail. The elephants from Further

Fig. 30.

Skull of the African Elephant (Elephas africanus). About ^ nat. size.

India and the Malay Peninsula are probably also a distinct
race which seems especially liable to produce albino forms,
the white elephants of Siam and Burma being well known.

The origin of the other living species of elephant, Elephas
qfricanus (fig. 30), is not very clearly known, owing to the want
of fossil remains in the • later Tertiary beds of Africa. It
appears probable that the narrow-toothed, straight-tusked form
of E. antiquus was either the ancestor, or nearly allied to the

ELEPHAS. 43

ancestor, of the African Elephant, a conclusion that is sup-
ported by the recent discovery in Zululand, in beds of probably
late Pliocene age, of a molar which is very similar to some
molars of E. antiquus. This Zululand species has been called
Elephas zulu, and was found associated with remains of other
mammals, the first Tertiary mammalian fauna hitherto dis-
covered in South Africa. ,

The African Elephant to-day ranges very widely over Africa
south of the Sahara, but fossil remains have been found in
Northern Africa and in the south of Europe. It is distin-
guished by its convex forehead, its very large ears, and by the
presence of two finger-like processes on the tip of the trunk
(see fig. 31 B). The molar teeth are considerably simpler than
those of the Indian species (see fig. 29 B), the ridges being

Fig. 81.

A B

Tip of trunk of (A) the Asiatic and (B) the African Elephant.

fewer in number and widening out in the middle in a peculiar
manner ; the teeth also are relatively smaller than in other
elephants. In fact if, as is almost certain, E. africanus is
derived from some such species as E. antiquus or E. zulu, then
the changes that the molars have undergone are not in the
same direction as those which gave rise to the Indian Elephant
and the Mammoth, but are to some extent retrograde.

It has lately been shown that although there is only a single
species of African Elephant, nevertheless, in different parts of
the continent, there are different local races which may perhaps
be regarded as subspecies, and are in fact species in the making.
Differences in the form of the skull of these different races can
be detected, but the most striking characters distinguishing
them from one another are the size and shape of the ears.

44 ELEPHANTS.

According to Mr. Lydekker the distinctive features of these
local races are : —

I. In the Addo Bush, or East Cape, Elephant (Elephas
africanus capensis) the ears are rather small, somewhat square
in shape, with rounded corners, and a small, sharply pointed
angular lappet at the lower angle. The forehead falls away
towards the temples, so as to appear highly arched.

II. The West Cape Elephant (E. a. toxotis) has the ears much
larger (4 ft. 5 in. in a female 8 ft. 8 in. high), longer, and
semi-oval in shape; although, as in the preceding race, sharply
inflected at the lappet.

III. In the Matabeleland race (E. a. selousi) the ear is much
less elliptical than in the West Cape Elephant, and approaches
more to that of the Camerun race, but agrees with that of the
former in that the lappet underhangs the jaw and chin.

IV. In the West African Elephant (E. a. cyclotis), typically
from South Cameruns, the ears are very large, but of quite
different shape, the contour being oval, and the lappet in the
form of half -ellipse. The skin has a mosaic-like appearance,
and its colour is paler grey than in most other races. The
Congo Elephant, which comes very close to this type, has long
and slender tusks.

V. The Masai Elephant (E. a knochenhaueri) , typically from
German East Africa, has small triangular ears, with the lappet
angulated and pointed. The exhibited specimen (which stands
11 ft. 4 in. in height, with ears measuring 4 ft. 2J in. by 3ft.
5 in.) may belong to this race.

VI. In the Aberdare Elephant (E. a. peeli), typically from
the Aberdare Mountains, British East Africa, the ears are pear-
shaped, with the lappet very long, although somewhat rounded
at the tip.

VII. The Lake Rudolf Elephant (E. a. cavendishi) is nearly
allied to the last, but has broader ears, in which the lappet is
shorter.

VIII. In the Abyssinian, or Sudan, Elephant (E. a. oxyotis)
the ears form an elongated triangle, with the upper border
rounded and the lappet very sharply pointed and angular. This
elephant attains very large dimensions.

RELATIONSHIPS. 45

IX. The N. Somali Elephant (E. a. orleansi), on the other Central
hand, is smal^ with the upper border of the small ears straight,

and the lappet short and distinctly denned.

X. In the West Sudan Elephant (E. a. rothschildi) the ears are
in some respects intermediate between those of the Abyssinian
and those of the West African race, although approximating to
the former in the shape of the lappet.

There is also a dwarf elephant from the Congo (E. a.pumilio).
The Albert Nyanza Elephant has been separated as E. a. albert-
ensis, and is characterised by the unusually short and broad
skull.

RELATIONSHIPS OF THE PROBOSCIDEA.

The discovery of the Eocene Proboscidea proves that, although
the elephants are no doubt rightly included among the Ungu-
lata or hoofed -animals, they are at the same time very widely
separated from the other members of that group, or at least
from the existing members of it, and seem to have formed an
independent series from the earliest Tertiary times, when they
probably arose from some quite generalised form of the primitive
group called the Condylarthra. A very early side-branch from
the Proboscidean stem is probably represented by the Sirenia
or Sea-cows, aquatic animals which, though now as unlike
elephants as possible, still possess a number of anatomical
peculiarities in common with them, so that this relationship
was recognised long ago. Recently a number of Sirenians have
been found in the Eocene strata of Egypt (pier-case 30), one Flor-
in the same horizon as Moeritherium. These early Sirenians caseji°
are much less specialised than the existing forms, having
the full series of teeth and a complete pelvis, and probably a
functional hind limb. In many points they resemble Moeri-
therium, as for instance in the form of the brain, teeth, and
pelvis ; and it seems likely that both they and Moeritherium
had a common ancestor in Lower Eocene times. It was pro-
bably a swamp-living creature, some of whose descendants
became gradually more and more exclusively aquatic in their
habits, thus giving rise to the Sireuia, while others became

46

ELEPHANTS.

exclusively terrestrial and gave rise to the elephants as described
above.

\
TABLE SHOWING THE RIDGE-FORMULA OP THE TRUE MOLARS IN THE

APPROXIMATE LINE OF DESCENT OF THE ELEPHANTS FROM

Moerith eriuti i.

Moeritherium
Palaeomastodon
Tetrabelodon a
k

Mastodon cautleyi

„ latidens

Elephas (Stegodoi

31 JJ

» ))

Elephas planifrons
7, hysudricus
,9 indicus
„ primiyt

A series of specimens illustrating the gradual increase in the
•number of ridges on the molars is shown in table-case 24.

Ml.

M2.

M3.

2

2

2

3

2
3

3
3

3

3

3

fti QJ'if/0'n t*

3

4—5

lit) 1 1 tie/to ...

5

4=5

4

4

5—6

girostns ~

4

6=6

/i 4

4

5

4

is 4

4
4—5

5
5—6

4

m)clifti . . . Y

4—5
6

"?

5—6

7—8
7—8

bombifrons . -J-

6—7
7—8

8—9

8—9

insiqnis . Ii^8

7—10

7—8
8^12

9—11
9—13

HS ^

8—9

10—12

7

8—9

10—13

'US . .

10—12

13—17

12

12—13
16

14—18
24

. . , . . . 12

16

24—27

iius 9~15

14—16

18-27

**• * ' ' • ' it— 15

14—16

18—27

Printed by TAYLOR and FRANCIS, Bed Lion Court, Fleet Street.

GUIDE-BOOKS.

{The Guide-books can be obtained only at tlie Museum. Postage extra.)
General Guide to the Museum, 8vo. 3d.
Guide to the Galleries of Mammals, 8vo. Qd.
- Great Game Animals, Svo. Is.

Horse Family, 8vo. Is.

Elephants (Recent and Fossil), Svo. Qd.

Gallery of Birds, 4to. 2s. Qd.

General Series of Birds, 4to. Qd.

Nesting Series of British Birds, 4to. 4d.

Gallery of Reptilia and Amphibia, Svo. Qd.

Gallery of Fishes, Svo. Is.

— Exhibited Series of Insects, Svo. Is.

— Shell and Starfish Galleries, Svo. Qd.

— Coral Gallery, Svo. Is.

— Fossil Mammals and Birds, Svo. Qd.

— Fossil Eeptiles and Fishes, Svo. Qd.

— Fossil Invertebrate Animals, Svo. Is.
Mineral Gallery, Svo.

Index to the Collection of Minerals, Svo. 2d.

An Introduction to the Study of Minerals, with a Guide to the Mineral
Gallery, Svo. Qd.

to the Study of Rocks, Svo. Qd.

— to the Study of Meteorites, Svo. Qd.
Guide to Sowerby's Models of British Fungi, Svo. 4d.

the British Myceto/oa, Svo. - 3d.

List of British Seed-plants and Ferns, Svo. 4d.

Special Guides: No. 1. Old Natural History Books, Svo. 3d.

— No. 2. History of Plant Classification, Svo. 4d.

No. 3. Memorials of Linnaeus, Svo. 3d.

Handbook of Instructions for Collectors, Svo. Is. Qd. ; or in eleven
separate sections, at 3d. or 4d. each.

CATALOGUES (Selection).

History of the Collections : — Vol. I. Libraries ; Botany ; Geology ;

Minerals. 1904, Svo. 15s.— Yol. II. Zoology. 1906, Svo. «£1 10*.

Catalogue of the Library of the British Museum (Natural History)

Vols. I., II. 1903-4, 4to. £1 each.

ReportonJ;he Zoological Collections made .... during the Voyage of
-TOTS. « Alert/ 1881-82. 54 Plates. 1884, Svo. £1 10s.
Report on the Collections of Natural History made in the Antarctic
Regions during the Voyage of the ' Southern Cross/ 53 Plates.
1902, roy. Svo. £2.

Reports on the Natural History of the ' Discovery ' National Antarctic
Expedition, 1901-1904 :— Vol. I. Geology. 10 Plates, 72 Text-
figures, 2 Maps. 1907, 4to. £1 10s.~ Vol. II. Zoology (Vertebrata :
Mollusca: Crustacea). 33 Plates, 146 Text-figs. 1 Map. 1907,4to. £3.
Vol. III. Zoology (Invertebrata) and Botany (Marine Algae: Musci).
51 Plates, 8 Text-figures, 1 Chart. 1907, 4to. £2 Ws.
Monograph of Christmas Island (Indian Ocean). Woodcuts, 22 Plates

and a Map. 1900, Svo. .£1.

First and Second Reports on Economic Zoology. Text illust. 1903-4.
A roy. Svo. 6s. each.

Catalogue of Monkeys. Lemurs, and Fruit-eating Bats. Woodcuts.
1870, Svo. 4s.

— Carnivorous Mammalia. Woodcuts. 1869, Svo. 8s. Qd.

Seals andWhales. 2nd edition. Woodcuts. 1866, Svo. 8s.

. Supplement. Woodcuts. 1871, Svo. 2s. Qd.

List of the Specimens of Cetacea. 1885, Svo. Is. Qd.

Catalogue of Ruminant Mammalia (Pecora). 4 Plates. 1872, Svo. 3s. Qd

CATALOGUES (Selection continued).

Catalogue of Marsupialia and Monotremata. 28 Plates. 1888, 8vo. £lSs.

Birds. Vols. VII.-XXVII. Woodcuts and Coloured

Plates. 1883-98, Svo. 14s. to 36s. a volume.
Hand-list of the Genera and Species of Birds. Vols. I.-IV. 1899-

1903, 8vo. 10s. a volume.

Catalogue of Birds' Eggs. Vols. I-IV. Coloured Plates. 1901-5,
8vo. £1 5s. to <£! 10s. a volume.

Chelonians. Woodcuts and Plates. 1889, 8vo. 15s.

Lizards. 2nd edition. Vols. I.-III. Plates. 1885-87, 8vo.

20s. to 26s. each.

Snakes. Vols. I.-III. Woodcuts and Plates. 1893-96,

Svo. 17s. 6d. to £1 6s. each.

Pishes. 2nd edition. Vol. I. Woodcuts and 15 Plates.

"TgSHTSvo. 15s.

Catalogue of British Echinoderms. Woodcuts and Plates. 1892, 8vo.
"12s. 6d. ~

Madreporarian Corals. Vols. I.-VI. Plates. 1893-1906,

4to. 18s. to 35s. a volume.

Illustrations of Australian Plants collected in 1770 during Captain
Cook's Voyage round the World in H.M.S. ' Endeavour.' Part I.
101 Plates. 1900, fol. £1 5s.— Part II. 14^ Plates. 1901, fol.
£1 15s.— Part III. 77 Plates and 3 Maps. 1905, fol. £1 5s.
Catalogueof African Plants collected by Dr. P. Wei witschin 1853-61 : —
Vol. I. 4 Parts, 1896-1900, 8vo. 4s. to 7s. 6d. e'ach. — Vol. II.
2 Parts, 1899-1901, 8vo. 6s. each.

Catalogue of Fossil Mammalia. Parts I.-V. Woodcuts. 1885-87,
8vo. 4s. to 6s. a volume.

Fossil Birds. 75 Woodcuts. 1891, 8vo. 10s. 6d.

T'ossli Keptilia and Amphibia. Parts I. -IV. Wood-
cuts. 1888-90, 8vo. 7s. 6d. a volume.

Fossil Fishes. Parts I.-IV. Woodcuts and Plates.

vo. 21s. a volume.

Catalogue of the Tertiary Vertebrata of the Fayum, Egypt. 98 Text-
figures and 26 Plates. 1906, 4to. £1 15s.

List of British Oligocene and Eocene Mollusca. 1891, Svo. 6s.

Catalogue of Australasian Tertiary Mollusca. 8 Plates. 1897, Svo. 10s.

Catalogue of Fossil Cephalopoda. Woodcuts. Part I. 1888, Svo.
10s. 6d. Part II. 1891, Svo. 15s. Part III. 1897, Svo. 12s. 6d.

British Fossil Crustacea. 1877, Svo. 5s.

the Jurassic Bryozoa. 22 Woodcuts and 11 Plates.

1896, 8vo. 16s.

Catalogue of the Cretaceous Bryozoa. Vol. I. 64 Woodcuts and
17 Plates. 1899, 8vo. 16s.

Blastoidea. 20 Plates. 1886, 4to. 25s.

The Genera and Species of Blastoidea. 1899, Svo. 3s.

Catalogue of Paleozoic Plants. 1886, Svo. 5s.

, Mesozoic Plants. Woodcuts and Plates. Svo. : —

The Wealden Flora. Part I., 1894, !0s. Part II., 1895, 15s.
The Jurassic Flora. Part I., 1900, 20s. Part II., 1904, 10s.

Catalogue of the Fossil Plants of the Glossopteris Flora. 51 Text-
figures and 8 Plates. 19U5, Svo. 12s. 6d

The above-mentioned Catalogues may be purchased of Messrs.
LONGMANS & Co., 39 Paternoster Row ; Mr. QUAEITCH, 11 Graf ton
Street, New Bond Street ; and Messrs. DULAU & Co., 37 Soho Square;
or at the NATURAL HISTORY MUSEUM, Cromwell Road, London, S. W.
A more detailed list may be obtained on application to the DIRECTOR
of the Museum.

14 DAY USE

RETURN TO DESK FROM WHICH BORROWED

•EARtH SCIENCES LIBRARY

This book is due on the last date stamped below, or

on the date to which renewed.
Renewed books are subject to immediate recall.

(B1321slO)476

General Library

University of California

Berkeley

BRITISH MUSEUM (NATURAL HISTORY).

DAYS AND HOURS OP ADMISSION.

The Exhibition Galleries are open to the Public, free, every week-
day in

January, from 10 A.M. till 4 P.M.

February 1st to 14th, „ „ „ „ 4.30 „

February 15th to end, „ ,, ,. ,,5 ,.

March, „ ,-, „ „ 5.30 .,

April to August „ ., „ ., 6 „

September, „ ,, „ „ 5.30 „

October, „ „ „ „ 5

November and December, ,, ,, „ „ 4 ,,

Also, from May 1st to the middle of July, on Mondays and Satur-
days only, till 8 P.M.,

and from the middle of July to the end of August, on Mondays
and Saturdays only, till 7 P.M.

The Museum is open on Sunday afternoons throughout the year.
The Museum is closed on Good-Friday and Christinas-Day.

BY ORDER or THE TBUSTEES.