CHAPTER X

HISTORY OF THE PROBOSCIDEA

Utterly foreign as the elephant-tribe appears to be to present-day North America, it was a very conspicuous element in the fauna of that continent from the middle Miocene to the end of the Pleistocene, and in the latter epoch it spread over South America also. Like so many others of the mammals which have, from time to time, flourished in the Americas, the elephants and their allies, the †mastodons, were immigrants from the Old World, and, until comparatively lately, the region of their origin was a complete mystery. They appeared suddenly and unheralded and at approximately the same time in Europe and North America and nothing is known from preceding geological formations of either continent which could with any plausibility be regarded as ancestral to them. The mystery was dispelled by the discoveries of Dr. C. W. Andrews in Egypt, which demonstrated that these strange and huge beasts had originated in Africa and had migrated thence through Asia to Europe, on the one side, and to North America on the other.

The proboscideans occupy a very isolated position among the hoofed mammals, and in structure they display a curious mingling of high specialization with an extreme conservatism of primitive characters, the specialization being exemplified in the teeth and head and the conservatism in the body and limbs, very much as in the †oreodont family of artiodactyls (p. 382). The most conspicuous of the external features in the order is the long trunk, or proboscis, which gives its name to the group, and is a great prolongation of the nose, with the nostrils at the end and a finger-like tip, which can be used to pick up minute objects.

[Illustration: FIG. 223.—Molar of the African Elephant (_Loxodonta africanus_) showing the oblique mode of wear. Heavy black lines indicate enamel, enclosing areas of dentine, cement covering the whole tooth.]

In the true elephants the dental formula is: _i_ 1/0, _c_ 0/0, _p_ 0/0, _m_ 3/3, × 2 = 14, though this formula is misleading, to the extent that the milk premolars, three in number in each jaw, take the place and perform the functions of the premolars, thus adding 12 to the effective number of teeth. The single upper incisor on each side grows into an immense tusk, which has enamel only on the tip, where it is speedily worn away; the lower jaw is without incisors and there are no canines above or below. The grinding teeth are very large and have a highly complex structure and a most exceptional method of eruption on coming into use. They are thoroughly hypsodont and each is composed of a large number of high, broad and thin plates of dentine covered with enamel and the spaces between the enamel ridges are filled with cement (see Fig. 47, p. 97); indeed, the whole tooth is so thickly covered with cement that, when unworn, it looks like a mere lump, with no ridges showing on the surface. The teeth increase in size and in the number of component ridges from before backward, and in the Indian species (_Elephas maximus_) the number of ridges in the six grinding teeth, including the milk premolars, is: 4, 8, 12, 12, 16, 24. In the African Elephant (_Loxodonta africanus_) the teeth are not so high and have fewer and thicker plates, the formula being: 3, 6, 7, 7, 8, 10. The teeth do not succeed one another vertically in the normal mammalian fashion, but come in successively from behind and the series moves forward, so that the foremost tooth is pushed out, when it is so worn down as to be of no further service. As these teeth are very large and the jaws are relatively short, only one tooth on each side, above and below, is in use at the same time, though part of a second may also be involved. The movement of the successive teeth is not directly forward, but oblique, an upper tooth coming forward and downward and a lower tooth forward and upward. In consequence of this arrangement the teeth are abraded obliquely, the anterior part first coming into use, and, by the time a tooth is fully in place, the front portion is worn down to less than half the height of the hinder part. All of these peculiarities in the dental system imply a very high degree of specialization and a notable difference from other mammals.

The skull is equally specialized, as is indeed required by the character of the teeth and the development of the long and heavy proboscis. The premaxillæ are converted into sheaths for the great tusks; the nasals are extremely abbreviated and the anterior nasal opening is shifted to the top, directly above the posterior opening, so that the nasal canal passes vertically downward through the skull. All of the bones forming the cranium are enormously thickened and at the same time lightened by the formation of an extensive system of communicating sinuses, and thus the brain-chamber is, as it were, hidden away in the middle of the huge mass of the skull. This explains the difficulty of killing an elephant by shooting it in the head; the shot must be so directed as to reach the brain, which requires knowledge and skill.

The neck is short, the body long and extremely massive, the tail of moderate length. The shoulder-blade is very large and has a prominent metacromion given off from the spine; the hip-bones are immensely expanded in correlation with the breadth of the thorax and abdomen. The limbs are long, massive and columnar, their upper segments, especially the thigh, are very long, so that the knee-joint is brought below the body and free from it to the position of the hock-joint in the Horse; hence, the hind leg appears to bend in the opposite direction from the bend in the legs of ordinary quadrupeds, in which the true knee-joint is concealed. The fore-arm bones are separate and, for most of its length, the ulna is far heavier than the radius, a wide departure from the proportions usual in hoofed animals. The femur has no pit in its head for the round ligament and no third trochanter; the shaft is broad and much flattened, having quite lost the normal cylindrical shape. The bones of the lower leg are also separate, but the fibula, though stout, is very much more slender than the ulna. The long bones have no marrow-cavities, but are filled with spongy bone. The feet are extremely short and broad and of columnar shape, the weight resting upon a pad of elastic tissue and the small, nail-like hoofs are mere excrescences upon the periphery. There are five digits in manus and pes, but not all of them have hoofs; in the Indian and West African species the number of hoofs is five in the fore foot and four in the hind, in the East African four and three respectively. In the adult the skin is quite hairless, though the young calf has a considerable quantity of hair.

[Illustration: FIG. 224.—Right manus of the Indian Elephant (_E. maximus_).]

At present, the Proboscidea are restricted to the warmer parts of Asia and Africa, where five species, four of them African, are recognized. This is a very great reduction in the number of species and in the area inhabited during the Pleistocene epoch, when they ranged through every continent, except Australia, and were adapted to every climate from the tropics to the shores of the Arctic Sea. Four distinct species of proboscideans existed in Pleistocene North America, three elephants and a †mastodon, though not all in the same areas, nor probably all at the same time, their ranges both in time and space overlapping to a greater or less degree, but not exactly coinciding in either respect.

[Illustration: FIG. 225.—Vertical section through the manus of the Indian Elephant. _U_, lower end of ulna. _L_, lunar. _M_, magnum. _III_, third metacarpal. _1_, _2_, _3_, phalanges. _E_, pad of elastic tissue. (After M. Weber.)]

The first species was an immigrant, the northern †Mammoth (_Elephas †primigenius_), which extended over the greater part of the northern hemisphere, both in the Old World and in the New. This is the species of which complete carcasses with hide and hair have been found in the frozen gravels of northern Siberia, its structure and appearance being thus almost as well known as those of any modern elephant. That the †Mammoth was perfectly adapted to life in a climate of severe cold is shown not only by the contents of the stomach, which are comminuted fragments of present-day Siberian vegetation, but also by the dense coat of woolly hair, covered by long, coarse outer hair, which afforded full protection against the cold. The tusks, with considerable variation of form, had a tendency to spiral curvature, curving first downward and outward, then upward and inward; the grinding teeth were characterized by their relative breadth and the numerous thin enamel-ridges which traversed them. The number of these ridges was very variable in different individuals, but may be expressed for the six successive teeth as follows: 3-4, 6-9, 9-12, 9-15, 14-16, 18-27. The skeleton was more like that of the Indian Elephant than of the other species, though with a number of small differences in the skull. In size, the †Mammoth was comparatively small, standing about nine feet six inches at the shoulders. In North America its range was from Alaska southeastward across the continent to New England.

The second species, the †Columbian Elephant (_E. †columbi_ Fig. 114, p. 198), was eighteen inches or more taller than the †Mammoth and rivalled the largest existing elephants in stature; its huge tusks curved first downward and then upward and inward, their tips crossing when full-grown. The grinding teeth had fewer and thicker enamel plates than those of the †Mammoth. The range of the †Columbian Elephant overlapped the southern border of that of the †Mammoth, but was, on the whole, much more southern; it crossed the continent from ocean to ocean and covered nearly the whole of the United States, extending down to the southern end of the Mexican plateau. The two species were very closely related and in some cases are so intergraded that it is difficult to distinguish them; the †Mammoth was an undoubted immigrant and the †Columbian Elephant was probably a local North American variant of it, adapted to a somewhat warmer climate. Nothing is known of the skin or hair in the latter animal, but, from the fact that it was not a tropical species and was exposed to very cold winters, it may be inferred that it had a hairy covering of some sort.

The third species of elephant (_E. †imperator_) was older geologically than the others, as it was more characteristic of the lower Pleistocene and uppermost Pliocene; its range coincided with the western half of the region covered by _E. †columbi_, extending far into Mexico, but not occurring east of the Mississippi River. It was an enormous creature, the largest of known elephants, with an estimated height of thirteen and a half feet at the shoulder (Osborn). The grinding teeth had thicker and more crumpled enamel plates than in either of the other species.

[Illustration: FIG. 226.—The American †Mastodon (_†Mastodon americanus_), Pleistocene. Restored from a skeleton in the museum of Princeton University.]

The fourth of the Pleistocene proboscideans of North America was a member of a different and much more ancient genus, _†Mastodon_, which in the Old World became extinct before the end of the Pliocene. The American †Mastodon (_†M. americanus_) was thus a belated survival of an ancient type, seemingly out of place even in the strange Pleistocene world, which had so many bizarre creatures. The distinguishing characteristic of the genus was in the simple, low-crowned and comparatively small grinding teeth, which had three or four prominent transverse ridges, covered with heavy enamel, and, usually, with no cement on the crowns. As these teeth were so much smaller than those of the elephants, as many as three on each side of each jaw might be in simultaneous use. In this species there was no vertical succession of teeth, but in some of the Tertiary †mastodons such succession has been observed. The long tusks were directed nearly straight forward and were almost parallel, with but slight curvature, the convexity downward. In the males there was a short single tusk or, less commonly, a pair of such tusks, in the lower jaw, which were probably not visible externally; these were the vanishing remnants of an earlier stage of development, when the †mastodons had a fully developed pair of lower tusks, nearly as large as the superior pair.

[Illustration: FIG. 227.—Last lower molar of the American †Mastodon.]

The skull, while essentially proboscidean, was yet much lower and flatter and less dome-like than in the elephants; the thickening of the cranial bones was less extreme. The remainder of the skeleton differed so little from that of the elephants as to require no description. In size, this species about equalled the †Mammoth, the larger individuals measuring nine feet six inches at the shoulder. Remains have been found which prove that the American †Mastodon had a covering of long, coarse hair, and that it fed upon the leaves, shoots and small branches of trees, especially of conifers. There is much reason to believe that the species outlived the elephants in this continent and persisted until after the establishment here of the American Indian, and it may well have been human agency which finally extinguished the dwindling race. The range of the species nearly coincided with that of the †Columbian Elephant, but did not extend so far into Mexico, and in the central part of the continent reached much farther north, even into Alaska.

In the Pliocene of Texas, Nebraska and Idaho lived the American representatives of a genus (_†Stegodon_) which was a connecting link between the elephants and the †mastodons, and which was especially characteristic of the Pliocene of India. The tusks, which were confined to the upper jaw, had lost their enamel and the last molar, above and below, had five or six enamel ridges, but the crowns, which in the Asiatic species were buried in cement, had but a small amount of this material. Several species of _†Mastodon_ occur in the same beds, but only isolated teeth have been found.

The †mastodons, in a broad sense of the term, have been divided into several genera and subgenera in accordance with different schemes; the simplest perhaps is to group into a second genus those species which had fully developed lower tusks. This four-tusked genus has received several names, of which _†Tetrabelodon_ is most commonly used in this country, but the term _†Gomphotherium_ is much older and, according to the law of priority, must therefore be employed. The lower Pliocene species of _†Gomphotherium_ had a pair of large lower tusks, of cylindrical shape, and both upper and lower tusks had longitudinal bands of enamel, and in order to support the weight of these great tusks the symphyseal, or chin, region of the lower jaw was greatly elongated; the molars had four cross-crests.

[Illustration: FIG. 228.—Head of upper Miocene †mastodon (_†Gomphotherium productum_) showing the chisel-like lower tusks. Restored from a skull in the American Museum of Natural History.]

In the upper Miocene is found another and more primitive stage of proboscidean development. In these species the grinding teeth were three-ridged; the upper tusks were quite short and curved downward, diverging somewhat from each other, and they had enamel bands. The lower tusks were still shorter and of depressed, flattened and somewhat chisel-like form and so worn as to show that they were regularly employed in cropping and browsing. The skull was low and broad and the symphysis of the lower jaw was greatly prolonged to carry the tusks.

A very important fact concerning these early †mastodons is that they had the normal method of tooth-succession, permanent premolars forming beneath (in the lower jaw, above in the upper) the milk-teeth and pushing them out at maturity.

Of the middle Miocene proboscideans not much is known beyond the mere fact of their presence in North America at that time and indeed little of the skeleton, other than the skull, has yet been found in the American Miocene; but well-nigh complete skeletons have been obtained from the middle Miocene of Europe, and these bring out the surprising fact that the body and limbs of these species did not differ in any noteworthy manner from those of the existing elephants; the modern skeletal structure of these animals had been attained at a time when the dentition and skull were still in a far less advanced stage of development. In size, however, there was a decided difference, the species of the American Miocene rarely attaining a height of six feet.

Proboscidea have been reported from the lower Miocene of the Great Plains, but the material is insufficient for a definitive judgment. There is no doubt as to their presence in Europe at that time, but in neither continent can the history be traced any farther and we must turn to Africa for a backward continuation of the story. In the lower Oligocene of the Fayûm, southwest of Cairo in Egypt, occurs the highly interesting genus _†Palæomastodon_, which was much more primitive than any of the genera described above, though it was an unmistakable member of the order and even of the family Elephantidæ. The dentition was already much reduced, giving the formula: _i_ 1/1, _c_ 0/0, _p_ 3/2, _m_ 3/3. The upper tusks were short, compressed, directed downward, and slightly divergent, and had a broad band of enamel on the outer side; the lower tusks were still shorter and procumbent, pointing straight forward, and were covered with enamel, which was very thick on the lower side and thin or wanting on the upper. All of the grinding teeth were in place and function at the same time, which was not true of any of the genera previously considered, and each of the premolars had its predecessor in the milk-series, which it succeeded and displaced in the normal vertical manner. The premolars were smaller and simpler than the molars, which were made up of three pairs of conical tubercles arranged to form a three-crested crown. The skull, as compared with that of the elephants, was long and narrow, the premaxillaries extending into a long snout; the nasals were shortened, though not so much as in the succeeding genera, and there was probably rather a long and flexible snout than a true proboscis. The skull had a long and well-defined sagittal crest, which none of the later genera had, and the development of sinuses in the cranial bones, though considerable, was much less than in the elephants. The occiput was relatively high and the thickened parietals did not tower above it to any such degree as they do in the elephants. The symphysis of the lower jaw was greatly prolonged, extending out beyond the ends of the upper tusks, and this implies that the lower lip had a corresponding prolongation.

The skeleton is still incompletely known, though it may be said that the neck was probably longer than in the subsequent genera of the family. The limb-bones were already proboscidean in character, differing only in details from those of the more typical members of the order, but the animal was more lightly built and had less massive limbs. The presence of the third trochanter on the femur, which is lacking in all of the succeeding forms, is an interesting approximation to other and still more primitive groups of ungulates. The several species of _†Palæomastodon_ represent a considerable range in size, from animals which were not much larger than a tapir to those which equalled a half-grown Indian Elephant.

It is possible to take another and very long step back from _†Palæomastodon_, so long, indeed, as to make it apparent that one or more links in the chain are still missing. The genus _†Mœritherium_ is found together with _†Palæomastodon_ in the lower Oligocene, but also occurs separately in the upper Eocene. It seems likely that it is a persistent middle Eocene type and that the known species of it were somewhat aside from the main line of descent, but that it very closely represents, nevertheless, a very early stage in the elephant genealogy. These known species were quite small animals, about the size of a tapir, and therefore not much less than the smaller members of _†Palæomastodon_. The dental formula of _†Mœritherium_ was: _i_ 3/2, _c_ 1/0, _p_ 3/3, _m_ 3/3, × 2 = 36. The first or median upper incisor was a relatively small and simple tooth, but the second was quite a large, downwardly directed tusk, which was much smaller and less curved than in _†Palæomastodon_, and was not capable of indefinite growth. The third incisor and the canine were small, spike-like teeth of no functional importance, but their presence is significant as approximating the primitive, unreduced dentition of the ungulates. The lower incisors were nearly procumbent, with a slight upward inclination; the first one was long and the second a thick, enamel-covered tusk, with a chisel-like edge, which was produced by wear. The premolars were smaller and simpler than the molars, which were quadritubercular, the four conical cusps arranged so as to form two transverse crests, giving a pattern like that of the early pigs and peccaries and of precisely the kind that might have been predicted from the teeth of _†Palæomastodon_.

The skull had an utterly different appearance from that of _†Palæomastodon_, the difference being much greater than between the latter and the Miocene _†Gomphotherium_. It was long and narrow, and, except for the very prominent zygomatic arches, of nearly uniform, tubular shape, the brain-case being of small capacity, though, as compared with other Eocene mammals, the brain was proportionately large. “It is possible that the early tendency toward a considerable cerebral development shown in these primitive Proboscidea is one of the causes why the group has survived and flourished through so long a period” (Andrews). The cranium was very long and the facial region extremely short, the premaxillaries not being prolonged into a snout, as they were in _†Palæomastodon_; the occipital bones formed nearly the entire posterior surface of the cranium and even encroached slightly upon the roof. There was a long, but not very prominent, sagittal crest, and some of the cranial bones were much thickened; in one species the hinder part of the cranial walls was distinctly inflated, a beginning of the enormous thickening which has culminated in the true elephants. The nasal bones were already much shortened, though they were twice as long as those of _†Palæomastodon_, and the animal would appear to have had an incipient proboscis.

The neck was of moderate length and the body very long, with at least twenty pairs of ribs, and there was probably a long tail. The hip-bone differed remarkably in its extreme narrowness from that of the later Proboscidea and the limb-bones were much more slender, though not dissimilar in shape.

At a very early period the order became divided into two main branches, one of which includes all the forms so far considered, and the other the very strange _†Dinotherium_. The †dinotheres entered Europe together with the †mastodons in the lower Miocene and continued into the Pliocene without much change and then died out, leaving no descendants. They never invaded North America, probably because they were of more or less aquatic habit, like the hippopotamuses, and therefore less likely to find suitable conditions in the narrow and unstable land-bridges which connected the Old World with the New, than were animals of purely terrestrial habitat. The †dinotheres were of huge size, equalling the larger elephants in this respect and closely resembling them in the skeleton of the body and limbs. As usual in this order, the generic peculiarities were to be found in the teeth and skull. There were no superior tusks, all the upper incisors and canines being lost, but there was a pair of large lower tusks, which were directed downward, with a strong backward curvature. The dental formula then was: _i_ 0/1, _c_ 0/0, _p_ 2/2, _m_ 3/3, × 2 = 22. The grinding teeth were relatively quite small and had, except the first molar, two transverse crests, giving a pattern singularly like that seen in the tapirs. The skull was remarkably long, low and flat, and no doubt these animals had a proboscis of some sort. That the †dinotheres were derived from the same ancestral stock as the †mastodons and elephants is perfectly obvious and is not questioned by any one, but it is not yet possible to trace the connection.

The proboscideans were late immigrants into South America, being known there only in the Pleistocene and late Pliocene times, and only the †mastodons entered the southern continent, where they gave rise to several peculiar local species in Argentina, Bolivia, Chili and Brazil; one of these (_†Mastodon andium_) had a deposit of cement on the crowns of the grinding teeth. Why the elephants, which extended to the northern border of the Neotropical region, should have failed to reach South America and maintain themselves there, is but one of many similar questions to which no assured answer can be given.

The evolution of the Proboscidea was, in a certain sense, very similar to that of the †oreodont family (p. 381) among the Artiodactyla, in that the developmental changes affected chiefly the dentition and the skull, the skeleton of the body and limbs having very early acquired a character which was afterward but little modified. Were the skull and teeth of the lower Miocene _†Gomphotherium_ not known, we should hardly hesitate to refer the skeleton to the genus _Elephas_, and even in the Oligocene _†Palæomastodon_ all the bones of the skeleton, other than the skull, were characteristically and unmistakably proboscidean. On the other hand, the transformations of the teeth and skull were very profound and far-reaching, very much more so than those which took place in the †oreodonts.

[Illustration: FIG. 229.—Evolution of the Proboscidea: on the right, a series of skulls; on the left, last lower molar. (After Lull, modified by Sinclair.) N.B. _†Tetrabelodon_ should read _†Gomphotherium_.]

In the dentition we may consider separately the development of the tusks and of the grinding teeth. The first step in the known series, as exemplified by _†Mœritherium_, was the enlargement of the second incisor in each jaw to form a tusk which, though actually quite long, was very small when judged by the proboscidean standard. The upper tusk was directed vertically downward and the lower one was procumbent, pointing almost directly forward; the third incisor and the canine were small and in the lower jaw already lost. In the next known stage, _†Palæomastodon_, all of the anterior teeth, except the tusks, had been suppressed; the upper tusks were longer and more curved and of an oval cross-section; they extended less directly downward and more forward, while the enamel was restricted to the outer side of the tusk; the lower tusks were more fully procumbent than in the preceding genus. The third stage, that of the lower Miocene _†Gomphotherium_, showed the upper tusks greatly elongated and directed more forward than downward, while the lower tusks were but little larger than before. From the middle Miocene two phyla may be distinguished by the tusks alone; in one, which was not destined to long life, the lower pair increased greatly both in length and in diameter, while in the other series they rapidly diminished and eventually disappeared. Even in the Pleistocene, however, the American †Mastodon had remnants of these tusks in the males. In the later †mastodons, the †stegodonts, and true elephants, the upper tusks, which alone remained, lost the enamel bands and attained enormous proportions, differing in the various genera and species in the extent and direction of curvature. An aberrant mode of tusk development was to be seen in the †dinotheres, in which the upper pair was suppressed and the lower pair enlarged and so curved that the points were directed backward.

The grinding teeth underwent much more radical and striking changes. At first (_†Mœritherium_) they were small, very low-crowned and of simple pig-like or quadritubercular pattern, making two interrupted cross-crests; all were in use simultaneously and the succession of milk-teeth and premolars was by vertical replacement, as in normal mammals generally. In _†Palæomastodon_ there were three pairs of tubercles on the molars and in _†Gomphotherium_ these coalesced into ridges, but in all the †mastodons there was more or less distinctness of the conical tubercles. In one or more phyla the three-ridged plan persisted for a long time, one such phylum terminating in the Pleistocene _†Mastodon americanus_. In the other series the number of ridges increased, first to four, then to five, six and more (_†Stegodon_); the crowns of the teeth became much larger and higher, and the ridges, as their number increased, became much thinner, and the valleys between them were filled with cement, and finally, in the true elephants, with their fully hypsodont, many-crested teeth, were thickly covered all over with cement. The vertical succession of milk-teeth and premolars was retained in _†Gomphotherium_, at least in some species, but the large molars, which could not find room to be exposed while the premolars were in place, came in successively from behind. This horizontal mode of succession is the only one to be seen in the true elephants, in which but one tooth, or parts of two, on each side of each jaw are in simultaneous use and the premolars have entirely disappeared, but the milk-teeth are retained.

The changes in the skull, which amounted to a reconstruction, were very largely conditioned by the great increase in the length and consequent weight of the tusks, in the size of the grinding teeth and the development of the proboscis. In the earliest known type (_†Mœritherium_) the skull had little about it that would, at first sight, suggest proboscidean affinities; it was long and narrow, with sagittal crest and occiput of normal type, very long cranial and very short facial region. The nasal opening was directed forward and the nasal canal was relatively long and horizontal in direction, but the nasal bones were already much shortened, indicating that the proboscis was probably in an incipient stage. The symphysis of the lower jaw was procumbent and somewhat elongated, but to only a comparatively slight degree.

While the skull of _†Mœritherium_ was not obviously proboscidean, that of its successor, _†Palæomastodon_, was unmistakably so, yet retained several primitive features, which were lost in all of the subsequent genera, such as the sagittal crest, the relatively low cranium and moderate thickening of the cranial bones, the forward direction of the nasal opening, etc.; the symphysis of the lower jaw was very greatly prolonged.

As the tusks enlarged and the proboscis grew longer, the weight of the head and its appendages necessitated a largely increased area of attachment for the neck-muscles, and this was attained by a very great thickening of the cranial roof, the occiput not increasing proportionately; at the same time, the thickened bones were honeycombed with sinuses, so as to reduce their weight without sacrifice of strength. In those species of the Miocene _†Gomphotherium_ which had large and heavy tusks, this thickening was not very much less than in the true elephants. The enlargement of the tusks had other consequences, as, for example, in lengthening and broadening the premaxillaries and, in the elephants, in their downward bending, so as to shorten still further the facial region of the skull. With the development of the proboscis, the nasal bones were reduced to a minimum and the anterior nasal opening was no longer directed forward, but obliquely upward, while the nasal passage lost its horizontal direction and became almost vertical. The lower jaw continued to elongate the symphysis, reaching a maximum in certain species of _†Gomphotherium_; but the reverse process of shortening this anterior region of the jaw began with the reduction of the lower tusks, and, when these had disappeared, nothing remained of the immensely elongated symphysis, except the short spout of the elephant’s jaw. As the grinding teeth increased in height, there was a concomitant increase in the vertical depth of the jaws for their lodgment.

It was an obvious advantage in the mechanical problem of supporting the enormous weight of head, tusks and trunk to shorten the neck and thus bring the weight nearer to the point of support at the withers, the lengthening proboscis rendering it unnecessary for the mouth to reach the ground in feeding or drinking. The other parts of the skeleton underwent comparatively little change, the degree of modification being greatest between _†Mœritherium_ and _†Palæomastodon_. Throughout the series the bones of the fore-arm and lower leg remained separate, and the feet very short and five-toed. In size also the great stature and massiveness were attained early. After the great migration of the Proboscidea to the northern continents, we find considerable differences of size between the various phyla, though all were very large, and even as early as the lower Miocene of France, there were species which rivalled the modern elephants in bulk. It was this rapid attainment of great size and weight which appears to have been the determining factor in the conservatism of the skeleton. After the skeleton had become fully adjusted to the mechanical necessities imposed by immense weight, and that adjustment, as we have seen, was effected at a comparatively early period in the history of the order, then no further modification of importance would seem to have been called for. No doubt the habits and mode of life of these massive, sedate and slow-moving animals underwent but little change from the lower Oligocene onward. There is reason to think that _†Mœritherium_ was semi-aquatic and a haunter of marshes and streams, but, if so, the change to a life on dry ground was complete in the lower Oligocene, for the structure of _†Palæomastodon_ gives no reason for supposing that it was anything but a dweller on solid land and a denizen of forests.

Although this book does not undertake to deal with the obscure problems connected with the marine mammals, it may be noted in passing that one of these problems has been brought near to solution, if not actually solved, by the discoveries in Egypt and that is the question concerning the origin of the Sirenia. The order includes the existing Manatee or Sea-Cow (_Manatus_) of the coast of Florida, northeastern South America and western Africa, and the Dugong (_Halicore_) of the Indian Ocean. These are mammals which are adapted to a strictly marine habitat and are incapable of existence on land, having lost the hind limbs and converted the fore limbs into swimming paddles. Unlike the whales, porpoises and other Cetacea, the Sirenia are herbivorous and feed upon seaweed and eel-grass and the aquatic plants of large rivers. The Egyptian discoveries tend very strongly to the conclusion that the Sirenia and Proboscidea were both derived from a common stock and that the genus _†Mœritherium_ was not very far removed from the probable ancestor from which both of the orders descended.