had been simple, there would have been no difficulty in considering Sivatherium as simply a gigantic four-horned Antelope, essentially similar to the living Antilope (Tetraceros) quadricornis of India.  The hinder pair of horns, however, is not only much larger than the front pair, but each possesses two branches or snags—­a peculiarity not to be paralleled amongst any existing Antelope, save the abnormal Prongbuck (Antilocapra) of North America.  Dr Murie, however, in an admirable memoir on the structure and relationships of Sivatherium, has drawn attention to the fact that the Prongbuck sheds the sheath of its horns annually, and has suggested that this may also have been the case with the extinct form.  This conjecture is rendered probable, amongst other reasons, by the fact that no traces of a horny sheath surrounding the horn-cores of the Indian fossil have been as yet detected.  Upon the whole, therefore, we may regard the elephantine Sivatherium as being most nearly allied to the Prongbuck of Western America, and thus as belonging to the family of the Antelopes.

[Illustration:  Fig. 245.—­Skull of Sivatherium giganteum, reduced in size.  Miocene, India. (After Murie.)]

It is to the Miocene period, again, to which we must refer the first appearance of the important order of the Elephants and their allies (Proboscideans), all of which are characterised by their elongated trunk-like noses, the possession of five toes to the foot, the absence of canine teeth, the development of two or more of the incisor teeth into long tusks, and the adaptation of the molar teeth to a vegetable diet.  Only three generic groups of this order are known-namely, the extinct Deinotherium, the equally extinct Mastodons, and the Elephants; and all these three types are known to have been in existence as early as the Miocene period, the first of them being exclusively confined to deposits of this age.  Of the three, the genus Deinotherium is much the most abnormal in its characters; so much so, that good authorities regard it as really being one of the Sea-cows (Sirenia)—­though this view has been rendered untenable by the discovery of limb-bones which can hardly belong to any other animal, and which are distinctly Proboscidean in type.  The most celebrated skull of the Deinothere (fig. 246) is one which was exhumed from the Upper Miocene deposits of Epplesheim, in Hesse-Darmstadt, in the year 1836.  This skull was four and a half feet in length, and indicated an animal larger than any existing species of Elephant.  The upper jaw is destitute of incisor or canine teeth, but is furnished on each side with five molars, which are opposed to a corresponding series of grinding teeth in the lower jaw.  No canines are present in the lower jaw; but the front portion of the jaw is abruptly bent downwards, and carries two huge tusk-like incisor teeth, which are curved downwards and backwards, and the use of which is rather problematical.  Not only does the Deinothere occur in Europe, but remains belonging to this genus have also been detected in the Siwalik Hills, in India.