here, and it must be enough to say that while it cannot be assigned to either group, yet in the distribution of hair on the muzzle, in the presence of a small suborbital gland, in shortness of tail and the light color of its horns, it is sheep-like; in the absence of interdigital glands, the shortness and stoutness of its cannon bones, and in the presence of a small accessory inner column on the upper molars, it is bovine. But in the coarse longitudinal striation of the bases of its horns it differs from both. The shape of the horns is also peculiar. Curving outward, downward and then sharply upward, with broad, flattened bases meeting in the middle line, their outlines are not unlike those of old bulls of the African buffalo.
At the present time the musk-ox inhabits only arctic America, from Greenland westward nearly to the Mackenzie River, but its range was formerly circumpolar, and in Pleistocene times it inhabited Europe as far south as Germany and France. The musk-ox of Greenland has lately been set aside as a distinct species. The most we can say is that Ovibos is a unique form, standing perhaps somewhere between oxen and sheep, and descended from an ancient ruminant type through an ancestry of which we know nothing, for the only fossil remains which are at all distinguishable from the existing genus, are yet closely similar to it, and are no older than the Pleistocene of the central United States; in earlier periods its history is a blank about which it is useless to speculate.
The last of our three anomalies, the white, or mountain goat (Oreamnos montanus), is not as completely orphaned as the other two, for it seems quite surely to be connected with a small and peculiar series consisting of the European chamois and several species of Nemorhaedus inhabiting eastern Asia and Sumatra. These are often called mountain antelopes, or goat antelopes. So little is yet known of the soft anatomy of the white goat that we are much in the dark as to its minute resemblances, but its glandular system is certainly suggestive of the chamois, and many of its attitudes are strikingly similar. In all the points in which it approaches goats it is like some, at least, among antelopes, while in the elongated spines of the anterior dorsal vertebrae, which support the hump, and in extreme shortness of the cannon bone, it is far from goat-like. The goat idea, indeed, has little more foundation than the suggestive resemblance of the profile with its caprine beard. It is truly no goat at all, and should more properly be regarded as an aberrant antelope, if anything could be justly termed "aberrant" in an aggregation of animals, hardly any two of which agree in all respects of structure. No American fossils seem to point to Oreamnos, and as Nemorhaedus extends to Japan and eastern Siberia, it is probable that it was an Asiatic immigrant, not earlier than the Pleistocene.
From this intricate genealogical tangle one turns with relief to the deer family, where the course of development lies reasonably plain. If the rank of animals in the aristocracy of nature were to be fixed by the remoteness of the period to which we know their ancestors, the deer would out-rank their bovine cousins by a full half of the Miocene period, and the study of fossils onward from this early beginning presents few clearer lines of evidence supporting modern theories respecting the development of species, than is shown in the increasing size and complexity of the antlers in succeeding geological ages, from the simple fork of the middle Miocene to those with three prongs of the late Miocene, the four-pronged of the Pliocene, and finally to the many-branched shapes of the Pleistocene and the present age. Now it is further true that each one of these types is represented today in the mature antlers of existing deer, from the small South American species with a simple spike, up to the wapiti and red deer carrying six or eight points, and still more significant is it that the whole story is recapitulated in the growth of each individual of the higher races. The earliest cervine animals known seem to have had no antlers at all, a stage to which the fawn of the year corresponds; the subsequent normal addition in the life-history, of a tine for each year of growth until the mature antler is reached, answering with exactness to the stages of advance shown in the development-history of the race. A year of individual life is the symbol of a geological period of progression. This is a marvelous record, of which we may say—paraphrasing with Huxley the well-known saying of Voltaire—"if it had not already existed, evolution must have been invented to explain."
The least technical, and for the present purpose the most useful of the characters distinguishing existing deer from all of the bovine stock, lies in the antlers, which are solid, of bony substance, and are annually shed. They are present in the males of all species except the Chinese water deer, and the very divergent musk-deer, which probably should not be regarded as a deer at all. They are normally absent from all females except those of the genus Rangifer. Most deer have canine teeth in the upper jaw, though they are absent in the moose, in the distinctively American type and a few others. The cleaned skull always shows a large vacuity in the outer wall in front of the orbit, which prevents the lachrymal bone from reaching the nasals. No deer has a gall bladder. There are many other distinctions, but as all have exceptions they are of value only in combinations.
The earliest known deer, belonging to the genus Dremotherium, or Amphitragulus, from the middle Tertiary of France, were of small size and had four toes, canine teeth and no antlers. Their successors seem to have borne simple forked antlers or horns, probably covered with hair, and permanently fixed on the skull. Very similar animals existed in contemporaneous and later deposits in North America. From this point the course of progress is tolerably clear as to deer in general, although we are not sure of all the intermediate details—for it must not be forgotten that a series of types exhibiting progressive modifications in each succeeding geological period is quite as conclusive in pointing out the genealogy of an existing group as if we knew each individual term in the ancestral series of each of its members. Thus we do not yet know whether the peculiar antler of the distinctively American deer, of the genus Mazama, is derived from an American source or took its origin in the old world, for the fossil antlers known as Anoglochis, from the Pliocene of Europe, are quite suggestive of the Mazama style, but as nothing is known of the other skeletal details of Anoglochis, any such connection must at present be purely speculative, but the element of doubt in this special case in no way disturbs the certainty of the general conclusion that all our present Cérvidae have come through distinct stages in the successive periods, from the simple types of the middle Tertiary.
The family is undoubtedly of old world origin, and for the most part belongs to the northern hemisphere, South America being the only continental area in which they are found south of the equator.
The analytical habit of mind which finds vent in the subdivision of species, is also exhibited in a tendency to break up large genera into a number of small ones, but in the present group this practice has the disadvantage of obscuring a broad distinction between the dominant types inhabiting respectively the old world and the new. The former, represented by the genus Cervus, has a brow-tine to the antlers; has the posterior portion of the nasal chamber undivided by the vertical plate of the vomer; and the upper ends only of the lateral metacarpals remain, whereas in all these particulars the typical American deer are exactly opposite. As there are objections to considering these characters as of family value, arising from the intermediate position of the circumpolar genera Alces and Rangifer, as well as the water deer and the roe, a broader meaning is given to classification by retaining the comprehensive genera Cervus and Mazama, and recognizing the subordinate divisions only as sub-genera.
The one representative of Cervus inhabiting America is the wapiti, or "elk" (C. canadensis), which is without doubt an immigrant from Asia by way of Alaska, and it may be of interest to state the grounds upon which this conclusion rests, as they afford an excellent example of the way in which such results are reached. It is an accepted truth in geographical distribution, that the portion of the earth in which the greatest number of forms differentiated from one type are to be found, is almost always the region in which that type had its origin. Now, out of about a dozen species and sub-species of wapiti and red deer to which names have been given, not less than eight are Asiatic, so that Asia, and probably its central portion, is indicated as the region in which the elaphine deer arose; in confirmation of which is the further fact that the antler characteristic of these deer seems to have originated from the same ancestral form as that which produced the sikine and rusine types, which
