direct knowledge, many myths have gathered around this mountain dweller, leading, as usual in our North American game animals, to an abundance of inappropriate names. The name “goat” is objectionable, but will have to stand until some better term can be found. The Stoney Indians in Alberta use the name “Waputehk,” and in Chinook, the universal jargon of the Northwest, the goat is called Snow Mawitch (white deer). Neither of these terms are likely to become common. It is not a goat, nor even closely related to them, but is the sole representative on this continent, of a very aberrant group of so-called mountain antelopes, known to science as the Rupicaprinæ, a Subfamily of the Bovidæ.

THE MOUNTAIN ANTELOPES.

The Rupicaprinæ comprise five widely scattered genera, extending from the Pyrenees of Spain, to the Rocky Mountains of the western United States, as enumerated below.

GOAT COUNTRY ON THE SUMMIT OF THE MAIN ROCKIES

EAST OF THE MAIN ROCKIES, INDICATING CHARACTER OF COUNTRY WHERE GOAT, SHOWN ON PAGE 14, WAS SHOT.

In western Europe we find first the chamois (Rupicapra), known in the Spanish Sierras and Pyrenees as the izard, and extending eastward through the Alps and Carpathians as far as the Caucasus. Throughout all this range only one species is recognized.

The next genus of this group is the goral (Cemas), with four species ranging throughout the Himalayas and parts of China, into Amurland.

In Tibet we have the third and decidedly most aberrant member of the Rupicaprinæ, the takin (Budorcas), the horns of which suggest those of the gnu. Only one species of this genus is known.

The fourth, and to Americans perhaps the most interesting Old World member of this Subfamily, is the serow (Næmorhedus), locally known as the forest goat. This genus is perhaps, more closely allied to Oreamnos than any of the preceding genera, and its horns resemble those of the mountain goat, but are shorter and thicker. The genus Næmorhedus inhabits the Himalayas, Tibet and China with outlying representatives in Burma, Sumatra, Formosa and Japan and it is divided into numerous species. The fifth genus is Oreamnos, the subject of this article.

All the members of these genera resemble the goat in tooth structure, but differ widely from them in the position and shape of the horns, face glands and other important details. The whole group is to be regarded as an early off-shoot of the Bovidæ, to some extent intermediate between the goats and the true bovine antelopes. The Rupicaprinæ must have pushed north, with their not distant ally the musk-ox, at a very early time and become adjusted to alpine and boreal conditions. At the close of the glacial period many of its members deserted the low country and retired to high altitudes so that in some instances, notably that of the chamois, we have an example of discontinuous distribution. Its sole American representative probably reached this continent by way of the Bering Sea land connection, during the middle Pleistocene, together with the other American genera of the Bovidæ.

ROCKY MOUNTAIN GOAT

KILLED, AUGUST, 1902, BY ANDREW J. STONE IN THE SCHESLEY MOUNTAINS, BRITISH COLUMBIA.

Measurements of the animal, in detail, are given on 35

HEAD OF THE GOAT SHOWN ON THE OPPOSITE PAGE.

GENERIC CHARACTERS.

Oreamnos as remarked above, while more closely related to Næmorhedus than to the other members of the group, has departed widely in structure from all of its relatives. Its most striking character is its almost pure white coat. This coloring is in perfect harmony with an environment of snow fields, but in some parts of its range it renders the animal unnecessarily conspicuous. Until white men appeared on the scene, it made very little difference to the goat whether his enemies could see him or not, as his home was beyond the reach of pumas, wolves, and for the most part of bears and until other game became scarce, the Indians did not hunt this inaccessible peak-dweller too closely. All the types of Oreamnos are characterized by this white coat and the only exception is the well authenticated occurrence of goat in the Selkirks of southern British Columbia, with a clearly-defined dark brown line extending along the center of the back and terminating in an almost black tail. This color variation appears to be fixed in both the summer and winter pelage, as the markings were found on the skins of goats killed both in July and November. Reports of goat with these characters are widespread along the upper Columbia River, so that it would seem as though toward the southern limit of its range, a color variation were just beginning to appear. In addition to its uniformly white color, Oreamnos differs from the serow in the prominence of its eye sockets, in the elongated shape of the muzzle and face, in the position and shape of the horns and more particularly in the cannon bones, which are exceptionally short and stout. In this latter respect Oreamnos departs widely from all the other members of the Rupicaprinæ. The most striking character however, of Oreamnos, is the presence, situated in a half circle immediately behind each horn, of a large, black scent-gland, as large as half an orange. This gland is sometimes so tough as to wear deeply into the base of the horn. A horn worn away in this manner was secured by the writer in British Columbia.

The comparatively short duration of time since the appearance of Oreamnos in America and the somewhat uniform character of its habitat, probably account for the absence of much type variation.

TYPES OF OREAMNOS.

The first specimens of the mountain goat to be described, came from the Cascade Mountains on the Columbia River in Oregon and of course now stand as the type of Oreamnos montanus, having been first described by Rafinesque in 1817. This subspecies is intermediate in size between the eastern form of American goat, O. m. missoulæ, and the large Canadian O. m. columbianus, and, is characterized by a short but broad skull. The true Oreamnos montanus extends from about the Canadian boundary, south through Washington into Oregon. In the '70's a considerable number were found on Mt. Ranier in Washington, and they still occur on Mt. Baker to the northward. It is absent, however, from the Olympic Mountains, from Vancouver Island and from the southern Cascades in Oregon. Nothing is known of the northern limits of this subspecies, but it probably does not extend very far into British Columbia, merging at that point into O. m. columbianus. The most southerly Oregon records that the writer has been able to obtain is Mt. Jefferson in that State, latitude 44° 40´ north, in approximately the same latitude as the Sawtooth Mountains in Idaho.