These two were the only described forms, until 1904, when the attention of Dr. J. A. Allen, of the American Museum of Natural History, was called by the writer to the great difference in bulk of body and size of horns of the goat of British Columbia, and those of the Bitter Root Mountains in Montana. Upon comparing a number of specimens from the Cascade Mountains, the type locality of _Oreamnos montanus_, from the Bitter Root Mountains of Montana and Idaho, from the main Rockies in southern British Columbia and from the Schesley Mountains of northern British Columbia, it was found that all these specimens could be divided into three easily distinguishable groups each of subspecific rank. [Illustration: ROCKY MOUNTAIN GOAT KILLED IN SPILLAMACHENE VALLEY, SOUTH OF GOLDEN, BRITISH COLUMBIA, NOVEMBER, 1903 Total length with tail, following convolutions of body, 73 inches; tail, 7 inches; hind foot, 12 inches; height at shoulders, 41 inches; measurements taken after mounting. On exhibition in the American Museum of Natural History.] The skulls of animals killed in the Schesley Mountains by Andrew J. Stone in 1903, were found to be in all respects identical with those killed by the writer and Mr. Charles Arthur Moore, Jr., in the main Rockies, near the Columbia River the following year. Animals from these districts were characterized by great bulk and by a long and relatively narrow skull. This was the third type described and it received from Dr. Allen the name of _O. m. columbianus_. This subspecies probably extends from the American border up through the Canadian Rockies, to the northern limits of goat in that region, which is west of the Mackenzie River at about north latitude 63° 30'. The goat in the northern Rockies, may possibly be found to be specifically distinct from the goat on the coast of southern Alaska. [Illustration: SIDE VIEW OF SPECIMEN SHOWN ON OPPOSITE PAGE.] In the midst of the distributional area of this large subspecies and in the vicinity of the Big Bend of the Columbia River, a very small goat is found. This animal, upon further investigation, may prove interesting. At present, however, all the Canadian goats must be provisionally assigned to _O. m. columbianus_. A curious break in the range of this subspecies is found just north of the Liard River, where, according to no less an authority than Andrew J. Stone, no goat are found for a distance of over a hundred miles. Probably the local topography, of which we have no knowledge, will explain the absence of goat from this territory. No goat have yet been found north of the Yukon River.