When we go back to the early Tertiary we find a forest, with trees that shed their leaves, interspersed with glades, in which already the grasses were beginning to be developed. This state of affairs had existed but for a comparatively short time, geologically speaking. It had come only in the latter part of the preceding era. Lake and swamp, meadow and forest intermingled to make a rich and varied scene. Slowly the land toward the western side of North America lifted itself into plateau and mountain range. Slowly the westerly winds began to be cut off by the barriers thus raised across their path. As they swept over the plateau and down into the eastern plain their moisture came to be diminished. Gradually a very different state of affairs set in. The ground became harder, the forest became sparser, the plants became higher and firmer, the grasses tougher and more wiry, and, by the time the Quaternary arrived, a condition probably even drier than that of to-day existed over our western highlands. Throughout this long change, spread over millions of years, a creature which has become our horse steadily persisted and steadily advanced. Side lines developed which finally disappeared, but the main line kept on, and when the Quaternary came the horse arrived with it. Many of the skeletons in this series were known before it was realized what they were. As time went on and intermediate forms were found, it became possible to recognize these as ancestors of the horse and to assign them their proper position in the family tree. [Illustration: THE EVOLUTION OF THE HORSE'S FOOT _After H. F. Osborne and Charles R. Knight. By permission of the American Museum of Natural History._]