Some of the Kirghiz breeds, in which the face is strongly bent downwards on the cranium, probably include this ancient Siwalik race amongst their ancestors. Some of the Eastern races which reached Europe[F15] in pre-glacial times found a congenial home in Tuscany and Umbria. Others, moving in a north-western direction, found their way into Britain, while others crossed by one or more land connections into North Africa. About the late Pleistocene descendants of the varieties and species which reached Europe before the Ice Age, a considerable amount of information has been gained from engravings and coloured drawings on the walls of caves occupied by Palæolithic man, and from fragments of skulls, teeth, and limb-bones found in Pleistocene deposits. Up to the end of last century naturalists, as a rule, assumed that the wild horses hunted during the Early Stone Age all belonged to the same species, the so-called (_E. fossilis_), and when about 1870 Prjevalsky discovered wild horses in Mongolia, it was further assumed that these wild herds were the descendants of _E. fossilis_, and hence represented the wild species from which all the modern domesticated breeds had sprung. [F15] The horses which reached Europe in Pliocene times are usually said to belong to one species (_E. stenonis_). Partly from fossil teeth and limb-bones, and partly from the engravings on the walls of caves and on pieces of horn, the conclusion was arrived at that _E. fossilis_, the assumed common ancestor of modern breeds, was characterised by a large coarse head, coarse limbs, and long-pillared molars (fig. 22). Prjevalsky’s horse when first discovered was said to be characterised by coarse limbs as well as by a large heavy head. As it was further assumed a generation ago that all domestic horses had long-pillared molars, and that the cannon-bones varied with the surroundings—being short and broad in some areas, long and narrow in others—there seemed no escape from the conclusion that all the horses now living under domestication are descended from one and the same wild Pleistocene ancestor. But though in Prjevalsky’s horse the head is coarse, the limbs are nearly as fine as in thoroughbred race-horses, and though in horses of the Prjevalsky or steppe type the pillars of the molars are long, they are not long in all modern horses. Moreover, though there is evidence of the existence in Europe in prehistoric times of a species with a coarse head and relatively fine cannon-bones, there is no evidence of the existence of a species which combined a coarse head with short broad cannon-bones. On the other hand, it has been ascertained that since Miocene times there have been living side by side in Europe big-boned and fine-boned species, and that, except by dwarfing, long narrow cannon-bones are rarely if ever transformed into short, broad cannon-bones.