What particular beast-shape the Centaurs most often affected need not much concern us. The analogy, on which my interpretation of the name Pheres rests, makes certainly for some shape more terrifying than that of a horse; and the word φῆρες itself also denotes wild and savage beasts rather than domestic animals. But the horse-centaur, though it monopolised art, was not the only form of centaur known, nor, if we may judge from modern descriptions of the Callicantzari, had it so firm a hold on the popular imagination as some other types. Possibly its very existence is due only to the aesthetic taste of a horse-loving people. Pindar certainly knew of one Centaurus earlier in date and far more monstrous than the horse-centaurs which artists chose to depict, and provided a genealogy accordingly. Moreover in the passage of Hesiod which I have quoted above and which, by its agreement with the _Iliad_ as to the human character of the Centaurs, is proved to embody an early tradition, there is at least a suggestion of a more savage form assumed by the Centaurs. Several of their names in that passage[655] seem to indicate various qualities and habits which they possessed. One is called Petraeos, because the Centaurs lived in rocky caves or because they hurled rocks at their foes; another is Oureios, because they were a mountain-tribe; then there are the two sons of Peukeus, so named because the Centaurs’ weapons were pine-branches. And why is another named Arctos? Is it not because the Centaurs assumed by sorcery the form of bears? There is some probability then that the equine type of Centaur, the conventional Centaur of Greek Art, was a comparatively late development, and that the remote age which gave to the Centaurs the name of Pheres believed rather that that tribe of sorcerers were wont to transform themselves into the more monstrous and terrible shapes of bears and other wild beasts.