There remains the argument from totems. Unfortunately their very existence in Europe is questioned, and this is not the place to discuss the question. It is safer not to meddle in European totems at present. Their appearance in Greek mythology, however, may fittingly here be made the subject of a brief allusion. The value, to the anthropologist, of ancient Roman customs and beliefs is that they show us the Italians at a much lower stage of civilisation than that in which the Vedas show us the Hindoos or the Homeric poems the Greeks. They show us an Aryan people having no mythology, and they warrant the inference that myths were unknown to the pro-ethnic Aryans. The Greek myths about the amours of Zeus in animal form cannot go back, therefore, to Aryan times. They may be the peculiar invention of the early Greeks, or it may be that the families which claimed to be descended from animals were pre-Hellenic, and that, when they joined the immigrating Greeks, they learnt the worship of Zeus, and were aided in their conversion by identifying Zeus with their animal ancestor. Against the instances of polyandry and the survivals of totemism, which may or may not show that the matriarchate was known to Aryan peoples, we may fairly set the evidence of comparative philology. The original Aryan language possessed terms for grandfather, father, son, and grandson; and these are just the direct ascendants and descendants who could compose a joint undivided family. There was a word for the paternal uncle, whom the children brought up in such a family would know; there is none for the maternal uncle, with whom they would not dwell. There were special designations for husband's father, husband's mother, husband's brother, husband's sister, and even for husband's brothers' wives—just the words which would be required if the wife left her own family to dwell in that of her husband. There were none for wife's father, mother, &c., which would be required if the husband became a member of his wife's family. And this—which is inconsistent with the matriarchal system—is in accord with the evidence afforded by wedding customs, viz., that the wife left father and mother, and was brought, by the _domum deductio_, to her husband's home.