The source of this difficulty plainly is the assumption that the original ancestor of the tribe had a family name, which was inherited by all his descendants. It is impossible to disprove or to prove this assumption. We may, however, note that the Teutons (according to Dr. Taylor[129]) rejoiced in only one name a-piece. An Athenian added to his own name his father's. And—to set assumption against assumption—we may conjecture that as patronymics are formed from personal names, so gentile names were developed out of patronymics. At first, a man's sons bore nothing in their names to indicate from what father they were sprung. In course of time the sons of Anchises were known as Anchisiadæ; and as long as the family group consisted only of parents and children, this system of nomenclature would suffice. It might even continue into times when the family group included three generations: Iulus, as well as his father, Æneas, might be an Anchisiades. And here we may note that if all the members of a joint undivided family bore the surname Anchisiades, an aversion to marriage in the near kin would forbid the marriage of any two Anchisiadæ. When, however, owing to natural growth, the joint undivided family of Anchises becomes so large that it is necessary for his younger (married) sons to go out into the world and start joint undivided families of their own, leaving Æneas and his children in possession of the old home, it is obvious that persons who once had belonged to the same joint undivided family, and therefore had possessed the same family name, and had been prohibited to intermarry, would now belong to different families, and (being named after the respective house-fathers of the newly formed families) would have different patronymics, and would be allowed to marry persons whom previously they were forbidden to wed. In these circumstances an extension both of prohibited degrees and of the family name might very naturally be the ultimate result. Iulus, who for years had worshipped Anchises as house-spirit, and had consequently been an Anchisiades, might, when Æneas became his house-spirit, come to be known as an Ænæades, but on the other hand the old patronymic might stick to him and to his children for ever. In the same way, the aversion to marrying women who belonged to the same joint undivided family might cease when they ceased to belong to the same family, but it might continue. Hence a continual tendency to extend the family name, and to enlarge the number of prohibited degrees.