also in Teutonic saga not later than the eighth century (Hildebrand and Hadubrand), and in Iranian saga not later than the tenth century (Rustem and Sohrab). Commenting upon these facts (Waifs and Strays, IV) I claimed the Irish tale of Cuchulainn and Conlaoch as the Celtic variant of a pan-Aryan incident. But M. D'Arbois goes much further than this. According to him the German version is dependent upon the Irish one, and is a result of that Teutonic and Celtic contact in central Europe which lasted throughout the fourth and third centuries B.C. Again, he maintains that the Iranian version, which, although only known to us in Firdusi's poem, is certainly ages older than Christianity, represents a younger and less perfect form of the story than does the Irish one. Sohrab, he points out, has to fight against an Amazon queen. On the part of the son this combat is meaningless, and the incident in Firdusi can only be a distorted reminiscence of the father's overcoming the Amazon who is to be the mother of the son, never to be seen again by him until the last fatal encounter. This is the form of the story in the Tochmarc Emer. If M. D'Arbois is correct the Irish tale is thrown back into prehistoric times, must indeed date as far back as any known portion of Hellenic saga. But the constancy of Emer (she refuses the chief of Munster's heir for Cuchulainn's sake) is an essential element of the story. The men and women of early Ireland were, then, on a somewhat higher level in love matters than the beasts of the field?
M. D'Arbois accords little space to the Finn or Ossianic cycle. As is well known, this has formed the subject of a revolutionary series of investigations by Prof. Zimmer. I have summarised these for English readers in The Academy of February 1891 (reprinted Waifs and Strays, IV, with additions and modifications), and can only deal briefly with the subject here. Traditional Irish history makes Finn a third-century warrior. Modern scholars have accepted this date. Some have considered the Finn saga to
