Again, we occasionally find the early records of the two great branches of the Celtic race, the Gaelic and the Cymric, throwing a mutual light upon each other. There exists, for instance, an ancient Irish saga, of which several versions have come down to us, a saga well known in Irish literature under the title of the Expulsion of the Dési,[2] which, according to Zimmer--than whom there can be no better authority--was, judging from its linguistic forms, committed to writing in the eighth century. The Dési were a tribe settled in Bregia, in Meath, and the Annals[3] tell us that the great Cormac mac Art defeated them in seven battles, forcing them to emigrate and seek new homes. This composition describes their wanderings in detail. Some of the tribe we are told migrated to Munster, whilst another portion crossed the Irish Sea and settled down in that part of South Wales called Dyfed, under the leadership of one Eochaidh [Yohy], thence called "from-over-sea." There Eochaidh with his sons and grand-children lived and died, and propagated themselves to the time of the writer, who states that they were then--at the time he wrote--ruled over by one Teudor mac Regin, king of Dyfed, who was then alive, and whose pedigree is traced in fourteen generations up to the father of that Eochaidh who had led them over in Cormac mac Art's time. Taking a generation as 33 years, and starting with the year 270, about the time of the expulsion of the Dési, we find that Teudor Mac Regin should have reigned about the year 730, and the Irish saga must have been written at this time, which agrees with Zimmer's reckoning, although his computation is based on purely linguistic grounds. That school of interpreters who decry all ancient Irish history as a mixture of mythology and fiction, and who can see in Cormac mac Art only a sun-god, would probably ascribe the expulsion of the Dési and other records of a similar nature to the creative imagination of the later Irish, who, they hold, invented their genealogies as they did their history. But in this case it happens by the merest accident that we _have_ collateral evidence of these events, for in a Welsh pedigree of Ellen, mother of Owen, son of Howel Dda, preserved in a manuscript of the eleventh century, this same Teudor is mentioned, and his genealogy traced back by the Welsh scribe; the names of eleven of his ancestors, corresponding--except for inconsiderable orthographical differences--with those preserved in the ancient Irish text.