Page:Folk-lore - A Quarterly Review. Volume 27, 1916.djvu/82

 has already been identified by the late Sir E. Anwyl (Z.C. ii. 130) with the magic castle where Pryderi was imprisoned, but I do not suppose that we must necessarily identify these two places, because there were different local Otherworlds; it seems, however, to me most certain that Gweir's imprisonment was something of the same kind as Pryderi's. Can we now presuppose something of the same kind about Mabon? I think not: we know of him that he was stolen from his mother, and after this time nothing was known of him till he was found in a horrible prison. This prison was in Caer Loyw (R.B. 836), and this tale might be consequently (but not necessarily) localised in Caer Loyw. Now, according to Sir E. Anwyl (C.Z. iii. 127) Gwri Wallt Euryn was a local hero of Gloucester, and so one could only suppose that both tales (that of Mabon and that of Gwri) have risen in the same locality; but if we accept the identity of Gwri and Mabon, we have to presume that the story told about Pryderi was originally told about Gwri, and we must accept then that both the birth and the imprisonment story related originally to Gwri Wallt Euryn, and what would remain then for Pryderi? . And yet Mr. Gruffydd postulates that Pryderi was originally a son of an immortal being who took possession of him [third night]. But in this case the kidnapping of the infant would be quite otherwise represented than the rape of Mabon (and of course that of Pryderi). These theories are simply contradictory, and consequently one of them at least is wrong, and, anticipating the later arguments, I may say that the identification with Mongan's story is not only improbable, but it is, from the point of view of method, hopelessly wild. But to return to the first part of this theory: the greater part is hypothetical; neither do we