throws back Bron considerably, and enhances his importance in the conversion of Britain legend, thus, to my mind, confirming rather than invalidating my hypothesis of an early confusion between Bron and Bran as one of the chief factors in the genesis of the Grail legend.
One point is well brought out in Prof. Rhys's studies, the close connection between Brythonic and Gaelic romance. Many of his most suggestive parallels involve Irish or North-Scotch influence upon the Arthurian romances. This is a theory I have always steadily advocated in these pages, and I am glad to find that such a cautious scholar as Mons. Gaidoz gives it the full weight of his support in a recent number of Mélusine.
At the back of all the questions connected with the assumed mythological texts of the Welsh and Bretons lies that of the condition of Britain during and immediately after the Roman occupation. M. Loth has striven to elucidate this by an examination of the words which the Brythonic languages have borrowed from Latin. His conclusions are most interesting. He maintains that outside a line drawn between Caerleon on Usk and York, passing by Chester, Roman influence was but little felt, that the Welsh laws of the tenth century in which the social organisation is still tribal, in which legal theory is still based upon composition, represent an unbroken tradition reaching back to pre-Roman times. Only in respect of the class of villeins (taeog) do Roman ideas make themselves felt; taeog organisation is obviously based upon that applied to the Roman colonus.
Considering the totality of Latin loan-words, M. Loth shows that material life among the pre-Roman Britains was rudimentary, all terms relating to building and domestic industry being borrowed; family life was probably communistic, as the words for "husband" and "wife" are borrowed; the medium of exchange was cattle, the Breton term for which is a loan from the Latin soldus; terms relating to reading and writing are borrowed, but those
