Davies in his British Mythology has hazarded the following sentence:—"The father of Sir Tristrem is here (viz. in the romance so entitled) called Rouland, which seems to be a mere French translation of his British name Tallwch, and the Irish Tuileach, a rolling or overwhelming flood."—p. 447.  Tal-lwch is the front or end of the lake, from llwch, pl. llychau, an obsolete word for a lake.  See Richards's Dict. in llwch, and Tal-y-llychau, ibid.  Owen's gloss, tallwch, the state of being spread, is quite in his style, but seems to make the surface, not the end, of the lake be its tal or front—unreasonably, if we may judge from that name of a place, Tal-y-llychau.  Armstrong, O'Reilly, and the Gaelic Society give tuil or tuile and tuilteach, a flood; but not tuileach, which is an adjective of similitude, flood-like.  Lhuyd gives dile and tuil for Irish, lyv for Cornish, and dilus for Armorican.  Rostrenen the Armorican gives diluich and deluch.  All these words (for tuil-teach is a compound, meaning a flooded habitation) are open to some suspicion of coming from diluvium.  The meanest philologist may see their want of connexion with the Welsh compound tal-lwch.  But, after all, what immense absurdity it is, to say that the participle roulant is "a mere translation" of the noun flood, because a flood rolls!  At that rate, roulant will be a translation of wheel and of ball; and all the epithets in the Gradus will be synonymous with their respective substantives.  It is almost a waste of reasoning to add—was Charlemagne's warden of the marches, appointed to repel the incursions of the Britons, likely to be a Briton himself?  Nor would such matter as this have been alluded to at all, had not the learned Chevalier Panizzi paid it the ill-merited compliment of quoting it. (The British Magazine, Rose-Maitland, v.24, 1843)