{4} Lord Strangford remarks on this passage:—‘Your Gomer and your Cimmerians are of course only lay figures, to be accepted in the rhetorical and subjective sense. As such I accept them, but I enter a protest against the “genuine tongue of his ancestors.” Modern Celtic tongues are to the old Celtic heard by Julius Cæsar, broadly speaking, what the modern Romanic tongues are to Cæsar’s own Latin. Welsh, in fact, is a _detritus_; a language in the category of modern French, or, to speak less roughly and with a closer approximation, of old Provençal, not in the category of Lithuanian, much less in the category of Basque. By true inductive research, based on an accurate comparison of such forms of Celtic speech, oral and recorded, as we now possess, modern philology has, in so far as was possible, succeeded in restoring certain forms of the parent speech, and in so doing has achieved not the least striking of its many triumphs; for those very forms thus restored have since been verified past all cavil by their actual discovery in the old Gaulish inscriptions recently come to light. The _phonesis_ of Welsh as it stands is modern, not primitive its grammar,—the verbs excepted,—is constructed out of the fragments of its earlier forms, and its vocabulary is strongly Romanised, two out of the six words here given being Latin of the Empire. Rightly understood, this enhances the value of modern Celtic instead of depreciating it, because it serves to rectify it. To me it is a wonder that Welsh should have retained so much of its integrity under the iron pressure of four hundred years of Roman dominion. Modern Welsh tenacity and cohesive power under English pressure is nothing compared with what that must have been.’