The non-Aryan language here referred to is the Euskarian, now spoken only by the Basques, but which seems in earlier times to have been the language of the Aquitanians and Spaniards, and may possibly have extended much further to the East.  Whether it has any connection with the Ligurian and Oscan dialects are questions upon which, of course, I do not presume to offer any opinion.  But it is important to remark that it is a language the area of which has gradually diminished without any corresponding extirpation of the people who primitively spoke it; so that the people of Spain and of Aquitaine at the present day must be largely “Euskarian” by descent in just the same sense as the Cornish men are “Celtic” by descent.

Such seem to me to be the main facts respecting the ethnology of the British islands and of Western Europe, which may be said to be fairly established.  The hypothesis by which I think (with De Belloguet and Thurnam) the facts may best be explained is this:  In very remote times Western Europe and the British islands were inhabited by the dark stock, or the Melanochroi, alone, and these Melanochroi spoke dialects allied to the Euskarian.  The Xanthochroi, spreading over the great Eurasiatic plains westward, and speaking Aryan dialects, gradually invaded the territories of the Melanochroi.  The Xanthochroi, who thus came into contact with the Western Melanochroi, spoke a Celtic language; and that Celtic language, whether Cymric or Gaelic, spread over the Melanochroi far beyond the limits of intermixture of blood, supplanting Euskarian, just as English and French, have supplanted Celtic.  Even as early as Caesar’s time, I suppose that the Euskarian was everywhere, except in Spain and in Aquitaine, replaced by Celtic, and thus the Celtic speakers were no longer of one ethnological stock, but of two.  Both in Western Europe and in England a third wave of language—­in the one case Latin, in the other Teutonic—­has spread over the same area.  In Western Europe, it has left a fragment of the primary Euskarian in one corner of the country, and a fragment of the secondary Celtic in another.  In the British islands, only outlying pools of the secondary linguistic wave remain in Wales, the Highlands, Ireland, and the Isle of Man.  If this hypothesis is a sound one, it follows that the name of Celtic is not properly applicable to the Melanochroic or dark stock of Europe.  They are merely, so to speak, secondary Celts.  The primary and aboriginal Celtic-speaking people are Xanthochroi—­the typical Gauls of the ancient writers, and the close allies by blood, customs, and language, of the Germans.

IX.

PALAEONTOLOGY AND THE DOCTRINE OF EVOLUTION.

(THE ANNIVERSARY ADDRESS TO THE GEOLOGICAL SOCIETY, FOR 1870.)