[Sidenote: Pytheas and the ethnology of Britain.]
But, apart from the deeds of Pytheas himself, perhaps the most interesting information which we owe to the fragmentary record of his voyage relates to the ethnology of Britain. He learned that it was called the Pretanic Island. Before his time the Gauls for the most part had come to change the original sound _qu_ into _p_; whereas certain tribes of Western Gaul[961] as well as all those Celtic-speaking inhabitants of the British Isles from whose dialect Gaelic, Irish, and Manx have been evolved retained it, though the latter afterwards modified it into _c_. On the other hand, wherever the Indo-European tongue from which Celtic was an offshoot had the sound of _p_, most of the Celtic-speaking tribes both of Britain and Gaul had let it disappear. The word _Pretanic_ therefore implied the existence of an earlier word _Qrtanic_; and supposing that Pytheas, as some believe, heard _Pretanic_ only in Gaul, it might be argued that _Qrtanic_ was still the British pronunciation. If so, none of the tribes who had changed _qu_ into _p_, from whose dialect Welsh, Cornish, and Breton descended, and who are commonly called Brythons, had yet invaded Britain. But if, as seems much more probable, Pytheas derived his information from Britons, the Brythons were already predominant at all events in those parts of Britain in which he conversed with them. Indeed, as we shall afterwards see,[962] it is morally certain that Brythonic tribes had been settled here at least half a century before he came.
The subject of the ethnology of the Celtic-speaking tribes of Britain is extremely difficult; and on nearly every important point Celtic philologists differ widely among themselves. It is almost an article of faith that the earlier Celtic invaders were Goidels, or tribes who had not changed _qu_ into _p_; but there are some who maintain that neither in the time of Pytheas nor even in that of Caesar were there any Goidels in Britain; and that those who were settled in Wales in the third century of our era were all of Irish origin. No direct evidence indeed can be adduced for the common view; but it is hard to conceive that the earliest Celtic immigrants, unless they set out from Spain or from North-Western Gaul, should have passed by Britain in order to settle in Ireland. Even those who admit the priority of the Goidels in Britain are not of one mind. While the foremost Celtic scholar of this country maintains that when Celts first reached Britain the distinction between the Goidelic and Brythonic dialects already existed, the foremost Celtic scholar of France insists that at that time the Celtic language was everywhere the same: according to him none of the Celts had then changed _qu_ into _p_: that change was made later by Celtic conquerors of Gaul, some of whose descendants afterwards colonized Britain; and the people with whom Pytheas conversed were not, strictly speaking, Goidels, but simply Celts who spoke a language from which the Goidelic dialects--Gaelic, Manx, and Irish--were subsequently evolved.