Page:American Historical Review, Vol. 23.djvu/182

172 the most difficult issues to be settled at the end of the war. The European nationalities are, with certain well-known exceptions, as the Belgian, Swiss, and Irish, virtually linguistic groups, and especially in eastern and southeastern Europe, where lie most of the outstanding problems, language is the only tangible and available criterion of nationality. Now, even with the fullest recognition of nationality as the basis of political independence or autonomy, it is not to be expected, not even possible, that political boundaries should coincide precisely with linguistic boundaries. But it is the first essential to know what these latter are. Yet the requisite information is scattered through countless statistical reports, local monographs, and articles in journals of diverse character, linguistic, historical, and geographical.

Mr. Dominian, who is a graduate of Robert College, Constantinople, and has the advantage of familiarity with the languages of Southeastern Europe, is conversant with this scattered literature, and has made the results available in what, so far as I know, is the only single work which combines sufficient detail with so broad a scope. Especially convenient are the many linguistic maps, and one would welcome still more of them, at the sacrifice of the profuse illustrations of scenery which have presumably been borrowed from elsewhere to adorn the book. For example, the "View of Dissentis in the section of Switzerland where Romansh is spoken" might well be replaced by a map of the Romansh speech area, and most of the illustrations are still less relevant to the discussion. A reduced reproduction of Cvijić's ethnographical (= linguistic) map of the Balkans from Petermanns Mitteilungen of March, 1913, would have been a valuable addition.

A full linguistic atlas of Europe is a desideratum, and the author has come so near to supplying it that one regrets he did not go further and include many more of the available but scattered linguistic maps of different sections. The areas of present Celtic speech are not discussed. True, they have no bearing on any present problem of nationality, not even the Irish question. But that is true also of several other boundaries which are discussed. The areas of Lithuanian and of Lettic are stated only in the most general terms. Tetzner, Die Slawen in Deutschland, a work not mentioned in the author's bibliography, contains the fullest information, with detailed maps, for the Prussian Lithuanians, Cassubians, Masurians, Wends, etc.

However, the sections of most general interest for the coming problems of reconstruction are those dealing with the areas of Polish, Bohemian (including Slovak), and the Balkan languages, and with the peoples of Asiatic Turkey. The treatment is objective and impartial. In the case of Macedonia, Servian and Greek critics will certainly accuse the author of having accepted outright the Bulgarian view, and will point out with truth that he has taken his statistics from Bulgarian sources (Brancoff, Tsanoff, Schopoff; Brailsford's Macedonia, which gives much the same conclusion, and which, despite an over-