Otto Jespersen: Anm. af Grandgent, English in America. 81 appreciation, Engl. Philologie, 2d ed. p. 331), his small, but useful book on German and English Sounds (Boston 1892), and several papers on special points of American pronunciation, printed in Modern Language Notes and elsewhere. But in the treatise now before us he has collected not only the essential contents of those previous works, but also a great many new observations, so that he has given us what many of us have long been anxious to have, a tolerably full and clear account of the most important peculiarities of the American pronunciation of English. Or rather, of the American pronunciations of English, for the author does not content himself with giving an accurate analysis of his own natural pronunciation, but he has personally examined a great many Americans belonging to various parts of the United States, and finally he has sent out printed questions to institutions of learning all over the country. «Five sets of circulars were distributed; to each one there were from 150 to 200 replies, representing nearly all the region east of Mississippi and six or seven states west of that river. In every case the answers are supposed to describe the natural, unstudied speech of a cultivated American. » The first chapter deals with the sounds in detail. Here I cannot help thinking that the alphabetical! arrangement employed by the author is an unlucky one: it separates phenomena that belong closely together, it necessitates numerous cross-references, etc. Useful, nay indispensable as the alphabetical order may be in an index, it is nevertheless the most unscientific one imaginable. See, for instance, the treatment of «long u» under #, #ù, iú, and iú. Under Καὶ we find the observation that «Length and strength are almost universally pronounced with ἃ Æ between the η and the /» (a pronunciation which is by no means unknown in England); under the heading p we read «A p is frequently developed between m and a voiceless spirant, as in warmth (where it is nearly universal), something, camphor», and finally, under ἐ, we are told that «many persons develop a í between 7n and a voiceless spirant: sense — sènts, answer τε æœntsəs, thirteenth = μᾶς tintþ.» Now, would not it have been better to have given these three facts in one place? In reality, und when we look at the articulations instead of at the letters, "we have here not three distinct phenomena, but one and the same phenomenon, a want of simultaneity in the movements of the velum palati and those of the other organs. Nor is this the only instance, where a more systematic order would have facilitated a thorough understanding of the facts observed. But these are in themselves so numerous and so interesting that we ought, perhaps, to be thankful 1 The alphabet used is that of Le Maître Phonétique, slightly modified. Nord. tidsskr. f. filol. 3die række. IV. 6