TRANSCRIPTION OF GEOGRAPHICAL NAMES IN CENTRAL ASIA. 657
Mongolian, Chinese, etc. Yet at Dr. Hedin's repeated request I finally felt impelled to take the work in hand, more particularly as much valuable time would inevitably have been lost had he been compelled to appeal to a philologist outside of Sweden to correct both text and maps in the minute way that was obviously required in the application of such a system as that which I had drawn up.
The sources which I had at my command for the study of the East Turkish dialects were: H. Vâmbéry, Cagataische Sprachstudien (Leipzig, 1867); SejX Sulejman Efendi, Cagataj-Osmanisches Wörterbuch bearbeitet von Ignaz Ktinos (Budapest, 1902); R. B. Shaw, A Grammar of the Language of Eastern Turkistan, in ,journal of the Asiatic Society of Bengal, Part I, vol. XLVI (Calcutta, 1877) and the same author's A sketch of the Turki Language as Spoken in Eastern Turkistan (Kàshghar and Yarkand), Part II: Vocabulary, Ibid., vol. XLVII (Calcutta, I 88o). Unfortunately I was unable to make the use that I expected of Shaw's work owing to considerable ambiguity in the sound-values that he assigns to the vowels. As to the relation between Radloff and Shaw in the matter of the passages which the former quotes in his great dictionary from the work of the latter I do not venture to express any opinion. For the Mongol dialects used in the Mongol regions visited by Dr. Hedin there do not exist any sources whatever; compare, for instance, what G. J. Ramstedt says about the Mongol dictionaries in journal de la Société Finno-Ougrienne, XXI, 2, (Helsingfors, 1903).
In revising the geographical names which Dr. Hedin had collected it was , open to me to proceed in either of two ways: either I might apply the system as above defined to the material just as it stood, without making any alterations in the forms of the words; or, when carrying out the transcription, I might at the same time reduce or »normalize» the word-forms into conformity with the written forms of East Turkish. Upon comparing the material which Dr. Hedin had printed in Pelerm. Mitteil., Ergänzhft 28, with the written language, I found that the differences were not important, and accordingly I felt justified in adopting the second of the two methods just indicated, for by this means the geographical nomenclature of these regions would gain immensely both in clearness and explicitness.
But owing to the plan of publication I was unfortunately not able to study the whole of the material at once, but had to revise the maps one after the other as they were successively finished in MS. by the cartographers, and it was only exceptionally that I had any opportunity subsequently of effecting such changes as for one reason or another seemed to be advisable, or of correcting any errors which I had made myself. As it happened, the first map of the atlas presented no very great difficulties, though even here there were many names that I failed to trace in the rather scanty lexicographical material that I had at my command; consequently these I was unable to normalize into agreement with the literary East Turkish language.
The normalizing of the place-names the meaning of which was clear consisted exclusively in the adoption of the above described alphabet, as also in the simplification of a number of consonants which in Dr. Hedins note-books and in his list of names in Peterm. Milted. were written double. This duplication of the consonant occurred partly at the end of the first syllable before another consonant (e.
IIIFカラー高解像度
