62 . THE MAPS [Chap. III
form of the name as actually pronounced by the local inhabitants or
Spelling of Turki local guides, refraining from any attempt to spell the name in accordance names.
with its presumed correct Turki etymology. I am aware that this method of record has implied in certain cases obvious inconsistencies. 6 But it appeared the safest in the case of a geographical student like myself who, though an Orientalist familiar with Persian and the spoken language of Chinese Turkistan, does not profess to be a Turkologist. It must also be remembered that the Eastern Turki spoken in the Tarim basin and adjacent regions has not evolved a standard literary form, and that any attempt to spell their local names after the fashion adopted in modern literary productions of the Turkish-speaking population of Russian Central Asia and regions further west would very often render them unrecognizable to the people on the spot.
In recording Turki and Îranian names I have followed the system of transliteration approved by the International Congress of Orientalists and used also,
System of translitera- in its simplified Hunterian form, for Indian Government publications.
The use of diacritical marks has, however, been restricted to the indispensable minimum. Hence no distinction has been made in the ease of words derived from Arabic between the several kinds of sibilants, gutturals, etc., which are pronounced alike by the people of Chinese Turkistan.
It has seemed to me convenient to separate the component parts of Turki and Iranian names, wherever clearly recognizable as distinct words, by the uniform
Composition of Turki employment of h hens and to use these, too, to mark the connection
and Iranian names. hyphens,
the names proper of such common terms as dawcin or pass, maxâr or shrine, köl (also kol, kul) or lake, etc. I must, however, confess in this matter of hyphens and similar points to occasional inconsistencies, hoping•that they will be the more readily excused in view of the prolonged period over which both the original surveys and their cartographic elaboration have extended. Finally, I trust that the convenient brevity and significance of such Turki terms as sai (gravel plateau or stony glacis), Or (ravine in alluvial soil), ycirdang (wind-eroded ridge), and a; few others, will be held to justify their use in descriptive entries of the map, even on ground where Turki is not the local tongue.