Sec. iii.] SYMBOLS AND LOCAL NAMES 61
numbers' with the addition of the letter a. On the third expedition different series of `camp numbers' in Arabic, large Roman and small Roman figures have been used to distinguish routes surveyed respectively by R.B. Lai Singh, Muhammad Yakab, and Afraz-gul or myself. 3
The method of lettering conforms generally to that adopted in 1909 for the 'Inter-
national Map of the World' on the 1/Million scale. Place names and
Method of lettering. other topographical information have accordingly been printed in
upright or Roman type (Doric) while Sanserif Italic lettering has been used for names of water-forms and communications.4 Hill names have been further distinguished by the use of Condensed Antique type. 6,
The wish to record in the map as much of topographical information as space would permit, together with certain limitations of available type founts, has led to the majority of village and similar local names being printed in a somewhat small Doric type (Diamond on Nonpareil), while the larger type sizes were reserved for more important names. From this rule a departure has, however, been made in some cases for the sake of better legibility. Thus along desert tracks where adequate space was available, Nonpareil type has been used for the names of certain localities, such as well-known roadside stations or camping places which, though not permanently occupied, are yet of some importance to the passing traveller. It may be noted here also that where local names are shown without any corresponding symbol, they refer to localities such as grazing grounds, small valleys, etc., closely adjoining the route line.
Throughout our common surveys I charged myself personally with the record of local
names. For routes which were followed only by my assistants I did Record of local names. my best to check and correct the record of local names they brought
back through an independent examination of natives acquainted with the ground. On some occasions, when an assistant had to be detached for prolonged survey work at a considerable distance from my own routes, I endeavoured to assure greater accuracy of record by sending with him a literate person capable of recording Turki local names more or less correctly in Arabic script. Nevertheless, I fear, it has not always been possible in the case of independent surveys by assistants to eliminate mistakes due to imperfect hearing, inadequate training in phonetic spelling, or occasional misapprehension of a language with which they could acquire but a limited colloquial familiarity.
In the case of Chinese local names in Kan-su the difficulties in the way of correct record were much greater than in that of Turkistan names. Not being
Chinese local names. a Sinologue myself I took the very necessary precaution of having them,
wherever possible, written down on the spot also in Chinese characters by my Chinese literaties. Where, as frequently happened on the third expedition, surveys had to be effected on Chinese ground without him, an endeavour was made to obtain a record of local names in Chinese writing by other means. That notwithstanding these precautions numerous mistakes are likely to have occurred, particularly as regards the names of small localities, will not surprise those familiar with the intricacies of Chinese phonetics and dialectic variations. The latter were bound to be often very puzzling to my literati whose home was distant Hunan and who, owing to the prevailing secretiveness of the Kan-su population, found it difficult at times to secure ready help from Mandarin-speaking local people. In transcribing Chinese names the Wade system as adopted in Professor Giles's Dictionary has been followed. For the transcription of all names recorded in Chinese characters I am indebted to the kind help of Dr. Lionel Giles.
With regard to Turki and Iranian names it was my aim to record in each case the