CHAPTEE I
HISTORY OP THE SURVEYS
 Section T.-GENERAL CHARACTER OF THE TOPOGRAPHICAL WORK
The main facts concerning the surveys effected on my three Ceutral-Asfan expeditions and many of the more notable incidents and results which attended them have been already recorded in the '' Personal Narratives ' and ' .Detailed Reports ' published by me of the first two journeys1 or, as - regards the third, in the fairly comprehensive preliminary account printed in the Royal Geographical Society's Journal.s I shall accordingly restrict myself here to notes on the character and range of the surveys made of each journey; on the assistance available for the topographical work of each, and on the routes along which this work was carried,* distinguishing the routes upon which the surveys were effected under my personal supervision from others where the operations were conducted by my assistants or myself alone.
Before however recording these details for each successive expedition it will be convenient here to indicate essential points regarding the methods which Methods in surveys. wore uniformly observed in all our surveys. As already stated above, these methods were the same as those employed by the Survey of India for 'reconnaissance survey'work. They implied continuous plane-tabling along all routes followed, supplemented throughout by astronomical observations for latitude anct by triangulation rendered possible in particular areas by proximity to previously fixed points, by the configuration of the ground, available time and similar considerations.
Except on the journey of 1900-01 when a scale of 8 miles to .1 inch was used, the surveys were on the scale of 4 miles to 1 inch, this having proved by 'Siealtable experience on the ground to he the most convenient for adequate record
of topographical detail under our conditions of travel. On mountainous ground no efforts were spared to place plane-table stations on commanding heights above passes and route lines, maximum elevations of nearly 20,000 feet being climbed by us in the K'un-lun for this purpose and of over 16,000 feet at numerous points of the Pamir and Nan-sham ranges. In the great plains of the Tarim basin and in the similar drainageless deserts eastwards, the fatness of the ground, the absence of recognizable landmarks and the peculiar dust-laden condition of the atmosphere, persisting for prolonged periods, made it very often impossible to fix positions by intersections or triangles. On this ground the exact measurement of distances which the use of the cyclometer invariably carried on the second and third expeditions permitted, was essential for the plane-table traverses.
For the purpose of securing points to check these traverses, astronomical observations for latitude were made by my assistants with a 6-inch transit theodo- Latitude observations, lite at frequent intervals when atmospheric conditions and available time permitted. These observations were beset with serious difficulties, both on account of climatic conditions and the rapidity of movement necessitated by other scientific tasks and the wide extent of difficult ground to be covered. Prolonged periods of