accustomed. But there was for me this novel feature in the work that the wonted operations were to be conducted at a site which had already been searched, at least partially, by an earlier European explorer. Dr. Hedin's popular narrative of his journey of 1899-1901 had, by its chapters on ` The ruins of ancient Lop-nor ' and ` Lôu-lan ', and particularly by the excellent illustrations accompanying them, familiarized me with the general features of the ruins which a lucky chance had led him to discover in March, 1900, on his first crossing of the desert from Altmish-bulak, and with the remains which a second visit, paid specially for this purpose, had allowed him to bring to light in March, 1901.1 He had made important discoveries, and though they had not yet received full expert analysis, the antiquarian evidence which they yielded was in many respects assured beyond doubt. But it was obvious that a thorough exploration of the site, or even of a portion of it, had remained beyond the range of the operations of its first discoverer. Dr. Hedin, out of a total stay of six days on his second visit, had been able to give only three to actual excavation at the eastern group of ruins, and a fourth at the western. He had the services of only five men besides himself, and not one in the whole party had previous experience of, or special training for, such work, while the ruins to be searched were numerous and widely scattered. The need of a systematic archaeological exploration of the site was thus clearly established from the first. But there remained the question how much the site thus ` researched ' would still furnish in new facts, observations, and finds.
The hope which my first rapid inspection of the eastern group of ruins (designated thereafter as L.A.) had raised was fully justified by the results of the work carried on here without intermission between December 18 and 23. In describing them, I propose to follow the chronological order in which the various structures were searched by us, and to add what observations I have to make regarding the general character of the ancient Chinese station represented by the ruins of L.A. Most of these structures had been examined by Dr. Hedin, and a number of them searched by his men either under his supervision or without it.
In Chapters XLIV, XLV of his scientific publication, Dr. Hedin has given a description of ` the ruined houses of Lôu-lan ' as he saw them, together with such measurements as he was
able to take, and a number of very instructive photographs.2 But as the survey and excavations carried out were affected by the limitations of time and labour already mentioned,
as well as by other obvious drawbacks, I have not thought it necessary to discuss the details
of his observations except where they contain evidence which was no longer obtainable on my visit. Nor have I felt it incumbent on me to examine the abundant inferences, except where
they might claim special antiquarian or geographical interest and could be supported by critically
admissible archaeological arguments. It has not been possible for me to compare in detail or otherwise utilize the valuable finds of MS. remains and other antiques brought back by Dr. Hedin
from this site, as the special section of his large work in which the late Professor A. Conrady and
Herr Himly were to have given the results of their examination of these materials has so far not been published.' In regard to them my information is restricted to the preliminary notes published
by the last-named scholar in 1902,4 and these are necessarily too brief and provisional in character to warrant detailed analysis here by the side of the abundant new materials which the site has since. furnished.
Our operations were begun at the ruin L.A. I (see Plates 23, 24), which lay nearest to the
I See Hedin, Central Asia and Tibet, ii. pp. z I z-5o ; 3 Cf. Hedin, Central Asia, ii. p. 62z.
also i. pp. 376-84. 4 See Himly, in Petermann's Millheilungen, 1902, part
! See Hedin, Central Asia, ii. pp. 621-648 with Plates xii, pp. 288-90.