National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

Sec. v)   RUINS BY THE SU-LO HO MARSHES   575

which had been pitched at a 'neighbouring spring, the character of this grand ruin remained puzzling.

It comprised three palace-like halls, with a total frontage of over 440 feet, as seen in the plan, Halls and Plate 41 ; the walls, built of stamped clay and fully six feet thick, still rose to a height of about twenty- enclosure. five feet, though badly broken in parts. The building occupied the top of a natural clay terrace, some fifteen feet high, which had been cut down steeply on all sides to serve as a base, and this added greatly to the appearance of height. There were remains of a:massive walled enclosure, with high towers at the corners as if guarding a palace court, and traces of a rampart outside this. Yet the position occupied by the whole clearly proved that this palatial structure could not have been intended as a fortified station. Its true character was not discovered until systematic exploration of the ruins became possible a month and a half later. But fortunately Hasan Akhan, my experienced head camel-man,' had carefully searched the ground at the foot of the ruin and picked up there two copper coins. They proved to be of the Wu-chic type of the Han, and thus furnished the first distinct indication as to the antiquity of this site.

Straight to the north and at a short distance extended a wide marsh, made up partly of salt- Line of

encrusted bog and partly of reed-fringed lagoons, where in ancient times, just as now, it would be towers

flanking

neither necessary nor possible to continue the line of wall. But to the north-west and north-east marsh. towers were in view, marking the line which had to be guarded. My glasses showed quite clearly that the nearest towers, T. xvii. a, xix, xx, were all built on small isolated clay ridges or Mesas, such as rose in numbers above the flat expanse of the marshy basin. Obviously the constructors of the line had been fully alive to the advantages which these commanding positions offered both for widened outlook and for safety, and had duly used them.

At the time my geographical interest was aroused even more by the striking resemblance which Rows of these clay ridges and terraces, generally ranged in rows running here from south-east to north-west Mesas in and further on from south to north, bore to the great array of Mesas which I had found in the dried-up riverinebasin. basin east of Bash-toghrak. It seemed like an exact reproduction of the aspect which that old terminal lake-bed might have borne before desiccation had removed water and vegetation from near those eroded formations. Since then my explorations of 1914 have given me an opportunity of visiting ground to the north-east of the Lou-lan Site where the surface conditions existing during the early centuries of our era, as attested by my archaeological discoveries on and around the high clay terraces of that area, must have exactly corresponded to those still observable along this part of the Tun-huang Limes.2 Another interesting illustration of physical conditions long past elsewhere was afforded by the rows of living Toghraks which closely lined the water-channels and lagoons visible from afar within the wide marsh belt. Their growth clearly betokened the presence of fresh and, at least periodically, running water. But it was not until my return six weeks later that I obtained ocular proof that the Su-lo Ho waters actually passed through and inundated this basin during the spring and summer floods. Meanwhile the view obtained from a distance sufficed to recall to my mind those lines of dead Toghraks I had crossed so often in the desert on my march to the Lou-lan Site.'

The fodder supply brought from Abdal for our ponies had by now been completely exhausted, Ruined

and this necessitated our gaining Tun-huang without any avoidable delay. So on the long march, watch-

stations on

which brought us on March to to the last halting-place with water before crossing the absolutely mesas. barren desert of gravel to the edge of the oasis, I had reluctantly to renounce all explorations off the route. This took us first for over fourteen miles through an unbroken belt of abundant jungle