SECTION II.—THE LOCATION OF THE `TOWN OF THE DRAGON'
On the morning of February 27th I was able to secure an early start without the effort usually needed ; for the men, in spite of the toil of the night, now all realized the importance of covering more ground, if a break-down of the animals was to be avoided before we could reach water and grazing. Our march was first directed upon the Mesa L.J., which the remains of the ancient tower discovered on the preceding day had clearly marked out as a safe starting-point. About half-way to it, Afraz-gul's keen eyes lighted upon two Wu-chu coins close to our track, still preserving their legend, though worn. I noted that the very last trace of ancient vegetation disappeared soon after we had passed the previously described dry river-bed. It was clear that we had reached here the extreme eastern limit of the area to which the waters of the Kuruk-daryâ. had once carried life. There would be no ruins to guide us along the ancient route beyond the tower of L.J., which had evidently once guarded it as an advanced watch-post. There was every indication that the desert eastwards had in ancient times been as devoid as it is now of any kind of plant or animal life. As we left behind us the withered and bleached fragments of the last tamarisk trunk lying on the salt soil, I felt that we were passing from the land of the dead into ground that never knew life—except on the route to be traced.
From the top of L.J. we sighted far away a long dark Mesa bearing N. 6o° E., and as this closely agreed with the direction in which the succession of sites from L.A. onwards had been discovered by us, I decided to steer towards it. We had crossed a belt of hard salt-impregnated clay and were moving among Yardangs only four to six feet in height, separated by small trenches in which coarse sand lay over shôr. There, at a distance of less than a mile from L.J., Tursun Akhùn, one of the camelmen, marching ahead with me, suddenly called my attention to some scattered coins within five yards or so of our track. I had given strict orders ever since our start from the Lou-lan station that any objects discovered on the march were to be brought to my notice, but left undisturbed until I could pick them up myself. I was thus able to satisfy myself that Chinese copper coins by the score strewed the sandy soil along a line parallel to the direction of our march for a distance of some thirty yards.
Rapid but careful examination showed that these coins, two hundred and eleven in all, were lying in groups or small heaps over a strip of ground nowhere more than three or four feet across. The coins were all Wu-chu coins of the large inscribed type (Pl. CXIX) and, with the exception of a few which had suffered breaks, were in perfect condition. They were all of uniform size and cast, and showing neither wear nor clipping seemed as if fresh from a mint. Examination with the compass proved that the well-defined line along which they lay ran from north-east to south-west. It was clear that all these coins had dropped from a caravan moving in the very direction in which I had supposed the ancient route to lie. They must have got loose from the string which tied them and gradually dropped out unobserved through an opening in their bag or case. The swaying movement of the camel or cart in which this receptacle was probably carried sufficiently explains why the line marked by the scattered coins had the width above indicated.
Any doubt as to the character of the convoy from which this ` petty cash ' had been lost was removed when Naik Shamsuddin, on arriving with the camels, for which he acted as rearguard, and searching the ground near by, came upon a scattered heap of bronze arrow-heads behind a small Yardang, at a point about fifty yards S. 5o° W. from where the line of coins ended. The arrow-heads, all in very fair preservation, lay close together over a piece of ground about a foot and a half in diameter. Two were found sticking together through corrosion, which suggests that they had been originally closely packed together in the same bag or box. All the arrow-heads, as seen in PI. XXIII, were uniform in shape, showing a solid triangular blade. Their type exactly agreed