Continuing in the same direction to the south-east for about a mile and a half, we came upon the remains of a much-decayed post, consisting of the foundations of a clay-built tower, T. XLIV. c, about 16 feet square, and of a room closely adjoining it to the north. The southern wall alone of the latter survived, standing to a height of about 3 feet and built of bricks 15" x 8" x 4" in size. From a small refuse heap close by three Chinese records on wooden slips were recovered, besides a wooden board, T. XLIV. c. 01, which had served as a fire-stick of the ` female ' type ; a bronze arrow-head of the regulation type, 03, and a Huo-ch`lian coin. Another arrow-head of the same type and a Wu-chu coin were picked up on bare gravel soil to the north of the post.
A fourth watch-tower, T. XLIV. d, was sighted by me at a distance of about three miles to the south-east on the evening of April 29th. But the examination of its remains had to be left to Naik Shamsuddin on the following day, as the search for the line of the Limes called me northward. He found there a tower built of bricks, in fair preservation, measuring 16 feet square at the base and standing to the height of 14 feet. The top had carried a guard-room, of which the walls still stood in places 3 feet high. Within it lay refuse covered by tamarisk branches which had belonged to the roofing. From this refuse and from some lying at the foot of the tower, ten wooden slips with Chinese records were extracted ; at the foot of the tower, moreover, two fragmentary Chinese documents on paper were recovered. Among miscellaneous relics found there were a bronze arrow-head of a type common on the Tun-huang Limes and several pieces of finely woven grey silk. About fifty yards to the south of the tower was picked up a copper ` cash ' of K`ang-hsi (A. D. 1661-1728), which had evidently been dropped by some late visitor to this ground. The tower is clearly visible from the caravan route to Su-chou, and only the length of the march in prospect kept me, when passing by on May 1st, from visiting it in person.
[M. Maspero's examination of the wooden documents has proved that T. XLIV. d. 08-9 is exactly dated on February 24th, A. D. 64, and contains orders concerning operations at an agricultural colony. This and two fragmentary records, T. xLly. d. 05, o6, which mention orders issued to military officers, are obviously connected with the administration of the Han Limes. On the other hand, the two paper documents recovered at the foot of the watch-tower appear to have found their way there at a much later period. Of T. xLly. d. 014, a long register concerning deliveries of coal to certain persons under military regulations, M. Maspero assumes on palaeographic grounds that it cannot date back farther than Sung times. In the case of T. xLIV. d. 015, which records a legal inquiry made concerning house property left by a deceased person, the writing according to M. Maspero points to the period intervening between the Sung and Yuan or Mongol dynasties.]
There still remains to be mentioned a small tower, T. XLIV. f, measuring about 16 feet square and 8 feet high, which was passed by us about half-way between T. xcly. c and the well of Ko-tach`üan-tzt . It is built of bricks, measuring to" x 7" x 4", a size different from that prevailing at stations of the Limes,' and appeared to me of late origin. No potsherds of ancient type could be
traced near it. The fact that it lies to the west, away from the practically straight line of the ancient watch-towers T. XLIV. a—d, would alone suffice to suggest that it had no connexion with these.
It is impossible to doubt, in view of the documentary and other finds, that the four watchtowers above mentioned go back to a period contemporary with the occupation of the Limes, i. e. to Han times. [This conclusion has been fully confirmed by the evidence of such dates as the Chinese records on wood examined by Professor Maspero have proved to contain.] This chronological fact invests with special antiquarian interest the question of the character and object of this line of posts, running from SSE. to NNW., in relation to the Limes, which stretches north of it from east to west. As we did not succeed in tracing a continuation of this line either towards the
4 See Serindia; ii. p. 737.