974 IN THE DESERT DELTA OF SISTAN [Chap. XXX
the Hàmûn, we come at a distance of about it miles from R.R. iv upon remains of what evidently was a defensible station of importance, R. R. xvii.2 They occupy a low gravel-covered plateau and comprise a badly decayed post (Fig. 485) measuring approximately 6o feet square, marked A in the sketch-plan, Pl. 59 ; ruined quarters, B, outside its SE. corner, and a large walled enclosure of quadrangular shape, all these approximately orientated. The interior of the post was found so deeply filled with debris that the walls dividing the several rooms could be traced only with difficulty and in parts. But there could be no doubt that the general disposition was the same as in the ruins already described, except that there appear to have been four long vaulted chambers on the ground floor instead of the usual three. The outer wall had a thickness of 5 feet, and the entrance passed through its south face. The bricks were of a large size, approximating to those found in the ruined quarters outside, about 20" X I 2" X 4". That there had been an upper story is certain. The shapeless masonry still rose at the SE. corner about 14 feet above the sloping ground outside, and subsequent clearing showed that accumulated debris and sand covered the original ground level near the entrance to a height of over 9 feet.
Walled When first examining R.R. xvii on December 27, 1915, my attention was attracted by straight
enclosures lines of whitish clay showing above the greyish gravel that covered the ground to the south. Close outside post
R.R. xvrt. inspection soon revealed that they marked the top of the low mounds into which walls enclosing a large area to the south and a smaller one to the north of the post had decayed. Excavation in a few places showed that these walls, which wind-erosion together with occasional rain had caused to crumble away, had an original thickness of about 4 feet. To the south the enclosed area measured about 596 feet by 536, the northern wall passing close to the quarters, B, outside the post. The other oblong enclosure north of this had the same dimension from east to west, but measured only about 202 feet across, as seen in the sketch-plan, Pl. 59. A wall about 5 feet thick was shown by subsequent excavation to connect the wall dividing the two enclosures with the western face of the post. Together with the block of quarters traced near the SE. corner of the latter, it evidently served to form a kind of outer court, c, before the entrance of it. Within the SW. corner of the northern enclosure I traced on the surface indications of walls, and subsequent excavation proved this corner to have been occupied by a small room, D, measuring 12 by 14 feet. As an entrance to it could be clearly made out both to the north and east, it was probably meant to shelter men guarding these faces of the smaller enclosure.
Clearing The large walled enclosures outside the ruin R.R. xvii and the presence of quarters by its
made out- side seemed obvious indications that this had been a station of special importance on the defensive side post.
line. So I returned to it a month later, before my departure from Sistân, with a number of labourers for the purpose of some brief trial excavations. Within the post, A, work proved difficult owing to the masses of hard consolidated debris of masonry filling the interior. But in addition to part of the easternmost vaulted chamber, enough was cleared of the south wall to prove that the entrance lay through the middle of it. By a trench opened outside this the foot of the wall was disclosed at a depth of over 9 feet. The top layer consisted of hard clay, the debris of fallen masonry. But in the fine sand below this, which had accumulated on the lee side of the building, we struck, as I had expected, a layer of refuse outside the entrance. It could easily be distinguished by colour and smell. But apart from rotten fragments of woollen fabrics the contents had all decayed. Here and elsewhere at these watch-stations it was unfortunately only too evident that there was little hope of recovering datable records such as the Tun-huang Limes had yielded from unprotected