Sec. in REMAINS OF PREHISTORIC SETTLEMENTS 953
xvii, to the NW., and the next, R.R. xviii, which continues the protective border line in the same direction, as seen on the map, both occupy the summit of mounds rising well above the level plain and showing marks of prehistoric occupation. It deserves to be noted that the triangular bronze arrow-head R.R. xviii. of (PI. CXVI) was found among the broken masonry of the small post and must hence be ascribed to the period when the latter with the rest of the border line was garrisoned. The ruined station R.R. xix, about 3 miles from R.R. xviii, was the last I succeeded in tracing at the north-western extremity of the line. Its remains, too, rise on a mound, but as the vicinity of the Hamûn here causes the slopes to be affected by salt efflorescence, but little pottery debris could be found. It deserves to be mentioned that none of the mounds or plateaus here carrying ruins of watch-stations rise quite as high above the plain as those previously described, R.R. v—xi, to the S.E. of R.R. v. I am inclined to explain this difference by the fact that the ground near the former generally shows a cover of light gravel, which had evidently retarded wind-erosion as compared with the bare clay surface farther north. It is evidently due to the same protective covering of gravel on the mounds themselves that the ruins which occupy them have not suffered more decay. In spite of their considerable antiquity, as indicated by the evidence discussed below,3 the undercutting of the exposed masonry through wind-erosion amounts nowhere to more than 4 feet, and at most points is much less.4
Along that portion of the ancient border line which was traced running to the south from R.R. xii. a, prehistoric relics proved distinctly less abundant. The explanation may possibly lie in the fact that the line here crosses a network of old outlets of the Rûd-i-biyaban, which, by their deltaic shifts and inundations, perhaps made occupation more difficult for the people of the chalcolithic period. At the ruined post R.R. )(Hi fragments of painted pottery and stone vessels were still to be found. But the sites occupied by the stations marked R.R. xiv, xx, xxi, lying on the banks of old river branches, showed no such relics. About mile to the SE. of R.R. xxi, however, abundant slags with early pottery debris lying on a gravel ridge marked a kiln. When subsequently I succeeded in picking up again a line of small ruined posts, R.R. xxii—xxv, running to the ESE. on the Afghan side of the boundary, no objects were found during our hasty search which could definitely be assigned to chalcolithic times.
There still remain to be mentioned two localities visited by me, both to the north of the border line, where chalcolithic settlements are definitely proved. About I- miles to the NW. of R.R. v a low mound, R.R. xv, bearing a small ruined structure on its summit, attracted attention. On examination the ruin proved to be of late Muhammadan times, probably representing the remains of a rest-house or the like. But the ground on which it stands is thickly strewn with pottery fragments. Most of these were of the plain unglazed chalcolithic type ; of the worked stones usually associated with this, a brief search of the surface sufficed to secure quite a representative little collection (R.R. xv. 01-22, 26, 28, 30—I, Pl. CXII). It is noteworthy that this includes, besides some well-worked arrow-heads of jasper and chert of neolithic type (03, 8, 26, 28), some very rudely chipped points and a small ribbed ` knife-blade ', of (Pl. CXII), exactly of the same type as those found in great numbers on wind-eroded ground in the Lop Desert.5 Such ` blades ' were known in the palaeolithic period, but apparently survived also into neolithic times .6 Yet side by side with relics of such antiquity and on the same surface level lay pieces of glazed pottery of modern appearance, such as the specimens with Chinese pattern, R.R. xv. 024-5. It would puzzle the
pp. 572, 66i sq.
s Cf. Serindia, i. P. 357 ; iv. Pl. XXX ; above, i. pp. 184, 206, Pl. XXII.
6 See Mr. R. A. Smith, Man, 1911, No. 52, p. 82.