obtain any such article from the Tatis on either side of Mokuila. Possibly the fact of their being situated on a high road, and close to populous villages, has accelerated the exhaustion of the few relics the soil may once have contained. In the absence of conclusive evidence, such as coin finds might afford, I am unable to express any definite opinion as to the period when the settlements once existing here were abandoned. Judging, however, from the extent of the erosion, down to a depth of i 5 feet, which the ground-level has undergone at Kakshal Tati, as shown by witnesses', the antiquity of this site at least becomes very probable. The only place where I met with evidence of erosion equally deep was the Niya River Site, abandoned in the second half of the third century of our era. But as already observed, the climatic conditions under which erosion proceeds may vary at sites so widely distant, and we have no means of tracing the changes which these conditions may have undergone during past centuries. The close vicinity of the Topa-Tim Stupa, on the other side of the Sughaz-yar, also points to a pre-Muhammadan origin for the remains of Kakshal Tati.
Since we have no certain clue to the date when Kakshal Tati changed from a cultivated Cause of and thickly inhabited area into a waste, it would be useless to formulate any definite opinion
Y Yment of site.
as to the cause. The change certainly presupposes a loss of the water-supply needed for irrigation ; but this loss might have resulted just as well from some political upheaval which reduced the population or otherwise interfered with the systematic upkeep of canals, as from some alteration in the natural conditions affecting the water-supply. Yet, judging from the great quantity of flood-water which the Sughaz-yâr annually carries, as proved by its remarkably deep-cut bed and the villagers' statements, it seems difficult to believe that natural causes could ever within historical times have prevented water from the Kilian river being brought to Kakshal Tati. However this may be, it is important to note that this ruined area extends along the actual high road, and thus furnishes additional evidence for the antiquity of the line followed by the latter.
OBJECT FROM THE VICINITY OF GIJMA.
G. ooi. Bronze octagonal seal. rr diameter ; purchased at Karghalik ; said to have been found near Giima. Design—an open lotus, with a snake on each side. Design precisely like a stencil; at back a large shank pierced. See Plate L.
OBJECTS FROM KAKSHAL TATI.
G. (Yangi-arik) oox. Fragment of coarse red pottery vessel, upper portion. Well baked and very hard. The curve from shoulder to neck and outwards again to rim, is very angular and strong. The rim is thickened, slightly chamfered and moulded. Width 3g"; height 2,1e; thickness â"-8". See Plate XLI.
G. (Yangi-arik) 002. Fragment of coarse red pottery vessel, very hard. Roughly decorated with incised patterns. A series of oblique notches form a
band at junction of neck with shoulder, and below this a rough lattice pattern. Height 3"; width 3r; thickness about â".
(Yangi-arik) 003. Eight fragments of red pottery, more or less coarse, unglazed. The texture of most of the fragments is fretted and sponge-like on the surface. One piece, clearer than the others, exhibits fine colour and texture.
Two fragments of black pottery of very fine texture.
Fragment of dark chocolate-coloured substance, resembling lac.
Piece of ' slag'.
. A. (Yangi-arik) ooi. Fragment of coarse red pottery in two pieces, having portion of an animal (dog?) roughly scratched in the surface while moist, with a coarse implement. Thickness varying from g" to $". See Plate XLI.