wall was traceable, nor was any such protection needed on the east and south, where the rock face falls away very steeply. The site is known as Kino-kôt, ` the Black Fort '. Plentiful debris of pottery and remains from kitchen middens indicated prolonged occupation, but local tradition was silent as to the date of its cessation. The masonry, rough but solid, showed a marked difference from that observed in the huts of the present inhabitants of this tract.
About half-way down the precipitous southern slope I noticed on a large boulder shallow incised marks, evidently intended to represent that ancient Indian emblem of ` Pddukds', so common in regions where Buddhism once prevailed. One of the two pairs of sacred footprints here outlined measured about 14 by 10 inches. Their pre-Muhammadan origin was attested by similar rock-markings of ` Chakras ', obviously rough representations of the Buddhist Wheel of the Law. Whether the ` figure of a woman ' said to be engraved on a rock high above the left bank of the Indus about half-way between Childs fort and Hôdar might also be of Buddhist origin, I regret not to have been able to ascertain, since I heard of it only in the course of my rapid descent of the river when a landing was declared impossible.
From the height of Kino-kôt I had a full view of the low sandy plateau extending to the west of the mouth of the Hôdar valley and bearing the name of Damôdas. It is known to have been occupied by the houses and cultivated plots of the largest Hôdar settlement until the big Indus flood of 1841 swept away all irrigation terraces and left this extensive area smothered under rubble and coarse sand.
We halted for the night at the hamlet of Balugush, hidden away under fruit trees, and found there coolness as welcome as unexpected at this comparatively low elevation of about 3,630 feet above sea-level. The track by which next morning we ascended the narrow valley between absolutely bare rock faces was impracticable for baggage animals, and men had thenceforward to be employed as bearers until we reached Yasin. All marching had, of course, also to be done on foot. Above the hamlet of Dar, a mile from Balugush, a projecting rocky spur (Fig. 6), difficult of access, was found to carry the remains of an old village fort resembling Kino-kat but very much smaller. Here the more massive and careful masonry of the walls, as compared with that of the modern dwellings near by, was distinctly noticeable. From Dar for about three miles a narrow but continuous strip of cultivation extended up the valley, as shown by the detailed survey carried out by R. B. Lal Singh from the point where we crossed the Indus. But farther up, except at one point, Hamdchéch, where the eye was gladdened by luxuriant groves of fruit trees and a few fields, the bottom of the valley was either a barren winding gorge or showed only traces of abandoned cultivation.
It was difficult to resist the conclusion that since this abandonment the water-supply available for irrigation must here have undergone considerable diminution ; for when after covering about twelve miles we left the main valley at an elevation of about 6,000 feet to ascend the Pakôra Nullah north-westwards, we found the well-marked stream bed in the latter quite dry, except for short stretches where the water supplied by springs farther up comes to the surface again. Conditions of obvious ` desiccation ', and that actually in progress, revealed themselves on reaching the little hamlet of Pakôra, at about 7,200 feet. The spring that supplies the area now under cultivation was said no longer to yield water sufficient to reach the terraced fields visible for some distance below. Its supply, moreover, had to be dammed up daily to form a small reservoir in order to reach the present cultivation. This instance of undoubtedly recent ` desiccation ' appears to deserve special notice just because it contrasts with the conditions observed, as we shall presently see, in the well-watered valleys of Darél and Tangir, which lie but a short distance westwards. I may add that the ` desiccation ' here noted, which may or may not be local, cannot be ascribed