96 EXPLORATIONS IN PERSIAN MAKRAN [Chap. III
QaPa-kôh, occupies the top of a steep rocky hillock rising about 250 feet above the area of palm-girt fields to the south of the fortified residence of the local chief, Sirdar Jan Muhammad Khan ( Fig. 29) . The narrow crest of the hillock, only about 12 yards across, bears remains of roughly built stone walls for a length of about 60 yards. The scantiness of potsherds suggests that the small circumvallation was occupied only occasionally as a place of refuge.
The place of another small fort, called Qalatuk, and situated about a mile to the north-east of the main village, is marked by a mound some 20 yards in diameter and about 15 feet in height, covered with stone debris. Among the coarse- pottery found on and around the mound, the rare painted fragments showed patterns of the simplest kind painted dark red on a light buff or black ground. Proceeding a short distance farther in the same direction, beyond the eastern edge of the hamlet of Bôghan, there was found a stretch of waste ground about a quarter of a mile across covered with debris from rough stone walls and coarse potsherds of the same type. On slightly raised ground I found remains of Muhammadan graves probably contemporary with the occupation of this area. It appears to have been irrigated from a small canal, of which traces could be followed to where the river emerges from the narrow valley leading down from Champ.
The most direct route towards Bampar would have led us up this valley, but there were two reasons which decided me against following it. On the one hand there was the risk of being held up in it by floods if rain, not unknown at this season, fell and rendered the river impassable for days in any of its numerous defiles. On the other hand, I wished to make sure by a visit to Geh that no early remains in this, the only other oasis of some size to the west before reaching the Bashakard territory, should escape me. Our journey resumed on February 10th towards Geh took us on the first march for some 15 miles down by the Kaj a river. At the hamlet of Hait, which was first passed at a bend of the river, ample irrigation is provided by a kccréz said to be very old and to have its head some 4 miles away at the Damba-gah site near Qasrgand. Then crossing to the left bank of the wide river-bed I found above the village of Bag an extensive stretch of ground known as Damba guh, covered with the debris of roughly built stone walls and abundant pottery. Most of the latter was plain coarse ware. Medieval occupation was proved by glazed potsherds, including some with graffito design under green glaze, and others with brown scroll patterns over yellow ground, also a fragment of Chinese blue and white porcelain. A fragment of lustred ware is ascribed by Mr. Hobson to a type of the tenth to twelfth century. The area stretching for about half a mile from east to west lies well above the level of the ground now capable of irrigation from the river. More cultivated ground and