36 ALEXANDER'S CAMPAIGN IN THE PANJAB [Chap. I
marched 10 miles from his camp opposite Haranpur to where he encountered Alexander, can only be due to some misapprehension.
The incidental reference made by me to the small village of Sikandarpur as situated at the apex of a triangular stretch of ground which remains on the left bank of the river owing to a north-western bend of the present river bed, has suggested to Professor Breloer that the name of this little place may have unconsciously played a 'bestimmende Rolle' in my location of the battle-field. In this respect he may rest assured that the many years spent by me in the Panjâb, where the name Sikandar is so common in personal as well as local nomenclature, would alone have sufficed to render me immune against being influenced, consciously or unconsciously, by the chance occurrence of such a modern name.
SECTION iv—THE PASS OF NANDANA AND ITS RUINS
General Cunningham already had called attention to a local tradition at Jalâlpur that the town in ancient times had extended in a west-north-west direction for miles along the hill slopes. I first took occasion to examine the ruins supporting this tradition, collectively known by the name of Girjhak, when paying a visit to the conspicuous Hindu shrine of Mangal Dév ( Fig. 9) . This occupies the top of a gently sloping shoulder of the hill chain which rises to the height of 1,833 feet north-west of Jalâlpur. A steep path, partly paved and evidently old, leads up to the temple past sharply tilted conglomerate strata eroded into fantastic shapes. Neither the extant temple nor the mosque close by, both situated about 400 feet above the plain, look old; but the terraces below them are covered with the debris of dwellings built with rough stones and completely decayed. Here were picked up some painted and incised potsherds ( see Gir. 4, 5, Pl. I) , decorated in a style closely recalling that found by me on pottery from sites on the North-West Frontier and in Northern Balûchistân, which can be tentatively dated between late prehistoric and early historical times.' A well-polished surface is found both on red and grey ware.
When I subsequently proceeded along the steep rocky slopes towards the spur descending to the hamlet of Chitti, about a mile farther to the west, I found remains of similarly built dwellings covering whatever tolerably level space was afforded by terraces or tops of ridges. The position of these ruins reminded me strongly of that occupied by the `Kâfirs' houses', so common on the hill-sides of the Peshawar and Swat valleys and dating from the Buddhist period. But here
' Cf. Stein, Archaeological Tour in Northern Balûchistân, pp. 8 sqq., 53, &c.