Sec. iii] NANDANA AND THE CROSSING BELOW JALALPUR 29
But from the middle of April until August it carries a large volume of water, inundating both banks, and is then quite unfordable, a ferry being regularly maintained at the mouth of the Kandar Kas. For the preceding three years the current had at the head of this river branch set increasingly towards Jalâlpur (see Skeleton Map 1), apparently as a result of exceptional floods having come down from Kashmir. The extent of the inundation caused was marked by large trunks of timber we found left on the opposite bank for a distance of about 550 yards before reaching the cultivated portion of the island. The island is occupied by the small hamlet of Admâna and scattered groups of homesteads reckoned with it. The fact of the whole of the cultivated area of Admâna being included within the boundary of the Gujrât District to the south suggests that at a time not very distant the main branch of the river lay approximately along the Halkiwâni bed.
The length of the island is shown by the one-inch map of the Survey, No. 43. H. 6 ( 1911) as slightly over 6 miles in a direct line. Our rapid plane-table survey of the island indicated practically the same length, a reduction of about half a mile at the lower end of the island where the Halkiwâni rejoins a southern channel of the river being compensated by a corresponding accretion at the upper. The maximum width of the island on a line passing Admâna hamlet is close on 1j- miles, and has remained unchanged since 1911. High tamarisk and other bushes cover a great deal of the ground left uncultivated, and in a `Reserved Forest' area, known as Mâjhi Rakh, a thick wood of fine trees has grown up.
It is certain that the island of Admâna as it exists now is by far the largest of any which the Survey maps show in the Jhelum within the whole length of its course that can come into consideration here. This point deserves to be specially noticed. Though islands in a river like the Jhelum are liable to changes, yet the general course of the river and the character of its bed just along this section are not likely to have changed greatly during the last two thousand years or so; for they are here largely determined by two permanent geographical features, the Jalâlpur-Dilâwar hill spur on the one side and the high ground at the end of the Pabbi range facing it on the other.
Curtius tells us that the island which masked Alexander's crossing was `larger than the rest, wooded and suitable for concealing an ambuscade'." If we then assume that in Alexander's time there stretched from below Jalâlpur an island much of the same size and type as the present island of Admâna, it is easy to follow the successive phases of the crossing as recorded by Arrian. Moving down a channel approximately corresponding to the present Halkiwâni, the
11 Cf. above, p. 12.