932 RUINED SITES WITHIN THE OASIS OF PERSIAN SISTAN [Chap. XXIX
paring the huge volume of the Helmand in flood with the very limited flow that it carries in the late summer and autumn, a competent judge considers this time-honoured method of creating a temporary main distributary head as that best suited to the hydrographic conditions and the nature of the ground in the delta. If under efficient administration the ` duty ' of the water received from the Helmand were raised to the level attained in Egypt and parts of India, works of the Band-iSistân type might well suffice to secure irrigation for all the cultivable soil in the delta.
I have thought it desirable briefly to point out here the paramount importance of such weirs on the Helmand for agricultural life in Sistân, not merely on account of its close bearing upon the historical past of the land, but also because it helps to throw light on the name of the river itself. The ancient form of this name as preserved by the Avesta, Haêtumant, means literally ` having dams ', the word haêtu showing this meaning in a passage of the Vendidàd (xix. 3o), just like the Sanskrit setu, its phonetic equivalent.4 This designation of the river becomes fully significant if considered in relation with the important part which the annual construction of the great weir in the main bed and the maintenance of the multitude of minor canal heads must have played in Sistân ever since its river was first harnessed to support settled agricultural occupation. This also may help, perhaps, to account better for the application of the name Haêtumant to the territory of Sistân in two Avesta passages (Vd. i. 13 ; xix. 39).
With regard to the later forms of the name it is interesting to note that the modern designation Helmand corresponds to the phonetic rule of Eastern Iranian which replaces the d of Western Iranian (Persian) by 1. We have seen before that the language areas of Eastern and Western Iranian seem to have met in SIstân, as proved by the double forms Zapdyyat : Apciyycu, &c.5 The character of Sistân as a kind of linguistic watershed is curiously brought out also by the varying forms of the name as found in classical texts. By the side of the 'Ervµav8pos of Arrian, IV. vi. 6, `Erotµâv8pos of Ptolemy, VI. xvii. 17, Ethymantus of Curtius, VIII. ix. 10,6 we have the 'EpvµavOos of Polybios, XI. xxxiv. 13, and Erymandos of Pliny, N. H. VI. 25. In the last two forms we clearly recognize an intermediate stage of the phonetic process which in Eastern Iranian first turned the tenuis between vowels, as usual, into a media, hence t>d, and then caused this d through r to change into l.'
SECTION III.—THE SITE OF ZAHIDAN AND LATER RUINS TO
The physical conditions prevailing within the actually occupied and irrigated portion of the Helmand delta make it clear that ruins of any antiquity can survive there only if the ground on which they stood has since their abandonment remained unaffected by the moisture and heavy accumulation of alluvium which necessarily accompany both regular irrigation and inundation by occasional floods. Great changes in the course of the river branches constituting the northern, i. e. the actually existing portion of the Helmand delta, and in the canal system dependent upon them, have repeatedly taken place during recent times. They are abundantly attested as regards earlier times by the extensive ruined sites to be found in Afghan territory to the east of the present
4 This meaning of llaêtumant is correctly indicated in Bartholomae, Alliran. Wörterbuch, p. 1729. The rendering ` furtenreiche ', Grundriss d. iran. Philologie, ii. p. 393, reflects an earlier interpretation which did not pay adequate regard to the attested meaning of Avestic haélu and of Skr. setu.
5 See above, ii. p. 906, note I.
6 The description which Curtius gives of the Ethymantus
curving in frequent meanders and being used up for irrigation applies so closely to the Helmand that the doubt expressed by M'Crindle (Invasion of India by Alexander, p. 184) as to the identity of the river with the Erymanthos of Polybios and the Hetymandros of Arrian was not justified.
Cf. Darmesteter, Chants populaires des Afghans, p. xxv.