2 ACROSS SWAT AND DIR [Chap. I
Swat in Alexander's Indian campaign.
Alexander's operations north of Kabul R.
My journey was to take me not only to distant regions but also far back in the ages. So it was doubly appropriate that its very first stages should lead over ground so full of ancient associations as the Swat Valley (Fig. t). The name of the river Suvastu from which the main valley with the adjoining territory derives its present designation, is already found in the Rigveda and Mahabhârata,t and figures with scarcely a change also in Megasthenes' Indika as well as in Ptolemy's Geographia.2 The old Sanskrit name of the territory, Udyeina, is often mentioned in classical texts from the Mahabharata downwards. On account of the many legends about Buddha's life which popular tradition has placed in the Swat Valley, as in the neighbouring Gandhara, it figures still more frequently in the literature of Northern Buddhism. But the total absence of definite topographical or historical data, which characterizes these quasi-indigenous references, makes it needless to discuss them here in connexion with the ancient geography of the territory.
Information, scarcely more definite, is derived from the earliest foreign records of this region found in the accounts of Alexander's Indian campaign. It is true that on general geographical grounds it may be taken for certain that the Macedonian's march of conquest through the mountains north of the Kabul River, which Arrian and Curtius describe at length, must have brought his columns into Swat and the chief valleys immediately adjoining it. But the accounts of the two historians mentioned and other abstract notices surviving in classical texts, show uniform vagueness in regard to those data which might help us to follow Alexander's operations on the map or to form a clear idea of the political and economic conditions prevailing. Even in respect of those incidents like the siege of Massaga or the capture of that famous rock fortress, Aornos, which on account of their romantic interest receive frequent, and in some sources elaborate, notice, the want of exact topographical indications leaves little hope that the sites will ever be identified with any certainty.3
The extant records of this portion of Alexander's Indian campaign have been discussed so
I See Lassen, Ind. Alterthumskunde, ii.2 p. 140 ; R. V. VItt. 19. 37; Mahabharata, vi. 333. Cf. Roth, Nirukta, Erlauterungen, p. 43 ; Zimmer, Altindisches Leben, p. 18 ; Ludwig, Der Rigveda, iii. p. zoo ; Imperial Gazetteer of India, xxiii. p. x87 ; Macdonell and Keith, Vedic Index of Names and Subjects, ii. p. 460 ; V. de Saint-Martin, Étude sur la géographie d'après les hymnes védiques, p. 53.
2 Ptol., Geographia, vit. i. 26, 27, names the 2o6avror rightly as an affluent of the Rear or Kubha (*Kuvd in its Prakrit form), i. e. the Kabul River. Megasthenes (see Arrian, Indika, iv. ix) mentions the Eo'aaror as a tributary of the Kbphen (Kophes) or Kabul River and along with it, as another tributary, the rappofar. There can be no doubt that by the latter is meant the same river which in Arrian's account of Alexander's campaign north of the Kabul River valley figures under the name of Toupaîor. It was long ago rightly identified with the Panjkbra which unites from the north with the Swat River before its entry into the Peshawar Valley. See Marquart, Geschichte von Eran, ii. pp. 244 sqq., where the ancient notices concerning the drainage of the Kabul River have been examined with critical care. The Guruhd or Garuhd, mentioned in Varahamihira's Brhatsamhita (xiv. 23), and the Gauri, which the Mahabharata names along with the Suvastu, are other early designations of the Panjkbra (Kern ; Lassen, loc. cit.).
It appears very probable that the Panjkbra is meant also
by the river Choaspes which Curtius mentions in a position corresponding to that of Arrian's Guraios (vtn. x. 2 1-2). Another reference to the Choaspes is found in a passage of Strabo's Geography, xv. 26, p. 697, which states that it flows into the river Kophes ... after traversing Bandobene and Gandaritis'. The mention here made of Gandaritis, i. e. Gandhara, renders it clear that the river intended is the Panjkbra, which after its junction with the Swat River enters the Peshawar Valley and flows into the Kabul River (Kophes) near Charsadda.
The fact that the united stream is still known by the name of Swat, as apparently in antiquity, lends support to Dr. Marquart's plausible explanation of Choaspes = *hu-aspa `having good horses ', as an Iranianized rendering of the Sanskrit name Suvastu. (For other rivers bearing the Iranian name xodo 9r, Av. hvaspa, see my note ' Afghanistan in Avestic geography ' in Acadeny, May 16, 1885, p. 349.) The transformation of the river name Vitastd by a similar ' popular etymology ' into `I'Vtam r — *vida-aspa, ' gaining horses ', offers a close parallel. For ingenious attempts to explain the name Bandobene and the references to certain towns which Strabo's passage mentions in connexion with the Choaspes, see Marquart, loc. cit., pp. 245 sqq.
3 For the difficulties besetting the identification of the site of Aornos, cf. my Report on Archaeological Survey work in the N. W. F. P. and Baluchistan, 1905, pp. 19 sqq.