^{page 277 note 1} For these revised readings I am indebted to the kindness of Mr. Allan. He reports that in no. 1 Agōdakē is also a possible reading, and that in the last syllables of no. 2 the vowels are uncertain, but the letters seem to be -ṣayānā.

^{page 278 note 1} Of instrumental thus used as adverbs of place we find in older Skt. dahṣiṇā, madhyā, nīcā, prācā, uccā, paṤcā, and tiraṤcā (Whitney, SG., 2nd ed., § 1112 (e)).

^{page 278 note 2} With this type cf. the forms constructed from adverbs ending in -tas, e.g. tatastya. So, too, paurastya implies an earlier purastya.

^{page 278 note 3} This reference to W. Panjabi, Sindhi, and Lahndi I owe to the kindness of Professor Turner. He compares Sindhi cuḍru “lunar”, from Skt. cāndra, and writes: “The long vowel must have continued throughout the Middle Indian (Pkt.) period

^{page 279 note 1} With this use of the ablative to denote the place of issue compare its frequent use for the same purpose in charters. Possibly, however, we should read Agōdakē in the locative (above, p. 277, n. 1). In either case the formula is parallel to that of the Ṥibi coin-legends, Majhamikāya Ṥibi-jana-padasa, where Majhamikāya, denoting the city of Madhyamikā, may be taken as either loc. or abl.

^{page 279 note 2} Though abhiṣṭhāyin does not seem to be extant in literature, it is quite regularly derived from abhi-ṣṭhā-, meaning primarily “stand upon”, thence “prevail over”, “dominate”, “be supreme”, which occurs in Vēdic texts (e.g. Rv. IV, iv, 9, abhídyumnā tasthivāṃsō jánānām, “standing supreme in glories of the folk”), and sometimes in later Skt., as well as in Pali (e.g. Dīgha-nikāya xx, 20, Sahassa-Brahmalōkānaṃ Mahābrahmā 'bhitiṭṭhati, “M. is the lord of the Thousand Brahma-worlds”). For the mixture of Sanskrit and Prakrit in the same legend there are parallels elsewhere.

^{page 279 note 3} JA., tome ccxiv (1929), pp. 314 ff. Xot all of his views, however, are convincing, and some seem untenable. In this connection it may be pointed out that the verse from the Mantru-pāṭha quoted by him (Yaugandharir ēva nō rājēli Sālvīr avādiṣuḥ | vivṛtta-cakrā āsīnās tīrēṇa Yamune tava, “The Sālva women, turning wheels, sitting on thy bank, o Yamunā, have told us that their king is a Yaugandhari”) is enough to prove that the Sālvas were regarded by orthodox Brahman ritualists as being well within the pale of Brahmanism (cf. Mahābhārata VIII, xlv, 14). It is a gāthā of more or less normal type for use in the sīmantōnnahana rite (see A. Hillebrandt, Ritual-Litteratur, etc., in Bühler's Grundriss, pp. 43 f.). But it is noteworthy as showing that at one time the Sālvas were settled on the bank of the Yamunā (probably the western bank), and that one of their kings bore the title Yaugandhari, i.e. “chieftain of the Yugandharas”, who, as we know from the Vrtti on the Cāndra-vyākaraṇa II, iv, 103, were one of the six tribes forming the Sālva confederation. This use of the derivative Yaugandhari from the tribal name Yugandhara is exactly parallel to that of ōdumbari (i.e. Skt. Audtimbari) from Udumbara, which was borne by the chieftains of the Udumbaras (see Mr. Allan's Catalogue, ut cit., pp. lxxxiv, 122 ff.). The Udumbaras, like the Yugandharas, were members of the Sālva confederation.

^{page 281 note 1} I am unconvinced by the ingenious theory which sees in most of these tribal names evidence of Austro-Asiatic speech and race. But that question is here irrelevant. Some of these tribes may have been more or less non-Aryan in blood; but the names by which the Aryans designated them are definitely of Aryan type, inasmuch as they bear specifically Aryan affixes.

^{page 281 note 2} There are many other names of this type in the North, e.g. Gōsal (a Jat clan in Jind), Gondal (a Jat clan in Shahpur, Multan, and Montgomery Districts), Vātal (nomad sieve-makers in Sialkot), Dājal (a tribe and village in Mianwali District); but it is not always clear whether the -al in the last syllable has originally short a or is abbreviated from -āl < -uāl < -pāla, as has happened in the case of Haṇḍal beside Haṇḍāl (also Jat clans). I am indebted to Mr. Lakshmi Chandra Khurana for drawing my attention to these Northern appellations.

^{page 281 note 3} The study of Indian names is a field in which little work has yet been done. A. Hilka in his Beiträge zur Kenntnis der indischen Namengebung: Die altindischen Personennamen (Heft 3 of Hillebrandt's Indische Forschungen) has some useful material, but he deals only with names of persons, and even within these limits is very inadequate; for example, he takes no notice of the duplication of names in the south where Skt. names are used side by side with vernacular derivatives, e.g. Dejja- Dēvarāja, Gōjjiga-Gōvindarāja, Bijja- Vijayāditya, Sōvi-SōmēṤvara, Vikki- Vikramāditya. Biṭṭi- Viṣṇuvardhana.

^{page 281 note 4} He states distinctly that after receiving the submission of the Siboi Alexander marched against “the conterminous tribes”, and of these the first whom he mentions are the Agalasseis.

^{page 282 note 1} Dr. Smith, Vincent in his paper The Position of the Autononmous Tribes of the Panjab conquered by Alexander the Great (JRAS., 1903, pp. 685 ff.)Google Scholar, and Chapter IV of his Early History of India, locates the Ṥibis N.E. of Jhang town on the map, not making allowance for their admitted presence in Shorkot.

^{page 282 note 2} The “Agalasseis” according to Diodorus mustered 40,000 foot (so also Curtius, ix, 4), and 3,000 horse.

^{page 282 note 3} So spelt by Diodorus. Smith, Vincent and others wrongly write Agalassoi.Google Scholar

^{page 282 note 4} Another instance in which Greeks have presented Indian names with a caseinflexion is that of two kings conquered by Alexander, “Mousikanos” and “Oxykanos” (v. 1. “Porticanus”). Whatever the stems here may have been, the ending -anos seems to have been attached to them because the Greeks on hearing a phrase containing a gen. plur., let us say something like Mūṣikāṇdrn raja, “king of the Miisikas”, wrongly understood it to mean “King Mūṣikana”. If the “Abastanes” are really Ambaṣṭhas, the ending -anes may have arisen out of some Pkt. Phrase equivalent to Skt. Ambaṣṭhānāṃ jana-padam.

^{page 282 note 5} Thus in the Sutta-nipāta we find samūhatāsē (14, 369), paccayāsē (15), samaṇabrāhmaṇāsē (1079 ff.), anāsavāsē (1082 f.), and samuppilavāsō (670)

^{page 283 note 1} This becomes clearer when we ask the question: Whence did the Pali poets derive their -āsē and -āsō ? The corresponding Vēdie form -āsas disappeared from Sanskrit literature at the end of the Vēdic Age, and never reappeared. The Pali poets did not borrow -āsō and -āsē from the Vēdas, which they were not allowed to hear, nor from writers of Skt. Kāvya, who never used them. Even if by a stretch of imagination we suppose that a poet might have learned them from Vēdic tradition, would he have stuffed them as literary embellishments into verses appealing to plain men to whom they were unknown ? Obviously not: they must have come in the first instance from a live dialect or dialects together with the rest of the forms of speech and the vocabulary which characterize the early strata of this poetry.

^{page 283 note 2} It has been suggested that the ē of -āsē is due to the influence of Māgadhī. This is quite unlikely and unnecessary, for even in the third century B.C. some north-western dialects had -e < -as in the nom. sing., as is shown by the Mansehra and Kalsi Edicts of AṤoka, representing the dialects respectively of Hazara and of the Lower Himalayas in the neighbourhood of Mussoorie; and still later the -e appears in a number of Kharōṣṭhī inscriptions, which indicate that it was a feature of the dialects spoken west of the Indus (see CII., vol. ii, pt. 1, pp. cxii f.).

^{page 283 note 3} In ZDMG., 46 (1892)Google Scholar.

^{page 283 note 4} In the Bhandarkar Commemoration Volume, pp. 117 ff.

^{page 284 note 1} The name of the town of Satavali in the Southern Konkan (IA., ii, 317, 319, 322) may be left aside, as its origin is rather dubious.

^{page 285 note 1} The story of the birth of Sāta as son of a Yakṣa and a Ṛṣi's daughter (Kalhāsarit-sāgam, vi, 87 ff.)Google Scholar has preserved in a fantastic mythical setting the genuine name of the founder of the dynasty and a vague reminiscence of their claim to brahmanhood. Buddhist legends know a Sātāgira Yakkha (Sutta-nipāta, i, 9, etc.); possibly some myth concerning him may have come to be associated with the real Sāta.

^{page 285 note 2} On the composition of names with Ṥrī and siri cf. Rapson's remarks in JRAS., 1901, pp. 99 f.Google Scholar