signation Kâśi1 or Kâsî. Whence it arose history has Something of the same sort is to be seen in the fifth chapter of the Kási-khanda. Father Vivien de Saint-Martin-the genesis of whose fictitious river I trace in note 2 to p. xxviii.,-began with being disposed to make the Asi an affluent to the Varânâ, with a Varâņasî below their confluence, and the city Varâṇast therefrom denominated. Mémoires sur les Contrées Occidentales, Vol. II., p. 361. Here "il serait trèspossible que l'un de ces ruisseaux se fût nommé Asi, et qu'après sa réunion à la Varânâ, la petite rivière eût pris le nom composé de Varâṇast qu'elle aurait communiqué à la ville." This, as speculation, will pass; but, to this writer, with his bias in favour of the theological or mythopeic method of geographizing, what are, at first, only suggestions, very soon ripen into indubitable certainties: "Cette rivière [the 'Epévveσis], la dernière de la liste d'Arrien, se reconnaît sans difficulté dans la Varânast, petite rivière qui se jette dans la gauche du Gange à Bénarès, qui en a pris son nom (en sanscrit Vârânâsi)." Etude sur la Géographie Grecque et Latine de l'Inde, p. 286. This author more than inclines to see Vârânâst in the words Erarasa (or Cragausa) metropolis, foisted into the Latin translation of Ptolemy. Ibid., pp. 227, 351. Here, very much as just above, having to do with a Latin interpolation, he sets out with describing it as such, and as offering "un reste de ressemblance qu'on entrevoit encore à travers la corruption du mot;" and, a little while afterwards, as if process of time necessarily stood for an accession of facts and reasons, persuades himself that he may speak of "une ville que Ptolémée énumère sous le nom altéré d'Erarasa,” and that he finds, therein, "la trace bien reconnaissable de Vârânâsî, forme sanscrite de notre Bénarès." I have everywhere scrupulously reproduced the varieties of spelling indulged in by the writer just cited. The final 4 and the initial a of two words coalescing into a compound might, possibly, yield a; and Varaṇa and Asi would, therefore, combine into Varanasi. But this form seems to be the peculiar property of a single recent and very indifferent lexicographer; and, moreover, the name of the second stream is, correctly, Asi, not Asi. In the Kast-khanda, XXX., 18, it is the subject of a pun, in connexion with asi, "a sword." This is the oldest form, and that recognized in the Haima-kosa and by Ujjwaladatta's commentary on the Unnddi-sútra. • Kâśi is not so markedly feminine as the more usual Kâśl, its derivative. Most Indian cities have feminine appellations. long forgotten; but conjecture may, possibly, unravel its etymology." Among the descendants of Ayus' was Kâśa, whose son is noticed under the patronyms of Kâśeya," Kâsîya, and Kâśi.' The regal successors of Kâsi, and Kasika is found in the Kási-khanda, XXX., 70, and elsewhere. Compare Avantikâ for Avanti, as in note 1 to p. xxxiii., infra. The vocabularists refer the word to kds, "to shine." And herewith agrees the Kási-khanda, XXVI., 67: काशते च यतो ज्योतिस्तदनाख्येयमीश्वर । अतो नामापरं चास्तु काशीति प्रथितं विभो ॥ In the stanza immediately preceding this, the city is called Muktikshetra Krishna is speaking; and he says that the radiance of Kasi emanates from Siva. If, where they interpret Kâst by "splendid," Colonel Wilford and his numerous followers intend to take the word from the adjective kasin, they have forgotten that the feminine is not kasi, but kasint. See the Asiatic Researches, Vol. III., p. 409. Professor Wilson has already written: "It seems probable. that the city [of Kâsi] was founded, not by him [Kshattravṛiddha], but by his grandson or great-grandson, denominated Kâsa and Kâśiraja." Mr. James Prinsep's Benares Illustrated, p. 8. It is meant, here, I suppose, to hint a derivative connexion of Kâst with Kâsa or Kâsiraja. The latter name Professor Wilson everywhere puts, erroneously, for "King Kâśi." See note 7 in the present page. See the English Vishnu-purâṇa, Vol. IV., pp. 30-32. Compare Mâṇḍûkeya, from Mâṇḍûka; and Swâphalki, from Swaphalka So reads the Harivamsa, sl. 1734, in the best MSS. accessible to me. • Gana on Pânini, IV., II., 90; and the Brahma-purdna. Bhagavata-purána, IX., XVII., 4. In the Vishnu-purána, he is called Kásirája; but the term, a compound, is there to be explained "Raja Kâsi." Differently, Kásirája, Kasipati, otc., descriptive of Ajâtaśatru, Divodâsa, Pratardana, and others, signify "Raja of the Kâsis." That may be the same as f: is clear from the Mahabharata, Anuśásana-parvan, śl. 1949 and 1952. equally their subjects, were called Kâsis. Though at first a masculine appellation, Kâśi, as applied to the city so styled, is feminine.' An exact parallel to this hypothetical evolution is not far to seek. The name of King Champa, femininized, became that of the metropolis of Anga, Champâ.' The term Kâśi, denominating, if not a city, a people 1 Kâsi's successors were likewise known as Kâśyas and as Kâsikas. These terms are, all, actually employed. The last is, also, applied to persons or things pertaining to Kâśi. • Kunti, a woman, was so called from Kunti, a man. Kâst, according to the Vishnu-purdna,-see the English translalation, Vol. IV., p. 159,-was the name of the wife of Bhimasena. The reading is, however, erroneous, most probably. I find, as a variant, Kâseyi. This, like the corresponding Kâśyâ of the Mahdbharata, Adi-parvan, él. 3829, is a derivative of Kâsi. * See the English Vishņu-purdṇa, Vol. IV., p. 125. I am not unaware of the gana on Pâņini, IV., II., 82. "In the Mahabharata, frequent mention of Kâst occurs," according to Professor Wilson, as quoted in Benares Illustrated, p. 8. I should be much surprised to find Kâsî mentioned even once in the Mahábhárata. Not till medieval times, it seems, do we read of the city of Kâst. To the authority, on this behalf, of the Purâņas may be added that of an inscription which I have deciphered and published in the Journal of the Asiatic Society of Bengal, for 1862, pp. 14, 15. The document in question, a land-grant, was issued by Vinayakapâla, Raja of Mahodaya or Kanauj, about the middle of the eleventh century, it may be. Kâsì is there indirectly described as in the vishaya of Vârâṇasi, in the bhukti of Pratishthâna. For Pratishthâna, vide infra, p. xxv., note 1. It is, in my judgment, very doubtful indeed that Ptolemy's Kaooida metamorphoses Kâsi, as has been confidently asserted by Colonel Wilford and very many others. See the Asiatic Researches, Vol. III., p. 410; Vol. IX., p. 73. Fa Hian may have intended to reproduce Káśirdjya, "kingdom of the Kâsis," in his words rendered by "le royaume de Kia chi." Vide infra, p. xxviii., note 1. and its chieftains, occurs repeatedly in Sanskrit works of all but the highest antiquity.' Of Kâśi, in whatever sense of the word, we cannot, however, collect, from indigenous records, materials from which to con The expression fag actual, in the Daśa-kumdra-charita, means “Vârâņasi, a city of the Kâśis." In the subjoined verse, from the Ramayana, Uttara-kâṇḍa, XXXVIII., VI., 17, Vârâņasî is qualified by an expression meaning, the commentator says, "a city in the country of the Kâsis:" तद्भवानद्य काशेयपुरीं वाराणसों व्रज । Finally, in the Mahabharata, Ádi-parvan, śl. 4083, 4084, we read of the king of the Kâsis as dwelling in the city of Vârâṇast. The oldest among them, probably, is Pâņini, IV., II., 116; with which compare IV., II., 113. Then come the Satapatha-bráhmaṇa, the Brihad-aranyaka and Kaushitaki-brahmana Upanishads, etc., etc. In some of these works, the substantive is involved in the adjective Kâsya. This word, like Kâśika,-for which see the Mahabharata, Udyoga-parvan, l. 5907,-means, etymologically, Kásian. But commentators on old writings explain it, and rightly, to signify "king of the Kâśis." Kâśirâja and Kâśya are used of the same person in the Bhagavad-gitd, I., 5, 17. The Rigveda affords no warrant for connecting with the Kâsis any person whom it mentions. It speaks of Divodâsa, and it speaks of Pratardana; but only in later literature are they called father and son, and rulers of the Kâśis; and, where Kâtyâyana, in his Rigvedánukramanika, characterizes the latter as Kásirája, he may have expressed himself metachronically, under the influence of a modern tradition which he and his contemporaries accepted. As to the former, we find, indeed, in post-vaidik books, two Divodâsas; into whom a single personage seems to have been parted. One of them is son of Badhryaśwa, as in the Rigveda; but it is the other, the son of Bhimaratha, and father of Pratardana, that is called king of the Kâsis. It may be added, that there is no ground for considering Badhryaśwa and Bhimaratha to be two names of one and the same person. See the English Vishnu-purána, Vol. IV., pp. 33, and 145, 146. Badhryaśwa, not Bahwaśwa, is the reading of the Vishnu-purána. Correct accordingly Professor Wilson's translation of the Rigveda, Vol. III, p. 504, note 1. See, further, the Mahabharata, Anuśásana-parvan, Chapter XXX struct anything approaching a history. The kingdom of the Kâsis, and its rulers, as is evinced by the frequency of reference to them, enjoyed, from distant ages, more or less of notoriety; and this is, substantially, all that the Hindu memorials teach us. The Purâņas specify but one dynasty of Kâśi kings; a goodly catalogue, beginning, in the most authoritative of those works, with the son of Kâśa.1 To Kâśa, by a lapse of perhaps two centuries, succeeded Divodâsa, in whose reign Buddhism seems to have been still acting on the aggressive. In this synchronism there is no discernible improbability; and, with some likelihood, it embodies an historic fact. A reflexion of actual events may, likewise, be afforded in the story of the burning of Vârânasî by the discus of Vishnu. Of the age of Ajâtaśatru, as of other very early leaders of the Kâsis, none but most vague indications have, as yet, 1 A Kâśa is named in the gana on Pâņini, IV., I., 10. According to my five wretched copies of the Vâyu-purâṇa, Kâśa was followed by Kâśaya (???), Râshtra (??), Dîrghatapas, Dharma, Dhanwantari, Ketumat, Bhimaratha, Divodâsa. The Brahmanda-purana has, in one place, Kâśa and Kâsiya, as sire and son, and, a little further on, instead of them, Kâsika and Kâśeya. Kâśika, as evolving Kâśeya, must be considered as an optional elongation of Kâsi. 'See the English Vishnu-purána, Vol. IV., pp. 30-40. दिवोदास इति ख्यातो वाराणस्यधिपो - भवत् । शून्यां विवेशयामास क्षेमको नाम राचसः ॥ Then follows an account of the expulsion of Divodâsa from Vârâpast. So far as we know, he was the only king of the Kâsi family that had to do with that city. 3 * See the Vishnu-puráṇa, Book V., Chapter XXXIV, |