^{page 765 note 1} I am indebted to the late Professor Kielhorn and Dr. Grierson for the reference to Patañjali, who gives as examples of a certain grammatical rule, the forms Ahichchhatrī and Kānyakubjī in the sense of a woman born at Ahiehchhatra and Kanauj respectively (Mahābhāahya, ed. Kielhorn, , vol. ii, p. 233, 1. 7).Google Scholar This use of the adjective formed from the name of the city or town is decisive proof that Kānyakubja was a well-known place in the second century B.C. Dr. Grierson has kindly examined for me the references in the epics. A list of tīrthas, or holy places, given in Mbh., iii, 8313, includes the words “At Kānyakubja Kauśika (scil. Viśvamitra) drank soma with Indra.” Böhtlingk and Roth also cite Mbh., i, 6651; iii, 11044; xiii, 216, for the form Kānyakubja as the name of a town or country.

In the Rāmāyaṇa the name Kānyakubja (v.l. Kanya-) occurs in only one passage, namely i, 3, in Schlegel's edition. The passage is wanting in the Calcutta edition, and probably is an interpolation. But chapters 32 and 33 of Book i, in the Calcutta edition, give as part of the story of Viśvamitra's ancestors, a long account of the well-known legend of the crippled (kubja) maidens (kanyā), the daughters of Kuśanābha, and this indirect reference may be understood to imply the author's knowledge of the town of Kanyakubja. Concerning the variant spelling of the name see subsequent notes. The statement made by Kalhaṇa (Rājatar., Bk. i, v. 117) that Kanyakubja was included in the extensive conquests effected by Jalauka, the son of Aśoka, cannot be relied on as good evidence of the alleged fact.

^{page 767 note 1} Bühler wrote Kānoj (Ind. Ant., vi, 181,Google Scholar etc.), presumably following a Western spelling. In Northern India the first vowel undoubtedly is short. Muhammadan authors write Ḳanauj This name is frequently confounded with that of Ḳinnauj a dependency of Multān—an error resulting in much fictitious history, which vitiates many passages in vol. i of Elliot's History, namely pp. 14, 21, 22, 23, 33, 87, 90, 91, 147, 153, 207, 208, 210, 405, ?409. Al Masūdi's detailed account (ibid., pp. 21–3) is reproduced in Bomb. Gaz. (1896), vol. i, part i, p. 518,Google Scholar as applying to Kanauj, whereas it is really concerned with Ḳinnauj. The proof is given by Raverty, (J.A.S.B., part i, vol. lxi (1892) pp. 206–8, 254;Google ScholarNotes on Afghanistan, pp. 509, 566, 571). See E. Hist. India, 2nd ed., corrigenda.Google Scholar

^{page 767 note 2} Watters, , On Yuan Chwang's Travels, i, 341.Google Scholar Kanyākubja (grant of Madanapāla, etc., Ind. Ant., xviii, 18);Google Scholar Kāanyakubja (Stein, , transl. Rājatar., Bk. iv, 237, and indexGoogle Scholar); Kanyakubja (grants of Chandradeva, etc., Ind. Ant., xviii, 13, 133,Google Scholar etc.). The Chinese form written by Watters as Ka-nao-yi or Kanoyi is spelled by Giles as Chi-jao-i, by Rémusat (Laidlay) as Ki-jao-i, and by Beal as Ki-jou-i.

^{page 767 note 3} Hiuen Tsang records the name of Kusumapura for both cities:—

Kanauj, (Beal, i, 207;Google ScholarWatters, i, 341);Google Scholar PāṬaliputra, or more accurately, an adjoining site (Beal, ii, 83, 85;Google ScholarWatters, ii, 87).Google Scholar

^{page 767 note 4} Gādhipura, (Rājatar., Bk. iv, 133);Google Scholar Gādhinagara, Gwālior Sāsbahu inscription of Mahīpāla (Ind. Ant., xv, 35).Google Scholar

^{page 768 note 1} Ep. Ind., v, 208, etc.Google Scholar

^{page 768 note 2} Ind. Ant., xv, p. 8, note 46; xviii, 13.Google Scholar

^{page 768 note 3} Bühler, (Ind. Ant., v, 181, 183).Google Scholar See also ibid., iii, 41; iv, 46. Mr. D. R. Bhandarkar agrees that KalyāṇakaṬaka denotes Kanauj itself (“The Gurjaras,” p. 15Google Scholar of reprint from J. Bo. Br. R.A.S., vol. xxi).Google Scholar

^{page 768 note 4} Hist. Nat., vi, 21.Google Scholar

^{page 768 note 5} Peheva (Pehoa) inscr., 1. 9 (Ep. Ind., i, 187).Google Scholar

^{page 769 note 1} Travels, ch. xviii, in Legge's version. The renderings of Beal (Buddhist Records of the Western World, i, p. xliii) and Giles agree substantially.Google Scholar

^{page 770 note 1} J.A.S.B., part i, vol. liii (1884), p. 156.Google Scholar

^{page 770 note 2} Ibid., p. 158.

^{page 770 note 3} Gupta coins in all metals are frequently found at Ajodhyā. Out of fifteen specimens of the scarce copper coinage in Sir A. Cunningham's cabinet ten came from Ajodhyā, and the five copper coins in the late Mr. Hooper's collection all came from the same placei Tregear's example of Kumāragupta's copper issues, at one time considered unique, also was obtained at Ajodhyā. The few specimens of Kumāragupta's copper coinage discovered in recent years all come, I think, from Ahichhatra. The evidence indicates that Ajodhyā and Ahichhatra both possessed mints for copper in the reigns of Chandragupta II and Kumāragupta I (J.R.A.S., 1889, p. 50).Google Scholar Paramārtha, a Buddhist author of the sixth century, describes Skandagupta as “king Vikramāditya of Ajodhya.” Skandagupta assumed the title Vikramāditya on certain silver coins (E. Hist. of India, 2nd ed., p. 292).Google Scholar

^{page 771 note 1} The Gaüdavaho, ed. by Paṇḍit, Shankar Pāṇḍurang, Introd., p. cxxix, note (Bombay Sanskrit Series, 1887).Google Scholar

^{page 773 note 1} Dr. Mark Collins has contributed to the discussion of the Maukhari problem in his dissertation entitled “The Geographical Data of the Raghuvaṃśa and Dasakumāracharita,” pp. 24, 49, and Table iii (Leipzig, G. Kreysing, 1907).Google Scholar He holds that Grahavarman was king of Kanauj, that the original Maukhari territory probably was Aṇga, to the east of Magadha, and that the Mālava of Bāṇa probably was the district so called near Fatehpur (pp. 24, 25, 49–54). Compare Tāranāth's “Mālava in Prayāga” (Schiefner, p. 251).Google Scholar In E. Hist. India, 2nd ed., p. 311, note 1, the words “Perhaps it was Mo-la-po” should be cancelled.Google Scholar

^{page 773 note 2} Ganjām plates dated 300 G.E. = 319–20 A.D., recording a grant by the mahārāja mahāsāmanta Mādhavarāja II, feudatory of the mahārājā dhirāja Śaśānkarāja (Ep. Ind., vi, 143).Google Scholar

^{page 775 note 1} The passages in the Chach-nāmah which Sir H. Elliot translated as referring to Kanauj in the time of Muḥammad bin Ḳāsim, early in the eighth century, really are concerned with Ḳinnauj a dependency of Multan (Elliot, , Hist., i, 153, 207, 208).Google Scholar Professor Dowson's note to p. 153 proves that he perceived the error, although he was not in a position to explain it. The territory of Ḳinnauj lay to the north-east of the kingdom of Sind, of which the capital was Alor (Raverty, , Notes on Afghanistan, pp. 509, 566, 571;Google Scholar“The Mihrān of Sind,” J.A.S.B., part i, vol. lxi (1892), pp. 207, 208, 254;Google ScholarE. Hist. of India, 2nd ed., corrigenda). For the story of the usurpation by Harsha's minister see E. Hist. of India, 2nd ed., p. 326.Google Scholar

^{page 775 note 2} Record of the Buddhist Religion, transl., Takakusu, , pp. liii, lv.Google Scholar

^{page 776 note 1} Yaśovarman = I-cha[sha]-fon [?fou-]mo, king of Central India [Madhyadeśa], who sent his minister Sang-po-ta to the Chinese court in 731 (Pauthier, quoted by Stein, , transl., Rājat., Bk. iv, v. 134 note).Google Scholar The Tch'a-fou-yuen-koei, ch. 964, p. 18 r, states that “La vingt et unième année k'ai-yuen (733), le quatrième mois, on conféra par brevet au roi de Kou-che-mi (Cachemire), Mou-to-pi (Mouktāpīḍa), le titre de roi de ce pays.” The historian proceeds to give a copy of the grant (Chavannes, , Les Turcs Occidentaux, p. 209).Google Scholar See Early Hist. of India, 2nd ed., pp. 334, 335, 343, 349, and corrigenda.Google Scholar

^{page 777 note 1} “Moi-même et le roi de l'Inde du centre, nous avons obstrué les cinq grands chemins des T'oa-po (Tibétains) et nous avons empêché leurs allées et venues; nous avons livré bataille et avons été aussitôt victorieux” (Tang-chou, ch. ccxxi, in Chavannes, , Turcs Occid., p. 167).Google Scholar The king of Central India (Madhyadeśa) referred to by the Chinese historian must have been Yaśovarman of Kanauj, who is called by the same title in another Chinese work (ante, p. 776, note).Google Scholar

^{page 777 note 2} Stein, , transl. Rājatar., Bk. iv, vv. 131–46, 366. The story of Lalitāditya, as told by Kalhaṇa, is a strange mixture of fact and romance. We do not know the Kanauj version, which might have differed materially from that of Kalhana.Google Scholar

^{page 779 note 1} Gaüḍavaho, Introd., pp. xx–xxxiii.Google Scholar

^{page 779 note 2} For the legend see Dowson, , Classical Dictionary, s. v. Hariśchandra. The name of Ayodhyā is not given in Dowson's version of the tale.Google Scholar

^{page 780 note 1} Reports, iii, 135; xv, 164.Google Scholar

^{page 781 note 1} In addition to the testimony of the Jain works analyzed by the editor of the Gaüḍavaho in his Introd., pp. cxxxv–clxi, the paṬṬāvalī of the Tapagachchha sect records the important statement that “at this time [scil. 800 Vikrama = 742–3 A.D.] BappabhaṬṬi, who converted king Āma, was born; died 1365 Vīra or Sam. 895.” The other legends show that Vākpatirāja lived at Lakshaṇāvatī, and that Yaśovarman reigned at Kanauj about the same time (Klatt, , Ind. Ant., xi, 253).Google Scholar It is not correct to affirm that the paṬṬavalī itself gives the date for Vākpatirāja and Yaśovarman. They are mentioned only in Klatt's note. But the year 800 Vikrama must fall within the limits of Yaśovarman's reign.

^{page 783 note 1} Waṇi grant (Ind. Ant., xi, 156, 160);Google Scholar Rādhanpur grant, dated 730 Śaka = 808 A.D. (Ep. Ind., vi, 240).Google Scholar

^{page 785 note 1} These mysterious coins are described and figured in my Catal. of Coins in the I.M., vol. i, pp. 91,Google Scholar note, 265, 268; by Rapson, , in Indian Coins, sec. 112, pl. iv, 22;Google Scholar and by Cunningham, in Coins of Med. India, p. 44, pl. iii, 11.Google Scholar See also his Reports, vol. ii, 159;Google Scholar iii, 138. The only specimen of which the exact provenance has been recorded is that inserted in the great Mānikyāla stūpa, but I do not think the Yaśovarman coins ever occur in the Kanauj country, or, indeed, anywhere to the south of the Panjāb. I never saw them in the United Provinces.

^{page 785 note 2} Gaüḍavaho, Introd., p. xxxix.Google Scholar

^{page 785 note 3} These limiting dates are determined by the known facts that Śaśśṅka was alive in 619, that, according to Hiuen Tsang, he died miserably when he heard of Pūrṅavarman's restoration of the Bodhi-tree (Beal, ii, 121),Google Scholar and that the pilgrim visited Bodh Gayā about 637. The narrative implies that Śaśāṅka predeceased Pūrṇavarman, who did not die until Harsha was in a position to offer Jayasena the revenues of eighty villages in Orissa (Beal, , Life of Hiuen Tsiang, p. 153). Harsha was not in such a position before 612 A.D. at the earliest, and probably not until a date considerably later.Google Scholar

^{page 786 note 1} The Jain books relate wonderful stories about Āma, king of Kanauj and Gwālior, who is described as the son of Yaśovarman (Gaüḍavaho, Introd., pp. cxxxvii, cxlv, cl).Google Scholar

^{page 786 note 2} The authority for the date 783 is the Jain Harivṃśa. The RāshṬrakūṬa and Parihār inscriptions are the principal sources of information concerning the annexation of Kanauj by the Parihārs (Gurjara-Pratihāras) of Bhīlmāl. I have discussed the subject fully in a separate essay on the Gurjaras, and given the results briefly in E. Hist. of India, 2nd ed., pp. 349, 350.Google Scholar Rājaśekhara writes: “To the capital of Vajrāyudha, the King of Pañchāla, to Kanauj” (Karpūra-mañjarī, iii, 5², ed. Konow, & Lanman, , p. 266).Google Scholar

Pañchāla, or the land of the Pañchālas, according to the Mahābhārata, as summarized by Cunningham, (Coins of Ancient India, p. 79;Google ScholarReports, xi, 11),Google Scholar was divided, after the great war, into two kingdoms, namely, Northern Pañchāla, with its capital at Ahichchhatra, and Southern Pañchāla, with its capital at Kāmpilya. The Chinese pilgrims do not mention Pañchāla as the name of a kingdom; and, no doubt, in the time of Harsha, as well as in the best days of both the Maurya and the Gupta empires, the whole of Pañchāla must have been comprised in the home provinces, and presumably administered by imperial officials. The Pañchālas are included by Varāha Mihira among the peoples of the middle country (Madhyadeśa = Āryāvarta), and the country of Pañchāla is reckoned by him as one of the nine great kingdoms (Bṛihat Saṃhita, various passages, especially xiv, 32; Ind. Ant., xxii, 186;Google ScholarCollins, , op. cit., p. 19).Google Scholar The list of nine kingdoms is repeated by AlberunI in 1030 A.D. with the remark that the names were not then in common use. Varāha Mihira wrote in the sixth century, but his lists may, and apparently do, refer to much earlier times. So far as I know, the kingdom of Pañchāla is not noticed again under that name until the beginning of. the ninth century, in the Pāla copperplates, and Rājaśekhara's allusion to it in the tenth century is the latest on record, except Alberūnī's. Nothing is known about the history of Kāmpilya (Kampil). Kanauj is situated in the Southern Pañchāla of the Mahābhārata. For Ahicchatra (Ahichhatra, Ahichchhatra, Adhichhatrā, Ahikshetra, Ahikshatra, ĀdikoṬ, ), see Cunningham, , Reports, vol. i (1871), pp. 255–65, pls. xliii, xliv;Google ScholarFührer, , Monumental Antiquities S. N.W.P. and Oudh (1891), pp. 26–9;Google ScholarProgress Rep. Archœl. S. N.W.P. for 1891–1892, pp. 1–5;Google ScholarEp. Ind., ii, 243Google Scholar (genealogy of early kings); ibid., iv, 210 (bhukti, ‘province’); Gatal. Coins in I.M., vol. i, pp. 97, 145, 184, 185.Google Scholar

^{page 788 note 1} Jain Harivṁśa (Bomb. Gaz., 1896, vol. i, part ii, p. 197 n.).Google Scholar

^{page 708 note 2} Bhāgalpur grant of Nārāyaṇapāla (Ind. Ant., xv, 304; xx, 188);Google Scholar Khālimpur plate of Dharmapāla (Ep. Ind., iv, 252).Google Scholar The position of the Bhojas in Berār has been determined by Dr. Mark Collins (Geogr. Data of the Raghuvaṁśa and Daśakumāracarita, Leipzig, G. Kreysing, 1907). The approximate positions of the other nations are fairly well known. The history of these transactions is examined more fully in my essay on the Gurjaras. Here it is dealt with only so far as it concerns Kanauj.Google Scholar

^{page 789 note 1} The leading authority is the Sāgar Tāl inscription from Gwālior, , ed. and transl. in Archœol. S. Annual Rep., 1903–1904, p. 277;Google Scholar discussed by Kielhorn, in Nachr. der K. Gesellschaft d. Wissensch. zu Göttingen, 1905.Google Scholar

^{page 790 note 1} Cambay plates (Ep. Ind., vii, 30, 43). Indra III reigned from February, 915 A.D., to about 917.Google Scholar

^{page 790 note 2} Konow, and Lanman, , Karpūra-mañjarī, p. 188.Google Scholar

^{page 791 note 1} Al 'Utbī and Alberūnī, in Elliot, vol. i.Google Scholar The name of Rājyapāla, erroneously read as Rāi Jaipāl in Al 'Utbi, has been recovered from inscriptions (Ind. Ant., xviii, 34;Google ScholarEp. Ind., ii, 233;Google Scholar see also Ep. Ind., i, 219).Google Scholar

^{page 791 note 2} Cunningham, , Reports, i, 279–93.Google Scholar

^{page 791 note 3} Rivett-Carnac, , “Archæol. Notes,” Ind. Ant., vol. viii (1879), pp. 100–104.Google Scholar

^{page 792 note 1} Pillar inscription at Belkharā, twelve miles S. E. of Chanārgaṛh (Chunar), roughly edited and translated by Cunningham, , Reports, xi, 128, pl. xxxviii. Lines 3 and 4 read. …śrīmatkanyahubjavijaya rājyt saṁvat 1253 vaisālcha sudi 11 bhaume. This record does not seem to have been properly edited by anybody.Google Scholar

^{page 792 note 2} Cunningham, , Reports, xi, 104;Google ScholarFührer, , Sharqi Archit. of Jaunpur, p. 64.Google Scholar

^{page 792 note 3} SeṬ-MaheṬ inscription, edited by Kielhorn, (Ind. Ant., xvii, 61).Google Scholar The record is dated simply in Saṁvat 1276, and Gopāla is described as Gādhipurādhipa. SeṬ-MaheṬ (SaheṬ-MaheṬ), I may note, certainly is not Śrāvastī, as Professor Kielhorn supposed it to be when writing twenty years ago. My opinion is not altered by the recent discovery of a well-preserved copper-plate inscription “in the foundations of a cell of the large monastery which occupies the south-western portion of the SaheṬ mound,” recording the donation of six villages by Gopālachandra, Rājā of Kanauj, “to the community of Buddhist friars residing in the Great Convent of Holy Jetavana” [Pioneer Mail, 15th May, 1908).Google Scholar The date is given as 1136 S., which may be a misprint for 1236 or 1286 S. The writer of the article assumes that this find is ‘conclusive proof’ of the identity of SaheṬ-MaheṬ with Śrāvastī, but I need hardly say that such a plate may have come from elsewhere. Its presence probably indicates official connexion between the SaheṬ-MaheṬ monastery and the Jetavana, but nothing more.