^{page 519 note 1} Arch. Rep., ii, 34.

^{page 520 note 1} Lassen, i, 657.

^{page 520 note 2} Wilson, V.P., 348 ff.

^{page 522 note 1} But not by so many generations as is sometimes supposed. For example, it is said that Bahukā or Bahu, the Solar prince who in attacked by the Tālajangha and Haihaya, is the thirty-fifth in descent from Ikshwāku, while Tālajangha is only eighteenth in the Yadu list; but the V.P. does not make them contemporary; the text speaks of a conflict between the tribes, i.e. several generations perhaps after the eponymous ancestors. One authority, indeed, makes Triśankhu and Tālajangha contemporary, but Bahukā is eighth in descent from Triśankhu.

^{page 523 note 1} As, indeed, is the way in which some statements regarding the supernatural origin of Krishna are grafted on to the narrative of his purely human family and exploits. It almost looks as if some later hand had made a very rude mosaic by letting pieces into another narrative (see especially the passage in Wilson's Ed., pp. 439–40). As Krishna's human life and exploits could not be got rid of, it seems as if an attempt was made to attach his family to the Signified Solar Line rather than to the outlandish Yadava, but feebly and tentatively by this one suggestion; it could not really be done.

^{page 523 note 2} “Tree and Serpent Worship,” p. 59.

^{page 523 note 3} On the supposition that the v.P. is in error in making Viśāla a late descendant.

^{page 524 note 1} I take the list as in Lassen, I, App., pp. xiii, xiv.

^{page 524 note 2} This is discussed further on. The tribe was certainly not an early one in the locality, but is placed there in the V.P.

^{page 525 note 1} This is well put in the Oudh Gazetteer, Introduction, xxxiii, xxxiv.

^{page 525 note 2} The Vāyu-P. makes Lava reign at S'ravasti. In the V.P. there is some apparent attempt to make Kuśa in some way connected with the Kosala beyond the Vindhyan, and to make him found a city on the brow of the Vindhyan hills.

^{page 525 note 3} It seems that these may either be subject tribes which regained independence by the destruction of Aryan overlords, or some new invasion of Indo-Scythians. There is a mention of a Maruṇḍa tribe in Samudragupta's inscription, but apparently it is on the North-West frontier.

^{page 526 note 1} As to the statement that the ‘Bhar’ (as a whole) are Kolarian, I cannot find that it is supported by any general evidence. It cannot be inferred from the connection of the (very uncertainly spelt) name of the Bhar people with ‘Baṛ'— the Indian fig-tree—because that derivation is quite arbitrary; and it is very doubtful if, the tribal name being really Bharata or Bhārata, that name could be derived from Baṛ (Skt. Vaṭa). It is true that Mr. Crookea says of some of the Bhar tribes (Crookes, ii, 9, 10) that they “propitiate evil spirits in the old figtrees in the village” and worship ‘five ancestors’—pāṅçonpīr. The existing Bhar are apparently a very mixed race, and have no distinct language; the presence of some Kolarian customs would not necessarily indicate that the whole race was Kolarian. On the contrary, if they are the remains of a larger race coming from the North-West and contemporary with the Aryans, their origin would more probably be Turanian.

^{page 527 note 1} Beames' Elliot, Gloss., i, 115 ff. They do not appear in history before the Muhammadan conquest of Kanauj, and only near Fatihpur on the Ganges.

^{page 527 note 2} Kalpasutra, (Sacred Books of the East, vol. xxii, p. 266)Google Scholar. Mr. Hewitt also refers to the Sumaṅgala Vilāsana, in the Pāli Text Society's edition, regarding the S'akya. This I have not seen.

^{page 527 note 3} Like some other tribes of the period, the Liççhavi are said to have been at first non-monarchical; when they accepted kingly rule, Jayadeva I originated the Nepāl era.

^{page 528 note 1} See Lasaen, i, 688, and the references in the note.

^{page 528 note 2} See Wilson, V.P., p. 373.

^{page 528 note 3} The whole case is stated in Crookes, (Tribes of the N.W.P.), vol. iv, 86 ff.Google Scholar

^{page 528 note 4} Crookes, vol. iv, p. 86.

^{page 529 note 1} Tod's statement is in vol. i, 177 ff. (reprint).

^{page 530 note 1} From some Mewār records Tod obtained the date ‘Samvat’ 201 for Kankasena, and assumed (the impossible) application of the Vikrama era; and so later writers speak of the date a.d. 144–5 as an ascertained fact! I do not know what is the authority (if any) for the Samvat figures; but assuming them, the earliest (if it is a S'akā date) would be circâ279 a.d. for Kankasena; but if the Valabhi era is meant (which is possible) it would be circâ a.d. 520. As several generations passed away before Kankasena's posterity founded Valabhipura, and the number is not given, it is impossible to compare with the actual probable date of the building of the city.

^{page 530 note 2} There are some instructive remarks about the word Bappa in Bom. Gat., vol. i, pt. 1, pp. 83 ff., where also a Gurjara origin is suggested. This does not commend itself to me in any way. The Bāla (Malli ?) were in India long before the Yu-chi, Tuṣara, or Gurjara, though very possibly of ultimately the same ethnic stock.

^{page 530 note 3} See Arch. Rep., ii, 34. There are still (ancient) ‘Vāla’ estates of Dhñk in Gujarāt, and Chamārdi in Kāthiāwār. See Bom. Gaz., vol. viii, p. 129, as to the Vāla and their admixture with the Kāthi (Kathaei of the Greeks).

^{page 531 note 1} “Which only arises from careless spelling of ‘Cushwāha,’ ‘Kusha,'etc.

^{page 531 note 2} A peculiarly ‘Solar’ institution, as the horae was sacred to the Sun. Is not this sacrifice (as Tod has illustrated) really of non-Brahmanic, Northern origin ?

^{page 532 note 1} “Ayodhya,” writes the author of the Oudh Gazetteer (Introd., cap. iii), “its eponymous city, was the capital of that happy kingdom in which all that the Hindu reveres or desires was realized as it never can be realized again, and the seat of that glorious dynasty which began with the Sun and culminated, after sixty generations, in the incarnate deity and perfect man Rāma.”

^{page 532 note 1} Just as the Krishna cult seems to have been encouraged, if not invented, by the Vaiṣṇava sect, as a rival to the S'aivaite worship; though naturally the populace accepted both together.

^{page 533 note 2} I have given instances in Ind. Vill. Comm., pp. 99, 277.

^{page 534 note 1} Many elans have long become Mussulmān, and then often claim descent from some Moslem source.

^{page 536 note 1} On this subject I may refer to Cunningham's, remarks in the Arch. Rep., vol. xxGoogle Scholar, quoted by Crookes (iv, 219).

^{page 537 note 1} In Lyall's Kangra Sett. Report, the fanciful legend of origin is given; showing that they have (naturally) forgotten any real ancestry, and believe themselves older than eitherSolar or Lunar! But it is quite possible that they may really be a relic of the Yādava king of Trigartta, preserved in that quiet retreat.

^{page 538 note 1} The ‘country of Avanti’ is still spoken of in Rudradāman's inscriptions, circû 150 a.d. Dhār and Ujjayini to the north, were in early times most probably approached from the Narbada Valley by the opening in the hills at Mhau, through which the modern railway passes. Ujjayini lies a little way east of the Chambal Valley. The route to and from Mālwa viâ Nīmach and Ratlam would come into use later on.

^{page 538 note 2} Wilson, V.P., p. 188.

^{page 538 note 3} V.P., p. 185. The Māla are mentioned between the Bodha and the Matsya. Wilson thinks this refers to Chattīsgarh. On p. 193 we have Mālava (var.lect. Mālaka, Mājava) and alsoMallava (var. led. Vallabha).

^{page 538 note 4} V.P., p. 177.

^{page 539 note 1} See J.R.A.S., 10 1897, pp. 860–882Google Scholar. The coins (not earlier than the fifth century) are described in Rapson (Bühler's Encyclopaedia of Eastern Research), pp. 12, 13, to which Professor Macdonell kindly called my attention, as also to a notice in Bom. Gaz., vol. i, pt. 2, p. 311.

^{page 539 note 2} Rapson, § 101.

^{page 540 note 1} Jats seem to have preferred the Panjāb and the Rohilkhand districts; they extended, however, to Rājputāna.

^{page 540 note 2} In at least one of the lists of the ‘thirty-six houses’ the Gurjara were admitted to a place. At any rate they were sufficiently influential to make their name adhere to a country over the whole of which they never had dominion, and over part of which their dominion was not very long lived. There is of course no early literary use of Gurjarāṣtṛa, nor am I aware of its use by a later writer of pure Sanskrit. Gujarāt (or Gujarāṭh) was constructed by Moslem historians on the same principle as Sauraṣṭra from the S'aura tribe. In the Ūdepur prasaśti King Bhimdev I of Anhilpur (a.d. 1021–1063) is still called ‘Lord of Lāṭa and King of Gurjara.’

^{page 541 note 1} It is to be wished that we had some more definite authority or knew the source of the alleged division of the Āndhra into the real family (the Āndhrajātika) and the ‘Āndhrabhrityā.’ It does seem as if there was some such distinction in this widely dominating race. Wilson, V.P., p. 472 (note).

^{page 542 note 1} What is the real origin of the inferior title Thākur, which is applied also to the whole class in parts of the North-West Provinces, while Thakar appears in the Panjāh as the name of an inferior group ? Among the Nāga princes of Gondwāna ‘Thākur’ meant the priest attached to the royal household (Cent. Prov. Gaz., Introd., p. lxvi).

^{page 542 note 1} In the late thirteenth century Karn, the unfortunate prince of the Vāghela division of the Çālukya, considered it beneath his dignity to give his daughter in marriage to a Yādava prince of Devagiri: he consented only as an alternative to seeing her married to a Moslem. But had the Devagiri chief (already called by the annalists Mahrātha) been of pure Yādava blood, such an objection would be unintelligible. Probably by that time they were of very mixed race. It is notorious in the pages of Tod, Malcolm, and others, with what facility Rājputs of rank form unions, mostly irregular, with women of any class, and what numbers of inferior race—‘Dāsa,’ ‘Goli,’ etc.—are produced. Men of this birth may yet be of superior character and energy, and may originate houses, who rise far above the rank their origin would suggest, and attain to wealth and influence.

^{page 543 note 1} The strict seclusion of females observed in the higher Rājput families (of which several extreme cases are given in the Rāsmālā for example) seems not so much to be the result of Hindu teaching as a measure of protection suggested by the conduct of the Moslems, at any rate during the earlier invasions and periods of local domination.

^{page 543 note 2} “Memoirs of Central India” (reprint), ii, 119. I am not aware whether such a custom is observed by the Rājput tribes in the North-West Provinces, Oudh, or the Panjāb, by those really entitled to the rank.

^{page 543 note 3} The Rājputana Gazetteer calls attention to the still rather frequent traces of Jaina religion in Rājputana (vol. i, p. 89).

^{page 543 note 4} Tod, vol. i, chap, vi (reprint), especially pp. 62 ff

^{page 544 note 1} Rāsmālā, p. 536. See also the remarks about Vaiśya at p. 553.

^{page 544 note 2} Viswāmitra, though himself an Aryan, is always connected in one way or another with non-Brahmanic tribes. Here he is introduced as taking away Vasiṣtha's Sacred Cow. The four heroes are produced in answer to Vasiṣtha's prayer to avenge the wrong. It is impossible not to suspect an allegorical meaning here.

^{page 545 note 1} Tod's account of it is in vol. i, p. 82 (reprint). Is the story older than Chand and the other bards ? See Rāsmāla, p. 536.

^{page 545 note 2} Curiously enough, one list of ‘the thirty-six’ gives a place both to Parihāra and Pratihāra (Rāsmāla, p. 535); Cunningham suggested that they might be the Poruaroi of Ptolemy; but Lassen, with more probability, referred that name to the much better known Pramāra. This, if substantiated, would take back the Pramāra to before the second century. DrOppert, Gustav (“Original Inhabitants of India,” pp. 22–3, note, and p. 92)Google Scholar has an altogether different suggestion. Chand the bard makes the Pramāra superiors of the Parihāra, and accounts for the settlement of the latter in ‘Mardeś’ as by grant of the former; but then he places a Pramāra as king of Ujjain and suzerain. The Parihāra were, in fact, driven out of Marwār when the Rāthor princes of Kanauj came to an end, in mediaeval times.

^{page 545 note 3} See J.R.A.S., 1894, p. 1 ff. and 1895, p. 514Google Scholar. The first inscription is probably of the ninth century. That from Ghaṭayāla (No. 13 in Kielhorn's list) is Vik. 918.

^{page 546 note 1} It is curious that Chand (if I may trust Tod, i, 84, note) calls the Pramāra ‘of Telingana,’ as if they had something to do with the Andhra and Chālukhyā. Tod's list of their branches includes the Mori (Maurya). It is true that relics of this old race appear long after the time of Aśoka, as princes in the Konkan, and also at Çithūr in the Vindhya Hills; but if there is any connection with the Pramāra, the latter must be a branch of the Maurya, not vice versâ.

^{page 546 note 2} See McCrindle, , “The Invasion of Alexander,” etc., p. 354, and the Rāsmāla, p. 227; but see the Sindh Gazetteer, p. 862. The Dhāt State is still held by Sodhā RājputsGoogle Scholar.

^{page 546 note 3} Ep. Ind., i, p. 222 ff. It has various interesting items; among them an attack on the Haihaya king of Çedi by ihe Pramāra vassal Vākpati at the end of the tenth century.

^{page 546 note 4} As reported in Beames’ Elliot, Gloss., i, 68.

^{page 547 note 1} Bhārgava, Cyavana, Apravana, Aurva, and Jamadagnya.

^{page 547 note 2} And there is a traditional ‘Veni Vach Rāya’ (Bach) who was the original founder of Īdar. Curiously enough, one of the early Çauhān ancestors (Anhil) is placed at Garhamaṇḍala, i.e. with the Vatsya-Cedi people. I cannot find any detail about this.

^{page 547 note 3} Ep. Ind., ii, 117. Harsha is a hill in the Jaipur State.

^{page 548 note 1} See also Bom. Gaz., i, pt. 2, p. 182.

^{page 548 note 2} Rāsmāla, p. 18 ff.

^{page 548 note 3} Ep. Ind., i, 266 (line 33).

^{page 548 note 4} Bom. Gaz., vol. i, pt. 2, p. ISO. Here the ‘Çuluk’ comes in as the handful of sacrificial water taken by the god Brahmadeva.

^{page 548 note 5} See Wilson, V.P., p. 369, note.

^{page 549 note 1} Bom. Gaz., Vol. i, pt. 1, p. 120; pt. 2, p. 386.

^{page 549 note 2} Observe that the uncommon name of ‘Abhimanyu’ is itself that of one of the Yādaya ancestors, and (much later) that of a son of Arjuna.

^{page 550 note 1} I have remarked on these names in the list of ‘Rājput’ tribes at the end.

^{page 550 note 2} When scattered parties in the course of time wandered as far as Hindustan, and even to the distant Panjāb, very naturally different clan-names would arise; and those who still remembered ‘Rāthor’ would forget the history, and invent vague fables about Kuśa and Rāma, and ‘Hiranya Kaśipu,’ which have no meaning whatever.

^{page 550 note 3} For the Satyāki descent see Ep. Ind., iii, 268 ft, and the Tuṅga, see Ep. Ind., iv, 286 (Karhād plate). Other references for the Yādava descent are given in a note to Bom. Gaz., vol. i, pt. 2, p. 194.

^{page 550 note 4} See Anc. Geog., p. 317 (where there is much that has since been corrected, but that does not affect the present point). Full details are given in Bom. Gaz., vol. i, pt. 1, p. 119 ff. In vol. i, pt. 2, p. 386, a somewhat divergent account is given. Dr. Bhagvānlāl considered the undated inscription to be of about 450 a.d. Dr. Fleet thinks it about the seventh century. Bhagvānlāl takes the “Mānapura’ city to be the capital of the ancestor (of Abhimanyu) = Mānankapura, perhaps the Mālkhad (below Sholapur) of the later R. kings. Dr. Fleet suggests some place in Central India (Mānpur, in Mālwa). This latter is very unlikely. Whether the ‘Raṭṭa’ were a Dravidian clan, or connected with the early Araṭṭa (Bāhikā) of Epic times, or an early Yādava Aryan (and no real reason is given against the latter suggestion), Mālva is a most unlikely place; their whole early history is closely associated with the Dakhan. It is true that the Yādava descent is noted (but independently,, and not with the vagueness that purely legendary accounts exhibit) in inscriptions of a date when Brahraanic caste was prevalent, and tribes began to desire an ‘orthodox’ origin; but this is not conclusive against all the natural probabilities of the case.

^{page 551 note 1} The Bom. Gaz., vol. i, pt. 1, p. 150 ff., has given details about the Chāvada princes (or, as a Nausārī inscription calls them, Chavoṭaka), whom the bards represent as successors of the Vālā; but has not co-ordinated them with the Rāsṭṛakūṭa successors. The former must have either ruled independently in the northern part of the peninsula, or been in dependence on the latter. The Ratnamāla (about a.d. 1230), speaking of the struggle between King Bhuvaḍa or Bhūwar (Calukhyā) and the Pançasar prince (early in the eighth century?), puts the king at ‘Kalyāni,’ because that was the only place he knew of, though it was not really a Calukya capital before the eleventh century. Perhaps, as conscious of this, he purposely confuses the matter by referring also to ‘Kanyakubjā,’ which, geographically, is nonsense, but suggests another general locale tor ancient kings. I think his ‘Bhuvaḍa’ is a colloquial contraction for Buddhavarmma (a.d. 713). I believe the Caura (or Chāvda) are the really original tribe of S'aurā who gave their name to the country, and are contemporaries of the earliest Yādava inhabitants, and that the name S'ura of the Yādava genealogy has a possible reference to them.

^{page 551 note 2} Vol. iv, p. 412.

^{page 551 note 3} Wilson, p. 196.

^{page 551 note 4} Lassen, iii, 897.

^{page 552 note 1} Vol. i, p. 31, in the original edition; this I have not seen, unless vol. i, p. 29, of the reprint is the same; but there is no specific mention of an inscription asserting Pāṇḍava origin.

^{page 552 note 2} Wilson, V.P., p. 422.

^{page 553 note 1} It is curious that the Waniyo (Baniya or Vaiśya) in Gujarāt and the corresponding caste in the Dakhan are always regarded as very superior.

^{page 553 note 2} See the note to p. 68, Ep. Ind., vol. i. The objection, I submit, is quite untenable, and the evidence clear.

^{page 553 note 3} There is a good account in his “Chief Families of the Roy Bareilly District” (Lucknow, 1870)Google Scholar. A curious mark of Dakhan origin is noted by Crookes (vol. i, 120), that the Bais women retain a garment in one piece, which is evidently the graceful ‘sāri,’ the characteristic female dress of the ‘West and South.

^{page 555 note 1} References are to Beames' Elliot's Glossary, to Crookes', “Tribes and Castes of the North-West Provinces and Oudh,” 4 vols.Google Scholar, and to Ibbetson's, “Outlines of Panjāb Ethnography,” 1881 (1 vol., 4to)Google Scholar.

^{page 556 note 1} i.e., other than Yādava.

^{page 557 note 1} There is a king ‘Dirghabāhu’ in the V.P. list, but not fifty-first. Whether ‘Dargbaṅsi’ may be Dirghabaṅsi, and refer to him, I cannot say.

^{page 557 note 2} See p. 530, ante.

^{page 557 note 3} See Rasmālā, 237, and cf Rājp. Gaz., vol. ii, p. 265. Of the Dābhi I can discover no trace beyond what is in Tod, i, 105 (reprint).

^{page 558 note 1} Crookes (iv, 86), alluding to a Kāṅhdeś inscription of the thirteenth century. See Bom. Gaz., vol. i, pt. 2, pp. 470, 521, and vol. xxiv, p. 65 note, p. 414.

^{page 560 note 1} May not these be the Hāra Hūṇa of the fifth century p