^{page 145 note 1} v.s. 1657–77, A.H. 1009–30. The Hijra date is added for convenience of comparison, since much of the following discussion is necessarily cast in terms of the Hijra era.

^{page 145 note 2} Edited with commentary entitled Bhaktisudhābindu-svāda by Pt. , Bhagvān Prasād. I have consulted also the readings of a version printed by the Bombay Press, undated; the with mūl in Devanāgarī and commentary in Gujarātī by Rāy Prabhurām Mehtā, Ahmedabad, v.s. 1960; and three MSS in the India Office Library, numbered in the catalogue H. 33, s. 2090, and s. 2736.

^{page 145 note 3} v.s. 1769, A.H. 1124.

^{page 146 note 1} The symbols ε and ɔ are used to transliterate the Devanāgari and

^{page 146 note 2} The most usual forms are kīyo, dīyo (rhymed with each other and with līyo ‘took’ in chapays 45, 51, 61, 66, 70, 89, 112, 115, 173, 185; forms with short first vowel, kīyo, etc., in ch. 79, 94, 95, 180); less frequently occur kīno, kīnī, dīno, dīnī (rhymed with each other and with līno in ch. 63, 71, 93, 154, 166, 181), and the form based on the present stem, karyo (ch. 167, 184). For these forms and their inter-relations, cf. Tedesco, Paul, ‘Geben und nehmen im Indischen’, JAOS, XLIII, 1923, 358–90.Google Scholar

^{page 146 note 3} cf. Dave, T.N., A study of the Gujarāti language in the 16th century (V.S.), London, 1935, 53–4.Google Scholar

^{page 146 note 4} The possible inferences from this are referred to later; ef. pp. 156–157.

^{page 146 note 5} The commentator is, to his credit, conscious of this: he adds somewhat pathetically, at the end of a long arabesque of commentary (I translate): ‘The meaning of many words in this mūl, and of the line beginning kamadhuja …, are not within the understanding of this unfortunate one; skilful great ones may with favour emend them’.

^{page 146 note 6} cf. Mahābhārata, II, xiv, 34–50; XII, cccxl, 87–9. A estimate of its importance is given in the Mir'āt-i-Aḥmadī, (GOS, Vol. L, p. 142): ‘they come from all parts and quarters of India, from the Deccan and from Sind, to worship’. The legend of foundation is also given:

I am indebted to my colleagues of the Department of the Near and Middle East, particularly to Professor Lambton, for their assistance with the Persian passages.

^{page 147 note 1} The Old Gujarātī absolutive, however, ends in -ī; cf. Dave, op. cit., 48–9.

^{page 147 note 2} cf. the passage quoted below, p. 154, from the Akbar-nīma: ‘On account of the heights and hollows, the brave men got off their horses and engaged …’.

^{page 147 note 3} cf. Scott, W. in Bombay Gazetteer (abbrev. BG), VIII, 1884, 593,Google Scholar s.v. Okhamandal. Another Vāɖhel genealogy is at variance with this; cf. p. 156.

^{page 147 note 4} cf. R. L. Turner, Nepali dictionary, s.v. māṛnu.

^{page 147 note 5} The texts and MSS write for ba and va; cf. also notes on vāɖha, below.

^{page 148 note 1} cf. Dave, op. cit., 172.

^{page 148 note 2} The makara is not, however, usually equated with the matsyāvatāra, and it is very doubtful whether N would have made this identification. While the may be associated with the decoration of icons (cf. Gopinatha Rao, T.A., Elements of Hindu iconography, I, 1, pp. 103,Google Scholar 254, et passim), at least as a member of the has the (ibid., 244); nevertheless in the northern tradition may be called ‘makara-decorated’, as a verse (No. 4 in the recension) in Bihārī Lāl's Satsai shows:

But it is perhaps more likely that dhuja refers here to a votive column, cf. Chanda, Ramaprasad, ‘Archaeology and Vaishnava tradition’, MASI, V (Calcutta, 1920), 162–3,Google Scholar and Banerjea, Jitendra Nath, The development of Hindu iconography. Second edition, Calcutta, 1956, 103—4.Google Scholar These writers refer, among other such columns, to a makaradhvaja capital discovered at Besnagar, and infer from this the dedication of a neighbouring shrine to Pradyumna. Why this particular vyūha should be associated with a temple is not clear.

It is unfortunate that the temples at Dvārkā have not been critically described; the note in BG, VIII, 601, is tantalizingly vague. W. Crooke's note in the Encyclopædia of religion and ethics, s.v. DWĀRKĀ, adds nothing to this; but his remark that the figure of carved over the entrance door ‘indicates a dedication to Śiva, which makes it difficult to assign the original building to the Vaiṡṇava cult of Kṛṡṇa’ is over-dogmatic: since ‘the worship of Ganapati without reference to any particular sect is practised by nearly all Hindus at the beginning of any religious ceremony’ [including, presumably, entering a temple] ( Bhandarkar, R.G., Vaiṡṇavism, Śaivism and minor religious systems, Strassburg, 1913, p. 150 Google Scholar); cf. Rao, op. cit., I, 1, p. 93 (description of the bhogafayanamürti ): ‘… on the south (apparently outside the shrine and in niches provided) there should be the figure of Gaṇēśa …’.

^{page 148 note 3} cf. Bhagavadgītā, XI, 46: … tenai 'ua caiurbhujena sahasrabāaho bhava vifvamūrte. is usually represented with four arms, with two (Rao, op. cit., I, 1, pp. 227, 209).

^{page 148 note 4} cf. BG, VIII, 160, s.v. Bhánds.

^{page 148 note 5} p. 152: ‘Caturbhūj [sic] in the muḥalla of Cangpol in the house of a wearer of the thread; its appearance is of black stone, and it is provided with four arms; and the reason of its being thus named is because of there being four arms, as they call the upper arm bhūj’. (This extract is incorrectly translated in GOS, XLIII, p. 129.)

^{page 149 note 1} Śabd-sāgar: . khodnā, khodkar girānā; ; zor se dabānā.

^{page 149 note 2} Dh., 10.181–2 (numbering as in Böhtlingk, Pāṇini's Grammatik); for the nasalization, cf. J. Bloch, L'indo-aryen, 45–7.

^{page 149 note 3} cf. BG, VIII, 110, 590.

^{page 149 note 4} e.g. Mir'āt-i-Aḥmadī, Mir'āt-i-Sikandarī, Akbar-nāma. The text-edition of the lastnamed (Calcutta, 1886) also, curiously enough, has the text wrongly divided: as though reading ‘took refuge in the locality of Sīvābād’ rather than ‘took refuge with Sīvā Bādhel’. ب for is a common North Indian corruption; and the text does not differentiate between and .

^{page 149 note 5} Paṇṭit, B.S., ‘Syntax of the past tense in Old Rājasthānī’, BSOAS, VIII, 2–3, 1936,Google Scholar 697.

^{page 149 note 6} i.e. taking kaaka as feminine. Since senā (and, presumably following it, fɔj) are feminine, transfer of the gender of kaaka to that of senā is not unlikely. Cf. Kurylowicz, J., Études indo-européennes, I, 248.Google Scholar

^{page 149 note 7}

(Quoted in BG, VIII, 593.)

^{page 149 note 8} BG, VIII, 590.

^{page 149 note 9} Tod, , Annals and antiquities of Rājasthān, I, 105.Google Scholar

^{page 149 note 10} asura here probably means no more than ‘non-Hindu’.

^{page 150 note 1} cɔka: ‘square, market-place’; there is such a ‘carfax’ in front of the temple at Dvārkā.

^{page 150 note 2} BG, I, 1, p. 295. I have not been able to trace the original source of this assertion.

^{page 150 note 3} The facts are uncertain, and different accounts of Isḥāq's derangement are given in the Mir'āt-i-Sikandarī and in ‘Muḥammad bin, Abdallah ‘Umar's Arabic history of Gujarāt (ed. Ross, E. Denison, 1910–1929), pp. 117,Google Scholar 150. The latter authority adds that the ‘Rajput chief of Dwarka … was loyal to the Sultan’ (cf. Commissariat, M.S., History of Gujarāt, I, 335–6Google Scholar).

^{page 151 note 1} The Mir'āt-i-Sikandarī gives a full account of the conquest of Jagat and ‘Sānkhodvārā’ (pp. 60–3 in Fazlullah's translation), which has provided the basis for most later accounts. It will suffice here to quote the Mir'āt-i-Aḥmadī, p. 143:

‘In the year eight hundred and seventy-eight of the Hijra, Maḥmūd Bēkada conquered [Dvārkān] and destroyed the images on the island of Sānkhō Dhār, and established a mosque’; and, ibid.:

‘In the early days of Islamic rule the temple idols, which were of his father and his brother, were transported here [Śankhodhar] to the temple; Maḥmūd destroyed them’. Cf. also Bayley's History of Gujarát, 195–7, Briggs's Ferishta, IV, 59–60, and Aḥmad's De's translation, III, 1, pp. 259–62.

^{page 151 note 2} cf. Samuel Butler, Hudibras, Part II, Canto I.

^{page 151 note 3} e.g., in J. Burgess, List of antiquarian remains in the Bombay Presidency (= ASWI, XI), 243,248,251,253,258; H. Blochmann in I A, IV, 1875,291,367; and Epigraphia Indo-Moslemica, 1925–6, 20; 1929–30, 4–5. The fullest version of his title appears as wa'l-ḥaq Nāṡiru'd-dunyā wa'd-dīn Abu'l-fatḥ Maḥmūd bin Muḥammad bin …, etc.

^{page 151 note 4} The Mir'āt-i-Sikandarī was compiled by Sikandar bin Muḥammad, alias ‘Manjhū’, in A.H. 1020; the Mir'āt-i-Aḥmadī by Mīrzā Muḥammad Ḥasan, entitled Alī Muḥammad Bahādur, between A.H. 1170 and 1174; and the Akbar-nāma by Akbar's minister and confidential adviser, Abu'l Faຓl, in A.H. 1010.

^{page 152 note 1} Translation by Fazlullah Lutfullah Faridi, ‘Education Society's Press’, n.d., pp. 326–7.

^{page 152 note 2} ‘This Vádhel Rạjput seems, no doubt, to have been a landowner of Bet, though the text is rather vague as to his identity’ [translator's footnote]; but Fazlullah's confidence is misplaced: cf. below, p. 155.

^{page 153 note 1} ‘The port of āramrāe is superior to most of its class. The inhabitants are of the Bāḍhel tribe. It musters 1,000 horse and 2,000 foot’ (Abu'l Faຖl's Ā'īn-i-Akbarī, Jarrett's translation, II, 248).

^{page 153 note 2} Translation by Bird, J., Oriental Translation Fund, London, 1834, 420–1.Google Scholar

^{page 153 note 3} ‘I cannot find this place on the map, but it is probably the same as Aranraw, opposite the Bate [i.e. Be] Pagoda’ [translator's footnote]. Arāmī0256ā appears spelt in many different ways in the MSS.

^{page 153 note 4} op. cit., 423.

^{page 153 note 5} Translation by H. Beveridge, BI, 138, 962–3.

^{page 154 note 1} Text:

^{page 154 note 2} Text:

^{page 154 note 3} Beveridge, op. cit., 965.

^{page 154 note 4} Beveridge adds a footnote to his mention of the name of : ‘The son of Moḥammad K. (Iqbālnāma)’; I have not been able to locate this reference. The Ā'īn-i-Akbarī (Blochmann's translation, I, 516) shows this as a ‘commander of two hundred’, and says: ‘He is not to be confounded with the author of the Ṭabaqát’. Only Aḥmad, the historian, is mentioned in Abu'l Faຓl's list (ibid., 528) of those who have been

^{page 155 note 1} pp. 566–93 in B. Be's translation, BI, 225.

^{page 155 note 2} His account of this incident is on p. 647 of De's translation.

^{page 155 note 3} In no other source is there any mention of the foundation of a mosque at Dvārkā; but a ‘ outpost’ would presumably have required one.

^{page 155 note 4} Text, p. 629:

^{page 155 note 5} BG, VIII, 164, 591.

^{page 155 note 6} ibid., 593.

^{page 156 note 1} Quoted from Khakhar, Dalpatrām Prāṇjīvan, Report on the architectural and archaeological remains in the province of Koch, Bombay, 1879.Google Scholar

^{page 157 note 1} Tod, op. cit., II, 1170; the Gujarātī commentary referred to on p. 146, n. 2, shows that the word is to be read as kābā: