^{page 314 note 1} Dialects current within the area in which standard Hindi is now the cultural language;the evidence relates chiefly to Braj Bhāṣā, Khaṛī Bolī, and to a lesser extent the Rājasthānīdialects.

^{page 314 note 2} Annual (from 1906 triennial) Report on Search for Hindi Manuscripts, 1900–; from 1926 Khoj mē upalabdha hastalikhit hindī granthõ ke traivārṣik vivaraṇ.

^{page 314 note 3} Grierson, G. A., Modern vernacular literature of Hindustan, Calcutta, 1889Google Scholar, introduction, XXII; Encyclopaedia Britannica, 11th ed., 1910, s.v. Hindostani.

^{page 314 note 4} op. cit.,22.

^{page 314 note 5} See further my forthcoming article ‘The rise of standard Hindi, and early Hindi prose fiction’ in a projected symposium on modern Asian literature.

^{page 314 note 6} Greaves, E., Sketch of Hindi literature, Madras, 1918, 87Google Scholar; Bailey gives a summary account of 32 writers of independent prose texts in Braj Bhāṣā and Khaṛī Bolī before Lallū Lāi of Fort College, William, BSOS, III, 3, 1924, 523–6Google Scholar.

^{page 315 note 1} A treatise on rhetoric by Gangā Bhāṭ, a bard attached to Akbar's court: Cand chand mahimā leā barnan, 8 folios. See Report on Search, 1909–11, 12 f.; appendix, pt. 1, 146 f. The MS is dated A.D. 1572. The language of this text is Sanskritized Khaṛī Bolī, with some admixture of Braj Bhāṣā; it can thus be considered an important antecedent of nineteenth-century Hindi prose.

^{page 315 note 2} IO Sanskrit MS 3318; Keith 7210. The MS was part of the collection of Gottlieb William Leitner, an Indian Civil Servant who was connected with the Punjab Education Department almost continuously from 1868 to 1886; presumably he acquired the MS from a source within the Panjab. The collection was purchased by the India Office Library on 27 June 1904. The MS escaped mention in J. F. Blumhardt's unpublished notes on the Hindi MSS at the India Office Library, and the existence of the Braj Bhāṣā commentary was first recorded by Keith.In 1950 Mr. A. G. Shirreff noted the commentary and its importance in his revision of the material left by Blumhardt. The only other mention of the commentary which I have been able to discover is by Kosambī, D. D. in his edition of the epigrams of Bhartṛhari, Śatalcatrayādi subhaṣitasaṃgraha,Bombay, 1948Google Scholar.

^{page 315 note 3} Iti ŚrīmatsakalanṛpatimaulimaṃḍanamaṇiŚrīmadhukarasāhinṛpatitanūjaŚrīmadeajidviracitāyāṃ vivekadīpikāyāṃ. bhartṛhariṭīkāyāṃ nītiŚatarṃ samāptaṃ, f. 25b, 11. 5 f. The third couplet of the Sanskrit introduction to the text, quoted hy Keith and again in this article, appears to refer inaccurately to the Vivekadīpikā as the Bhartṛhari text itself, ‘with a commentary’(saṭippanīṃ); presumably more than the NītiŚataka alone is meant, despite the fact that this MS is complete in itself.

^{page 316 note 1} F. 26b is ruled with side lines, entitled, and numbered, but the text itself is clearly complete half-way down f. 26a.

^{page 316 note 2} Quoted by Keith, together with other information on the MS not given here.

^{page 316 note 3} Such a correction in the text as that of the compound verb first written cora letu hai, f. 4a,1. 1, to normal cori letu hai points to general careful transmission of the received text by the scribe (as well as to a weakened or whispered pronunciation of final -i in his own speech). This is the necessary interpretation, since grammatical inconsistencies in the language of the extant text are too many for this particular correction to be considered as the work of a normalizing scribe. (The full phrase is ju vidyādhanahī cor(i) letu hai. It is not possible to assume that the form cora is in fact correct here and represents the substantive of this form, perhaps under the influence of the wording of the Sanskrit text, hartur yāti na gocaraṃ, since this would imply that the relative pronoun ju had been understood as associated with the hĩ-extended sentence object vidyādhanahī; according to the normal grammar of this text, however, the oblique form of the pronoun (jā) would be required in this ease. The phrase vā vātahi deṣatu hai, f. 4a, 1. 2, shows the corresponding usage of the correlative pronoun. We must therefore assume verbal stem cora- here, used as the basis of a compound verb; it recurs in the following line, vastu jau cori jatu hai, f. 4a, 1. 2. This stem is not normal in modern standard Hindi, which has curānā and corī karmā, nor is it listed by Platts or the Hindī Śabdasāgar, both of which cite numbers of Braj Bhāṣā forms.)

^{page 316 note 4} Abu'l Faẓl, Akbarnāma, tr. H. Beveridge, Calcutta, 1910, vol. III; Ā'īn i Akbarī, tr.H. Blochmann, Calcutta, 1873, vol. i; Niẓāmuddīn Aḥmad, Ṭabaqāt i Akbarī, tr. B. De, Calcutta, 1936, vol. II. A ‘History of western Bundelkhand’, translated in part by Silberrad, C. A., JASB, LXXI, Pt. I, No. 2, 1902, 99–135Google Scholar, from a Hindi original, seems to be based mainly on KeŚav Dās's poetic chronicle Bīr SiḤha caritra (A.D. 1604), whose historicity is suspect in detail; the other works of KeŚav Dās, notably his Kavipriyā, ed. KeŚav granthāvalī, i, Allahabad, 1954, are the chief literary sources, apart from the anthologies used by Grierson and others. Indrajit is not mentioned in the Bhaktamālā of Nābhā Dās, nor in Priyā Dās's commentary on that work.

^{page 316 note 5} Beveridge, op. cit., p. 324 ff.

^{page 316 note 6} Beveridge, op. cit., p. 803 ff.

^{page 317 note 1} Śiv Siṃha saroj, Lucknow, 1878, appendix, p. 64.

^{page 317 note 2} op. oit., 59.

^{page 317 note 3} Miśrabandhu vinod, Khaṇḍvā and Allahabad, 1909, p. 404.

^{page 317 note 4} Abu'l Faẓl refers to this person as both Rām Śāh and Rām Singh in the Akbarnāma, and as Rām Śāh and Rām Cand in the Ā'īn-i-Akbarī; Niẓāmuddīn calls him Rām Cand.

^{page 317 note 5} Kavipriyā, ed. eit., p. 97. The grant of Kachovā is also referred to in the Bīr Siṃha caritra; Silberrad, loc. cit., Ill, identifies Kachovā as Nad-Kaehuwa, 27 miles south-east of Jhānsī.

^{page 317 note 6} Śiv Siṃha Sengar, op. cit., p. 134; a verse by him is there given. The MiśVa brothers, loc. cit., describe him as a pedestrian poet. The Ratnabāvanī of Keśav Dās mentions naridra [sic] madhukara sāhi putra in its concluding verses; Report on Search, 1906–8, 178.

^{page 317 note 7} Ed. cit., p. 100.

^{page 317 note 8} Blochmann, op. cit., p. 487. Blochmann records no mention of Indrajit.

^{page 317 note 9} They are preceded in the MS by the invocations and title: Śrīrāmāya namaḥ. Śrīparamātmane namaḥ. Atha nītiśataṃ likhyae. Keith gives all four couplets, but it will be convenient to cite them again here. I am grateful to Professor J. Brough for his assistance with their elucidation.

^{page 318 note 1} Misprinted vārghato in I.O. Library catalogue.

^{page 318 note 2} An emendation is necessary here to meet the needs of grammar, though it is not certain that the mistake does not originate with the composer of the introduction himself, who was no Sanskrit scholar; cf. the faulty metre tolerated in thefirst couplet of the introduction.

^{page 318 note 3} A reference in the Bīr Siṃha caritra (1604) to Madhukar Śāh as nṛpatimukuṭamani madhukara sāhi (Bombay ed., 1904, p. 72) at once recalls the phraseology of the conclusion of this text; see p. 315, n. 3. Elsewhere in the same poem Bīr Singh is referred to as nṛpatimukuṭamani vira siṃha deva (ed. cit., p. 180). The phrase is clearly a current formula, and suggests that the composition of the conclusion (and hence also the introduction) of this text is to be related in some way to literary activities at Orchā in the early years of the seventeenth century. Interest in the genealogy of the Orchā royal house is a further characteristic feature shared by the introduction to this text and those to the works of Keśav Dās.

^{page 319 note 1} F. lb, 11. 6–12. In view of the date of the languageof this text, and the others from which extracts are given in this article, the inherent vowel has been everywhere represented in the transcription, although the extent to which it may have been realized in the speech of the time, as distinct from the recitation of poetry, is a matter for further linguistic investigation. It isnormally held that the inherent vowel was realized both medially and finally in older Braj Bhāṣā;see Dhirendra Varmā, Braj Bhāṣā, Allahabad, 1954, 40.

Anusvāra occurs frequently in the texts discussed in this article in conjunction with syllables containing or preceding a nasal consonant. Its appearance in these circumstances is accounted for by the assumption that nasal consonants induce marked nasalityin neighbouring vowels in the language of the texts; there are four such occurrences of anusvāra in the present extract, and one other for which there seems to be no justification.

Candrabindu is not used in this extract.

The akṣara , which has been represented as ṣa in the transcription wherever it occurs, is frequently though not invariably used to represent the phoneme /kh/.

^{page 319 note 2} Corrected from anavacchinna.

^{page 319 note 3} Kirātārjunīya, II, 30.

^{page 319 note 4} op. cit., introduction, pp. 39–55.

^{page 319 note 5} Verses 65 and 66 have been transposed in this text.

^{page 320 note 1} op. cit., no. 465, p. 153.

^{page 320 note 2} Ed. K. P. Parab, 4th ed., Bombay, 1905, p. 95 (SRB).

^{page 320 note 3} op. cit., introduction, p. 61.

^{page 320 note 4} cf. the reading daṃṣtrāṃtarāt in verse 3, prasahya maṃim uddharen, which is given in Kosambī's apparatus (op. cit., p. 4) as occurring in verse 4 of most of the MSS versions belonging to the Southern Recension, and in no Northern version. This reading taken in conjunction with Indrajit's verse 1 would suggest general Southern influence on the opening verses of his text at some stage in its transmission.

Other readings in early verses are characteristically Southern, though not without support in Northern versions also: saṃpūrṇaḥ kalayati in verse7, parikṣīṇaḥ kaścit, and sudhiyas tv arthaṃ inverse 7, śāstropaskṛta (op. cit., p. 6). These, however, are no evidence for Southern influence at these points in the text, since in all versions of the Southern Recension they occur in positions quite different from 7 and 8. The problems of thetransmission of Bhartṛhari texts are too complicated for affiliations to be' established over any length of text on the basis of textual variants alone.

^{page 320 note 5} Keith 3395. This is Kosambī's text F4, one offive MSS collated and grouped together loosely by him as ‘pseudo-version F’, showing ‘many common features without sufficiently close śloka agreement to form a version’; their verse order is therefore not noted in Kosambī's synoptic chart, but since text F4 is readily accessible its verses were included in the comparison.They belong fundamentally to Kosambī's Northern Recension, but show some Southern influence; op. cit., introduction, p. 24.

^{page 320 note 6} Kosambi's no. 74*, parivartini saṃsāre. The omission of this verse cannot be explained by a similarity of content or wording between the preceding verse, sprhayati bhujayor, and the following, lajjāṃ guṇaugha (Indrajit's vv. 98 (MS, 97) and 99 (MS 98) ). It represents a genuine lacuna in Indrajit's text.

^{page 320 note 7} Probably explicable as due to correction of a version D text against a version I text are the omission of parivartini saṃsāre,, the inclusion of verse 106 (MS, 105), siṃhaḥ śiśur api, which follows parivartini saṃsāre in version I, and the inclusion of verse 101 (MS, 100), diggajakamaṭha.It is of interest that the single MS which represents version Icontains a Rājasthānī commentary;see Kosambī, op. cit., introduction, p. 29.

^{page 321 note 1} Bhandarkar Oriental Research Institute, Poona, Descriptive cat. of the government collections of MSS, XIII, Pt. 1, 1940, no. 350, a recataloguing of P. Peterson's no. 357, Report for 1892–5, p. 257. See Kosambī, op. cit., introduction, pp. 74, 14. The MS is noted by Aufrecht, who gives the name of the commentary as Bālāvabodha, Cat. catalo., in, 86a. A fuller description of the MS is given than seemed necessary in the case of IO 3318; I am indebted for the information on which this is based to Śrī P. K. Gode and to the present Curator of the Bhandarkar Oriental Research Institute.

^{page 321 note 2} Candrabindu is represented together with anusvāra as ṃ in the transcription. In the extracts cited it occurs only in the syllable oṃ and in the word śrīṃdrajid in the introduction.

^{page 321 note 3} Kosambī's verse 632, not his separate verse 294, Bhoge rogabhayarri kule; its variant readings agree mainly with those found in his text F4, although there are several errors.

^{page 322 note 1} Verses 49 (MS, 16), re kaṃdarppa śaraṃ, 50 (MS, 17), mātar lakṣmi, and 51 (MS, 18),gaṃgātīre, which recur as verses 108 (MS, 84), 77 (MS, 57) and 53 (MS, 35) respectively.

^{page 322 note 2} Verse 14 is not numbered; the following verse is numbered 15 correctly.

^{page 322 note 3} Only one of these verses, no. 16 (MS, 7), asaṃpādayataḥ, is given in SRB, p. 84, § 5; the Subhāṣitaratnākara, ed. K. S. Bhāṭavaḍekar, Bombay, 1872 (SRK), attributes it to Māgha.

^{page 322 note 4} Verse 33 (MS, 14), phala(ṃ) svecchālabhyaṃ, which occurs only in Kosambi's Southern Recension, and verse 85 (MS, 22), eko bhrāṃtas tatdḥ; the order of the half-lines of this verse is not paralleled in any of the MSS whose variant readings are noted by Kosambī, and it may be incomplete.

^{page 322 note 5} All nine of these are found in Kosambī's MS ISM Kaḷamkar 195, in the same order as in this text. ISM Kajamkar 195, verses 90 and 91, and 117, 118, and 119 correspond to interpolations of two and three consecutive verses in this text: verses 69 and 70 (MS, 19 and 20) and 117–19 (MS, 25–7). Clearly there is some historical connexion between BORI 350 and ISM Kaḷamkar 195.

^{page 322 note 6} Verse 128 (MS, 101), bhoge rogabhayaṃ, occurs as the final verse of text F4. See p. 321, n. 3.

^{page 322 note 7} Verse 115 (MS, 91), gātraṃ samhucitam, is a likely borrowing from the Southern Recension, as in both the Northern versions (A and D) which contain it, it occurs beyond the point at which the BORI 350 text seems to have beencomplete. See also n. 4 above.

^{page 323 note 1} The one Khaṛī Bolī text noted is a fragmentary Nītiśataka commentary; see Report on Search, 1935–7, no. 146.

^{page 324 note 1} The Hindī Sāhitya Sammelan, Allahabad; the Rāṣṭrabhāṣā, Pariṣad, Paṭnā; the Hindī Vidyāpīṭh, Āgrā.

^{page 324 note 2} op. cit., introduction, p. 56.

^{page 324 note 3} These sources were: MSS in private possession at Bīkaner and Jaipur (catalogues seen by courtesy of Śrī N. R. Khaḍgāvat, Director of Archives, City Palace, Jaipur); the MS collection of the Rājasthān Prācya Vidyā Pratiṣṭhān, Jodhpur; MSS at the Government Museum, Alwar, and Copāsanī, Jodhpur (information by courtesy of Śrī G. N. Bahurā, Assistant Director, Rājasthān Prācya Vidyā Pratiṣṭhān); records of Hindi MSS preserved in Rājasthān (Rājasthān mē hindī Ice hastalilchit granthõ kī khoj, vols. I-IV, 1940–54); MSS at the Government libraries in Koṭāh and Bharatpur; the libraries of the Mahārājā of Bharatpur, the Mahārājā of Būndī, the Vidyāvibhāg, Kānkrolī, the Nāthdvārā Temple, near Udaipur, and the Raghunāth Temple,Śrīnagar (these libraries proved to be not accessible, nor did I see catalogues of them; I had to be satisfied with an assurance that they contained no text of interest to me). I was not, of course, able to think of visiting individual Jain bhaṇḍars, nor could I visit the abbeysnear Amritsar listed by Kosambī, loc. cit., although this would have been desirable in view of the likely provenance of 10 3318.

^{page 324 note 4} Rājasthān Prācya Vidyā Pratiṣṝhān, Jodhpur. It had been acquired by the Pratiṣṭhān(formerly the Rājasthān Purātattva Mandir) from private ownership in the neighbourhood of Bīkaner only 14 months before my visit to Jodhpur. I am indebted to Śrī Bahurā for allowing me to examine the MS and for arranging to have it photographed.

^{page 324 note 5} These texts are: Vairāgyaśataka; Nītiśataka; Sāracandrikā; Nāmavirudāvalī; Praśnottarajṇānaratnamālā; Rāmarasāyana; Hitopadeśa (pamcākhyāna), in that order.

^{page 324 note 6} This text seems to have no organic connexion with its predecessor in the codex, and to stand next to it only because its Sanskrit portion is by Bhartṛhari and its bhāṣā portion is in prose, whereas the following five texts have verse commentaries. Its text is arranged in paddhati groupings of from eight to ten verses and ends with the verse bhīmaṃ vanaṃ bhavati, which facts place it in Kosambī's Southern Recension; its commentary is briefer and much inferior to that of Indrajit, and the terms of its conclusion (there is no colophon) are quite different from those used in 10 3318 and BORI 350. It has little value except as supporting evidence for a tradition of vernacular comment in prose on Bhartṛhari's ‘Centuries’ at an indeterminate date.

^{page 325 note 1} The modified akṣara , which the scribe of this text frequently uses to represent the phoneme /y/, is represented as ẏa in the transcription. Modifie, for the phoneme /v/, occurs once in the extracts cited; it i s represented as a.

^{page 325 note 2} The verse shows the same textual variants as its counterpart in BORI 350, again with errors.

^{page 325 note 3} The syllable ṣai follows closely on an erasure of a partly completed syllable vi, by dittography from the preceding syllable.

^{page 325 note 4} Verses 50–2 (MS, 16–18): re kaṃdarpa saraṃ, mātar lakṣmi, and gaṃgātīre. They recur with commentary as nos. 108 (MS, 84), 77 (MS, 57) and 54 (MS, 35) respectively.

^{page 325 note 5} No. 36 is omitted; nos. 6 and 7 are both repeated.

^{page 325 note 6} Verse 1, yāṃ ciṃtayāmi, occurs as VI only in text F4 and in one other text used by Kosambī (also from Jodhpur; op. cit., p. 122). It is probable that Indrajit's V text had undergone influence of an analogue to text F4 before he received it, but this influence may not have extended to the imposition of yāṃ, ciṃtayāmi as verse 1 on the pre-Indrajit V text; yāṃciṃtayāmi does not occur in BORI 350, and seems the more likely to be the result of a separate contamination in Jodh. 10956a since the following verse 2, bo(d)dhāro, is introduced in the text with the words tāṃhẵ bhartṛhari prathama hī loka hī kartavyatā deṣi pichatāta hãi, f. lb, 1.10 f.The phrase prathama hī, it will be remembered, introduces the very first phrase of the IO 3318 commentary, preceding verse 1, while the adverb tāṃhẵ, refers most naturally back across the intervening verse and commentary to the Sanskrit introduction to the text. It may be noted further that the verse bo(d)dhāro is only very indistinctly numbered; the figure could be read 1 almost as easily as 2 (f. lb, 1.12). A scribe seemsto have been reluctant to admit yāṃ, ciṃtayāmi to the V text as he knew it.

Verse 4, parigrahotiduḥkhāya, is not included in BORI 350; it occurs as verse 82 in ISM Kaḷamkar 195.

^{page 325 note 7} These verses all occur, in the same order, outside the main series in BORI 350; see p. 322, n. 3–5.

Verse 85 (the MS numbers it 42, but the commentary is correctly numbered 22), bhāvyaṃ.bhuktaṃ, is complete in this text, showing the half-line order of Kosambī's text F3 and versions H and I.

^{page 326 note 1} The first six seem numbered at random; the eighth (verse 25) is not numbered; the ninth(verse 28) is illegibly numbered; four others show discrepancies between their verse and commentary numberings.

^{page 326 note 2} e.g. BORI 350 ekāgra hvai kanirirbhaya …: Jodh. 10956a aikāgra kari nirabhaẁa …(verse 101, commentary); BORI 350 nṛpatimuuliḍimaṃḍanamani … Jodh. 10956a nṛpatimaulimaṃḍanamani…(conclusion).

^{page 326 note 3} F. lb, 1.15–f. 2a, 1.1. The mistake is of course explicable palaeographically.

^{page 327 note 1} Dewhurst, R. P., ‘The Shṛngāra-shataka of Bhartṛhari with an old commentary in Hindi written early in the seventeenth century’, Journal of the U.P. Historical Society (Calcutta), I,1, 1917, 60–151Google Scholar, where the MS is described and the text printed, together with notes which deal mainly with the variant readings found in the Sanskrit verses; there are also eight pages of grammatical notes. These exemplify the main forms found in the text, but unfortunatelygive no idea of their relative frequency or the contexts in which they occur. The summaryaccount of the MS given above is drawn from Dewhurst's article, efforts to trace the MS itself having so far proved unsuccessful.

^{page 327 note 2} Saṃbatu 1683 varṣe māgha badi 3 somvāre.

^{page 327 note 3} Svapaṃtu sukhaśrāṃtakāṃtā, not listed by Kosambī or in SRB or SRK. One other verse unidentified by Dewhurst, śikhini kūjati, appears in Kosambī's version D and his texts Fl, 2, and 4.

^{page 327 note 4} Sukla, R. C., Hindī sāhitya kā itihās, 11th ed., Banāras, 1957, p. 373Google Scholar, quotes a passage from a Śṛngarāśataka commentary of the eighteenth century Vikram. which agrees completely in its wording, though not in all details of its spelling, with a passage in Dewhurst's text (loc.cit., 101 f.). The minor differences between the versions may indeedbe taken as errors of transmission tending to confirm the greater age of Dewhurst's missing MS, but in the light of the similarities and of the unprecedented age claimed for the MS it would be reassuring to be able to check the date given in the scribe's colophon.

^{page 328 note 1} The similarities of language between this text, which is said to be carelessly written, and the text of 10 3318 do not detract from the authority ofIO 3318 readings; Dewhurst's strictures on the capabilities of the scribe of his text bear chiefly on the errors in the Sanskrit verses, and in so far as they apply to the great variety of Braj Bhāṣā grammatical forms which the text contains reflect an unwarranted assumption that medieval Braj Bhasa was fully standardized as a literary medium. But the variety evident in the Braj Bhāṣā forms of the text was probably not introduced by the scribe. Judging by the language of contemporary Braj Bhāṣā poetry most of the forms found in the text were genuinely current in the early seventeenth century.