^{page 48 note 1} 1 The terms “Primary”, “Secondary”, and “Tertiary“ Prakrits are explained in Chapter II.

^{page 48 note 2} e.g. by Hoernle, in his Comparative Grammar of the Gaudian Languages.

^{page 48 note 3} The term “Indo-Aryan” distinguishes those Aryans who settled in India from those Aryans who settled in Persia and elsewhere, just as “Aryo-Indian” signifies those inhabitants of India who are Aryans, as distinguished from other Indian races, Dravidians, Mundās, and so on. “Gaudian,” meaning non-Dravidian, therefore connotes the same idea as “Aryo-Indian”. These two words refer to the people and their language from the point of view of India, while “Indo-Aryan” looks at them from the wider aspect of European ethnology and philology. See Encyclopcedia Britannica, 11th ed., 1910, art. “Indo-Aryan Languages”.

^{page 49 note 1} Including the mixed Khāndēśī dialect.

^{page 49 note 2} Including the mixed Bhīl dialects.

^{page 49 note 3} Nearly all the speakers of this language inhabit Nēpāl, a country which was not subject to the Census of 1911, and to which the Linguistic Survey did not extend. The figures here given refer only to temporary residents in India.

^{page 49 note 4} In the Census nearly all the speakers of Central Pahārī were classed as speaking Hindi.

^{page 50 note 1} Cf. Amīr Xusrau in Elliot, History of India as told by its own Historians, iii, 539. “Whatever live Hindū fell into the king's hands was pounded into bits under the feet of elephants. The Musalmāns, who were Hindīs (countryborn), had their lives spared.”

^{page 51 note 1} In the Linguistic Survey the term “Western Hindī” is employed instead of “Hindī”, in order to distinguish it from the altogether different “Eastern Hindī”. The word “Western” is here dropped, as being hardly necessary for the class of readers for whom this work is intended.

^{page 51 note 2} Not “Hindūstānī”, as often written by Europeans. See C. J. Lyall, Sketch of the Hindustani Language, Edinburgh, 1880, p. 1.

^{page 51 note 3} The South being a Dravidian country, the soldiers and rulers who came from various parts of Northern India and conquered it did not acquire the local language, but adhered to their own lingua franca picked up in the Delhi bāzār.

^{page 52 note 1} Lyall, op. cit., 9.

^{page 52 note 2} The two principal writers in this style were Rangīn and Jān Sāhib. Their works are valuable for students of the women's dialect.

^{page 52 note 3} A translation of the tenth book of the Bhāgavata Purāna.

^{page 53 note 1} Blochmann, Āīn-ĕ Akbarī, tr., 352.

^{page 53 note 2} The one exception is the fact that the termination of strong masculine nouns with a-bases ends in ā, not in au or ō, thus agreeing with the vernacular Hindōstānī of the upper Dōāb and with Panj¯bī, both of which owe it to the influence of the Outer languages.

^{page 54 note 1} Such, for instance, as the plural of the personal pronouns.

^{page 55 note 1} This word has nothing to do with the word Lahndā, which means “West”.

^{page 55 note 2} See Grierson, “The Modern Indo-Aryan Alphabets of North-western India”: JRAS., 1904, 67.

^{page 55 note 3} See Grierson, JRAS., 1911, 302.

^{page 56 note 1} These were first noted by T. Grahame Bailey. See his Panjābī Grammar as spoken in the Wazīrābād District, Lahore, 1904. For particulars, see § 152 below. I believe that no one has hitherto noted that the Vedic udātta corresponds to a Tibeto-Chinese “high” tone, while the visarga corresponds to the “entering” or “abrupt” tone, like it, also, being the result of the partial or total elision of a final consonant.

^{page 56 note 2} The differentiation of Gujarātī from the Mārwārī dialect of Rājasthānī is quite modern. There is a poem by Padmanābha of Jhālōr, a town only 80 miles from Jōdhpur, the capital of Mārwār, entitled the Kānhadadēva-prabandha. It was written in 1455–6 A.D. At the beginning of the year 1912 there was a lively controversy in Gujarāt as to whether it was in Gujarātī or Mārwarī. Really it is in neither, but is in the mother language, which in later years differentiated into these two forms of speech. Cf. Tessitori, JRAS., 1913, p. 553, and his “Notes on the Grammar of Old Western Rājasthānī, with special reference to Apabhraṁśa and to Gujarātī and Mārwārī”, in IA. xliii-v (1914–16), reprinted in one volume, Bombay, 1916.

^{page 57 note 1} V. Smith, JRAS., 1908, 768.

^{page 57 note 2} Id., 1908, 789; 1909, 56.

^{page 57 note 3} Tod, Rajaathan, Annals of Mewar, ch. ii.

^{page 57 note 4} Ib., Annals of Amber, ch. i.

^{page 57 note 5} Ib., History of the Rajput Tribes, ch. viii.

^{page 57 note 6} Ib., Annals of Mewar, ch. i.

^{page 57 note 7} V. Smith, Early History of India ², 377. See also the following: Tod, Rajasthan, Introduction; Elliot, Memoirs on the History, Folklore, and Distribution of the Races of the North- Western Provinces of India, ed. Beames, i, 99 ff. and index; Ibbetson, Outlines of Indian Ethnography, 262; Jackson, in Gazetteer of the Bombay Presidency, i, pt. i, app. iii, Account of Bhīnmäl, esp. pp. 463 ff.; V. Smith, “The Gurjaras of Rājputānā and Kanauj,” JRAS., 1909, 53 ff.; D. R. Bhandarkar, “Foreign Elements in the Hindü Population,” IA. xl (1911), 7 ff., esp. 21 ff.

^{page 57 note 8} V. Smith, Early History of India ³, 412, and IA. xl (1911), 86.

^{page 57 note 9} D. R. Bhandarkar, JASB. V(N.S.), 1909, 185. Cf. contra, Mohanlal Vishnulal Pandia, ib. viii (N.S.), 1912, 63 ff.

^{page 58 note 1} Sachau's tr., i, 202. Cf. D. R. Bhandarkar in IA. xl, quoted above.

^{page 58 note 2} Pth. kuttsī, G. kutśē, he will strike.

^{page 59 note 1} According to Nagēndranātha Vasu, JASB. lxv, pt. i, 1896, 114ff., these Brāhmans gave their name to the alphabet. In Al-Birūnī's time the Nāgara alphabet was used in Mālwā, which is close to Gujarāt. Sachau's Eng. tr., i, 173.

^{page 59 note 2} It is worth noting in this connexion that Old Mārwārī in some respects agrees with Kāšmīrī, e.g. in possessing a genitive postposition handō.

^{page 59 note 3} A Bardic and Historical Survey of Rājputānā has lately been set on foot by the Government of India, under the superintendence of the Asiatic Society of Bengal. It is in charge of Dr. L. P. Tessitori, who has already discovered a number of important works. See JASB. xiii (N.S. ), 1917, pp. 195 ff., and reports and texts published, under Tessitori's editorship, by the ASB.

^{page 60 note 1} Eastern Pahārī, as an independent language, is of very modern origin, the Indo-Aryan migration from the west into Nēpāl dating only from the sixteenth century A. D. The language is strongly influenced by the surrounding Tibeto-Burman dialects, and has changed considerably within living memory. It appears to have superseded another Indo-Aryan language akin to the Maithilī dialect of Bihārī, now spoken immediately to the south of Nēpāl. A specimen of this old dialect was published by Conrady in 1891. It is a drama, entitled the Hariścandranrtya.

^{page 61 note 1} The whole question is worked out in detail in vol. ix, pt. iv, of the LSI. dealing with Pahārī. It is impossible here to give more than the general results and a few of the principal references. Those desiring the full proof must refer to the volume of the LSI.

^{page 61 note 2} Cf. Cunningham, “Archaeological Survey Reports,” xiv, 125 ff.; Ibbetson, Outlines of Punjāb Ethnography, 268; Atkinson, Himalayan Districts of the North-Western Provinces of India, ii, 268–70, 375–81, 439–42, and index; Stein, tr. Rājataraṅginī, note to i, 317; ii, 430, and index.

^{page 62 note 1} In the Śatapatha Bráhmaṇa (I, vii, iii, 8), the Báhīkas, with whom the Khaśas are associated in Mahábhárata, viii, 2055 ff., are still within the pale, and worshippers of Agni.

^{page 62 note 2} Cf. Visnu Purāna (Wilson-Hall), I, xxi; Bhg. P., II, iv, 18; IX, xx, 29; Mārk. P., lvii, 56; Manu, x, 44; Bharatanātyaśāstra, xvii, 52.

^{page 62 note 3} Cf. Imperial Gazetteer of India (1907), i, 386.

^{page 62 note 4} Hodgson, “Origin and Classification of the Military Tribes of Nepal,” JASB. ii (1S33), 217 ff.; Vansittart, “The Tribes, Clans, and Castes of Nepal,” JASB. lxiii (1894), pt. i, 213 ff.; S. Lévi, Le Népal, i, 257 ff., 261–7, 276 ff.; ii, 216 ff., and index.

^{page 62 note 5} Such are the tendency to drop an initial aspirate (ōṇū for hōṇū, to become); to disaspirate sonant aspirates (bāī for bhāī, brother); to harden sonants (jawāp for jawāb, an answer; ōkhatī for ōkhadhī, medicine); to change c to ts and j to z (tsazarō for cajarō, good); to change t to ts (khēls for khēt, a field); to dropmedial r (katā for kartā, doing); to change a sibilant to χ (χuṇnā for śuṇnā, tohear), or to h (brās or brāh, a rhododendron); and many others.

^{page 63 note 1} Tod, Rajasthan, introduction; Elliot, Memoirs, etc., as quoted above, i, 99, and, index; Ibbetson, op. cit., 262 ff.; Jackson, Gazetteer, as above, i, 463; V. Smith, The Gurjaras, etc., as above, 53 ff.; “The Outliers of Rājasthānī,“ IA. xl (1911), 85 ff.; D. R. Bhandarkar, “Foreign Elements in the Hindū Population,” IA. xl (1911), 7ff., esp. 21 ff.

^{page 63 note 2} It is worth noting that the Rājā of Garhwāl claims descent from Kaniṣka, who is said to have come to Gaṛhwāl from Gujarāt or Western Rājputānā; Atkinson, op. cit., 449.

^{page 63 note 3} I have not considered here the question of Western Rājasthānī and Gujarātī. Gujarāt may well have been conquered by Gurjara tribes coming from the northwest. The Western Rājpūts had their centre of dispersion near Mt. Ābū, but whether the Gurjaras of Ābū came from the east or from the west I cannot say. All that can be said is that the agreement between Western Pahārī and Western Rājasthānī is very striking.

^{page 63 note 4} In Outliers, etc., as above.

^{page 64 note 1} See LSI. vi, 2 ff.

^{page 66 note 1} Grierson, ZDMG. lxvi, 75.

^{page 66 note 2} Pischel, Pr. Gr., 27, 28; Grierson, JRAS. 1902, 47.

^{page 67 note 1} Such are, e.g. the existence of a broad å, sounded like the a in “all”; the change of ai to ä; of k to c, and of c to s; the frequent confusion between dentals and cerebrals; an oblique case in ā; and a past participle formed with the letter l.

^{page 68 note 1} See B. A. Gupte, IA. xxxiv (1905), 27.

^{page 69 note 1} See Turner, “The Indo-Germanic Accent in Marāthi”: JRAS. 1916, 203 ff.

^{page 69 note 2} These are the change of s to ś, and the termination ē of the nominative of a-bases. In writing at the present day, ś is invariably written for both ś and s, though in modern times the pronunciation is s, not ś. The change of pronunciation is due to political reasons. See Languages of India, 72. In Bengali the ś-sound is retained. In Old Bihārī poetry, when, for metrical reasons, it is necessary to lengthen the final vowel of the nominative singular, this is done by making the word end in ē. Thus Vidyāpati Thakkura (A.D. 1400) has sinānē for snānam, paragāśē for prakāśah, pārē for pāram, dhīrē for dhīram, and hundreds of others, which will be found in any edition of the poet's works. In Hindī poetry such words would end in an, not in ē. The Old Eastern Hindi of Tulasī Dāsa, corresponding to Ardhamāgadhī Prakrit, occupies an intermediate position, and uses both ū (for an) and ē, as in parivārū for parivārah, and sanāyē for sajñānaḥ. It should be noted that both these ē and ū terminations are used indifferently both for the nominative and for the accusative, thus following the example of Apabhraṁśa, in which (Pischel, Grammatik der Prakrit Sprachen, p. 247) the accusative has the same form as the nominative.

^{page 70 note 1} Piachel (Prakrit Grammatik, p. 25) considered that there is no connexion between Magahl and Māgadhī Prakrit. With all respect for this great scholar, I am unable to agree with him on this point.

^{page 70 note 2} The dialect is named from the ancient town of Bhōjpur, on the southern bank of the Ganges, in the District of Shāhābād. For the history of Bhōjpur and its traditional connexion with the famous Bhōja of Mālwā, see Shāhābād Gazetteer (1906), 132. For an account of the character of the Bhojpurīs, see ib. 21.

^{page 71 note 1} This is an English word, derived from “Bengal”. The Indian name is Bãglā or Baṅgabhāṣā.

^{page 71 note 2} See Dinēś Candra Sēn, History of Bengali Language and Literature, Calcutta, 1911.

^{page 72 note 1} Such words can be found on every page of OBg. poetry.

^{page 73 note 1} See Grierson, “Piśāca = ’Ωμοφ⋯γος,” JRAS. 1905, 285 ff.; The Piśāca Languages of North-Western India, Introduction; “Piśācas in the Mahābhārata” in Festschrift für Vilhelm Thomsen (1912), 138 ff.; on the other hand, Konow, “The Home of Paiśācī,“ ZDMG. lxiv, 112 ff., maintains that Piśāca Prakrit was an Aryan language as spoken by Dravidians of Central India. The whole subject is again discussed in Grierson, “Paiśācī, Piśācas, and Modern Paiśāci,” ZDMG. lxvi, 49 ff. Paiśācī Prakrit and the Pāli of the Buddhist scriptures have much in common, and my own opinion is that the latter was originally a kind of literary lingua franca, based on Māgadhī Prakrit, which developed in the great university of Takṣaśilā, situated in the heart of Këkaya, the nidus of the former. Its development is exactly paralleled by that of literary Hindï, the original home of which was Delhi, but which took its present form in Benares far to the East. See my “Home of Literary Pāli” in R. G. Bhandarkar Commemoration Volume, 117 ff.

^{page 73 note 2} Cf. the Kaikëya and Vrācada Paiśācī Prakrits of the Indus Valley (Pischel, Pr. Gr., 27).

^{page 73 note 3} ZDMG. lxvi, 76, 77.

^{page 73 note 4} He is, however, contradicted by Mārkaṇḍēya, xix, 9, in which some words are quoted from the Brhatkathā, the work supposed to be He.'s authority, as examples of Kēkaya-paiśācikī, i.e. of the Paiśāci of North-Western India.

^{page 74 note 1} Such e.g. are the very un-Indian treatment of the letter r; the change of ⋅m and sm to ś and s, respectively, of ty and tm to t, and of t to l or r; the not infrequent retention of intervocalic consonants and hardening of sonant consonants; a weak sense of the difference between cerebrals and dentals; the tendency to aspirate a final surd; the frequent palatalization of gutturals, cerebrals, dentals, and l; and the regular retention of a short vowel before a simplified double consonant.

^{page 74 note 2} e.g. the treatment of the vowels; the non-development of cerebral letters; the preservation of numerous consonantal compounds; the change of d to l, of dv to d, and of ṣk (šk) to c.

^{page 74 note 3} JRAS. 1911, 45. I differ here, see ib. 195.

^{page 74 note 4} e.g. the Avesta change of sm to hm, and the preservation of s.

^{page 74 note 5} Wilson Philological Lectures on Sanskrit and the Derived Languages, 94.

^{page 75 note 1} Pīśāca Languages, 5.

^{page 75 note 2} Pīśāca Languages, 20.

^{page 76 note 1} LSI. II, i, 34. The only parallel that I have been able to find in an Oriental language is the Chinese sound, which in Southern Mandarin is pronounced like an English r, but in Pekin as ž (Mateer, xviii).

^{page 76 note 2} e.g. the aspiration of a final surd, the change of ṅg to n, and the elision of medial m.

^{page 77 note 1} JASB. vii, 783.

^{page 77 note 2} Tribes of the Hindoo Koosh, cxvi.

^{page 77 note 3} e.g. there can be little doubt but that they owe the presence of the cerebral n to the influence of Paštō

^{page 77 note 4} See E. Kuhn, Die Verwandtschaftsverhältnisse der Hinduktish Dialekte, in Album Kern, 29 ff.

^{page 77 note 5} Published in 1895. This work would have been more valuable if the author had consulted his predecessors, Biddulph and Leitner.

^{page 78 note 1} Mr. Grahame Bailey informs me that the word is pronounced with a cerebral ṇ and with the accent on the last syllable. The presence of the cerebral n is surprising, as I have never come across that letter either in the language itself or in the closely related Kāśmīrī.

^{page 78 note 2} See also, for important information regarding Brōkpā, or Highland, dialects, Shaw, “Stray Arians in Tibet,” JASB. xlvii, pt, i, 26 ff.

^{page 78 note 3} McCrindle, Ancient India as described in Classical Literature, 51.

^{page 79 note 1} Published by the ASB., under the editorship of the present writer, in 1898.

^{page 80 note 1} This was first shown by Leitner in The Bashgeli Kafirs and their Language, reprinted from the Journal of the United Service Institution of India, No. 43, Lahore, June 10, 1880.