^{page 5 note 1} ‘Notes on the grammar of OWR, with special reference to Apabhramśa and to Gujarati and Marwari’, Indian Antiquary, XLIII–XLV, 1914–1916.Google Scholar

^{page 5 note 2} ‘ Gujarātī phonology’, JRAS, 1921, 329 ff., 505 ff.

^{page 5 note 3} Turner, op. cit., 333.

^{page 5 note 4} The phonetic observations of Chatterji, S.K. in his Rājasthānī bhāsā (Udaipur, 1949)Google Scholar are unfortunately inaccurate and of little value.

^{page 5 note 5} ‘Aspiration in the Hārautī nominal’, Studies in linguistic analysis (forthcoming special volume of the Philological Society).

^{page 5 note 6} ibid., § 29.

^{page 6 note 1} Sometimes written as cf. Sitaram Lalas, Rājasthānī ryākaran, p. 14 and in Marū-bhāratī, IV, 3, 62 ff.

The phonetic developments here discussed provide an example of the ‘empty space’ principle in historical structural phonology (cf. Martinet, Économie des changements phonétiques, ch. 3). Thus in SMe. ٭c(h) > S/٭S > h/٭h > ɦ; in NWMā. the first of these developments is not yet complete. One may compare the Greek development of ٭k^w{i) ɦ r/٭t(i) ɦ σ/٭s(V) ɦ h.

^{page 7 note 1} Beside the alternative forms Me. kərεla, -ga, Mā. kərεla, H. kərзgO, -lo (f. gi, -Ii).

^{page 7 note 2} cf. Turner, , ‘The Sindhi recursives or voiced stops preceded by glottal closure ’, BSOS, III, 2, 1924,Google Scholar 301 ff.; Bailey, T. Grahame, ‘The Sindhi implosives ’, BSOS, II, 4, 1923, 835.CrossRefGoogle Scholar

^{page 7 note 3} is written as , [b] as and [U] (non-initial only) as : cf. Rājasthān, Vol. I), §§ 15–19. The non-implosive [d], etc. may be written as etc: cf. Lalas, Vyākaran,, loc. cit. (), and in Marā-bhāratī, IV, 3, 66 f.

^{page 7 note 4} For such an exception as H. bhəlo ‘good’ see Turner, , BSOS, VIII, 2–3,1936, 809;Google Scholar Nep. dict, s.v. bhalo; cf. Tessitori, , IA, XLIII, 1914,Google Scholar 87 ff

^{page 8 note 1} For which Mā. has the special symbol is written as or .

^{page 8 note 2} Giving rise to similar reflexes in regard to vowel-length as are noted in (i) (a) above: cf. Turner, ‘Middle Indian -d- and -dd-‘, Festgabe Jacobi, 34 ff.; BSOS, VIII, 2–3, 1936, 809.Google Scholar

^{page 8 note 3} cf. 1 (i) above.

^{page 8 note 4} LSI, IX, ii, 35 (cf. On the modern Indo-Aryan vernaculars, 101).

^{page 8 note 5} cf. , § 8

^{page 9 note 1} cf. ‘Aspiration in the Hārautī nominal’, § 22 n.

^{page 9 note 2} A phonetically similar articulation is found in Me. as a result of the vowel-lengthening (of [ə]) described in 1 (iv) (b) above. Corresponding to H. chз ‘6’ also, phonetically similar forms are found in both Me. and NWMā. (but Jais. [Chɔ], Bik. [chəo]).

^{page 9 note 3} cf., for a similar state of affairs in MHG, Fourquet, Word, VIII, 2, 1952, 133.Google Scholar

^{page 10 note 1} cf. § 38 Similar considerations apply in regard to the indication of aspiration: thus khato ‘catechu’ may be found written either as , or as (katho) after Hi. kəttha.