PREFACE XV person has got a standarised spelling (as in Chinsurah, Howrah, and Burdwan), I have thought it fit to retain it; but in all other cases, the transliteration is done in the mode indicated with the only exception of using cha for 6. The words are, however, always rendered, not phonetically, but. according to the recognised spelling, although widest possible divergence exists between the historical spelling and the actual pronunciation of Bengali words. I have, therefore, always rendered 4 by ya, whether occuring singly or in compound letter, 4 by na distinguishing them respec- tively, although they are not so distinguished in pronun- ciation, from & (ja) and 4 (za). Similarly the three consonants *, 4, and ¥ are distinguished by different signs (8, s and s) although they not often thus discriminated in pronunciation. The same remark applies to compound letters: I have rendered, as in Sanscrit, by ksa, & by ja, and so forth. Partly on account of this divergence between spelling and pronunciation, which makes it impossible to apply Sanscritic transliteration in fofo to the case of the living vernacular, I have been forced to make one or two important exceptions. I have not distinguished between a (va) and 4 (4a), for this distinction is hardly recognised in Bengali, either in spelling or pronunciation; I have therefore used 4a indiscrinimately for them. The final & (a) presents some difficulty, for very often it is passed over in pronunciation. We write a7" (Nila-darpana) but we read it as #7" (Nil-darpan). In these cases, I have generally dropped the & (a). This, on the whole, is not a very satisfactory method; but in the absence of a better one, I have tentatively followed it here, leaving the whole question, which is indeed one of great practical importance, to the consideration of expert scholars. In the task of collecting materials for the present volume, I have met with considerable difficulties known
