STATISTICAL SURVEY OF INDIA. 319 



spelling of these names, and his plan was eventually adopted for 

 the gazetteer and for general use. No practicable scheme cotild also 

 combine absolute precision. The Roman alphabet has, for instance, 

 but one letter for the consonant n ; the Sanscrit has four letters for 

 it in its various modifications as a dental, lingual, palatal, and guttural. 

 Again, the Indian alphabet has two separate letters for d, 

 two for t, and three for s and sh. Indian names could, therefore, be 

 represented only in an approximate manner in our tongue without 

 the manufacture of a new Roman alphabet, with additional letters, 

 by means of accents over the vowels, dots under the consonants, 

 italics, or the like. In the system laid down by Sir William Hunter 

 and adopted by the Government of India, clotted consonants are 

 rejected, as few accents as possible over the vowels are used, and, 

 generally speaking, everything is avoided which would give the 

 alphabet an un-English look. Moreover, names of important 

 places which had attained a historical or literary fixity of spelling 

 were retained in their popular form, such as Calcutta, Madras, 

 and Bombay. The method did not attempt to reproduce such 

 fine distinctions as the four Sanscrit oi'a or such consonants as the 

 dental and lingual t and d. But a uniform value was assigned to 

 each vowel, namely, a and u as in rural ; e as in grey, mediant ; and 

 i and o as in police. The accented d, i, and u represented the long 

 forms of the same vowels in Sanscrit, or the sounds in the English 

 far, jyier, and lure. 



The process of adoption of the uniform spelling of the geographical 

 names in all Indian Government publications has necessarily been of 

 slow growth. But the existence of a standard work of reference like 

 Sir "William Hunter's " Imperial Gazetteer " has been a great 

 step towards uniformity of spelling, and all new maps issued by the 

 Surveyor-General's department follow the prescribed spelling, so in 

 process of time we may not despair of seeing absolute uniformity 

 attained wherever India is mentioned. 



The second edition of the" Imperial Gazetteer," produced in 1885, 

 took as its starting point the census of 1881, which was also the 

 first complete and fairly synchronous census of India. Its adminis- 

 trative statistics refer chiefly to the years 1882-84, but in certain of 

 the larger questions dealt with the facts were brought down to 

 1885. 



