PROBLEMS IN INDIAN LANGUAGES 121 



awaits solution on the part of those who are thorough Chinese as 

 well as Sanskrit scholars. It will consist in ascertaining on scientific 

 principles the phonetic laws according to which, in different centuries 

 and in different dialects, the Chinese language has reproduced the 

 corresponding Sanskrit sounds. Mr. Nanjio, the Buddhist scholar 

 I have already mentioned, and another Japanese who studied San- 

 skrit for three years at Oxford, and is now Professor of Sanskrit in 

 the University of Tokyo, are at present engaged on investigations of 

 this character. I hope that in the course of two or three years the 

 results of their labors will be published and materially advance our 

 knowledge of the history of India down to about 700 A.D. 



We now come to the third period of the languages descended from 

 the earliest form of Sanskrit, the beginning of which about syn- 

 chronizes with the Muhammadan invasion of India and with the 

 conquest of England by the Normans in the eleventh century. Down 

 to the end of the second period the Prakrits, though phonetically 

 and inflectionally much worn down, were still synthetic languages. 

 But from the eleventh century onwards we find that the tertiary 

 Prakrits, the literatures of which date from the thirteenth and later 

 centuries, have assumed an analytic character, or in other words, 

 replace inflection by the use of prepositions and periphrastic forms, 

 much as modern English has done in comparison with Anglo-Saxon. 



At the present day these Indo- Aryan tongues, spoken, as I have 

 already said, by 220,000,000 of people, consist of nine main lan- 

 guages. The most numerously spoken is Hindi, with 63,000,000; 

 then comes Bengali with 45,000,000; Bihari with 35,000,000; Oriya, 

 Rajasthani, Gujarati, with about 10,000,000 each; Marat hi with 

 18,000,000; Panjabi with 17,000,000; and the group of which Sindhi 

 is the principal tongue, with 8,000,000. 



A comparative grammar of the chief languages was written as 

 much as thirty years ago, and has proved a useful pioneering work; 

 but it is no longer up to the knowledge or scientific standard of the 

 present day. One of the main problems in the study of modern 

 Aryan languages of India is the production of a thoroughly scientific 

 comparative grammar based on a more scholarly investigation of the 

 individual languages than has hitherto been made. Grammars and 

 dictionaries of all the principal languages have been compiled, but 

 most of them, though often of much practical value, are the work 

 of untrained scholars and therefore leave a good deal to be desired as 

 a basis of research. They deal, moreover, for the most part, only 

 with the literary form of the language. The non-literary dialects of 

 the uneducated, which are linguistically of great importance, have 

 been hitherto almost entirely neglected, and thus offer practically 

 a virgin field to the philologist. They are all the more important 

 owing to the extreme lengths to which the introduction of Sanskrit 



