oct. — mak. 1858-59.] for the Indian characters. 257 



yet met with are to be found in the large government maps of 

 India. In that part of the map which includes the district of Tin- 

 nevelly, the Tamil word for " tank," a simple dissyllable, is writ- 

 ten in thirteen different ways ! Though there is much confusion 

 at present, owing to the absence of an authorised system, it by 

 no means follows that this confusion would continue after a sys- 

 tem had been fully considered and definitively introduced, and 

 people generally had become aware of the existence of the system. 



Even at present, there is no difference of importance amongst 

 Orientalists respecting the way in which Indian words should be 

 written. As regards the letters by which the Indian vowels should 

 be represented, which is the most important point of all, English, 

 German, and French Orientalists are now perfectly agreed. The 

 points respecting which differences of system still exist are of 

 small moment and might easily be settled. Still I fully admit 

 that they require consideration, for any attempt to introduce into 

 general use a system which had not been thoroughly thought out 

 and digested in all its details, would probably end in failure and 

 throw the scheme back another 30 years. 



I will refer hereafter to minor details, but I may here mention 

 some general principles which will require to be considered. (1) A 

 system which takes the Sanskrit alone into account and contents 

 itself with meeting the wants of the Sanskrit, will not suit the pur- 

 pose in view, which is to apply the Roman character to all the 

 languages of India. The wants of the Hindustani and of the 

 Dravidian languages will also have to be provided for. For in- 

 stance, Sanskrit is destitute of the short sounds of e and o. Con- 

 sequently Sanskrit scholars, who distinguish the other long vowels 

 by accents, leave these two long vowels without any distinguish- 

 ing sign. This would, however, introduce confusion into the Dra- 

 vidian languages, which have short forms of e and o, as well as 

 long. It will be necessary, therefore, to deal with the Indian lan- 

 guages as a whole, in order that there may be but one system in use 

 everywhere, and also to study separately the peculiar necessities of 

 each of the Indian languages, in order that the peculiarities of 

 each may be taken into account in the general arrangement. 

 (2). Another difficulty which must be taken into consideration re- 



