No. 43 —1892.] ETHNOLOGY OF CEYLON. 



231 



of sculpture at Dondra and the difficulty of finding Sinhalese 

 derivations for the names of many towns or villages on the 

 western coast, a difficulty which is solved by adopting 

 Dravidian derivations. 



In a visit in England to Triibner's establishment at Ludgate 

 Hill I had an opportunity of comparing the printed characters 

 of Telugu, Canarese, and Malayalam, which all showed a 

 greater affinity to the Sinhalese than to the Tamil printed 

 forms. Compare, for instance, the printed texts of Bibles 

 in Tulu, Konkani, and Canarese. 



I had also been struck by the fact that the people along 

 the western coast and throughout the Southern Province, 

 upon inquiry after any offender who had absconded to the 

 Kandyan Provinces, always reported that he had run away 

 to " Sinhala," or the Sinhalese country, thus distinguishing 

 themselves from the Sinhalese. 



There has been no complete investigation, no careful 

 comparison, of costume, language, customs, and physical 

 resemblances of the Sinhalese with the Dravidian races of 

 South India. The present Paper, therefore, does not 

 pretend to be a complete treatise. It is strictly a sugges- 

 tive Paper, which seems to point to the conclusion which 

 will at no distant date be arrived at and proved by others 

 more competent to do so than myself. At present that 

 conclusion is not so clearly applicable to the inhabitants of 

 the central region of Ceylon as it is to the Southern and 

 Western coasts of the island and to the Northern and 

 Eastern Provinces. 



In the Maldives we have a race, in their physical char- 

 acteristics, language, and former practice of demon worship, 

 identical with the people of the southern coasts of Ceylon, 

 and they may be assumed to be identical with the natives 

 of Lanka and the subjects of Ravana, with a possible con- 

 nection with such a race as the Shanars. 



When we come to investigate the Yijayan period, the 

 theory of an Aryan descent gives way for want of evidence 

 in favour of a Dravidian origin, but this last does not shut 



