particularly the one regarding the Marathas. He has called the Marathas descendants of the Shakas or the Scythians. I think Risley is mistaken in his surmise. Moreover I do not approve of affiliating a well-known race like the Marathas to a race like the Scythians of which we know very little regarding most essential points for ethnology.
For example: It would not be difficult to settle whether the present Marathas are more closely allied to the Caucasians or Mongolians. On the other hand, whether the Scythians were Caucasians or Mongolians is yet a matter for controversy. The only advantage one can find in using the term "Scythians" is that this word is more known to the Europeans, and is found in the classical works; and to adopt that word would be in conformity with that usage which prefers terms like the Teutonic race and the Latin race. But we should remember that accuracy is of greater importance than uniformity.
The words "Aryan" and "Dravidian" also are very dangerous to use. I would much prefer geographical terms, or terms based on the present races or tribes.
To sum up my view in short, I approve of seven divisions of India which Risley has made, but I do not approve of the names he has given, and I regard his attempt to discover the ancestry of the races as one very imperfect.
My view of the question of races in India is as follows: All the people of India to-day with the probable exception of the Bengalese and the northeastern frontier tribes are peoples of the Caucasian race. I consider the Dravidians as a branch of the Caucasians.
