THE DISTRIBUTION OF BUTTERFLIES IN INDIA. 597 



that, and why were the lemurs practically the only creatures to make 

 the trip ? Certainly as far as butterflies are concerned there is nothing 

 to suggest such a connection. 



You would have thought Dr. Wallace had knocked the bottom out of 

 this kind of reasoning long ago, but he hasn't, for I see the Indian geolo- 

 gical surveyors still take Lemuria for granted, and quite recently were 

 excited about another continent cutting across it joining Trichinopoly 

 to Patagonia ! 



Dr. Wallace lays down a general principle that we are not justified 

 in supposing any part of the earth's surface, which is now over 1,000 

 fathoms below the sea, was formerly dry land, and certainly practically all 

 the facts are easily explained without going beyond the 500-fathcm line. 



To return to our Indian island the chief fact about its geology is the 

 antiquity and regularity of its formation. Outside Gondwana, where 

 there were slight changes of level in early times causing some alteration 

 of the principal water sheds, there was practically no alteration (and no 

 evidence that any part had ever been submerged under the sea except 

 small areas near the mouths of the Nerbudda and Cauvery) until the end 

 of the cretaceous epoch when a gigantic upheaval of volcanic or trap rock 

 took place. There are small areas of this rock in Assam, Chota Nagpur 

 and Sind, but these are quite separate from, and not necessarily contempo- 

 raneous with, the main overflow which covered Gujerat, Malwa, Berar and 

 the Bombay and Hyderabad Deccan as well as reaching about 200 miles 

 to the west over what is now sea. This stupendous eruption, lasting at 

 intervals for perhaps thousands of years, of course completely blotted out 

 the fauna and flora of the N.-W. third of the island and is sufficient to 

 account for the poverty of the Konkan region as compared with Malabar 

 to this day. 



Now to come to the insects themselves. Butterflies are divided into 

 7 families, and I propose to take each in order. (In doing so I ought to 

 premise that I am only giving the number of forms which I consider 

 truly entitled to be considered species ignoring many described varieties 

 which do not warrant the distinction, and I have treated a whole host of 

 ephemeral generic names which have no characters to justify them, in 

 the same way.) 



Of the first family, the Nymplialidm, we have 71 species, of which 10 

 are peculiar to the region. One of these belong to a distinct genus of its 

 own and is apparently confined to Travancore. 



