VALIDITY OF SOME FORMS OF MIMICRY. 747 



no time would an insectivorous bird, or i"eptile, find any difficulty 

 in procuring its daily sustenance. 



Bourbon and Mauritius are very largely under sugar cultivation, 

 and this necessitates constant manuring of the fields, with a 

 consequent abundance or superabundance of flies of all sorts 

 throughout the year. The rainfall, though greater in the wet 

 season, is not infrequent duiing the dry season, and this also 

 favours insect life ; and if we add to this the consideration of 

 the practical absence of butterfly-eating birds and i-eptiles, we 

 can estimate the difficulty of believing that these changes are the 

 outcome of natural selection in these two islands. 



In Ceylon Mr. E. Ernest Green, who has lived thirty years in 

 the island and who knows every part of it, writes as follows : — 



" Though insects are more abundant at certain seasons, I have 

 never experienced anj' part of Ceylon where there was anything 

 approaching a dearth of them. I know that I am always busy 

 pinning and setting throughout the year. 



" I sometimes wish that there ivas a short dead season, when one 

 could devote oneself to other work without being distracted Vjy the 

 constant accumulation of material. 



" I doubt if Melanitis is ever subject to much worry from birds. 

 It lives in the shade and never moves during the daytime, unless 

 flushed by some big animal. I am now receiving (16.8.10) both 

 dry and wet season forms of M. f.amhra from Kandy." 



An allied species Melanitis leda occurs also in Bourbon and 

 Mauritius, and it is to be remarked that the dry season forms 

 begin to appear before the advent of the dry season, that is to say 

 before any form of stress would tend to make itself felt. 



In explanation, it may be suggested that the butterfly Avas in- 

 troduced from the locality where natural selection pi'oduced these 

 changes and tlmt it is simply carrying on an inherited tendency. 

 That it is an introduced species is highly probalile, but it has been 

 known to entomologists in Bourbon and Mauritius for at least 

 sixty years, and it difiers in no way now than in the time of 

 Boisduval. It is difficult to believe that the factor which pro- 

 duced this cryptic defence being removed and no longer required 

 would not have led to some other form of colouring, or a return 

 to that ancestral type from which these forms were evolved. The 

 above remarks apply equally to Mycalesis narcissus. Precis rhad^ivia 

 (introduced 1858), and Terias Jioricola, and I have made a further 

 study of Terias hecahe in Ceylon. It is veiy frequently the case 

 that the wet form continues to appear well into the dry weather 

 and vice versa, but to a less extent ; this has been remarked on 

 fi-equently, but so far as I know no exact observations have been 

 made. In Colombo there was no I'ain from November 19th till 

 December 10th, 1908, thence to January 6th, 1909, -70 of an inch, 

 but of this no less than •57 fell on one day (Dec. 19th) ; such an 

 absence of rain in a tropical country at once causes a general drying 

 up of vegetation and the assumption of diy weather conditions. 

 At weekly intervals I captured all the Terias I could, which were 



