BUTTERFLIES OF SAMOA AND SOME NEIGHBOURING ISLAND-GROUPS. 33 



The difference in the proportions of the sexes of H. bolina in Western 

 and in American Samoa is extremely interesting ; in almost all the Pacific Islands 

 males are much commoner than females. Tonga and Tutuila conform to this 

 general rule, while Simmonds has shown that in certain parts of Fiji the 

 females preponderate. In no part of its range, however, does the disproportion 

 between the sexes seem to be so great as in Western Samoa, where the males 

 are much less than 1 per cent, of the total ; this is the more remarkable, in 

 that there do not appear to be any differences in conditions between the two 

 parts of Samoa to account for this disparity in sex-proportions. At the nearest 

 point Upolu and Tutuila are only about 40 miles apart, and the climates seem 

 to be as nearly as possible identical. Although I did not find the early stages 

 in Tutuila, Sida (which it is safe to assume is, as in Fiji, Tonga and Western 

 Samoa, the food-plant) is common. Nor does the lack of males in Western 

 Samoa appear to be in any way disadvantageous to the species, for here, as 

 in the other islands, H. bolina is one of the commonest insects. Another 

 apparently inexplicable point is the cause of the difference in size of H. bolina 

 in the two parts of Samoa ; the average expanse of females in Tutuila is 81 mm., 

 and of the same sex in Upolu or Savai'i only 7 mm. The small size of H. bolina 

 in Western Samoa finds a parallel in the case of several other species of butterflies, 

 including D. melissa melittula and P. villida samoensis, and it has been suggested 

 by Poulton that this general feature of small size in Western Samoan butterflies 

 may be correlated with the occurrence of hurricanes. In Tonga, however, 

 where the butterflies are not noticeably small, hurricanes are far more frequent 

 than in Samoa, and most of the islands, being low, are much more exposed to 

 the force of the wind than are the mountainous islands of Samoa. Moreover, 

 as pointed out above, the small size of the butterflies of Upolu and Savai'i is 

 not shared by the same species in Tutuila, where conditions as regards 

 hurricanes are identical. This latter point has been entirely obscured by the 

 fact that hardly any of the older Samoan records give exact localities, so that 

 specimens from all the islands in the group are mixed in collections. 



in. 1 



3 



