638 Mr. C. F. M. Swynnerton on the 



witnessed a good many attacks on butterflies by several such 

 apparent " specialists " at times when their specialty happened 

 to be present in less abundance. 



5. Were I to say that during the past year, the first during 

 which I have paid special attention to this particular point, 

 small Diptera, excluding house-flies, have been on the average 

 five hundred times as numerous on Mount Chiriuda as but- 

 terflies, I should^ I firmly believe, be guilty of a very con- 

 siderable understatement of what has actually been the case. 

 But, to be well on the safe side, I will place it at that figure. 

 The law of probabilities, then, would lead us to expect that 

 for every butterfly found by a collector during the past 

 year in the stomachs of our local birds, he ought, other 

 things being equal, to have found five hundred Diptera. 

 This argument might well be extended to embrace some 

 of the harder Orders. Tlius grasshoppers are here several 

 times more numerous than butterflies all the year round. 

 The same may be said of beetles during at least a few 

 months of the year. And Hymenoptera, including as they 

 do both ants and honey-bees, are vastly more numerous 

 here than the diurnal Lepidoptera. 



G. Although I did not use Diptera in the special 

 experiments I have referred to, I have on several other 

 occasions taken their wings whole and undamaged from 

 jiellets and excreta in which the wings of Lepidoptera had 

 been reduced to the minutely fragmentary condition described 

 above. 



7, I have noted the pellet-habit on the part, I believe, of 

 every carnivorous or purely insectivorous bird that I have 

 kept in captivity : Owls, Hawks, Passerines, large and small, 

 and Picarians. More or less frequent in any case, it seen)s 

 to come into play most when food is so continuously 

 abundant that, if the bird is to use that abundance to the 

 very best advantage, the intestines must be specially aided to 

 get rid of the niasses of chitin by which it is so commonly 

 accom j)anied. It seems likely, therefore, that where in nature 

 the food-supply is both abundant and comparatively uninter- 

 rupted, it may be no unusual thing for several pellets to be 



