1889.] Dohovtj— Notes on Assam Butterflies. 119 



chief difficulty with this and other species resulted from my constant 

 change of abode. The dimorphism of many other species yet remains 

 to be demonstrated by breeding. But in these cases, the evidence of 

 the prehensores, in itself irresistible, added to the thousand signs and 

 tokens of identity obvious to any unbiassed field-naturalist, settles 

 beforehand what the results will be. 



Another group in which much may be done by breeding is that of 

 the Euphms included by Mr. Butler in Salpinx. Mr. Moore has de- 

 scribed a number of Assam species forming the majority of his genus 

 Pademma. These are not local varieties. In a single glade in the 

 forest one may find them all, together with dozens of intermediate 

 forms. I have collected species of Salpinx in various regions from 

 Malabar to Timor, and even in the smaller islands, whore Euplmas of 

 other types are among the most constant of insects, they varied to a 

 remarkable extent. In Upper Assam, where midamus is the dominant 

 Euploea, the Pademnias more or less resemble that species, and after 

 various fluctuations in the border-country between midanms and core, 

 they finally assume, in Lower Bengal, a tolerably fixed and constant 

 form (Jxollari) as a mimic of the latter. 



In general, however, it seems to me that the results expected from 

 breeding will never be realized, and the formula now much in vogue 



among lepidopterists— " these forms may be only varieties of ■ — but 



till careful breeding has proved that such is the case, wo prefer to keep 

 them distinct" — is about as dangerous as can well be imagined. The 

 burden of^ proof is thus thrown on the unfortunate breeder, and the 

 describer feels that he can safely neglect the ordinary means of proving 

 a species, examination of a series of both sexes from more than one 

 locality, and an attempt to associate some structural peculiarity of form, 

 sex-mark or prehensores, with the proposed species, The Pieridm have 

 especially suffered in this regard, and in particular Terim; Ixias, and 

 Teracolus have been thus reduced to confusion thrice confounded. 

 Breeding can probably accomplish little with such genera. A group of 

 these butterflies, sitting together on the wet sand, will, in three cases out 

 of four, consist of but one variety, varying little ; or, if there is another 

 present, it will usually appear quite distinct. The next stream-bed may 

 contain quite another breed. There is no reason for supposing that those 

 varieties would not, in a majority of cases, brood true. It would be 

 only now and then that the breeder, having overcome the great and 

 greatly underrated (at least in the moistnr parts of the Eastern tropics) 

 difficulties of rearing a new generation, will find any great variation in 

 his specimens. But even this will prove nothing, because hybridism, 

 which undoubtedly takes place occasionally between butterflies unquos- 



