IN DEFENSIVE COLORATION 365 



forms between the groups now sharply separated had 

 not been obHterated.^ 



5. A History Inferred from Mimicry may be Confirmed 

 by other Evidence, 



[The followhig facts and arguments were first brought 

 forward in the Leeds lecture, and pubHshed in Nature 

 for October 2, 1890. They have been revised for the 

 present work.] 



Evidence for the evohition of Mimetic Resemblance has 

 been forthcoming as the result of recent [in 1890] work, 

 of which the general results are set forth below, but the 

 details still remain unpublished. 



Many moths have lost the scales which are charac- 

 teristic of the Order of insects to which they belong, so that 

 their wings become transparent, and they mimic stinging 

 insects such as humble-bees, wasps, and hornets. This is 

 the case with two British Hawk Moths [Haemorrhagia 

 fuciformis^ and H. tityus"^ = bombyliformis). It is known 

 that when these moths emerge from the chrysalis, the 

 transparent parts of their wings are thinly covered with 

 scales which are shaken off during their first flight. The 

 loss of the scales is due to the rudimentary nature of the 

 stalk at the base of the scale and of the socket in which 

 the stalk is inserted. A closely allied Eastern and Ethiopian 

 moth [Cephonodes hylas *) is still more completely denuded 

 of scales, but in it also the rudimentary sockets are found 

 to be thinly scattered over the transparent wings. These 

 facts suggested that other moths with transparent wings 

 would be found to repeat, in the course of their own 

 individual lives, the history of the change by which 

 transparency has been attained by the species. Investi- 

 gation has supported this suggestion. The examination 

 of two British Sesiid moths which resemble hornets 

 or wasps w^as especially instructive. In one of these, 

 Trochiliiun (Sesia) apiforme'^, the mimicry is not so 

 perfect as in the other, and is therefore presumably of 

 more recent date ; in this moth the rudimentary scales 

 which fall off are comparatively perfect, while in the 



^ Trans. Ent, Soc, Lorid., 1902, pp. 482-4. The mimics alluded to 

 are HypoUninas misippus, female, and Acraea encedon. 



