( Ixxiii ) 



characters, characters for insuring specific segregation 

 (Wallace's " recognition marks "), etc. Then also there is 

 the possibility of functional change, such for example as the 

 ocellated spots on certain Sphinx larvas which may have 

 been originally developed for terrifying enemies and after- 

 wards, in some species, converted into protective mark- 

 ings, or rice versa* With regard to colour and pattern as 

 a means of recognition — I mean, of course, by the insects 

 themselves and not by systematists — a wide field for observa- 

 tional entomology has been opened up by the suggestion. 

 That such characters are of value for the purpose of recog- 

 nition would appear probable from certain well authenticated 

 observations which indicate a temporary failure of the pro- 

 cess. I refer to those cases where a male butterfly has been 

 seen pursuing a mimetic female of another species until he 

 discovered his mistake. f This seems to show that the male 

 is guided at first sight by visible characters. 



At the risk of digressing from the main topic of my address, 

 I cannot avoid dwelling a little longer upon this all-important 

 distinction between diagnostic characters and those of utili- 

 tarian significance from the selectionist's standpoint. In the 

 class of insects most particularly is it necessary to bear in 

 mind that our systematic methods are more or less artificial. 

 The " species " are generally described without reference to 

 life-history, without a knowledge of habit, often in profound 

 ignorance of the function of the structural characters which 

 are made use of for diagnostic purposes. He must, indeed, 

 be a bold systematist who unhesitatingly declares that this 

 or that specific character ?»u.sf/^t? non-significant, and therefore 

 cannot have been produced by natural selection. This atti- 

 tude has always appeared to me as much wanting in scientific 

 caution as that opposite and illogical rashness with which the 

 supporters of the theory of evolution by natural selection are 

 sometimes credited. It is much about the same attitude as 



* Weismann's "Studies in the Theory of Descent," English Edition, 

 p. 365. The tailed wings and ocelli in certain species of Thecla may pos- 

 sibly come under this heading. See Poulton's work on '' The Colours of 

 Animals," p. 207. 



t Ann. Mag. Nat. Hist., 1S78, p 158. 



H 



