480 Mr. G. A. K. Marshall on 



conspicuousness. When I realized that it was the mid- 

 rib-like stripe — the character which more than any other 

 gives meaning to tlie cryptic resemblance to a dead leaf; 

 that it was this very character which, transformed into the 

 " warning band," became the conspicuous feature of the 

 wet phase — the operation of natural selection seemed as 

 clear in the one case as the other. 



When we examine the species of the whole genus 

 Precis and those of the genera allied to it, the conclusion 

 is forced upon us that the dry cryptic phases arc ancestral 

 as compared with the conspicuous wet phases, I do not 

 mean to imply that the cryptic forms have not altered, 

 but that the original form of the species possessed a cryptic 

 under-side, which has been handed down with more or less 

 change as the cryptic under-side of the existing dry phase, 

 while the conspicuous under-side ^of the existing wet phase 

 is a new and comparatively recent development. This 

 ({uestion of the relative age of the two forms is most im- 

 portant and interesting, and from the very first occupied 

 Mr. Marshall's attention. Thus the following passage is 

 extracted from a letter written a few weeks after his 

 discovery : — 



" SalisbiLvy, June 5, 18D8. — I should be most interested 

 to learn your ideas as to the reasons for the singular 

 seasonal change in this species, for I must admit that I 

 cannot arrive at any really satisfactory conclusion on the 

 subject as yet. The blue scales of sesamus are my chief 

 stumbling-block, for I certainly cannot perceive what 

 utility they can possess, and considering its protective 

 under-side colouring there seems no reason why it should 

 not have retained its wet-season colours above, as in the 

 case of P. artaxia or P. ccrijnc. I suppose you will agree 

 that scsamus is a later development ^ " 



For the reasons I have indicated above it is difficult 

 to doubt that the cryptic character of the under-side 

 of sesainus is ancestral and the conspicuous under-side of 

 7iatalcnsis relatively recent, but with regard to the upper- 

 sides this conclusion is by no means so evident. Indecl 

 on comparing the species with antilopc and other allies, it 

 seems probable that the upper-side of natalcnsis is more 

 ancestral than that of sc.srr?/(?ts, having been chiefly modified 

 in tint, thus falling into MuUerian association with the 

 larger Acra^as. The upper-side of sesamns probably 

 shows cryptic changes iu the acquisition of the lar darker 



