( cxxxvii ) 



passage, iu fact, from race to race and from species to species 

 is commonly a gradual one. On the other hand, we do from 

 time to time meet with cases that may come under the head of 

 discontinuity ; thus, while the scent-scales from the West 

 African form of Leuceronia argia show very slight variation in 

 structure from those of the Southern race or subspecies, the 

 distribution of these scales on the surface of the wing is com- 

 pletely different in the two forms. On the other hand, the 

 scent-scales in Leuceronia hcquetii and in Belenois gicUca, con- 

 sidei'ed in relation to their congeners, may be said to be 

 discontinuous in form though not in distribution. The 

 sporadic cases of absence of scent-scales in the midst of genera 

 usually possessing them, for instance Teracohhs auxo and 

 T. incretiis, also apparently Tachyris ada, may likewise be 

 accounted as instances of discontinuity. 



In the genus Pereute the scent-scales constitute a fairly con- 

 tinuous series. But it is to be observed that if the species of 

 Pereute were arranged according to the character of their 

 scent-scales, the arrangement would not coincide with that 

 founded upon the modifications of the colour-pattern. Thus, 

 the plume-scale of P. swainsonii is very like that of P. 

 callinice, but that of P. antodyca is quite different. P. 

 swainsonii and P. antodyca are, however, like one another in 

 general aspect, and unlike P. callinice. If we had only the 

 scent- scales of the two former species to go upon, we should 

 say, " here is discontinuity" ; but another species, P. callinira, 

 though discontinuous in aspect, supplies a link in the matter 

 of scent-scales. Many reasons might be adduced in explana- 

 tion of this phenomenon, which after all is only one instance 

 of a principle constantly to be met with in all taxonomic 

 investigation. One feature in the case may here be noticed : 

 there can be little or no doubt that mimicry in one form or 

 another has exercised a disturbing influence on the general 

 aspect of the species of Pereute. From this influence the 

 scent-scales must be free ; and they are therefore, wherever 

 this and similar factors are concerned, likely to be more 

 trustworthy guides than the general aspect to the true 

 afiinities of the species. 



Pereute swainsonii and P. antodyca, though so near to one 



