( 285 ) 



the elements in the pattern develop jihylogenetically according to a tixed " law." 

 The phenomenon shows, on the contrary, that some agency foreign to the allied 

 insects decides which of the many possible liuesof development an element (dot, line, 

 sjiot, i)atch, band, streak) of the pattern must follow. 



The variability in the pattern — apart from " sports " — is in some species rather 

 slight, relating only to minnte diflferences in the size of the spots, the relative 

 positions of the bars and general tint of the wings, while in others the individual 

 differences are so great that some authors have considered them specific. In C/i. 

 poli/xemi, for instance, both sexes are so extremely variable that ten odd specific 

 names have been bestowed upon individuals from North India and Burma alone. 

 The occurrence of very different iudividnals in the same locality may sometimes be 

 due to meteorological factors acting npon the individual chrysalis. But we have no 

 evidence of that in regard to Ckaraxes. Nor is there any evidence that the cases of 

 variability put down by Butler as seasonal variation, and those queried as such by 

 Moore and Anrivillius, are really seasonal. In many instances we have found that 

 individuals belonging to two supposed seasonal forms occur together at the same 

 time of the }'ear, the one form not being apparently more plentiful than the other. 

 Only such varieties should be called seasonal of which it is shown that they are 

 seasonal. Considering differences of undated specimens as seasonal is unjustifiable 

 — because arbitrary — as is treating those of not-localised individuals as geographical. 

 The only suggestion we can here make as to seasonal variation among Ckaraxes is 

 based on dated specimens of Ch. pobjxena from North India ; these specimens show 

 that the individuals emerging from hibernated pupae are rather smaller than the 

 average summer specimens. 



We recognise 103 distinct species, of which 30 are divided into 90 subspecies 

 (= geographical forms). To classify such a great number of species of a genus 

 satisfactorily is not an easy matter. Though the relationship of any two species 

 may be recognised with certainty, it remains often a purely arbitrary matter in 

 which order the groups of species have to follow one another in the linear arrange- 

 ment of a list We have as guides in the classification of Ckaraxes some structural 

 characters, such as the dentition of the costal margin and the length of the ceil of 

 the forewing, the sexual armature of the 6, and the pattern. Of these characters, 

 those of the sexual organs can be employed only with great caution. 



These organs exhibit often but very slight differences in closely related species, 

 and hence sometimes give important hints as to the relationship of species, if there 

 are any special structures observed ; but we must keep in mind that similarity in the 

 penis or clasper does not necessarily indicate relationship, as the similarity may be 

 the outcome of the obliteration of the different special structures. In this category 

 of unreliable characters belongs, for instance, the occasional absence of denticula- 

 tion from the penis. On the whole, there are few structural characters available 

 in the classification of Ckaraxes, for the simple reason that the constant structural 

 differences of groups of species, relating to both sexes, are employed to characterise 

 genera. The species dealt with in this monograph being divided up into genera 

 according to the structure of the body and wings, it is not possible to further divide 

 these genera on purely structural characters. The colouration of the wings must, 

 therefore, remain one of the chief guides in the arrangement of the species. But 

 here again one must take care not to be misled by analogies, which are not rare 

 among Ckaraxes as regards the phyletically younger pattern of the npperside of the 

 wing. There is a whole series of species which bear on the upperside a j«mafkable 



