{ 505 ' 



remains quite iucomplote as long as we do not know tLie limits of variation of the 

 forms in question ; as these limits, however, can be made out, to a certain degree of 

 correctness, onlv by comparing- many inilividnals, it becomes self-evident that the 

 question whether a certain jieculinr character found in the genital armature of an 

 individual or a number of individuals indicates specific distinctness or not, can only 

 be decided with a certain degree of correctness when the variation of the sexual 

 armature of the allied forms is known, i.e. after the examination of a great material. 

 When we said above that the degree of relationship between a number of closely 

 allied forms could ver\' well appear different according as we take the sexual 

 armature or the wing-pattern as a guide in the arrangement of the forms, we did 

 not give illustrations, because we had to recur to the jihenomenon. If it really is 

 true that there are cases in which two forms are in colour and markings the most 

 closely related, while the structure of the genital armature stands in opposition 

 to this relationship, pointing in quite a different direction, a comparison of the 

 characters of the wings and the characters of the apparatus of copulation of the most 

 closely allied forms, wliich are always geogra.i)liical representative forms, with the 

 aim of deducing the different relationship as indicated by similarities in the one or 

 in the other organ, must throw a highly interesting light upon those questions which 

 relate to the jirobable history of the geographical distribution and the origin of those 

 forms. 



B. Phenomena in the Variation' of the Organs of Copulation Kelating to 

 Some Questions of the Geogeaphical Distribution of Aximals. 



When we speak here of representative forms, we do so regardless of their being 

 specifically distinct from one another or not ; whether a geographical representative 

 has reached the degree of divergent develojiment which we call specific, or is still 

 a subspecies, is of no importance for the following discussion ; the imjiortant point 

 is, that there are differences between the representative forms. 



Now, when we see that some Indian and Australian representatives are very 

 similar in colour to one another and dissimilar to the forms inhabiting the inter- 

 jacent countries, does that mean that the Indian and the Australian forms have 

 separated after the more dissimilar representatives had branched oft'? When we 

 find the (Jelebensian mountain form of P. sarpedon to bear a much greater resem- 

 blance in colour to the form inhabiting the Sunda Islands than to the form found on 

 Celebes at lower elevations, is it correct to conclude that the Celebensian mountain 

 form is a descendant of the Sunda Island form and iuilicates that there was at a 

 former period a closer connection between Celebes and the Sunda Islands ? When 

 we observe in a great number of cases tliat the Javanese forms are similar to those 

 from Malacca, while the forms from Sumatra and Borneo again are similar to one 

 another, have we to infer from this fact that there was at one time a land-connection 

 between Malacca and Java independent of Sumatra and Borneo ? When West Africa 

 and Madagascar are inhal)ited by two representatives with a narrow and interrupted 

 baud, while the East African form has a broad and uninterrupted band, can we 

 conclude that the Madagascar form is the ancestor (or descendant) of the West 

 African form ? When, finally, we find (Central America, the Greater Antilles, the 

 Lesser Antilles, and Venezuela inhabited each by a rc]iresentative bird or Fapilio, 

 and notice that in colour the forms from th(^ first and third anil tiie forms from the 

 second and fourth localities are respectively similar, are we justified in so inter- 

 preting the fact as to say the Lesser Antilles have received the bird or insect from 



34 



