( 508 ) 



it to have developed from that form of P. aristeus to which it comes nearest in the 

 before-mentioned points, namelj' from P. aristeus hermocrates. Au examination of 

 the valve of the male shows, however, that there is also very good reason to deduce 

 rhesus from P. aristeus aristeus or pannatus. The ventral dentate ridge is in the 

 eastern forms (rhesus, aristeus, pannatus) half-ring-shaped (f 83, 84), and at a 

 glance distingnishable from the smaller, onl}' slightly curved, ridge of the western 

 forms (anticrates and hermocrates, f. 73, 74) ; as intergradations in the two kinds 

 of ridge are unknown (akhough we have examined a great many individuals), the 

 constant difference in the ridge would be regarded by many systematists as justi- 

 fying a specific separation of the western forms from the eastern ones, in which case 

 P. rhesus agreeing in the ridge very well with the eastern forms, apart from minor 

 differences, would ajjpear to be closer allied to aristeus and parmatus than to 

 hermocrates and anticrates, and hence could be considered a descendant of P. aristeus 

 aristeus from the Moluccas. 



If we take A, B, C as three geograpliical representatives, «, and a-, as the 

 different degrees of development of the wings of these forms, and i, and b-, as the 

 different degrees of development of the organs of copulation, the several cases above 

 adduced, which we think sufficiently illustrate the questions we are to deal with, can 

 be put diagrammatically as follows : — 



A \ B \ C 



b, I b., 1 b. 



The dilemma arising from the contradiction of the characters of geographical 

 representatives allows a satisfactory solution, if we take into account, firstly, lluit 

 according to the theory of evolution the peculiar modifications of the organs of 

 a certain animal are partly inherited and partly acquired, and that therefore a 

 similarity between two forms, and a dissimilarity, a priori neither prove nor 

 disprove a close phylogenetic connection of the forms ; and secondly, that when 

 the similarity between two forms is due to inheritance the character common to 

 the two forms is inherited either by both independently from the common ancestor, 

 from which also other forms which have lost that character have descended, or 

 by one of the two forms from the other. The conclusions generally deduced in 

 systematic and other works from the similarity and dissimilarity of geograi)hical 

 representatives in respect to the Geographical Distribution of Animals as a science 

 (not as a mere statement of facts) are to our mind mostly based on the erroneous 

 assumptions, firstly, that every similarity is due to inheritance, and secondly, that 

 the presence of a pecnliarity in two forms must necessarily be due to the common 

 character being inherited by one of the two forms from the other. Let us, then, 

 briefly inquire into the question of the probable origin of the similarities and dis- 

 similarities of allied forms, and try to arrive at a correct estimate of their actual 

 value in our judgment of the relationshij) of the forms in which they are observed. 



If we recall to mind that the develojiment of a species into more species is 

 possible only by means of isolated transforming factors associated with a more 

 or less complete prevention of the affected portion of the species from interbreeding 

 with the original stock, the distinguishing characters of locally separated allied 

 forms, which we must regard as the outcome of the transformation of a common 

 ancestral form, must be due to the eftect of isolated evolutionistic factors of any 

 kind present in the district inhabited by each single representative form. It is quite 

 jpossible that when a species separates thus, first into subspecies and then into 



