( 450 ) 



degree. The three Mahiyau subspecies of Papilio caunus, inhabiting the first 

 Malacca and Sumatra, the second Borneo, and the third Java, differ from one 

 another by closer examination in the size of the white markings ; the differences 

 are, however, so slight that they do not affect the general aspect of the specimens; 

 in fact, the Java caunus resembles the Bornean rhadamanthus just as much as the 

 Bornean caunus does. The difference between the three caunus4orm?, are certainly 

 Dot such that if the differences were altogether absent the specimens would be less 

 protected; there are, to be sure, very few cases of mimicry in which the resemblance 

 of the mimetic and imitated species is greater than that of any of the three caunus- 

 forms witli any of the three /■/ladnmrinthus-forms. If, therefore, mimicry is of value 

 to the imitating species in all cases where the resemblance is of a much more 

 superficial kind, we cannot see why it was necessary at all to have the white mark- 

 ings of the Borneo caunus, compared with the Malacca and Java caunus, a little 

 reduced. Further, though the resemblance between the species in question is very 

 great, there are still differences in markings between the mimetic and the imitating 

 species in each locality which are greater than the differences between the three 

 caunus-ioxmi. It certainly reqtiires a great deal of faith in the omnipotency of 

 Natural Selection to believe that the slight reduction of the white markings in the 

 Borneo form of Papilio caunus is due to a survival of those specimens of caunus in 

 which the white markings were a little smaller than usual; the enemies of caunus 

 to which we attribute the execution of the selection then must have been possessed 

 of a mncb keener power of discrimination of markings than the entomologists who, 

 until the appearance of Mr. Walter Rothschild's Revision of the Eastern Papilios, 

 treated the Bornean caunus as identical with the Java caunus as figured by 

 Westwood ! If we, however, admit that the slight distinguishing characters of 

 the three Llalayan caunus-iorm?, cannot possibly be due to the action of Natural 

 Selection, but must be the effect of some other transmuting factor, it is evident 

 that also in all other cases of mimitc distinguishing characters we need not refer to 

 Natural Selection as the cause of the minute ditl'erence. The presence of minute 

 distinguishing characters allows, therefore, a restriction of the possible causes of the 

 divergency of the respective forms; and, as we thus have to admit the importance 

 of insignificant distinctions in respect to evolution, it will be obvious that the 

 importance increases as the degree of distinctness decreases. 



We have said above that we take as the lower limit of the application of the 

 term " subspecies " such cases in which about half of the individuals are characterised 

 liy some peculiarity : which is the upjx'r limit? or, wlieu have we to begin to call 

 a form specifically distinct ? According to our definitions of the terms " species " and 

 " subspecies," the distinction between subspecies and species is a biological one, the 

 presence of which, as mentioned on p. 442, we systcmatists are not able to directly 

 prove or disprove from the material we are working with. As we know now from 

 experiments and careful field observations that mori)]u)logically very different forms, 

 connected or not by intergradations, can, in spite of the conspicuous dilferences, 

 be one species (individual and seasonal ])olymorphism, heterogenesis, etc.), it is 

 a priori evident that also geographically separated forms, in spite of their being 

 morphologically distinct and in spite of their not behig connected with one another 

 by intergradations, can very well l)e subspecies of one species, i.e. can under lavour- 

 able circumstances fuse into one form. The actual jmiof of specific distinctness the 

 systematist as such cannot bring ; we species-makers do, in fact, not pretend, at 

 least manv of us do not, that in everv case the fi>rm whicli we ])ronounce to be 



