( 438 ) 



approached the subject with tlie same delight in ' opposition ' and contradiction as 

 Erich Haase." Would it not be better to say that he who does not accept without 

 critical examination wliat is told him is the trne scientist ? But let us now examine 

 the classificatory results of the researches laid down in Artbildung. 



Eimer says that his researches reveal the phyletic connection between the forms 

 treated upon; licnce we must accept the "groups" of species formed by him to 

 contain only snch species which are more nearly related with one another than with 

 the members of other groups, except if one group has directly developed from a 

 species of another group. If we bear this in mind, the enigma embodied in the 

 Icosthenes-anticrates-ajax {Artbildung I. p. 156) and the ujax-policenes groups 

 {ibid. p. 193), and in the turnus {Artbildung II. p. T'J) and the (isterias groups 

 {ibid. p. 118), is insolvable, and we must ask what profound mystery is at the 

 bottom of the association of P. ajax -and turnus each with two groups of species. 

 Is the hlAiik fenmle {glaucus) of turnus more nearly related with P. asterias than 

 with its own male ? Is icalski, though a seasonal form of a)ax, less closely connected 

 with its sister forms, which i)artly originate every year from the eggs of the same 

 mother individual, of which also walshi is an offspring, than with the African 

 P. police?ies? The offspring of one female belong to two different groups of species ? 



In his first volume Eimer deals with four " groups " of Papilios, which all 

 belong to that section of Papilioninae to which Haase gave the name Cosmodesmus. 

 (Jronp characters, that is distinguishing characters of each group, are not given; 

 iu fact there are no such characters common to the members of one and absent 

 from the members of the other groups, with the exception of the antiphates group, 

 which contains only the geograi:>hical representatives of anti^jhates and some close 

 allies. But it is obvious that the reason why out of the great number of Asiatic, 

 African, and American species of Cosmodesmus just those particular forms were 

 united into groups regardless of all characters except wing-pattern, was that the 

 wing-pattern of the forms jiut together in each group demonstrated a connection 

 between the species iu accordance with Elmer's views. Species, however, put 

 together and treated as relatives because their wing-patterns exhibit certain (real 

 or sujiposed) iiffiuitics, cannot be brought forward as demonstrating that wing- 

 ]iatteru exhibits tlie phyletic connection admirably, and sliows the lines of develop- 

 ment by which each species, by which each character, has arrived at its present 

 state in mutation. If one intends to demonstrate the kind of variation of a particular 

 organ or character by a comparison of this particular organ or character in closely 

 allied species, it is absolutely necessary to ascertain that tlie species to be compared 

 are related to each other, not because they are similar in that jiarticular organ, but 

 because other characters, which are independent of that organ or character in their 

 variation, establish the relationship. Eimer rejects other characters than wing-pattern 

 as being of inferior value. Haase* pointed out that Elmer's classification was faulty, 

 because strnctural cliaracters, especially a very striking one in ueuration, had been 

 neglected. The lirst subcostal branch is, namely, in a number of species of Elmer's 

 groups I., II., III., and IV. invariably anastomosed to the costal nervnre (for instance, 

 in glycerion, paphm, agetes, antipltatea, aristeus, rhesus, etc.), while in other species 

 of these groujis that vein is free (for instance, iu podalirius, leosthenes, ajax, 

 protesilaus, etc.). The dcvelojiment of the neuration in the pupal wings shows con- 

 clusively that all nervnles were originally free, and that fusion and obliteration are 

 specialisations. Hence the species with the first subcostal branch free are in this 



* Haase, Untermchuiigai iiler Mimicry I. 1893. 



