336 NATURAL SCIENCE. May, 



figures of the varying forms assumed by these harpes in different 

 individuals of the same species and sub-species are given, and enough 

 material has been examined to lead to the conclusion that there is no 

 correspondence between the variation of these organs and the variation 

 of the wing-patterns on which the sub-species are founded. The 

 relative length of the prehensile section of the harpe differs to the 

 extent of 73 per cent, in Papilio machaon. The uncus — or dorsal hook 

 belonging to the tenth segment — of the British forms of this butterfly 

 is closely like that of the Japanese sub-species and different from the 

 forms found in continental European and Asiatic insects. Similar 

 differences are noticed among the races of eastern species which Dr. 

 Jordan has examined, and a division of a species according to the 

 genital armature would, in most cases, give quite a different set of 

 sub-species from that which has been established on the patterns of 

 the wings. 



Though the amount of variation in the harpes is generally con- 

 siderable, there are for each species limits which are not overstepped, 

 and the author therefore considers these organs of the greatest use for 

 the determination of true specific differences. It is certain, however, 

 that many of the various forms within certain species figured by Dr. 

 Jordan would be considered of specific value by systematists unac- 

 quainted with the amount of variation which may occur. An investi- 

 gation into the limits of divergence seems therefore necessary before 

 the use of these organs as specific characters can be relied upon. 

 But the comparative constancy of the genital armature within each 

 species has an evident bearing on the development of the race. Dr. 

 Jordan has made some observations on the corresponding organs of 

 female Papilios and finds a set of variable ridges and furrows on an 

 intersegmental membrane, in each case adapted to be held b}^ the 

 harpe of the male clasper. He infers that, by a process of mechanical 

 selection, the variation of the genital armature m both sexes must be 

 kept within limits, because individuals with abnormal organs would 

 be incapable of copulation. And it follows that any considerable 

 modification of the armature, generally affecting both sexes of a race, 

 would lead to practical isolation and accelerate the formation of a 

 new species. The author quotes a passage from the " Origin of 

 Species " in which Darwin dwelt on the difficulty of accounting for 

 the infertility of " natural varieties, as soon as they have been 

 permanently modified in a sufficient degree to take rank as a species." 

 The facts dealt with in the present memoir show at least one way in 

 which such segregation as Romanes' theory of physiological selection 

 was proposed to explain, might be produced ; though " the cause of 

 infertility would not be physiological diversity but mechanical 

 impossibility of fertilisation." The importance of the principle here 

 invoked has been long recognised by various naturalists, and on p. 48 

 of the current volume of Natural Scienxe, Mr. O. H. Latter quotes 

 an observation of Dr. Marchal, bearing on its importance in wasps. 



Dr. Jordan is careful to point out that mechanical selection can 

 only work on variations produced by other factors. No direct 

 evidence as to these other factors is furnished by his researches, but 

 some incidental remarks are worth notice. There is an Indo-Malayan 

 butterfly, Etiplaa rhadamanthus, mimicked by a Papilio, P. caunus, and 

 both these species have distinct forms in Sumatra, Java, and Borneo, 

 separable only by minute differences in the extent of the white 

 markings on the wings. As the author points out, it is hard to 

 believe that such minute variation in the mimicking species can be 

 the result of natural selection ; even one who considers natural 



