( ni'T ) 



his short stay on that island ; these 'males of delphonte/i Feld. do not differ from 

 those from the Northern Moluccas in any way. The Buru female of deijjholnw L., 

 however, has the tails much narrower tlian all my other specimens; their greatest 

 breadth does not exceed 3J, mm., while the tails have normally a breadth of at least 

 .) mm., even in my smallest example, which has the forewing of a length of only 

 65 mm. The same Buru female has also rather more lilue on the upperside of the 

 hindwings. 



Now I come to a question which perturlis me considerably : Are P. dcip/ioljiis L., 

 ileijihontes Feld., and delpyl us Feld. distinct species, or must we consider tlu-ni to be 

 local forms of one insect ? 



The oeciu'rence of the tailed and tailless Papilio on Burn (proxided that the 

 locality is correct, which I strictly believe) is rather in the affirmative, namelj', that 

 we have here constantly different species; but it is no proof, as we know from manv 

 I'ajiilios, that specimens agreeing with a localised subspecies occur occasionallv 

 (sometimes regularly) among the specimens of the typical form (Staudinger's law of 

 variation : a creature can be at the same time localised variety and not localised 

 aberration ; Elmer's law : local forms differ from the typical form in the same way as 

 the not localised aberrations do, lint exhibit the characters of the latter in a higher 

 degree). The tailed deipylus and the tailless deiphonies are in colour almost exactly 

 identical, and they differ only in the development of a tail; deipjylus and dcipkohus 

 are both tailed, and differ only in colom\ Thus it appears to me that it is impossible 

 to unite either P. deipylus with deipjhohvs, which agree in being tailed, and to 

 treat the tailless deiphontes as a distinct species; or P. deipylus with deiphontes, 

 which agree in pattern, and to take P. deiphohus as a separate insect ; but that we 

 must treat them either as being all three distinct, or as belonging all three to one 

 species. The latter view, which I was first inclined to adopt and which may turn out 

 to be correct, I dare not take, in consequence of the rule which I must always follow- 

 in this paper, that I treat an insect as a subspecies only if it is connected with the 

 tyj)ical form by intergradations. In the present case, however, the tailed and tailless 

 forms are not connected with one another by specimens with short tails ; we know at 

 present only of specimens with a long tail, and of .specimens with a short tooth to the 

 hindwing. Both tail and tooth are in fact somewhat variable, the former chiefly in 

 breadth, as I have said above; but no specimen I have seen has the tooth so much 

 prolonged, or the tail so much reduced, as to form a kind of connecting link between 

 the tailed and tailless insects. 



The tailed deipylus and deijjlwhus differ in the mule sex rather conspicuou.<ly 

 in colour; true connecting links are again wanting; the /ortrt/e of deipylus mu.<t 

 be left out of consideration (see under P. deipylus). 



Though the differences lietween P. deiphol/us, deiphontes. and deipylus are of 

 no great importance, they are, to our knowledge, constantly met with, and hence 

 I must eninnerate the three insects as distinct .species. 



Note. — In the scaling the males of P. deiphohus L. on the one hand, and of 

 P. deiphontes and deipylus Feld. on the other, exhibit a distinguisliing character 

 which is rather easily recognisable. The bluish grey streaks in the marginal region 

 of the forewings above are composed in P. deiphobus of long and thin, almost hairlike, 

 scales, while in the other two species these scales are much l)roader. Tliis shows 

 again that the tailed deipylus is less clo-^ely allied to the likewise taile<l deijihohus 

 than to the tailless deiphontes. The females seem to me to have no easily traceable 



24 



