( 468 ) 



111 all the specimeus examined the blade of the iuirjie is coiispiouuusly longer 

 iind slenderer than in all the other subspecies of 1\ pobjtes, and reaches close to the 

 apical martfin of the valve (f. 31, taken from a Mindoro examjile); tlie narrow 

 apical portion is especially long, and often rather strongly hook-shaped, the hook 

 jirojecting above the level of the raised vent ri-apical edge of the valve. 



The harpc of the average example from the Philipjunes is represented in f. 82; 

 it has a very different ap])earance from the harpe represented in f. 18 to 26, while 

 the difference between f. 32 and 29 is less conspicuous; the variation of the Philippine 

 ixamples takes place in such a direction that some individuals a]iproach f. 35. 



Of the males from the t^ulla Islands Mr. Rothschild says that they are not 

 exactly identical with the Pliilippiuc alphenor in wing-pattern, but approach a little 

 P. poh/tes perversus. The harpes of the Sulla 7nales (f. 33) are, on the contrary, 

 still slenderer than in Philippine alphenor, thus indicating that a relationship in 

 wing-jisittern is not necessarUy corroborated by the development of tlie genital 

 armature. 



From the Southern Moluccas we had uufortimately only one individual at our 

 disposal ; the harpe of this sjiecimen is shaped nearly as in pobjtes pertersus 

 (f. 35). 



In the harpes, therefore, the males of alphenor from the tliree Incalities 

 (Philippine Islands, Sulla Islands, and Southern Moluccas) . are I'airly well ilistiu- 

 guishable, though there is no distinct line of sejiaration. 



h. P. pobjtes nicanor from Halmaheira and Batjan; f. 36. 



The blade of the harpe is much higher than in pobjtes alphenor, and in this 

 respect nicanor comes nearest to the Indian pobjtes polytes; the apex of the harpe is 

 produced, nearly as in alphenor. F. 36 is taken from a Halmaheira individual. 



6. Papilio ambrax ; f. 37 and 38. 



The outline of the valve (f. 3^) of this species varies individually. The harpe 

 is generally formed as in f. 37, but is sometimes somewhat slenderer and at the 

 apex less hooked. 



We merely give the two figures of the valve and harpe of an ambrax individual 

 (from German New Guinea) in order to enable the reader to compare them with 

 f. 18 to 36. The differences between f. 37 (aml/raj) and 36 or 34 {/wli/tes), 

 belonging to two species, are not so conspicuous as the divergeuc'y exhibited by 

 various harpes within the species pobjtes. 



The facts of variation observed in P. pob/tes are as follows : — 



(1) The valve is variable in the individual s]iecimens, but does not exhibit in 

 the specimens examined any obvious variation according to hjcality. 



(2) The conspicuous individual aberration /'. pobjtes borealis J-ab. thihetanas 

 has no peculiarity in the genital armature. 



(3) The individual variation in the blade of tlie har]>e of the individuals from India 

 is such (f. 18 and 111) that the extremes stand further apart tlian many individuals of 

 poljte.'S pob/tes do from many individuals o{ pob/tes ihescus. 



(4) The harpe of pob/tes theseus from Java and the lesser Snnda Islands is well 

 distinguished from that of the Indian pob/tes pob/tes, but there is no parting line 

 on ac'count of the intermediate form of the harpe of pobjtes theseus from Borneo 

 and Sumatra. 



