( I'S ) 



Kvery iudividual specimen examined exhibited some peculiarity in the number, size, 

 and ibrm of the teetli of the dorsal ridge. 



The lateral (ventral) fold (/;) of the dorsal riilge (f. 08) is in f. lOT (Taiping, 

 Malay Peninsula) absent, in f. 99 (Sikkim) very short, in f. 106 (Sbau States) 

 complete but very slightly raised, aud is more distinct in f. 98 (Kumaou). In 

 f. 100 (Sikkim) the fold runs up to a small tooth, which has a somewhat transverse 

 ])osition to the longitudinal ridge ; in f. 108 the fold is raised to a low but olivious 

 dentate ridge, which we shall call " transrerse ridyeT 



In a Mindoro specimen caught by Mr. A. Everett in December 1894 the 

 additional transverse ridge is raised above the level of the dorsal ridge, and forms 

 a conspicuous, broad, tooth-like prominence; in f. Ill the dorsal ridge with the 

 transverse prominence is represented in a view vertical ujjon the j)lane of the valve; 

 and in f. 112 we give a view of the same organ from the apical side (with the eye 

 a little above the valve). 



Still more aberrant is a specimen from the Shan States caught by Mr. Roberts 

 in the same district where the individual was obtained from which f. 100 is taken. 

 The increase in the size of the transverse ridge, as illustrated by f. 107, 99, 106, 

 98, 100, 108, 111, and 112, reaches in f. 101 to 103 the maximum. The transverse 

 ridge h is higher, strongly dentate, and more extended than the dentate portion of 

 the dorsal ridge (/; in f. lOl the organ is seen from above, in f. 102 from the 

 ventral side, and in f. 103 from the ajiical side (comj)are f. Ill aud 112). 



We have examined more than a hundred specimens of P. sarpedon sarpedon, 

 and found only one that has the additional transverse ridge so extraordinarily 

 developed; the significance of this variation is obvious if we compare the sjjecial 

 structure of the dorsal ridge of /'. sarpedon anthedon, milon, choredon, teredon, etc. 

 AVe shall have to refer to this particular case again. 



b. P. sarpedon semifasciatus from (!hina. 



The Chinese subspecies of sarpeil.on is in the markings not always distinguish- 

 able from sarpedon sarpedon, but the greater number of the individuals from 

 Central and Western CUiina have a very remarkable character in the band of the 

 liindwing being more or less obliterated. Snch specimens with almost entirely black 

 hindwings are, in res])ect to jiattcrn, quantitatively more dillVrent from sarpedon 

 sarpedon than the individuals of any other subsjiccies are. Though we dissected a 

 long series of Chinese individuals, we did not perceive any character in the genital 

 iirmature by which they could be differentiated from sarpedon sarpedon ; and this 

 concerns the examjiles which are most aberrant in j)attern,as well as sjiecimens with 

 the ordinary sarpedon sarpedon pattern. The fact is of high interest, as it distinctly 

 shows that a great external discrepaiu:y of a localised form does not necessarily 

 imply that there is also a peculiarity in the genital armature of the form. 



c. P. sarpedon adonarensis from Adonara and Sambawa ; f. 127. 



The interesting external features of this form are pointed out by Mr. Rothschild 

 on p. 324 of this volume. In pattern it comes near the subspecies from the 

 neighbouring islands of Sumba, Timor, and Wetter ; and we were rather surprised 

 when we found tluit in the armature of the valve the specimens from Sambawa and 

 Adonara disagree with those sid)species, and agree much better with the Indo- 

 Malayan sarpedon sarpedon, as the dorsal ridge is dentate and has the transverse 

 ridge as feebly developed as it normally is in Indian examples. F. 127 represents 



