( 477 ) 



subdorsal tootli corresponds to that of cloanthan, liut is here much less chitiuised, 

 nsuiilly l>r(jaJenod at the ti[), and is iu fact a dilatation of the fold e partly rolled 

 in so as to form a half-cylinder ; thongh this tooth is variable, we shall not refer 

 to it under each subspecies. The dorsal ridge corresponds to the dorsal ridge of 

 cloanfliKS, but is here of a more complicated structure. Iu a view from the dorsal 

 side (f. 97) it will be seen that the dorsal ridge off. 96 consists of a dentate higli 

 basal ])ortiou and a free rod-like apical process (,7), the dorso-apiml jiroct'.s.s, 



which is curved ujjwards— that means towards median plane of the abdomen is 



denticulate at the tip, aud protrudes beyond tlie tip of the dorsal lobe. At the 

 ventral side of the dorsal ridge a longitudinal fold (f. 96, /, and 98, i) will be 

 noticed which is continued to the apex of the dorsal lobe ; tliat vertical portion 

 of the fold (f. 98, h) which runs up to the upper edge of the ridge is of hio-h 

 importance, as it develops in most subspecies to a peculiar organ. 



a. P. sarpeilon sarpedon ; f. 96 to 114. 



This form occurs all over India (except S. India aud Ceylon) to Java, the 

 Philippine Islands, and Japan ; in China it is replaced by another subspecies. There 

 are some external (characters by which the individuals from Java, Borneo, and the 

 Pliilijipines can lie distinguished from the individuals from N.W. India and 

 N. India, but these characters are very slight ; besides, the Malayan individuals 

 lead over to the more different forms from the lesser Sunda Islands, and hence 

 remain best included in sarpeilon S'lrpcdon. The genital armature is entirely in 

 accordance with this statement. 



We have examined specimens from N.W. India, Siiddm, Assam, Burma, 

 Shan States, Tenasserim, Cochin (Jhina, Sumatra, Nias, Java, Natuua Islands, 

 Borneo, Palawan, Mindoro, Imzon, the Bin Kin Islands ( = Loo Choo Islands), and 

 Japan. The specimens from all these localities agree so well with one another, 

 apart from individual peculiarities, that we did not succeed in finding in the genital 

 armature any character by which the specimens from one or the other j)lace could be 

 recognised. 



Tlie outline of the valve normally met with in sarpcdon sarpedon is represented 

 by f. 96, which is taken from a Kumaou individual. The siuus is about i mm. 

 deej) ; the dorsal lobe is rounded at the apex ; the ventral one is also rounded, its 

 ventral edge feebly incurved. The dorsal ridge is iu the dentition very variable ; 

 the imjiortant featnre is that normally the lateral fold h of f. 98 is very slight and 

 incomplete, as in f 99, or even absent. 



The variability in the form of the valve is illusti-atcd by f 96, 97, 11:!, and 

 114. In f. 97, taken from an example from the Shan States, tlie sinus is narrow, 

 and the apex of the ventral lobe also very narrow. F. 113 and 114 re[)reseut two 

 other Shan States specimens : one has the ventral lobe much jirodnced, aud the 

 sinus accordingly deep ; in the other the same lobe is very short and broadly 

 riiundeil. In the individuals from the JVIjilayaii region the ventral lobe is often, but 

 by no means regularly, more produced aud slenderer than iu the average Indian 

 specimen ; aud this is not surprising, as in the various subspecies from the lesser 

 Sunda Islands the lobes are constantly long and sleii<ler. 



The variation of the dentition of the dorsal ridge is very great, as will be seen 

 from f. 98 to 100 and 104 to 110. Sikkim specimens are represented by f. 99, 

 100, and 108; individuals from tlie Shan States by f. lO.) and 106; a Malacca 

 specimen by f. 107; two Bornean examples (from Mount Mulu) by f. 109 and 110, 



