( ^«-' ) 



specimen as f. 139 and gives a view of the dorsal ridge from tlie apical side, it 

 will be observed that the dentate ridge differs essentially from the transverse ridge 

 of f. 103, 130, 137, 142, 144, in so far as it is not homologous to the fold h of 

 f. 98, but to the middle portion of the dentate dorsal ridge to which that fold is 

 joined. In a Batjan example the ridge, which in f. 140 is tridcntate, is much 

 reduced. The apical process of the dorsal ridge is long. 



m. 1'. mrpedon antltedon IVoin the Southern M(i]u<Tas (Amboina, Ceram); 

 f. 141 and 142. 



The sinns of the valve is twice as broad as in mrpedon dodimiensis, bnt not so 

 deep as in that form. The ventral lobe of the valve varies a good deal individually, 

 and has generally the outline of that of choredon or dodingcnsis ; the dorsal lobe is 

 broader than the ventral one, and rounded at the apex. The dorsal ridge has the 

 apical (rod-like) process prolonged, as is the case in dodingensis. The transverse 

 ridge is always strongly developed, and nearly always so obviously detached as in 

 f. 142, which is taken from a Ceram individual of which f. 141 represents the 

 valve and harpe from above. 



In external features anthedon is much more closely allied with dodhu/ensis than 

 with choredon; in the valve and its armature the difference between the former is, on 

 the contrary, much greater than that between anthedon and choredon. Anthedon is 

 distinguished from choredon only in the free apical process of the dorsal ridge 

 being longer, and in the transverse ridge leaning more evidently over basally, as 

 will be seen from comparing f. 135 and 141, and 137 and 142; while it differs 

 from dodingensis in having a broader dorsal lobe, a much wider sinns, and a differ- 

 ently shaped and differently situate transverse ridge. The relationship indicated by 

 pattern is therefore not the same as that which we must deduce from the structure 

 of the genital armature. 



u. P. sarpcdon monticolus from Bonthain Peak, S. Celebes ; f. 14<) and 147. 



The discovery of a mountain form of sarj)edon in Celebes which is in general 

 ajipearance very different from the form inhabiting the lower districts of the 

 island is highly interesting. The subspecies resembles on superficial examination 

 Indo-Malayau individuals of sarpcdon sarpedon ; the actual affinities of the sub- 

 sjiecies are, in respect to the external characters, as follows: In the green colour 

 of the markings monticolm — it ought to be monticola! — agrees with sarpedon 

 sarpedon, sarpedon jugans, sarpedon adonarensis, and sarpedon tinwrensis, and 

 differs conspicuously from its compatriot 7nilon and the Moluccan races. In 

 the shape of the wings it resembles sarpedon sarpedon from the Sunda Islands. 

 In the shape of the median band of the forewing it comes nearest to sarpedon 

 dodingensis from the Northern Moluccas, except in the third spot being larger 

 than the fourth, in which character it agrees with fimorenxis. The underside of tlie 

 hiudwing is in monticolus, milon (Celebes), and dodingensis (Jsorthern Moluccas), 

 in opjiositioii to all other subspecies, provided with a red mark before the median 

 cell between veins and 7. And, lastly, the genital armature of monticolus stands 

 intermediate in structure between that of milon and that of dodingensis, and hcMcc 

 is obviously different from that of sarpedon sarpedon. 



We have to lay great stress upon the mixture of characters found in 1'. sarpedon 

 monticolus. As the relationship to 7nilon from Celebes, .mrpedon from the Sunda 

 Islands, and dodingensis from the Northern Jfoluccas is equally great in the 



