( 481 ) 



and lias no teeth, except at the tip of the apical rod-like processus, which is longer 

 than in all the preceding subspecies. F. 137 represents the dorsal ridge with the 

 transverse ridge of a Eedscar Bay individual in an ajncal view. Another extreme 

 in the structure of the dorsal ridge is represented by f. 130 to 132, which are 

 taken from an example from Cairns, North Queensland. F. 130 gives a view 

 from the a])ical side ; the dorsal ridge d is provided with teeth, and the transverse 

 ridge /; is joined to the dorsal ridge, and is not a sei)arate structure as in f. 136; 

 f. 131 gives the ridges in a ventral view, while f. 132 is taken from above. Tiie 

 intergradations between the extremes figured here are eipially aliundant in Australia 

 and New Guinea. 



The individual variation of cliorrdon in the direction from f. 130 to 137, iVom 

 the transverse ridge forming one piece with the dorsal ridge (as in sarpcdoti 

 sarpedon, tefedoti,ju(/ans), to the other extreme where the transverse ridge stands 

 isolated, will serve to comprehend the still more exaggerated development of the 

 transverse ridge in some of the following subspecies. 



h. P. sarpedon imparilis from New Britain ; f. 138. 



The external difl'ereuces between this subspecies and the preceding one, though 

 slight, are prominent enough to enable ns to distinguish all our New Britain indi- 

 viduals from clioredon. In the genital armature imparilis comes again very close 

 to ckoredon ; the lobes of the valve are, however, more pointed (f. 138) ; the 

 transverse ridge is smaller, stands less obli([ue, and leans over towards the base of 

 the valve ; the dentition of the transverse ridge is as variable as in clioredon. 



The specimens with additional spots on the forewing have no character in the 

 valve and the armature peculiar to them. 



i. P. sarpedon impar from New Georgia, Solomon Islands. 

 The male of this insect is unknown; i]ie, female is in pattern midway between 

 the preceding and the following form. 



k. P. sarpedon isunder ivom Guadalcauar and Bougainville, tSolomou Islands: 

 f. 148. 



Though this form is so very aberrant in markings that it has been described 

 by Godman & Salvin as a distinct species, and has also been kept separate from 

 sarjjedon by Mr. Rothschild in his Revision (who, however, informs me now 

 that, in consequence of the receipt of more specimens, he must sink it to the rank 

 of a subspecies of sarpedoti), the valve and its armature are only slightly different 

 from that of imparilis. The ventral lobe of the valve is broader, blunter, and 

 shorter. The transverse ridge is not sc])arated from the dorsal ridge ; it is high, 

 mostly simple, and seldom notched or faintly dentate. F. 148 is taken from an 

 individual from Guadalcauar. 



1. P. sarpedon dodinyensis from the Northern Jloluccas (Halmaheira and 

 Batjan); f. 13U and 140. 



The sinus of tiie valve (f. 13ii) is very deep and especially narrow, being about 

 twice as deep as broail. The ventral lobe of the valve is shaped nearly as in 

 imparilis. The dorsal ridge is bifurcate as in clioredon, but the bifurcation takes 

 place not far from the apex of the valve; the ventral part of the ridge is raised and 

 forms the usual transverse ridge. From f. 140, which is taken from the same 



