( 487 ) 



the longest, most bas.al, tooth (right-hand side in figure) is nearl}- horizontal. The 

 subdorsal process is simple, without denticnlation. 



Tliere is no difference in the genital armature between the darker and the less 

 dark specimens from Sikkim and Assam. 



1). P. aristcus hermocrates from Burma to Timor, Borneo, Palawan, and the 

 Philippine Islands ; f. 74, 7fi to 80. 



We have examined individuals from the Shan States, Borneo, Palawan, the Philip- 

 piine Islands, Sumba, Kalao, Wetter, and Timor. The individual variation in the 

 jiattern of the wings and in the size of the specimens is considerable, bnt we cannot 

 find distinguishing characters which would necessitate a division of hermocrates into 

 more subspecies ; the Wetter and Timor individuals, from which islands we have 

 only three altogether, are faintly different in the sha]ie of the forewing, and may 

 perhaps, on receipt of more material, be separable from hfrtnocrates. An examina- 

 tion of the genital armature likewise did not furnish us with characters by which 

 specimens from the various localities could be recognised, except the Shan States 

 individuals, which are more similar in jiart of the genital armature to the North 

 Indian anticrates. 



The chief distinction in the genital armature of hermocrafes concerns the form 

 of the dorsal lobe, which normally is shaped as in f. 78, and the form of the 

 ventral ridge, which has the basal tooth (f 74) less horizontal than it is in 

 anticvates. 



We have two individuals from Muong Gnow, Shan States, one darker than 

 many Philippine Islands specimens, one intermediate in the develojjment of the 

 bands befween antienites and ordinary Philippine hermocratfis. The dorsal lobe of 

 the dark specimen (f. 70) is scarcely different from that of anticratfs, being mucli 

 broader tlum it normally is in /lermocrates (compare f 76 to MJ) ; in the whiter 

 individual, which in the wing-pattern is scarcely different from dark Sikkim 

 examples o{ anticrates, the dorsal lobe is considerably narrower (f. 77). 



The series of f. 76 to SO represents the variation of the dorsal lobe. F. 78 

 is taken from a specimen from Kudat, North Borneo; the lobe is still narrower than 

 in f 77 ; the next two figures (79 and 80) represent the lobe of two Palawan 

 individuals. The specimens from Kalao, Timor, Wetter, and Sumlia have the lobes 

 as in f. 78 to 8o. 



The yradudl decrease in the size of the dorsal lobe from f. 7;") (aiiticrates) to 

 f. 76 {henmc rates), and from there to f. So, is evident; and it is further obvious 

 that the difference between the extreme form of the lobe of hermocrates (f. 80) and 

 the lobe of aristevs, as well as the difference between the other extreme of liermo- 

 rrateti (f. 76) and antirrates (f. 75), amounts (|uautitatively to much less than the 

 difference lietweeu the extremes oi hermocrates (f. 7() and %0). 



c. P. ar/fteus ar/steiis from the Southern and Nortlieni Jloliiccas ; f. 81 to 83. 



The individuals from the various islands of tlie Jloluccas, though individually 

 variable, exhibit neither in the wing-i)attern nor in the genital armature any 

 character confined to one or the other of the islands, or group of islands. This sub- 

 species is in jiattern most nearly related to hermocrates. bnt is easily distinguished 

 by the groiuid-colour of the underside being obviously darker. 



The dorsal lobe of the valve (f. 81 and 82) is in all sjiecimens examined 

 slenderer than in hermocrates ; tlie individual variation is slight. The ventral ridge 



