40 



INSECTS OF SAMOA. 



followed by a lunular line and narrow terminal and sub-terminal borders, fused 

 anteriorly; a small angular line basally in interspace 1. Hind wing, a trans- 

 verse series of small postdiscal spots, absent in interspaces 1 and 6, and 

 obsolescent in interspace 4, followed by a lunular line and narrow terminal and 

 sub-terminal lines. Underside much paler, apex of forewing pale purplish- 

 brown ; markings more or less as on upperside, but very faint except the post- 

 discal spot in interspace 1 of the forewing, which is large and black, and the 

 postdiscal spots in interspaces 2 to 6 of the hindwing ; the latter are ringed 

 with deeper tawny and placed on a pearly-white band, beyond which is a line 

 of pearly lunules. 



Female (PI. II, fig. 14). Similar to male, ground-colour paler and all 

 black markings more extensive. The purple iridescence is absent. 



Maximum, mean, and minimum expanse of twenty- two males 50, 47, and 

 42 mm., of six females 53, 51, and 48 mm. 



Type male from Malololelei, Upolu Island, 2,000 feet, ll.x.25, female from 

 same locality, 10.vii.24 ; paratypes from same locality, various dates, and one 

 female from Mata Vanu, Savai'i, 1,500 feet. 



Variation is very slight and specimens from Savai'i do not differ from those 

 from Upolu. A short series of bred specimens are very small, and are not taken 

 into consideration in the above figures of the expanse. 



This very distinct form will possibly prove to be a race of A. alcip])e 

 (described from a specimen from Amboina), but is widely separated from it 

 geographically. The only Atella hitherto known to occur in Polynesia is A. 

 gaberti (not very closely related to the present form), of Tahiti, so that this 

 record is a notable extension of the known range of the genus. A species is 

 also known to exist in Papua. 



Unlike Issoria bowdenia, this species is strictly confined to the upper parts 

 of the islands ; it does not occur below about 1,000 feet, and is commonest 

 at about 2,000 feet, so that the two species have quite separate ranges, which, 

 however, overlap slightly at an elevation of about 1,000 feet. Common at 

 Malololelei throughout the year, and at an elevation of about 2,000 feet in 

 Savai'i (E. J. Bryan) ; I also took it in Savai'i at 1,500 feet, on Mata Vanu. 

 There are no records from Tutuila or from any localities outside Samoa. 



In habits it much resembles I. bowdenia, being generally found round the 

 food-plant, a small tree, Melicytus s^., probably M. ramiflorus Forst. (Violaceae) ; 

 it also frequents the flowers of Lantana and of Micania scandens (Compositae). 



