BUTTERFLIES OF SAMOA AND SOME NEIGHBOURING ISLAND-GROUPS. 31 



Probably this is less than the true disproportion between the sexes at that 

 time, for a special watch was kept for females, and I hunted by preference in 

 the shady places where these were to be found, rather than in the more open 

 and sunny localities were the males were more common. Armstrong, in March 

 1925, saw males only, and of four specimens collected by Mrs. Cockerell at 

 Nukualofa, in July, three are males, the only. female being f. thomsoni. 



The maximum, mean and minimum expanse of twenty-four females from 

 various Tongan localities are 96, 83, and 72 mm. 



An interesting point in the habits of this species in Tonga (true also in 

 Tutuila) is that the female, unlike that of H. bolina inconstans, is of rather shy 

 and retiring habits. Specimens of this sex were never to be seen sitting high up 

 on a bush like the males, but always close to the ground and often quite in the 

 interior of the bush ; on several occasions one was seen to fly into the interior 

 of a patch of bushes, and emerge several yards away from its point of entry, 

 but males were never observed to do this. 



I failed to see much evidence of protective resemblance m Tonga ; the 

 male-like form can sometimes be mistaken in flight for E. eleuiho, but the 

 resemblance is greatest in old, worn specimens, whose value to the species is 

 presumably little or none ; moreover, the difference is usually readily seen on 

 account of the different manner of flight, and it is only very occasionally that 

 it is possible to mistake the two. At least one female was captured bearing an 

 obvious beak-mark on one wing ; this was of the right size and shape to have 

 been made by the large king-hunter (Halcyon sacra), which is common 

 in Tonga, and I do not know of any other Tongan insectivorous bird with a 

 beak large enough to have made the mark. 



A pale female was watched ovipositing on Sida ; the eggs were green. 



(ii) Forms occurring in American Samoa 



Of fifteen female specimens from Tutuila, which I have been able to examine 

 (not including those in the Kellers collection, which are dealt with separately), 

 eleven (one Dec. 1924, five Aug. 1925, and two Dec. 1925, Buxton and Hopkins ; 

 one Nov. or Dec. 1892, Bourke, and two April 1903, M. J. Nicholl) are of the 

 thomsoni form, the remaining four (all Aug. 1925) are of the form with dark- 

 suffused postdiscal band (murrayi). The maximum, mean, and minimum 

 expanse of these females are 89, 81, and 75 mm. A single female from Tau 

 Island, Feb. 1926 (A. F. Judd), is of the thomsoni form, and is 83 mm. in expanse. 



