BUTTERFLIES OF SAMOA AND SOME NEIGHBOURING ISLAND-GROUPS. 37 



border of all the wings, and in the whole region of the inner margin of the fore- 

 wings." The form differs from P. villida villida chiefly in its smaller size 

 (Table II), the mean expanse of males being 5 mm., and that of females 8 mm., 

 less in specimens of this race from Upolu than in those of the type form from 

 Tutnila. Specimens from Savai'i do not differ in any way from those from 

 Upolu, but, as I have only a few examples from the former island, I do not 

 quote the numbers ; the means quoted for specimens from Upolu and from 

 Tutuila are based on forty -three males and seventeen females from the former 

 island, and eleven males and seven females from the latter. In addition to 

 the difference in size, however, a form occurs in this race, more particularly in 

 the female, which is not mentioned by Rebel and has not been seen by me from 

 any locality outside Western Samoa ; in this form the ground-colour is strongly 

 suffused with orange-tawny scales of the same colour as the broad sub-terminal 

 band, so that in extreme specimens of the form the contrast between band and 

 ground-colour is almost entirely lost, and the two merge into one another. 

 Nearly all females from Western Samoa show some tendency towards this 

 form, but in the males it is less common and less well-developed. The orange 

 sub-terminal band of the hindwing, usually broken in P. villida villida, is generally 

 complete in the race samoensis. 



This insect provides the best example of seasonal prevalence that I was 

 able to observe in a Samoan butterfly ; it was very uncommon in the Apia 

 district in 1924, except from September to December, when it was rather 

 common ; it was abundant in the Aleipata district and in several localities 

 in Savai'i at the same time. It remained common in the Apia district until 

 towards the end of January, after which only odd specimens were seen there 

 until August, when it became common (sometimes abundant) in many localities 

 in this and other districts, and remained so until our departure in December. 

 It was never seen much above 1,000 feet, and only at that height where there 

 were large cleared spaces. It is attracted to the flowers of Lantana, Stachy- 

 tarpheta, Mimosa pudica, etc., but much more commonly is seen basking in the 

 sun on a road or other open space, frequently flying up to chase a passing rival. 

 On the two or three occasions when pairs were seen flying in copula, the female 

 carried the male. 



I was not able to obtain any of the earlier stages. Oviducal eggs, 

 however, were green and marked with twelve prominent longitudinal ridges ; 

 their height was about - 70 mm., and the diameter about 0*68 mm. Scaevola, 



