BUTTERFLIES OF SAMOA AND SOME NEIGHBOURING ISLAND-GROUPS. 47 



not seem to occur north of Samoa. The Tongan form appears to differ con- 

 stantly from other races, and may be described thus : 



Male (PI. I, fig. 5). Upperside white. On forewing a rounded black 

 spot at apex of cell, apical area and termen black, with a subterminal series 

 of seven white spots, the first three of which are elongate and the four posterior 

 rounded, the second spot (that at the apex) being very small and narrow. 

 Hindwing ; a black terminal band, narrowing posteriorly and bearing a series 

 of spots of the ground-colour, which appear yellowish by transparency ; the 

 black along the veins of the underside also shows through. Underside as 

 upperside, but all the black markings more extensive. Forewing with costa 

 and discocellulars black. Hindwing, all the veins broadly edged with black, 

 the areas of ground-colour along the costa and inner margin, and the subterminal 

 series of spots all strongly tinged with chrome-yellow, a small area of the same 

 colour in the cell. 



Female (PI. I, fig. 6). Similar to male, but all black markings more 

 extensive, and sooty-black instead of jet-black as in the male. Upperside 

 creamy white ; forewing costa edged with dusky black, and a broad curved 

 streak of black along the discocellulars ; hindwing, veins narrowly (dis- 

 cocellulars broadly) edged with dusky black. Underside as in male, but 

 ground-colour yellowish- white, subterminal spots of forewing yellowish, and 

 black markings more extensive. 



Maximum, mean, and minimum expanse of ten males 66, 62, and 58 mm., 

 of six females 68, 63, and 59 mm. Types and six paratypes (five male and one 

 female) from Neiafu, Vavau, Tonga, 7.iii.26 (J. S. Armstrong) ; I have also 

 two very poor males, captured at the same locality on 12.ii.25 by myself, and 

 have taken into consideration two males and two females from Vavau, collected 

 by the Eclipse Expedition in April or May 1911, and two females from Tonga 

 (exact locality not given), 1889 (J. J. Lister). I have named the race after 

 Schmeltz, who was the first to record it and point out some of the distinctions 

 between it and the Fijian race. 



This race is evidently very closely related to B.j. micronesia Fruhst., of Fiji, 

 but almost all Fijian specimens (as pointed out by Schmeltz) have the sub- 

 marginal spots on the underside of a greenish-yellow, totally different from the 

 deep chrome-yellow of B. java schmeltzi ; a few Fijian specimens have the 

 underside spots yellow instead of greenish, but never of the deep colour typical 

 of Tongan specimens ; the area of the colour is also less than in Tongan 



