378 MESSRS. DixET, BUE.B, AND pickard-cambridge ON [May 3, 



The variations of the underside of the hind wing in both forms, 

 B. iliihyia and B. gotzins, have been well described by Trimen 

 {loc. cit. pp. 265, 266). They are undoubtedly seasonal, as 

 pointed out by Barker (Trans. Ent. Soc. Lond. 1895, p. 415), and 

 by Marshall in the MS. notes and labels accompanying the series 

 in the Hope Collection above referred to, as well as in the 

 Annals & Mag. Nat. Hist. 1896, xviii. p. 333, &c. The deeply 

 ferruginous hind wing, on which the three creamy bands stand out 

 conspicuously, belongs in each case to the dry-season form, and 

 there are several intermediate grades leading up to the dull 

 ochreous yellow of the wet-season form. In addition to the 

 points noted by Trimen, it may be remarked that in the wet- 

 season form of B. gotzius the black submarginal band of the hind 

 wing is relati\ ely broader, and tlie proximally adjacent strip of 

 ochreous ground-colour narrower, than in the wet-season form of 

 B. ilithyia. In the former, indeed, the band of ground-colour is 

 often reduced to a mere chain of fulvous dots with dark edging, 

 forming a proximal border to the dark submarginal band. The 

 pairs of whitish internervular spots on the dark band are also 

 much less regular and conspicuous than in B. iliihyia. In the 

 dry-season forms the veins crossing the median creamy band are 

 in B. gotzius often traced out with the deep ferruginous tint of 

 the ground-colour, which marking has the effect of dividing the 

 median creamy band into spots ; this is not seen in B. ilithyia. 

 In B. gotzius also the dark submarginal band seems never entirely 

 to disappear, even in extreme dry-season forms, as it may do in 

 B. ilithyia. It soon, however, loses the whitish internervular 

 spots, which in the wet-season form are already less distinct than 

 in B. ilithyia. 



The Socotran B. hoydi resembles most specimens of B. gotzius 

 from the West African subregion in having the dark costal bar of 

 the fore wing continued rather heavily across the wing to join the 

 submarginal baud. This is also more or less the case with two 

 females of B. gotzius from Abyssinia and specimens of the same 

 from Somahland and Aden in the British Museum ; but in 

 examples from South and East Africa the connection between the 

 costal and the submarginal dark bands is often slight or absent. On 

 the other hand, in the submarginal series of spots of the fulvous 

 ground-colour on the upperside of the hind Aving, the Socotran 

 form comes nearer to specimens of B. gotzius from Somaliland, 

 Aden, and the Galla country than to any I have seen from West 



African coast-line at a remote period." Messrs. Sclater and Hartlawb (Proc. 

 Zool. Soc. 1881, p. 167) point out that Z'r^/y/^a?ca hcesitata, one of Prof. Balfour's 

 Socotran birds described by them, is most closely allied to a form inhabiting 

 Madagascar. Col. Godwin-Austen (Proc. Zool. Soc. 1881, p. 252) considers 

 that the laud-molluscan fauna of Socotra affords '" strong evidence that the 

 island was once directly connected with Madagascar to the south"; and adds 

 that "it is not unreasonable to suppose that in Socotra, the Seychelles, 

 Madagascar, and Rodriguez we have the remnants of a very ancient more 

 advanced coast-line on this western side of the Indian Ocean." 



[HI 



