( 26 ) 



tion, that frobenia was first isolated, dumetorum second, 

 ■comorarnm third, and mayottensis fourth. "We must if possible 

 distinguish in each case between ancestral characteis which 

 have persisted from a period previous to the isolation, and 

 recent characters which have been evolved since its beginning. 

 The former are of special interest in that they suggest to us 

 some of the features which saclava has now lost in ^Madagascar. 

 a. Ancestral Features. — 1. The Band Crossing the Hind- 

 Wing. — Saclava in Madagascar has a considerably broader 

 band than on the mainland of Africa. It is probable that 

 the Madagascar saclava also formerly possessed a narrower 

 band, which still persists in three out of the four island forms. 

 It is probable that the band of saclava has been widened in 

 consequence of synaposematic approach to the other species of 

 Neptis in Madagascar, and perhaps partially under the influ- 

 ence of Amaur'is nossinia, Ward, in which the white markings 

 are so strongly developed. The synaposematic sensitiveness of 

 Nepiis to the African species of Amauris as well as to other 

 Danainx in other regions has been referred to in Trans. Ent. 

 :Soc. Lond. 1902, pp. 467-8. Nei^tis hiJddeli, Boisd., of Mada- 

 gascar, described as a common species by Mabille, is distin- 

 guished by a great development of the white markings, and 



Mr. Grose-Smith informs me that he possesses an allied un- 

 described species from the same island with even broader 

 bands, especially upon the posterior wings. Furthermore 

 Neptis metella, D. and H., is represented in Madagascar by 

 the form gratilla, Mab., in which, as in mclava, the white 

 markings ai'e larger than in the African form. Another black 

 and white species, i^^. sextilla, Mab., is vinknown to me. There 

 is therefore clear evidence of the existence in Madagascar of 

 a powerful combination in which the white markings are 

 especially well developed. 



2. The Taivny Markings of the Island Species. — It is diffi- 

 cult to decide between the two alternative interpretations of 

 the difference in colour between saclava and the four species 

 in the outlying islands. If white markings are ancestral in 

 saclava it follows that the tawny colour has been evolved 

 independently in the north-western and eastern species. 



