The Bionomics of South African Insrcts. 425 



side of the wet phase the case would be strong and con- 

 vincing for adaptation, and an interpretation based on the 

 principles of warning colours or mimicry, Batesian or 

 Miillerian. But tlie under-side does more than this ; it 

 differs from the upper-side, and so far as it differs, it 

 becomes more cooisjacuous. The following details render 

 the case for adaptation in the direction of conspicuousness, 

 as it seems to me, overwhelming. 



In comparing the upper- nnd under-side of the wet 

 phase of the species to which the chief attention of 

 naturalists has been directed, it is of special interest to turn 

 to the accurate descriptions of Mr. Roland Trimen, F.R.S., 

 written long before Mr. Marshall's discovery was thought 

 of, a time when nakdcnsis and scsamtis were not only 

 considered distinct but were even separated by ceryne. In 

 '^ Rho'paloccra AfriciB Austrcdis" (London, 1862-66) we read, 

 on pages 130, 181, of Jnnonia odavia [Foxcis sesamus^] : 

 " Under-side. — Much paler, more creamy in tint, with a, 

 glistening pinlcisli, tinge." Again, on the under-side of 

 the fore-wing the author speaks of " the row of spots 

 parallel to hind-margin commencing distinctly from coster, 

 the first two spots increasing the number to seven ; 

 double row of bluish lunules more conspicuous than on 

 upper-side, whitei'." And the under-side of the hind-wing 

 is thus described : " Basal black containing four rather 

 large, very conspicuous spots of the ground-colour, and dusted 

 Vvith blue scales, which form a transverse streak between 

 costal and subcostal nervures near extremity of black ; 

 whitish-bluish lunules, in hind marginal border, large and 

 very conspicuous." The fact that the ground-colour of the 

 under-side is much paler than the salmon-red of the upper 

 and thus affords a far more effective contrast with the 

 black markings is seen when Fig. 1 on Plate XIII is com- 

 pared with Fig. l/> on Plate XII, Figs. 1 and 2 on Plate 

 XII represent worn specimens, and the comparison with 

 them is invalid. Fig Ih, Plate XII, possesses unusually 

 heavy black markings, but the representation of the depth 

 of the red ground-colour is normal for a fresh individual, as 

 is that of the undei^-side in Fig. 1, Plate XIII. The more 

 complete row of black spots and the greater conspicuous- 

 ness of the border, owing to the larger white and blue 

 markings in it, as described by Trimen, are also well seen 

 when the figures are compared, but allowance must be 



