Combinations of Trojncal American Butterfiies. 601 



iinder surface of the fore-wing, while these markings are 

 absent or occasionally just visible in examples of the same 

 species from Peru and Ecuador. Here the ancestral 

 feature, obsolete in more northern localities, is preserved 

 in the south. In some Bolivian specimens these spots 

 extend round the hind margins of both wings. 



Other species showing a closer resemblance within the 

 association are seen in the moths, Ferico'pis liydra (Figs. 

 18 and 19), Castnia 'pellonia, and the Papilio (Figs. 16 and 

 17). In all these, yellow markings appear at the costa of 

 the fore- wing extending more or less completely along the 

 outer margin of the oblique tawny band, and, except in the 

 Castnia, yellow spots are developed along the hind margin 

 of the hind-wing. These features are generally wanting 

 from Combination IV, although, as regards the fore-wing, 

 JEresia ithomiola ^ (Figs 12 and 18) approaches the Papilio 

 and the moths. The resemblance of the Castnia ioPapilio 

 hachus is much closer than that of the Pericopis. The 

 yellow outer border of the orange-tawny oblique band of 

 the fore-wing in the above constituents of Combination 

 IV, as also in the majority of the specimens of Napcogenes 

 aclixa, is undoubtedly transitional towards Blandford's 

 "7. Central Colombian modification" of "6. Ecuador 

 Type," — the latter name being applied by him to the 

 association now being considered. From the evolutionary 

 point of view, however, the yellow fore-wing marking of 

 the Central Colombian association is certainly ancestral, 

 and its absence in Ecuador, etc., a comparatively recent 

 modification. The relationship between these two com- 

 binations, distinguished by the presence or absence of the 

 yellow margin to the oblique fore-wing band was clearly 

 pointed out by H, W. Bates in the historic memoir on 

 Mimicry (Trans. Linn. Soc, Lond., 1862, vol. xxiii, PI. 

 Ill, p. 514) : — " Some of the close resemblances amongst 

 the Heliconidie themselves seem to be kept up by their 

 varying in a precisely similar way. There is a very 

 singular instance in three species of three different genera, 

 Melinxa, Medianitis and Heliconius, which are all in East 

 Peru, orange and black in colour, and in New Granada 

 orange, black and yellow. This seems to be a case of 

 coincident, simple variation ; for if three forms are quite 

 alike in colours, it is conceivable that they may vary alike 

 when placed under new conditions by migration. Our 

 Leptalides have been shown not to vary precisely like their 



