group of Btdter flies, Potaro District, British Guiana. 429 



the whole of Central and South America with Cuba and 

 Haiti, It is true most of the forms are known as distinct 

 species, but if we only could ascertain where the forms 

 overlapped such as we know atergatis and ceres overlap in 

 Trinidad, there could be no possible doubt that the species 

 was continuous from Mexico to Southern Brazil and from 

 Peru to Para, and that we should find the forms merginor 

 one mto another. With such a species as this it should 

 be possible to decide by experiment whether all different 

 climates and different climatic conditions tend to produce 

 different geographical races as we know them, or whether 

 the action of mimicry by selection is sufficient to account 

 for these changes. That such a species should adopt 

 so many tones of colour cannot be, because the forest has 

 such a varying aspect, and that the insect when on the 

 wing is variously concealed. The species in several of its 

 phases (? all) is extremely conspicuous in flight, and in 

 Guiana at least it is very conspicuous at rest on the 

 flowers of Bupatoriwn macrophylhim. This Eupatorium 

 is a widely distributed and common plant, and it is quite 

 likely that in other parts of the continent it is frequented 

 by this same Lycorea under various guises. 



Lycorea pasinuntia. 



This species is certainly commoner than the previous 

 but its variation is not so considerable. And with regard 

 to the complete darkening of the area of the band in the 

 hind- wing it is a comparatively rare feature. 



The variation works out thus with 27 examples : — • 



* This is the extreme form, but it is not nearly so complete as in 

 L. ceres. 



t Under this class are placed all those that are not completely 

 banded, but the average is not so great as with L. ceres. 



