86 THE THEORIES OF EVOLUTION 



low a certain level, it is not evident; above that level, 

 the development may be exaggerated and even de- 

 feat its own purpose. We have cited examples in 

 which an inadequate resemblance could render no 

 service; in other cases it is too perfect and involves a 

 wealth of superfluous details. A large number of 

 South American and Indian butterflies, living in the 

 -^ forests, mimic very closely the leaves of the various 

 trees. Their mimicry of colouring is remarkable; 

 more remarkable, however, is their mimicry of the 

 general shape and venation of leaves. The wings 

 are relatively more developed than the body which, 

 at rest, is hardly visible. The tips of the wings taper 

 down to a petiole, the venation is very delicate and 

 concealed by a sort of artificial venation much 

 coarser in appearance and streaked to imitate the ribs 

 of a leaf. The resemblance is only noticeable when 

 the insect is at rest; for the markings do not extend 

 to the part of the wing which is hidden in this posi- 

 tion. They look as though they had been painted 

 over the folded wing with one stroke of a brush. 

 From the many cases of mimicry observable we will 

 select two very striking ones. 



In the Coenophlebia Arclii&ona, a Bolivian butter- 

 fly, the petiole is formed by the tips of the anterior 

 wings ; a long middle rib and two lateral ones divide 

 the surface of the wings. In the Kallima jmrallecta, 

 a butterfly of the Malayan Archipelago, frequently 



