IN DEFENSIVE COLORATION 331 



from A and also from B. It follows that A* + B* will 

 incur 20,000 losses out of 1,000,000, or 2 °/ o ; while the 

 two different patterns A and B will each lose 20,000 out 

 of 500,000, or 4 °/ o . 



The examples are of course unnatural in a high degree, 

 especially in the sharp separation of the patterns into 

 distinguishable and indistinguishable. But they enable 

 us to understand the advantage of a close over a rough 

 resemblance and thus to realize the causes which have 

 encouraged the growth of a Miillerian likeness through 

 a long series of generations. 



3. Striking Examples of Miillerian Resemblance. — 

 a. The Neiv World. — The statements in Section 1 on pp. 

 328-9 apply to the finished product, — the cases of perfect 

 resemblance which are supposed to be indistinguishable 

 to the inexperienced enemies. Although the mode of 

 flight certainly facilitates discrimination, 1 we must admit 

 that the patterns of Melinaea mneme and Heliconius 

 numata figured by Mr. W. J. Kaye 2 are almost exactly 

 alike. Their likeness reaches the climax of perfection as 

 nearly as it is ever reached by these superficial resem- 

 blances between members of distantly related groups. 

 The plate shows seven examples of the Ithomiine and 

 seven of the Heliconine. Each set is arranged to exhibit 

 the gradual transition from a barred to a black hind wing. 

 Each of the seven varieties of the Ithomiine is matched 

 by a corresponding variety of the Heliconine. When 

 the two patterns are compared it is at once clear that the 

 Melinaea {Ithomiinae) has acted as the model. It still 

 presents an entirely characteristic Melinaea pattern which 

 the Heliconius has adopted. There is no reason for 

 believing that the latter is more palatable than its model. 

 In addition to their mimicry of the Ithomiinae the Helico- 



1 H. W. Bates pointed out in the original memoir on Mimicry that the 

 Heliconines ' move along in a sailing manner, often circling round for a 

 considerable time, with their wings horizontally extended'. The Itho- 

 miines, on the other hand, ' for the most part, keep near the ground, and 

 have a very slow irregular flight, settling frequently.' Trans. Linn. Soc, 

 Lond., vol. xxiii (1862), p. 499. A small but distinct means of discrimi- 

 nation between the patterns is mentioned on p. 350 of the present work. 



3 Trans. Eni. Soc., Lond., 1906, pi. xxvi. 



