COCKROACHES, ETC. 133 



prospect of food as well as fear of attack and it is 

 significant that the same appears to be the case with 

 poisonous serpents. The Cobra spreads its hood to 

 warn off attack, and behaves in exactly the same way 

 if a rabbit is introduced into its cage, but both with 

 snakes and Mantidce the primary meaning of the display 

 is warning. On the part of the Mantidce it is a case 

 of " bluff," for these insects have no poison to inject 

 into their foes, as wasps and bees have, nor is there 

 any reason at 'all to suppose that they are unpalat- 

 able as so many insects undoubtedly are ; but such 

 examples of "bluff" are common enough amongst 

 insects, as any -one who will read Professor Poulton's 

 delightful work, The Colours of Animals, can find out 

 for himself, and they need not cause us any surprise. 



It is possible to trace fairly completely the evolution 

 of these "displays" from their early beginnings to 

 the wonderful exhibition given by an African flower- 

 mimicking species Idolum diabolicum. A green Hiero- 

 dula, at the approach of prey or when irritated, will, 

 as already stated, raise the prothorax and draw up 

 the fore-legs against its sides, displaying to view the 

 ochre-yellow of the coxae and femora. I do not 

 regard this in the nature of a warning attitude at all. 

 It is usual in green insects that parts not exposed to 

 view when the insects are at rest are coloured less 

 deeply than the exposed parts, and the attitude of 

 the Hierodula just described is assumed because it is 

 the most favourable one for making a sudden rapid 

 snatch at its prey, but here is the germ on which 

 natural selection has worked. In Deroplatys desiccaia 

 the same attitude is assumed at times of excitement, 

 but here the inner side of the front femora is con- 



