TRACHEATES 2 1 9 



is suspended over the victim by a thread, so that the 

 convulsive movements of the caterpillar may not injure 

 the larva, which can always retreat by the thread. 

 Here again we must discard the supposition that the 

 wasp executes all these details because it knows what 

 will be good for the larva. And the instinct cannot 

 have been perfected by practice, because they lay very 

 few eggs in the course of their life. 



There are, moreover, plenty of instincts that only 

 act once in the lifetime. In these cases it is frequently 

 clear that the Lamarckian principle entirely fails, because 

 here there is no practice whatever and consequent 

 improvement. 



An insect only passes into the pupa stage once in 

 the course of life, so that this act and the preparations 

 for it have not been practised. But these preparations 

 are amongst the most wonderful and most complex 

 instincts. The matter is comparatively simple in many 

 of the day butterflies, such as the common white ones. 

 These, while in the caterpillar stage, spin a thread 

 round themselves, and hang down by it from a wall. 

 The thread must be just as long as the thickness of 

 the pupa, otherwise it may either be pressed or fall 

 out. How can the caterpillar know how thick it will 

 be in the pupa stage ? And how can it practise spinning 

 its thread when it only does this once in life ? 



The larva of the large stag-beetle passes into the 

 pupa stage in a hollow ball of clay prepared by itself 

 with polished walls. In the insect that will become 

 the male this is much longer, because of the large 

 antler-shaped jaws, than in the case of the future females, 



