354 THE VARIATION OF ANIMALS IN NATURE 



evidently limited and adjustment is a slow process. Apart 

 from functional adaptation within the lifetime of the indi- 

 vidual, the change requires at least one generation to modify 

 a whole population. Variation and the multiplication of 

 variant individuals are therefore, in regard to structure, the 

 main method of response, which is necessarily slow. 



On the physiological, and especially on the psychological 

 plane, functional adaptation becomes more and more important. 

 We mean that deficiency in one respect is made up for by 

 a compensatory change elsewhere. The co-ordination of an 

 animal's physiological activities essentially consists in keeping 

 a balance, within certain wide limits, between all the separate 

 activities, so that the internal environment of the organism is 

 stabilised. The psychological activities or behaviour (we are 

 not at present considering consciousness) of an animal are 

 even less fixed, because the number of ways in which the 

 problems can be answered are so much greater. It is a common- 

 place that the behaviour of all the more specialised animals 

 has an element of unpredictability. This element is perhaps 

 fundamental and not due to a mere temporary lack of data. 

 The frequency of any one type of behaviour may be recorded 

 without arriving at the possibility of prediction for a par- 

 ticular case. Thus, in Reinhard's experiment (1929, pp. 128- 

 130) on a wasp (Philanthus gibbosus) a female was confined 

 in the centre of three concentric glass funnels standing on 

 sand. On her first attempt she burrowed under the edge of 

 the inner one and ran up between it and the second ; on trials 

 2 to 15 she burrowed under all three funnels ; on trial 16 she 

 behaved as on the first occasion ; while on trials 17 to 22 she 

 ran straight up the neck of the inner funnel. After each trial 

 she was recaptured and placed in the centre again, till, on the 

 twenty-second escape, she eluded capture. 



Even the most specialised behaviour (e.g. oviposition) 

 involves to a greater or less extent the whole organism. A 

 living organism is an exceedingly flexible instrument and has 

 many ways of attaining the same end. Very similar ideas have 

 been expressed by Elton (1930, p. 31), who sees two processes 

 at work in at any rate the higher animals : ' the selection of 

 the environment by the animal,' as well as ' the natural selection 

 of the animal by the environment.' Elton emphasises the 

 ability of nearly all animals to wander, often to migrate over 



