35 8 THE VARIATION OF ANIMALS IN NATURE 



animal depends on a great variety of physiological processes, 

 each of which, considered in vacuo, can be carried out most 

 efficiently in a particular environment. The optimum en- 

 vironment is, therefore, a statistical conception involving 

 a compromise between a number of conflicting ideals. Even 

 in an unvarying environment the compromise is likely to be 

 an unstable equilibrium, and in a state of nature, where all 

 factors are undergoing big fluctuations with a period relatively 

 short compared with the developmental period of the species, 

 it is doubtful if any real equilibrium can be reached. In these 

 circumstances there will be a wide range of conditions under 

 which the species will be as well adapted as it ever can be. 

 On the one hand adaptation can rarely and only for short 

 periods be very close, while, on the other, selection will have 

 a permanent effect only when the maladjustment to the 

 environment has become unusually gross. 



Actually, in the course of evolution, increase in organisa- 

 tion makes 'the conception of optimum conditions more and 

 more precise, but this results from the organism making its 

 own environment which is ipso facto optimum. In recent 

 years man has made great progress in the art of maintain- 

 ing the atmosphere of his houses at the proper temperature 

 and humidity, and an essentially parallel process can be seen 

 in evolution. The establishment of approximately similar 

 optima for the various bodily processes is an important step 

 which has been made by the homoiothermic animals in which 

 the blood-stream has a relatively uniform constitution. In the 

 insects this stage does not appear to have been reached, and 

 only a very broad definition can be given to the optimum. 

 Thus each stage (egg, larva, pupa and adult) may have 

 different requirements, as found by Headlee (19 17, 1921) in 

 the bean-weevil (Bruchus obtectus) which lives, nevertheless, in 

 a much more constant environment than most species. Again, 

 the optimum will differ according to which stage of activity is 

 regarded. Thus Weber (1 931), in the whitefly {Tnaleurodes 

 vaporariorum), finds that the optimum temperature for the sur- 

 vival of the last larval stage is 22 C, while the optimum for 

 oviposition in the adult female is 2 5 o - 3 C. Maclagan 

 (1002a) in the spring-tail (Smynthurus vindis) finds that the 

 optimum temperature for growth is 16-7° C, while for egg- 

 production it is 7 C. A comparable temperature effect is 



