92 VARIABILITY 



of microbic species at once probable and in accord with all we 

 know of nature. Studying these changes, it becomes evident that 

 Natural Selection is operative only under certain conditions. The 

 germ - plasm must be variable ; but it must be spontaneously 

 variable, not variable under the direct action of the environment. 

 Spontaneous variations, occurring all round the specific mean, 

 afford materials for Natural Selection and tend to enable the species 

 to meet all contingencies. But variations caused by the direct 

 action of the environment have an opposite effect. Under condi- 

 tions nearly the same for all the individuals exposed to them, they 

 cannot occur all round the specific mean, but in one general 

 direction only. Therefore they cause a drift, which may or may 

 not be adaptive, but which, since a new environment is usually 

 unfavourable and therefore injurious, can be adaptive only as a 

 most unlikely coincidence. Even if such a drift were favourable 

 at first, the accumulation, the exaggeration of it during succeeding 

 generations would be sure ultimately to ruin the co-adaption of 

 parts and functions and so render it unfavourable ; for there is no 

 imaginable reason to suppose that if such a drift occurred in 

 ancestors it would not continue in descendants. Virulence is not 

 the only adaptation found in the microbe. The whole organism 

 is compounded of adaptations, every one of which is co-adapted to 

 all the others. Except by the theory of evolution through Natural 

 Selection it is as hard to account for any of them as it is to account 

 for virulence. As we see, Natural Selection cannot control a drift 

 in which every individual participates. Only in one way can it 

 be prevented by the germ-plasm becoming highly insusceptible, 

 highly resistant to the direct action of the environment. We are 

 driven^ then, to the conclusion that the germ-plasm is both spon- 

 taneously variable and highly resistant to the direct action of the 

 environment. In other words, we must believe that in any species 

 that is not undergoing extinction, spontaneous variations greatly 

 preponderate over those which are caused by the direct action of 

 the environment. 



148. The evidence is conclusive that the germ-plasm of all 

 persistent species possesses a high power of resisting enforced 

 change. For example, some probably all human diseases are of 

 great antiquity. Descriptions of them written two or three 

 thousands of years ago are accurate for the present day. During 

 all that time, though exposed to enzymes and other potent 

 influences, the microbes, minute atoms of naked protoplasm, 

 have not altered appreciably. Fitted closely to a stable environ- 



