CAUSES OF VARIATION loi 



Vivendi between an extremely complex and changeful animate 

 system and the extremely complex and changeful environment 

 in which it lives and moves and has its being. In all viable 

 organisms this equilibration has been established, and it is plain 

 that those organisms which could secure an entailment of this 

 equilibration would be the organisms to survive. The producers 

 of survivable descendants survive in them — an obvious economy 

 of successful experiment, if such a point of view can be enter- 

 tained. 



We have seen that during the early stages of development 

 there is often a visible segregation of a lineage of germ-cells which 

 do not share in body-making, but continue like the fertilised 

 ovum. This distinction between somatic cells which undergo 

 differentiation and germ-cells which retain the heritable qualities 

 intact is obviously an advantageous method of entailing on suc- 

 cessive generations that valuable asset which we have called 

 organic equilibration. It also economises and facilitates the 

 process of reproduction. 



But in spite of this almost universal device, the general tend- 

 ency of which is to secure persistence, continuity, and complete 

 hereditary resemblance, there is abundant opportunity left for 

 the assertion of that variability which we believe to be a primary 

 quality of vital units. Thus an inquiry into the causes of varia- 

 tion seems to us to be in the main an inquiry into the oppor- 

 tunities for the reassertion of a pristine tendency which the 

 continuity of the germ-plasm has to some extent restricted. The 

 stream of life passing through a continuous lineage of germ-cells 

 is, so to speak, hemmed in, but it continually tends to deviate 

 from this course, and there are not a few opportunities — some 

 normally recurrent, some more accidental — which allow of this 

 or even prompt it. In some cases, as we have said, it is impossible 

 to distinguish offspring from parent, or brother from brother, 

 or cousin from cousin. On what does this completeness of heredi- 

 tary resemblance {i.e. the absence of variation) depend ? 



