78 BASIS OF THE PEOBLEM 



the average germinal constitution of the progeny. It is only in 

 so far as selection is a selection of strains that there will be any 

 change. If the strains among the beans referred to above can be 

 isolated, then the strain with the predisposition towards the 

 greatest weight can be isolated, the other strains eliminated, and 

 the average weight brought up to that found in this strain. If 

 the strains cannot be isolated, then an approximation only to 

 this result can be made by the continued selection of strains of 

 the greatest weight, and this is what occurs in nature. 



This leads to the most important conclusion of all. Selection 

 can only act upon what is given. We may put aside the possible 

 but quite unproved influence of selection upon the direction of 

 mutation. Apart from this, selection can have no other effect 

 than to eliminate certain strains and to favour others. Supposing 

 that under changing circumstances a certain character is favoured, 

 all that can happen through selection is that the strain exhibiting 

 this character will be favoured and all other strains gradually 

 eliminated. Further than this change cannot proceed until 

 mutations have occurred in the direction of further increase in 

 this character. As has been said, selection is no more than 

 a sieve ; as a rule it is occupied in separating out the extreme 

 • types and favouring the mean, but at times under certain circum- 

 stances it will favour some type that deviates from the mean, but 

 it can do no more than favour what types are at any given moment 

 in existence.^ 



10. We set out to answer two questions regarding quality. 

 We asked in what change among species in a state of nature 

 consisted. The answer to this is clear ; it consists in changes 

 in the germinal constitution. If we can speak of history in 

 connexion with species in a state of nature, then their history is 

 based solely upon germinal change. We also asked how change 

 came about. It is less easy to answer in a few words. Obscure 



' A certain amount of confusion has been introduced into the whole question 

 of evolution by the manner in which in recent biological writings the function of 

 selection has been depreciated. The followers of Darwin undoubtedly at times 

 exaggerated the importance of selection. In particular selection has even been 

 spoken of as though it caused variation and, as it were, drew organisms along 

 certain paths. It is essential to remember, as has been emphasized above, that 

 it can only act upon what is given. But without doubt there has been recently 

 a tendency to underestimate the importance of selection. Though the process 

 of selection is nothing more than a process of sifting, that process is of fundamental 

 importance ; unless due weight is given to it, evolution cannot be rendered 

 comprehens ible. 



