450 DIRECTIVE FACTORS IN EVOLUTION: 



Types may have their waywardness gradually sifted out of 

 them. The uniformity of the flow of cartridges from a test- 

 ing machine gives a fallacious impression unless we dis- 

 cover that they have passed through three siftings which 

 reject the too heavy and the too light, the too long and the 

 too short, and those whose calibre is too broad or too 

 narrow. 



On the other hand, one of the impressions that we get 

 from Prof. D'Arcy Thompson's magistral work on Growth 

 and Form is that the variability of organisms runs on lines 

 laid down by the conditions of the inorganic. Variations 

 must conform to the trammels of surface-tension, minimal 

 areas, stability, and so on ; there is not an indefinite number 

 of ways in which an aggregate of cells can be arranged ; 

 one skull or leaf often differs from a related form in a way 

 which might be described as a general deformation due, 

 for instance, to a tilting of axes. The same general impres- 

 sion of definiteness we get from considering what we have 

 alluded to as temporal variations: one species often seems 

 to differ from another in rate or tempo, and this fits in with 

 Prof. D'Arcy Thompson's morphological illustrations, for dif- 

 ferences of form depend in great part on different rates of 

 growth in different directions. 



But even mutations and definite orthogenetic variations 

 cannot dispense with the criticism of Natural Selection. 

 There is ever a risk that they may go too far. It is easy to 

 have too much of a good thing. The antlers of the Irish 

 Elk which hastened the doom of their possessors are diagrams 

 of the evolutionary adage Nequid nimis. If we accept De 

 Vries's view that evolution is often effected by mutation, 

 by sudden considerable jumps, this is contrary to the idea 

 of Natural Selection working as an accumulator of small 



