— 408 — 



in the dune populations an accordance to the extreme conditions 

 can be recognized: in the small transpiration surfaces of the leaves, 

 in the somewhat fleshy leaves, in the large and deep going roots, 

 in the plentiful anthocyanin in stems and leaves, in the extensive 

 ramification in the earth's crust, by which they become perennial, 

 and in the transversely geotropic stems, they exhibit genotypically 

 conditioned characters that are of great importance to the piants 

 at these localities. 



The older L am ar ckian interpretation regarded these "adapt- 

 ations" as a direct eiTect of the extreme conditions. The Selec- 

 tionism saw an effect of the continued selection between the 

 fluctuating variations. That there is a fitness to function can- 

 not be denied. The question is merely in which way it has come 

 into existence. 



The selection cannot create new types (Johannsen) but it 

 can select the fittest among already existing types or among 

 types that arise af ter crosses, in other words: it can select 

 the fittest combinations of the existing genes, but it 

 cannot create new genes. 



As to the supposed direct effect of the conditions, we neither 

 have been able to trace such an effect nor disprove it. It agrees 

 badly with our present apprehensions concerning genotypic struc- 

 tures and genotypic constancy. The only instance we know of 

 genotypic inconstancy is the mutations. In which way they 

 take place and by which influences they are induced we do not 

 know. And as to the nature of mutations we know nothing. The 

 only way in which we might suppose that the conditions could 

 aet upon the genotypic structure is through the mutations, either 

 by increasing the number of mutations or by affecting the direction 

 of the mutations. 



But we need not a supposition of a direct effect of the external 

 conditions to explain the parallelism between the extremeness 

 of type and the extremeness of locality. This might be explained 

 as an effect of mutation, crosses and selection solely. The muta- 

 tions, crosses and segregations mainly provide the material for 

 the selection, and the selection itself brings about the ''adapta- 

 tion". Occasionally, crosses between different species occur 

 and supply one species with a character from another that enables 

 it to live in piaces, where it previously could not liA^e. In Viola 

 tricolor and arvensis this is unusually frequent {tricolor on basic 

 soil and arvensis on acid soil). 



