1913.] OF THE UNITED STATES. 275 



ally sound to study the relations which obtain between vegetational 

 areas and climatic areas. This sort of relation is truly only a spatial 

 one, however, and must not be assumed to be causal. The proba- 

 bility that such a relation is a causal one is of course increased as it 

 is found to hold in a large number of cases. With agricultural plants 

 the historic factor need not be considered; the actual experimental 

 test as to wdiether a given form will thrive in a given area is some- 

 what readily made and the results are clear enough. 



Johns Hopkins University, 

 April i8, 1913. 



