104 



The Plant World. 



same phenomenon. Further, it appears that the optimum areas 

 of scores and hundreds of species nearly coincide. In brief, 

 actual plant distribution through its lack of uniformity, and 

 homogeneity, its tendency to concentric dispersal, and the co- 

 incidence of the optimum areas of many species, seems to demand 

 larger units in harmony with the processes, structure, composi- 

 tion, and origin of its components. Whether we choose to call 

 them "centers of dispersal" is of small moment as compared 

 with the recognition of the fact that zones and their subdivisions 

 are not natural organizations of plants or plant formations. 

 Of course this criticism does not refer to the use of the very 

 convenient expression of certain spatial and temperature con- 

 cepts, viz., torrid, temperate and polar zones. 



The unsatisfactory character of zones as a basis for classi- 

 fication is felt also by students of climatology. Especially is 

 this true of the classification of the climatic types prevailing on 

 continents. The schemes of Hult (1892-3), Koppen (1900), 

 Supan (1903) and Herbertson (1905) are especially interesting 

 in this connection.* Further, the provinces pointed out by 

 Supan for North America show a remarkable coincidence with 

 the natural vegetation centers. If future work both on climate 

 and plant distribution should bring these fundamentally different 

 view points into essential agreement we should have the possi- 

 bility of a completely dynamic and genetic system of vegetation 

 and climatic units. 



* Ward, R. De. C.. Classification of Climates. Bull. Am. Geog. Soc. 38: 401-412. 465- 



477. 



