354 PLANT SOCIOLOGY 



lishing ''botanical regions." Their regions were founded upon purely 

 floristic materials, namely, the distribution of the families, genera, and 

 species of the vegetable kingdom. The son of the last-named botanist, 

 Alphonse de Candolle, subjected these floral regions to a sharp and 

 trenchant criticism. He recommended also the division of the large 

 and unwieldy regions into subregions, provinces, and districts, "down 

 to the localities, the smallest recognizable geographic units." 



The leaders of the plant-geographic movement during the last 

 decades of the last century — Engler and Drude, Flahault, Beck, and 

 others — followed the course outhned by De Candolle and extended his 

 nomenclature. l^ 



A division of the earth on a purely floristic basis leaves wholly 

 untouched a great many very important questions of plant sociology 

 and biogeography. The vegetational zones of Grisebach, physiog- 

 nomic in nature, are today easily located. But they are far too 

 superficial. They leave out of consideration the very important data 

 of plant sociology resulting from the investigations of the past decade. 



The plant communities themselves furnish the one natural basis for 

 a division of the earth for plant sociology. The associations and their 

 subdivisions — the alliances, orders, and classes of communities — are the 

 essential units. 



But since all phytosociological classification rests upon a floristic 

 foundation, it is exactly in this classification that systematic botany is 

 of the highest service. All geographically limited species, genera, and 

 famihes are related to certain communities, and these communities 

 indicate the boundaries of vegetational regions. In dehmiting these 

 regions, both the purely sociological and the systematic viewpoint can 

 and must be used. Thus it comes back indirectly, by way of the 

 plant communities, to the position taken by Drude and others: the 

 necessity of combining the floristic and phytosociological bases of 

 classification. 



In discovering and bounding vegetational regions thus conceived, 

 the following characteristics must be considered: 



1. The presence of unique or almost unique plant communities, their taxo- 

 nomic rank, and their ecologic and floristic speciaUzation, i.e., degree of organi- 

 zation of the communities and the relation of these communities to the chmax. 



2. The presence of extensions of foreign vegetational territories and the domi- 

 nance, suppression, or absence of certain communities or ecologic units (synusiae, 

 formations). 



3. The taxonomic position, number, and degree of development or ecological 

 speciaUzation of the unique, or almost unique, species of the region (endemism). 



4. The presence of less minutely locahzed, disjunct species and the dominance, 

 suppression, or absence of certain species or races. 



