THE CLASSIFICATION OF COMMUNITIES 363 



As the basis of comparison, one may use the entire list of species of 

 two associations or the Hsts without the accidentals or only the char- 

 acteristic combination of species. 



Supposing that the number of species in association A is 100 and in 

 B 150 and that 60 species are common to both; the coefficient of the 

 community is: 



fin 

 Ttq X 100 = 40 per cent. 



The application of this method is limited, because only two species 

 lists can be compared at once.^ 



It must be noted that the species are not to be reckoned merely 

 as figures in a statistical comparison. Because of their entire sociologic 

 relation they must be evaluated as social units of differing importance. 



Evaluation of the Characteristics of Communities. — Universal 

 rules for the systematic evaluation of floristic characteristics of com- 

 munities cannot be laid down. As in systematic botany, now one, now 

 another, character takes a leading place. Undoubtedly, fidelity is of 

 supreme diagnostic importance. But in every case the summation of 

 the floristic characteristics must determine the systematic relations. 

 In doubtful cases the summation of all the floristic, ecologic, syngenetic, 

 and synchorologic characteristics of the community must be decisive. 



If one holds the usually accepted view that the association is 

 defined by its combination of species and that it is the fundamental unit 

 in the classification of communities, the affinities of the communities 

 should be derived from a study of the association tables. The vast 

 labor involved in the exact floristic analysis of associations proves 

 useful, therefore, for community systematics. 



1. Higher Community Units. — With the acceptance of the floristic 

 principle of classification which forces itself irresistibly upon modern 

 research, the general plan of the system is determined. 



Alliance. — The most closely related associations are united into an 

 "alliance." The floristic relationship is shown especially by the 

 presence of a rather large number of species characteristic of the 

 alliance. The alliance corresponds approximately to the genus of 

 systematic botany. Koch (1926) has admirably arranged the most 

 important alliances of water and swamp communities of Central 



1 GuYOT (1924) proposes to use the coefficient of community for determining 

 the floristic correspondence between the individual stands of an association; 

 The stands would be compared with a "standard," which would be the complete 

 combination of characteristic species (c/. Jaccard, 1928). The significance of the 

 generic coefficient has been discussed by Maillefer (1929); 



