306 PLANT SOCIOLOGY 



based entirely on the concept of the "cHmax formation," a concept 

 which has been much emphasized by the two authors. It must not be 

 forgotten, however, that the areas today occupied by cHmax stages 

 have become greatly reduced and sometimes have become almost 

 completely obliterated. The developmental stages, on the contrary, 

 still occupy immense areas. It seems therefore unfortunate to burden 

 terminology with the introduction of special terminations to distinguish 

 stages which appear to be climax (-ation in association) and stages 

 which are developmental (-ies in associes). 



The principal criticism, however, to be directed against the 

 Clements- Weaver system is that it makes no adequate provision for a 

 graduated scale for designating vegetational units of different degrees. 

 The climax formation embraces several climax communities (in our 

 sense) to which are applied the term "association." This term thus 

 acquires a meaning quite different from that given it by the Brussels 

 Congress (1910) and from that in which it is used by the Montpellier 

 school. It is, moreover, very poorly circumscribed, a fact which may 

 perhaps explain why no American botanist has given a detailed tabular 

 analysis of a single association. This unit is large and difficult to 

 grasp, and the subordinate units are separated from it by a veritable 

 hiatus. Any unit corresponding to our association is lacking. The 

 "consociations" and "societies" of Clements and Weaver are based 

 entirely upon the dominance of certain species; they are thus quite 

 incapable of replacing our association in any system of classification. 

 The "aspect societies," "families," and "colonies" of Clements and 

 Weaver play very minor parts as systematic units. They are either 

 concepts of periodicity or expressions of sociability. 



It is therefore not surprising that no investigator has yet tried a 

 detailed sociological description of any group of plant communities of a 

 definite region or of an "association" d^Hmited according to Clements' 

 system. In an excellent paper, which may be considered a model of 

 the Clements- Weaver school, Steiger (1930) has given a careful quanti- 

 tative floristic and ecological study of a portion of the prairies. These 

 valuable ecological data cannot, however, serve to individualize the 

 sociological units. 



This seems to furnish the best proof that the most exact ecological 

 data remain sterile for sociological evaluations when assembled accord- 

 ing to the system of Clements and Weaver. 



From our present state of knowledge a few general laws may be 

 deduced. 



Early Colonization. — The first step in the development of vegeta- 

 tion is "migration." Next comes the estabhshment of the invading 



