108 CLIMAX FORMATIONS OF WESTERN NORTH AMERICA. 



stemon. The striking feature of a society, that of a control within a control, 

 often apparently greater than that of the consociation itself in the case of 

 grassland at least, is due to the difference in vegetation-forms and hence in 

 ecological requirements. A society may conceivably belong to the same 

 vegetation-form as the consociation or association in which it occurs, but such 

 cases are practically unknown. Apparent examples of this have all been 

 readily referred to successional causes, as where a localized area of Hordeum 

 jubatum occurs in prairie, or one of Aristida purpurea on the plains. The 

 almost invariable rule is that the society belongs to a vegetation-form of lower 

 requirements than that of the consociation. The forest will have societies of 

 shrubs, herbs, mosses, etc., the chaparral of undershrubs, grasses, and herbs, 

 and the prairie of herbs principally. In this connection, it is especially 

 important to recognize that savannahs do not represent tree or shrub societies 

 in grassland, but are an incomplete expression of the next stage in succession. 



The degree of control exerted by the society clearly depends upon the life- 

 history relations of the dominant and subdominant concerned. This is largely 

 a matter of the height and extent of the shoot, but the root also plays a large 

 part. In forest the societies of varying rank, from shrub to moss or lichen, 

 are wholly and obviously subordinate to the trees. In chaparral this is also 

 true to a large extent, but societies of undershrubs and grasses often play a 

 conspicuous part. As to grassland, the societies are frequently much more 

 conspicuous than the dominant grasses, and at times they appear to be in 

 control. In such cases the control is seasonal. Each subdominant reaches 

 a maximum in spring, summer, or fall, when it seems to dominate, but the 

 real relations are disclosed at the other seasons. This tendency of societies to 

 appear during a particular season further explains the relation of dominants 

 to subdominants. They not only make different demands by virtue of their 

 vegetation-forms, but these demands are also made at different times. Socie- 

 ties exhibit a similar seasonal relation in forest and scrub, in which their time 

 of appearance is almost wholly controlled by the dominants. In most cases, 

 this relation is so striking that it is possible to distinguish two or more aspects 

 during a season, marked by particular societies (Clements, 1905 : 296 ; 1916 : 

 132). From the preceding discussion, it is clear that various kinds have 

 already been distinguished (Clements, 1916 : 132), and it is highly probable 

 that still others will demand recognition as the study of vegetation becomes 

 more detailed and accurate. 



The society is not a subdivision of the consociation in the same way that this 

 is of the association. The latter consists of its consociations, grouped or 

 single; they occupy its total area. The consociation does not consist of 

 societies, but the latter merely occur in it or through it to a larger or smaller 

 degree. This is readily seen to be due to the basic difference in vegetation- 

 forms and to the seasonal nature of societies. As a consequence, a particular 

 society may occur not only in two or more different consociations, but also 

 in two or more associations of the same formation. It may extend more or 

 less continuously over wide stretches, or it may recur as successional processes 

 or physical factors determine. A typical example of this is Psoralea tenuiflora, 

 which occurs in nearly every association and consociation of the grassland 

 climax, while the closely related society of P. argophylla is restricted to the 

 prairie associations. Gutierrezia sarothrae forms similar communities of even 



