If all species were immobile, the family group would be characteristic of vegetation. Since the great majority are more or less mobile, aggregates of this sort are the exception rather than the rule. Mobility not only decreases the number of offspring in the family group, but it also spreads disseminules broadcast to enter dissimilar groups. It leads directly to mixed aggregation, by which individuals of one or more species invade the family group. Once established, the newcomers tend also to produce simple groups, thus causing an arrangement corresponding essentially to a community. Such collections of family groups are extremely variable in size and definition. This arises in part from the nature of simple aggregation, and in part from the varying mobility of different species. Mobility alone often produces similar communities by bringing together the disseminules of different plants, each of which then becomes the center of a mixed group. In the case of permobile species, several disseminules of each may be brought together. The resulting area, though larger, is practically the same. At present, it is difficult to formulate the law for this method of grouping. It may be stated provisionally as follows: mixed aggregation is the direct result of mobility, and the greater the mobility the more heterogeneous the mixture. The constitution of all the major areas of a formation is to be explained upon the basis of aggregation by the two methods described. The relative importance of family groups and communities differs for every formation, and the exact procedure in each can be obtained only by the detailed study of quadrats. The problem is further complicated by competition and reaction, particularly in closed vegetation. For this reason, aggregation can be studied most satisfactorily in a new or denuded area, where these processes are not yet in evidence. _Kinds of Association_ =253. Categories.= In the analysis of association, it must be kept clearly in mind that the concrete examples from which all generalizations must be drawn are often in very different stages of development, and are of correspondingly different ages. For this reason it has seemed best to consider the primary relations of association in general in this place, leaving the treatment of the effects of invasion, succession, alternation, and zonation to be taken up under these topics.