Various categories of association may be distinguished, according to the dominant physical factor concerned or the point of view taken. These will fall into two series, as we consider the relation of plant to plant with reference to some object or characteristic, or the grouping of plants together in response to some dominant factor. In the first series may be placed association with reference to substratum, to the ground (occupation), and to invasion; in the second belong light and water-content association. It should be noted that these are all actual associations in nature, and not concepts such as the vegetation form, within which plants from widely different associations may be classified. Naturally, it does not follow that it is not logical or valuable to group together those plants, such as hydrophytes, sciophytes, hysterophytes, etc., which have a common relation to some factor, but belong to different formations. =254. Stratum association.= Plants manifest independent or dependent association with reference to the stratum to which they are attached and from which they derive food or support. Independent association is exhibited by those holophytic species of a formation which are entirely independent of each other with respect to mechanical support or nutrition. It is characteristic of the greater number of the constituent species of formations. Dependent association is manifested in the relation between host and parasite, stratum and epiphyte, support and liane. Warming[24] has distinguished six kinds of associations: parasitism, helotism, mutualism, epiphytism, lianism, and commensalism. Commensalism corresponds to the primary principle of association which has given rise to vegetation. Homogeneous commensalism is the term applied to social exclusive plants, in which the patch is composed of a single species. Such association is extremely rare in nature, and if the most minute forms be considered, probably never occurs. On the other hand, heterogeneous commensalism, in which individuals of more than one species are present, is everywhere typical of vegetation. Warming regards saprophytism merely as a specialized kind of parasitism, an opinion that may well be defended. Helotism, however, is also a mere modification of parasitism, if it is not indeed parasitism pure and simple. Mutualism is an altogether vague concept, including parasites, epiphytes, and endophytes of doubtful physiological relation. Pound and Clements[25] treated lianes, parasites, and saprophytes as vegetation forms, relating herbaceous creepers and twiners to the lianes, and dividing the fungi and lichens into nine groups. Whatever the value of these divisions may be from the standpoint of vegetation forms, they represent the same relation between plant and nutritive stratum, and with respect to association should be merged in one group. Schimper[26] was the first to perceive the essential similarity of all such groups from the standpoint of association.