896 PLANT COMMUNITIES AND FLORAS. 



such as scrub or plain, the occurrence does not in the least confuse the definition and 

 description of the carpet. It only shows that for purposes of determination and 

 description, whilst attending primarily to the stage of the community when develop- 

 ment has reached its zenith, we must also take into account in each case the stages 

 of incipience and decadence, and the relations to other communities. 



Wherever the coniiguration and composition of the ground favour the formation 

 of various kinds of habitat within narrow limits of space, there the particular plant- 

 communities which correspond to those habitats develop in great variety close 

 together. The boundary -lines of adjacent communities are disposed in a multitude 

 of different ways in such cases. In lowlands where gentle hills alternate with 

 shallow depressions, and where from sand one passes to clay, and thence, perhaps, to 

 ground covered with saline efflorescence, the communities are not infrequently pieced 

 together like the components of a mosaic. In other places those which cover a small 

 area are interspersed like islands in the midst of the more extensive communities; 

 and, again, in other cases the different areas are in the form of contiguous strips 

 and zones. The last mode of distribution occurs chiefly along the margins of still 

 or running water, and is explained by the uniform diminution of moisture in the 

 ground as the distance from the water's edge increases, and in the case of still water 

 aiso by the tendency of plants to advance from the margins towards the middle of 

 the expanse of water where they grow, or, in other words, from the continuous trans- 

 formation of the ribbon types which flourish in water into reeds, of reeds into plains, 

 and of plains into carpets or into forests. 



It very often happens that two or more plant-communities are intermingled, and 

 that the whole exhibits a kind of stratification. A pine forest may exist by itself, 

 that is to say, it may include nothing but dominant pines with, perhaps, a sprinkling 

 of a few other kinds of tree, and the ground may be bare except for a bed of dry 

 fallen needles. On the other hand, a carpet may have developed on the forest floor, 

 whilst, in addition, a Bilberry scrub, a low scrub of Galluna vulgaris or Erica 

 carnea, and a high scrub of Juniper may have obtained a footing, aU. of which 

 communities are capable of existing independently without the shelter of the pines, 

 and are often met with thus alone. But although the presence of the one com- 

 munity is not indispensable to the existence of the other, the fact of their occurring 

 together shows that no injury is suffered on either side in consequence of the com- 

 panionship, and it is much more likely that communities growing on the same 

 gi'ound are mutually helpful and protective. In many cases there is no doubt of 

 this, as, for example, when a community of tall plants develops on soil previously 

 prepared by a community of low plants without completely ousting them. We 

 gather, then, that the conjunction of several communities is by no means fortuitous, 

 that the association is always between certain particular communities, and that even 

 here we find strict orderliness and subjection to definite laws. 



Unions of communities formed in the manner above described have been termed 

 plant-formations, probably from analogy with the combinations of strata of earth 

 and stone which geologists call formations. The selection of this name is not quite 



