10 Introduction 



decide with certainty^ in others we may never be able to 

 decide, where to draw the limits of the original formations. 

 For instance, much of our heathland is undoubtedly de- 

 generate forest, but in other cases it is probably primitive, 

 existing on soil which does not naturally bear woodland 

 at all. Let us suppose that a tract of primitive heathland 

 (which must be counted as a separate formation) adjoined 

 a tract of woodland on a light sandy soil. If the wood- 

 land is extensively cleared the heath community will 

 undoubtedly invade its area, and the original limits of 

 the two formations will be obliterated. The same is true 

 of chalk pasture. There is evidence that some of our chalk 

 downs are primitive grassland, that is, were never covered 

 by forest; but there are many tracts of chalk pasture 

 which certainly occupy the sites of old beech and ash 

 woods that have disappeared under the axe or from other 

 causes. It is difficult or impossible in the present state 

 of our knowledge to draw the limit between the two cases, 

 which may be represented by identical associations. 



When habitats and their corresponding plant-com- 

 munities are separated by characters less important than 

 the master-factors determining formations, the term sub- 

 formation is employed to designate a division of a 

 formation, based on these less important factors. A sub- 

 formation however exhibits the same features (of succession 

 etc.) as a formation and is to be distinguished carefully 

 from an association. 



The Plant-association 



The plant-association is the vegetation-unit next below 

 the plant-formation. Plant-associations are in general 

 the most obvious plant-communities that we recognise in 

 the field. Thus each of the types of vegetation, woodland, 

 scrub and grassland, within a given formation, is a plant- 

 association, and so is each definite phase in the primary 

 development of a formation. 



