CH. V] HINDER THE DISPERSAL OF SPECIES 51 



chalk downs south of Cambridge is Festuca ovina. Suppose 

 however that this plant in its dispersal had not yet reached th^ 

 downs. They would, none the less, be covered by a society of 

 plants which might be very numerous, and which we mi^ht 

 thmk closed. Yet when the fescue appeared, there can be little 

 doubt that it would soon secure a foothold. 



An association of plants ultimately passes its zenith and be- 

 comes gradually superseded by another, the process being known 

 as succession (16). "The pine... gave place at length to the oak 

 and the oak... yielded in its turn to the beech, the periods M-hen 

 these three forest trees predominated in succession taUyino- 

 pretty nearly with the ages of stone, bronze, and iron in Dent 

 mark" (68, p. 372). 



The more closed an association is, probably so much the more 

 difficult will a newcomer find it to obtain any foothold and by 

 so much will its dispersal be retarded. One will expect that 

 most newcomers will find it quite impossible to gain a footing 

 at all, but that every now and then (as in the case of Elodea the 

 famous "American water- weed" of the last generation which 

 spread so rapidly through the waters of Western Europe, though 

 only the female plant was introduced) one will do so knd will 

 spread, more especially to those places which the association 

 concerned already reaches. 



In many instances, of course, a plant in its tra^•els will come 

 across a type of vegetation into which it cannot spread at all 

 and which may thus, if broad and wide enough, form a complete 

 barrier. If an ordinary herb, accustomed to a good water supply, 

 and to hfe in the open sunshine, comes across a stretch of 

 country which is either a forest or a desert, it will be held up 

 in this manner, and whether it can cross will depend upon its 

 mechanism for dispersal, upon the width of the barrier, and 

 upon other factors. Forest trees arriving at a desert will un- 

 doubtedly be stopped, but when they meet a herbaceous asso- 

 ciation, in a country where the rainfall is sufficient, will probably 

 spread at the expense of the herbs, and cover previously open 

 ground with trees. One may see this going on at the edge of a 

 pine wood, or on any small clearing made by a peasant in a 

 tropical forest. There is hi lie evidence for the occurrence of I lie 

 reverse process, without the aid of desiccation of the climate or 

 something of the kind. 



Changes of conditions Avill make great differences to the rate, 

 or even to the possibility, of spread, often by the effects they 



4—2 



