HARPER: CLIMAX VEGETATION 523 
It is pretty well known that long-leaf pine, after it is four or 
five years old, is less affected by fire than almost any other tree 
we have,* and in southern forests periodically swept by fire little 
else can grow but this pine and a great variety of more or less 
xerophytic, mostly perennial, herbs, among which various grasses 
are usually most abundant. 
An island would ordinarily be exempt from fires except such as 
might originate on it, and consequently the smaller the island the 
less the chances of fire. On a peninsula, whether bordered by 
open water, marshes, or narrow streams, fire could approach from 
only one direction, and there would be many chances-of its being 
stopped by rain, fallen logs, etc., before going very far down the 
peninsula, so that the frequency of fire in such places would be 
very much less than in unprotected forests where fire could ap- 
proach from any direction. 
On this theory it is easy to account for the origin of insular 
and peninsular hammocks. About half the species of plants listed 
above have fleshy fruits, and are presumably disseminated mostly 
by birds; and several others have nuts which may be carried con- 
siderable distances by rodents, such as squirrels. The seeds of 
such plants are doubtless being dropped all through the fruiting 
season in all sorts of places, and occasionally one will germinate, 
if it falls in the right kind of soil and does not encounter enemies 
or too much competition. In this way an island or peninsula 
which we might suppose for the sake of argument to have been 
originally covered with pine-barren vegetation would gradually 
become seeded with hardwoods.t The absence of fire would in 
time allow sufficient humus from the pioneer vegetation to accu- 
mulate to give the hardwoods a start, and the latter would grow 
up and finally make enough shade to prevent the reproduction of 
the pines and other pioneer plants, which are what foresters term 
intolerant, and will not germinate in dense cenoe ts en 
* This point is discussed on pages 71-87 of a little book by G. Frederick Schwarz, 
entitled ‘‘ The long-leaf pine in virgin forest,’ published in 1907. 
+ The ratio of fleshy-fruited trees to oaks and hickories should be somewhat 
greater on islands than on peninsulas, on account of the greater accessibility of 
the latter to mammals, but I have not yet investigated enough islands quantitatively 
to say whether this is true or not. : 
t In this connection Professor Bray’s observations on the growth of Pinus 
palustris in eastern Texas (U. S. Bureau of Forestry Bull. 47: 22, 23. 1904) are of 
interest. 
