HARPER: CLIMAX VEGETATION 521 
which are so narrow and so far from the mainland that they have 
no area sufficiently protected from the winds to allow humus to 
accumulate and hardwood forests to develop, or are so low that 
they are swept by the waves. Such beaches are found along the 
south shore of Long Island, on the coast of North Carolina, in the 
east and west Florida coast regions,* along the coast of Mississippi, 
and in various other places. 
Now for the explanation of the phenomenon above described. 
Of the several hypotheses which readily suggest themselves, the 
least probable will be taken up and disposed of first. 
In the case of islands and peninsulas bordered by large bodies 
of water (either fresh or salt), particularly some of the northern- 
most ones here discussed, it has been supposed by some that the 
peculiarities of their vegetation, and their agricultural advantages, 
are due to the temperature being kept more uniform by the 
proximity of the water, and especially to the greater exemption 
from frost enjoyed by such places. But the climax plants, or 
many of them at least, are not known or believed to be any more 
sensitive to heat and cold than the pioneer plants of the same 
regions are. In fact nearly all the hammock plants of central 
Florida are perfectly at home several hundred miles farther north,t 
and remote from large bodies of water. Furthermore, the tem- 
perature hypothesis would hardly explain the marked contrast 
between the peninsular hammocks of the Florida lakes and the 
pine forests which border the more gently curving shores of the 
same lakes; or the stream-fork peninsulas, where the width of the 
streams is often insignificant. 
It might also be claimed that the islands and peninsulas are 
better watered than the pine forests near by; but this is not always 
true, and besides, the hammock plants are not especially moisture- 
loving, decidedly less so, for example, than bog plants, which are 
mostly of the pioneer class. 
It might be true in some cases, such as on the'sea islands and 
stream-fork peninsulas, that there is some fundamental difference 
in the soil. But in many or most of the instances cited, whatever 
difference there is in the soil can be ascribed to the vegetation 
* See Ann. Rep. Fla. Geol. Surv. 3:218, 226. I91l. 
+ This statement does not apply, however, to the hammock vegetation of south 
h 
J aVvuyve-. 
