524 HARPER: CLIMAX VEGETATION 
The relation between fire exemption and climax vegetation is 
reciprocal, for when the hardwoods are once well established the 
herbaceous vegetation under them is very sparse, and the humus 
is usually too damp or too thoroughly decomposed to burn readily. 
(There are of course all gradations between pioneer and climax 
forests, but I refer now to the extremes, and not to such inter- 
mediate stages as the common deciduous oak-hickory forests of 
the East and South, in which burning of the dead leaves in the 
fall, mainly through human agency, is a rather frequent occur- 
rence.) 
Outside of the long-leaf pine region or southern pine barrens 
(and south of the great northern coniferous forests, where fires are 
much more spectacular, and present a different set of problems) 
forest fires from natural causes, although they still occur, are less 
frequent, and the contrast between open and protected forests is 
consequently less marked. Peninsulas as wide as those of ‘‘ Tide- 
water Virginia’”’ would probably offer little hindrance to fire if 
their vegetation were of the pine-barren type, but they are: just 
outside of the range of Pinus rigida on the one hand and Pinus 
palustris on the other, and the various hardwoods and short-leaf 
pines which prevail there* grow densely enough to make fires 
rather rare. 
Although it is a little beyond the scope of this paper as indi- 
cated by the title; there is another piece of evidence in favor of 
the fire-protection theory of climax vegetation in the coastal plain 
that ought to be mentioned here. This theory furnishes a satis- 
factory explanation of another class of hammocks which has never 
been fully explained before: namely, those of the Altamaha Grit 
region of Georgia.t These are invariably located along creeks or 
rivers, and nearly always on the sandhill side of the stream. Their 
vegetation is very similar to that described above for the peninsulas 
of Lake Tsala Apopka. In former years I sought for the cause of 
these hammocks in the subsoil or the ground water; but it is clear 
now that protection from fire is the principal cause. For the 
pine-barren side of a stream, with its close carpet of wiregrass and 
See Bull. Torrey Club 37: 422, 423. 1910. The differences in vegetation on 
opposite sides of Hampton Roads are now a little easier to explain than they were 
before. 
t See Ann. N. Y. Acad. Sci. 17: 98-102. pl. 16. 1906. 
