54 HARPER: VEGETATION OF THE PINUS TAEDA BELT 
Just how much a census taken at the time of the discovery 
of America would have differed from this it is impossible to say. 
It is very likely, however, that Pinus Taeda has replaced P. 
palustris to a considerable extent through the influence of civiliza- 
tion, as claimed by Ashe and others; one of the reasons for the 
change probably being the diminishing frequency of fire, for P. 
Taeda is more sensitive to fire than P. palustris is. 
The percentages of evergreens are for Virginia 69.4, North 
Carolina 73.6, South Carolina 75.7, and for the whole region 74.0. 
The correspondence between these figures is close enough to give 
confidence in the method used, and at the same time the differences 
are just what one would expect from the slight differences in soil 
fertility in the three states already pointed out.* 
So many of the Ericaceae and Leguminosae are not readily 
recognizable from a moving train that it is hardly worth while to 
attempt to give statistics for these plants for each state separately, 
but for the whole region the proportion of Clethraceae, Ericaceae 
and Vacciniaceae among the shrubs seems to be about 12 per cent. 
and of Leguminosae and allied families among the herbs about 
4 percent. The figure for Ericaceae is the same as that already 
‘obtained for northern Floridat while that for Leguminosae is 
lower. It is reasonably certain that the Ericaceae are most 
abundant in North Carolina and the Leguminosae in South 
‘Carolina. 
By contrasting the more characteristic plants of each division 
we can get some interesting facts. In the following table the 
species which are most abundant in the Virginia portion are listed 
in the first column, the same for North Carolina in the second, 
and for South Carolina in the third. They are arranged in order | 
of abundance, just as in the preceding table. Those starred seem 
to be at least twice as abundant in the state indicated as in either 
of the other two. 
This table confirms in a general way observations of a similar 
nature made in passing through the same three states in 1906, 
but taking in a larger and more diversified area.t The proportion 
* Previous estimates of evergreen percentages for parts of the Carolina coastal 
plain (summarized in Bull. Torrey Club 41: 563-564. 10914) are now believed to 
be too low. ‘ 
f Ann. Rep. Fla. Geol. Surv. 6: 395. 1914. 
¢ Bull. Torrey Club 34: 363-366. 1907. 
