360 HARPER: COASTAL PLAIN OF THE CAROLINAS 
the way from the coast to the fall-line, and that in the remaining 
third (exclusive of the fall-line sand-hills) oaks and other angio- 
spermous trees predominate, just as in the upper third of the coastal 
plain of Georgia. On the way from Augusta to Yemassee, although 
Pinus palustris can be seen in scattered groups or individuals nearly 
the whole distance,* the real pine-barrens only begin about the 
inland edge of Hampton County, and seem to terminate near its 
coastward edge.t Between Yemassee and Charleston the railroad 
passes mostly through the maritime or littoral region (very similar 
to that of Georgia t), where Pixs palustris is rare and P. Taeda 
common. Beyond Charleston the limits of the pine-barrens are 
less clearly defined. 
The illustrations in Bulletins 43 and 56 § of the U.S. Bureau 
of Forestry give an excellent idea of the general appearance of the 
South Carolina pine-barrens, and the forests of the North Carolina 
coastal plain have been so well described by Mr. Ashe in Bulle- 
tins 5 and 6 of the North Carolina Geological Survey that it would 
be useless to attempt to improve on his observations with so little 
preparation ; but there is still room for a few notes on the herba- 
ceous vegetation, and for some statistics of distribution, which may 
serve as suggestions for future work in this region. 
Pine-barren vegetation, if we may judge by the number of 
species in a given area, seems to center at present in Georgia and 
Florida,|| and as a rule grows gradually poorer and less typical 
with increasing. distance from this center, the proportion of older 
species from the highlands at the same time increasing, as I had 
excellent opportunity to observe on this trip. Pinas El/iottii, whose 
distribution in Georgia coincides almost exactly with the pine-bar- 
rens, extends only a short distance into South Carolina.{] Pinus 
oo ete Georgia it seems to skip a good deal of the Eocene reek See Bull. Torrey 
Club 31: 15. 1904; 32: 456. 1905; Ann. N. Y. Acad. Sci. 17: 305. 1906. 
} This is corroborated by Hammond’s agricultural map of aach Carolina in the 
Tenth Census report already cited, and by Glenn’ s physical outline map in his South 
Carolina sp ga ent to Redw yay & enews S geography. 
t See - N.Y. Acad. Sci. 17: 20. 1906 
git coal be observed that pas 7 in Bulletin 56 is from the same photograph as 
plate r2 in Bulletin 43, which purports to show Pinus Elliottii among other things, 
and therefore was not taken in Berkeley Count 
|| See Torreya 7: 43; Science II. 2 1907. 
{| For details see the latter part of this paper. In the other direction it is not 
