16 ON THE TIMBER LANDS TRAVERSED BY A 
which have a very limited extent of bottom land (hemlock: 
will not grow on bottom land at all), and where the surround- 
ing hills come down to the water edge, forming more or less. 
ravines and precipices. The consequence is, that while the 
head-waters of the Kentucky river on the one hand, and of 
the Cumberland on the other, penetrate into the very heart. 
of the hemlock region, and are the mountain streams along 
which this timber grows to the greatest perfection, yet the 
Kentucky river does not carry it far northward, nor the Cum-_ 
berland river far westward. ‘The seeds will be carried down- 
ward and deposited by these streams, and will take root and. 
grow, just so long as the above conditions are complied with;. 
but, whenever the streams become large enough to have a 
belt of bottom lands along them, the possibility of a further 
spread of the hemlock ceases in Kentucky. The conditions 
of growth of that timber may be different elsewhere. 
It is worth while, in speaking of special timbers, to call 
attention to a somewhat remarkable forest of beeches, which 
occupies a belt of country eight or ten miles wide, beginning 
about three miles from Greensburg, and extending to within 
about the same distance of Campbellsville, and lying in Green: 
and Taylor counties. The extent of the belt in other direc- 
tions I could not determine. In this belt, beeches form the 
forest timbers to the almost entire exclusion of other growths. 
They not only occupy the valleys, but extend to the tops of 
the highest hills. The reason is to be found in the formation, 
which is a reddish, very much decayed St. Louis chert, out. 
of the very top of which the water oozes, and which is there- 
fore always wet. Inasmuch as height above drainage is the 
principal determinant of beech growth, it is natural that these 
hills should be covered with such a heavy forest of that tim- 
ber. 
As to the distribution of the magnolias, the so-called Amer- 
ican laurels (rhododendron and kalmia), and the linden trees, 
I confess that I see no reason why they should be confined to 
the eastern part of Kentucky, unless it be the purely topo- 
graphical one, that high mountains and deep and ragged 
186 
