13 101 
[Cope. 
tion parallel to the strike of the strata, and have their greatest diameter 
in the direction of the dip. Their depth is determined in some measure 
by. the softness of the stratum, whose removal has given them existence, 
but in thinly stratified or soft material, the roofs or large masses of rock 
fall in, which interrupt the passage below. Caves, however, exist when 
the strata are horizontal. Their course is changed by joints or faults, 
into which the excavating waters have found their way. 
That these caves were formed prior to the postpliocene fauna is evident 
from the fact that they contain its remains. That they were not in ex- 
istence prior to the drift is probable, from the fact that they contain no 
remains of life of any earlier period so far as known, though in only two 
cases, in Virginia and Pennsylvania, have they been examined to the 
bottom. No agency isat hand to account for their excavation, compara- 
ble in potency and efficiency to the floods supposed to have marked the 
close of the glacial period, and which Prof. Dana ascribes to the Cham- 
plain epoch. An extraordinary number of rapidly flowing waters must 
have operated over a great part of the Southern States, some of them at 
an elevation of 1500 feet and over, (perhaps 2000) above the present level 
of the sea. A cave in the Gap Mountain, on the Kanawha river, which 
I explored for three miles, has at least that elevation. 
That a territory experiencing such conditions was suitable for the 
occupation of such a fauna, as the deposits contained in these caves re- 
veal, is not probable. ‘The material in which the bones occur in the south 
isan impure limestone, being mixed with and colored by the red soil 
which covers the surface of the ground. It is rather soft, but hardens 
on exposure to the air. 
The question then remains so far unanswered as to whether a submerg- 
ence occurred subsequent to the development of the postpliocene mam- 
malian fauna. That some important change took place is rendered prob- 
able by the fact, that nearly all the neotropical types of the animals have 
been banished from our territory, and the greater part of the species of all 
types have become extinct. Two facts have come under my observation’ 
which indicate a subsequent submergence. <A series of caves or portions 
of a single cave once existing on the 8. HE. side of a range of low hills 
among the Allegheny mountains in Wythe Co., Virginia, was found to 
have been removed by denudation, fragments of the bottom deposit only 
remaining in fissures and concavities, separated by various intervals from 
each other. These fragments yielded the remains of twenty species of 
postpliocene mammalia.* 
This denudation can be ascribed to local 
causes, following a subsidence of uncertain’ extent. In a caye 
examined in Tennessee the ossiferous deposit was in part attached to 
the roof of the chamber. Identical fossils were taken from the floor. 
This might, however, be accounted for on local grounds. The islands of 
the eastern part of the West Indies appear to have been separated by 
_ submergence of larger areas, at the close of the period during which they 
were inhabited by postpliocene mammalia and shells. The caves of 
* See Proceed. Amer. Phil. Soc, 1869, 171 
