350 FOSSIL PLANTS. 
ruption is easily produced by a hard stroke. From the great bedof — 
sandstone overlying the, Pittsburg coal, near Greensburg, I have — 
received, from Rev. W. D. Moore, large specimens of fossil wood, 4 
most of them long, irregularly broken, much decayed pieces, evi- — 
dently, representing sections of trunks broken lengthwise. These d 
were found in various positions in the sandstone, and were mostly — 
broken before they were imbedded in it. But among them thereis 
one which bears, attached to a short stem, three diverging branches _ 
of its roots, a proof that it has been buried in its original stand- ! t 
ing mer and this one has its top horizontally broken and flate 
m these data and a number of others, which it is useless to ] 
mah being all of the same kind, and bearing the same evi- 4 
dence, it appears that the fracture of the fossil wood is of two 
kinds: irregular, for trunks fossilized after prostration or in a dez 
caying state, as they are generally found in our Tertiary and Cre: 
taceous strata ; and horizontal, by splits perpendicular to the natu- 
ral direction of the stems and the roots. If the cause of fracture 
in the first case is, without doubt, essentially due to atmospheric 
agency, that of the second, which has acted upon the vegetable 
while it was still subjected to the process of petrification, is cer- 
tainly different, and can be explained, I think, by the difference of 
density of both the surrounding mineral matter and the imbedded” 
vegetable. Evidently, all the stems in the process of fossiliza- 
tion have been subjected to a softening process of their whole 
mass. The outside pressure of the surrounding mineral ie 
must have been felt, and can have acted only in one way, that is, 
verticaily, as it happens in the forcing of a body of less density. 
out of water; and the result of that action cannot but have been 
a tendency to dislocation, and therefore to splitting of the rae 
in a horizontal direction. It might be supposed, perhaps, that 
gradual accumulation of sand or other mineral matter 
standing trees, in burying them, has formed layers wed differ 
density, whose action may have produced, in the fossil ve 
_ zones of petrification also varying in density, tending, theref 
to cleave from each other, and horizontally separable. But 
roots of fossilized trees which tend downwards in an 
