WARD] PETRIFIED FORESTS OF ARIZONA. 3829 
but most of them seem to be as yet only partial and probably do not 
extend entirely through the trunk. There is one, however, near the 
left bank of the canyon which has the appearance of doing so, and the 
trunk is probably only kept from parting at this point by the mechan- 
ical adjustment which causes the adjacent faces to perform thé office 
of a keystone to an arch. Any considerable shrinkage due to climatic 
or other causes would overcome this influence and the entire bridge 
would crash to the bottom of the canyon and roll down the escarp- 
ment in a number of huge segments. 
An examination of the relations of the Natural Bridge to the gulch 
which it spans shows clearly that the trunk was primarily entombed 
in the sandstone bed covering this entire region, and that, with the 
progress of erosion, which ultimately. carried away the entire plain to 
the north as well as in other directions, leaving this small mesa, it 
was at last exposed and lay for a great period near the rim of the 
escarpment. At first it was only partially buried and later came to 
lie on the surface of the ground. As the land rises somewhat to the 
south of it rills were formed above, and in times of floods or heavy rain 
it obstructed the flow of the water, forming a sort of dam. The water 
lying against the trunk long after it had ceased to overflow it, tended to 
disintegrate the rock upon which the trunk lay, until eventually it found 
its way through beneath the trunk at some one point. The smallest 
opening of this nature would soon become a free passage for the water, 
and a simple continuation of this process of local erosion would ulti- 
mately result in the formation of the entire gorge as it exists to-day. 
The other case which I observed of the presence of the conglom- 
eratic sandstone within the general petrified forest area occurs near its 
center, about midway between the upper and lower forests, along the 
narrow portion of the valley of the creek above described, on both 
sides of the canyon.and near the level of its bed, at an altitude of about 
5,300 feet. The exposure is typical in all respects, and logs were seen 
projecting from the canyon walls, from one of which specimens were 
collected. As this exposure is 400 feet below that in which the Natural. 
Bridge occurs and 450 feet below that on the southwestern mesa, its 
presence there can be accounted for only on one of two hypotheses— 
either that of the existence of another exactly similar stratum at this 
horizon, or that of a fault, or what would amount to the same thing, a 
slide or slipping down of a large block of the uppermost beds in such 
a manner as not to disturb their stratigraphical arrangement. 
The first of these hypotheses is rendered improbable by the fact that 
a careful study of the beds at the same horizon in other places revealed 
no such stratum, and it could scarcely be so local as not to be found 
elsewhere. The second hypothesis seems in every way probable, as in 
such a much-disturbed region it would be easy for the erosive agencies 
