russell.] LAVA CAVES. 9U 
although certain steep-sided depressions are there present which per- 
haps represent such caverns, whose roofs and sides have fallen in. 
These sunken areas vary in length up to 200 or 300 feet, and are from 
50 to perhaps 70 feet wide and 40 to 50 feet deep. Their bottoms and 
sides are strewn with large blocks of their fallen roofs. Lava caves of 
the nature here considered occur at several localities in the older sheets 
of Snake River lava, but are not known where the Snake River and 
its tributaries have excavated canyons in the nearly leA r el sheets. One 
was visited by me, on the eastern slope of Notched Butte, about 4 
miles south of Shoshone, and several of large size are reported to 
occur about 6 miles northwest of the same town, where lava descended 
from the mountains before reaching the adjacent plain. The cavern 
on the side of Notched Butte was entered through an opening formed 
by the falling in of a part of its roof. The portion open to inspection 
is 300 to 400 feet long, 25 to 30 feet wide, and is partially filled with 
blocks that have fallen from the under surface of its cover. The 
height of the opening at present is from 6 to 10 feet. In continua- 
tion of the portion of the cavern still remaining, there is a depression 
about 15 feet deep which is crossed at one locality by a natural bridge 
formed of a portion of the roof still in place. 
The interiors of such caverns present various details, due principally 
to the drip of the lava while } T et hot and mobile. The falling in of 
the thick roofs of this variety of lava caverns is evidently not in all 
cases due to a sudden collapse. The cave referred to, on Notched 
Butte, has been enlarged upward by the falling of blocks from the 
under side of its roof, and 1 think none of the original under surface 
of the arch remains. This process continuing, openings will finally 
be made to the surface, and more or less gradually the entire roof will 
fall. The cause of the displacement of blocks from time to time from 
the roof of such a cavern is difficult of explanation. The only agency 
that can be appealed to as generally active is change of temperature, 
but this can not be considered as important in caves with roof 20 to 40 
or more feet thick, as is frequently the case, until openings have been 
made. The temperature of caverns situated at such depths is gen- 
erally uniform. The only other agency that I can suggest is earth- 
quake shock. That earthquakes may have occurred in the region 
where Notched Butte is situated is shown by the presence of faults 
in the Snake River lava, as will be described later. 
From what is known concerning lava caves due to outflow of still 
liquid or plastic material from beneath a rigid crust, it appears that 
the controlling condition favorable to their formation is a sufficient 
gradient to permit the subsurface lava to flow, but yet not of such 
steepness as to cause a current which would shatter the crust in the 
earlier stages of its formation. A delicate adjustment between rate 
of cooling and rate of flow is also necessarv. The caves should be 
