104 SNAKE RIVER PLAINS OF IDAHO. [bull. 199. 
Where a tongue of lava extended up a lateral valley it frequently pre- 
sents a ridge of aa on its immediate border, and, in numerous instances, 
all along its margin there are ridges with open fissures in their crests, 
produced by lateral pressure on a thick and nearly rigid crust. In 
numerous instances, also, the lava, after entering the mouth of a lat- 
eral valley and perhaps extending 3 or ± miles from the border of the 
main flow, and hardening at the surface, receded, there being an out- 
ward underflow. In such instances broad, shallow basins, with rough 
surfaces, were formed, margined on their border adjacent to the older 
rocks (mostly quartzite) by ridges 10 to 15 feet high. These ridges 
are usually arches, or inclined blocks, produced by lateral thrust, but in 
addition there has been a withdrawal of the lava from beneath the 
floor of the basins. The explanation of this phenomenon seems to be 
that the lava after entering a lateral valley, and cooling so as to form a 
crust usually 3 to 5 or more feet thick, receded, being allowed to flow 
out, because the main lava flow across the mouth of the valley had been 
lowered by its onward progress. 
In some instances lava streams originating in the mountains, as 
described on a preceding page, and flowing down lateral valleys before 
reaching the Snake River Plains, expanded in the broader portions of 
their courses, and were later, in part, drawn off by an underflow, 
leaving their hardened crusts stranded. An example of this nature 
occurs in the valley of Lava Creek, about 1 mile above Martin. The 
lava stream which came down the valley of Lava Creek, as already 
described, formed cascades in the upper part of its course, but farther 
down expanded and became about 1 mile wide. About the western 
margin of this broader portion of the lava stream there is now a 
ridge consisting of pressure domes and tilted lava blocks, bordering a 
broad, flat-bottomed depression. In this instance also there was a with- 
drawal of the still -molten rock from beneath the rigid crust, the edge 
of the crust being left stranded. The phenomenon here referred to is 
similar to that which produces lava caves by the outflowing of liquid 
lava from beneath a rigid crust, but the area of hardened lava left 
unsupported from beneath was so great that it could not sustain its 
own weight, and it collapsed and became broken. 
The basins just described might become occupied by water and thus 
be transformed into lakes, but I am not aware of an instance in which 
this has happened. Apparently the bottoms of the basins are too much 
fissured to retain water, unless a considerable amount of silt were 
deposited on them. 
The encroachment of recent lava flows on the bases of the mountains 
has exerted an important influence on the surface drainage. The lava 
entering the mouths of valleys sometimes formed dams, which held 
their waters in check and caused lakes to form. An instance of this 
nature is furnished by Lava Lake, situated at the entrance of Fish 
