russell.] LACTJSTKAL CONDITIONS. 57 
which the fossil bones just referred to were obtained, fossil wood 
occurs in abundance. The wood is in fragments, and was evidently 
waterworn, as if rolled up and down a beach before being buried, 
and replaced by silica. Mingled with the fossil wood are occasionally 
found silicified pine or spruce cones in an admirable state of preser- 
vation. 
The meager data furnished by the fossils referred to above seems 
to indicate that the lower portion of the lacustral deposits cut through 
in forming Snake River Canyon belong to the Miocene and the upper 
part to the Pliocene, but at present it is impracticable to draw a line 
of division between the two. A summary of the evidence furnished 
by previous collections from the lacustral deposits of southwestern 
Idaho has been given by Lindgren" and need not be reviewed at this 
time. 
SUMMARY OF LACUSTRAL CONDITIONS. 
The general succession of events recorded by the sedimentary beds 
considered above seems to be as follows: A large lake or possibly two or 
more successive lakes in the same depression, the boundaries of which 
are known onlj r in part, existed in the Snake River Basin in Tertiary 
times. In this lake several hundred feet of fine, evenly laminated cla\ T 
or silt and volcanic dust and lapilli were laid down. The fact that loose 
sand in places 300 feet or more thick rests on the clays previously 
deposited suggests that the lake became shallow, and the presence of 
beds, usually thin, of well-worn pebbles consisting of a variety of rock 
fragments indicates the occurrence of strong currents in a shallow 
water body, or else that the lake was drained and its exposed bed 
crossed by streams. These sediments, which in the most typical 
instances are clearly lacustral, contain beds of coarse, well-rounded 
gravel, such as is laid down by streams, and it is impracticable at 
present to determine just what part is the result of still-water and 
what part of flowing-water deposition. The occurrence of 150 feet 
or more of line, evenly laminated sediments above the medial sand 
and gravel, as at Shell Mountain, shows that the lake after its low- 
water stage again became broad and deep. So far as known, the lava 
flows which have produced such marked features in the walls of the 
Snake River and its tributary canyons were poured out during the 
middle and later stages in the history of the old lake. Some of these 
lava streams entered water and flowed over an irregular bottom, as 
will 1 e des-ribed later, while others show no evidence of such an 
occurrence. These facts seem to indicate that in some instances the 
lava entered a lake, filling the inequalities in its bottom, or flowed over 
an exposed and eroded lake bottom and into streams which occupied 
"Eighteenth Ann. Kept. V. S. Geol. Survey. Part III, 18X9, p. 628, and Twentieth Ann. Rept., 
Part III, 1900, pp. 98-99. 
