84 SNAKE RIVER PLAINS OF IDAHO. [bull. 199. 
have eroded their banks, the instance just cited is the only one that 
has come under my notice where the evidence of such an occurrence 
was conclusive. Evidently a stream of lava in this respect partakes 
of the character of both a river and a glacier. When the lava is liquid 
it erodes like a stream of water, with the added advantage of greater 
specific gravity; and when viscous, it floats away rock masses that fall 
on its surface in a manner analogous to the way morainal material is 
carried by a glacier. No evidence was obtained to show that lava 
from the Cinder Buttes melted or changed in any conspicuous way the 
rocks with which it came in contact. Such action ma}^ have occurred 
during the earlier stages of the eruption, but does not seem probable. 
The rupturing of cinder cones by outflowing lava is a common fea- 
ture of the Cinder Buttes, and is well illustrated by an isolated volcano 
which built a complete crater, situated about a mile northeast of their 
highest peak. From this volcano the lava flowed westward in a strong 
stream, which made a wide breach in the wall of the crater from which 
it came, leaving the remainder intact. The lava descended a steep 
slope before reaching the plain, and is now covered with soil and vege- 
tation. In this instance, as in so many others near at hand, it is evi- 
dent that the eruption was at first violent and characterized by steam 
explosions which threw out lapilli and scoria 4 , but did not give origin 
to a lava stream; later, when a perfect cone with a deep crater in its 
summit had been formed, lava rose from the conduit within, bleached 
the crater's walls, and flowed quietly away without a renewal of the 
explosive eruptions. This isolated cone with a single lava flow 
extending from it and expanding on the plain recalls forcibly some 
of the illustrations contained in ScropeV* classical book on the extinct 
volcanoes of central France. 
At the locality where the northwest lava stream came to the sur- 
face there is an amphitheater, with rough walls, embracing a generally 
fiat but undulating and corrugated area about 200 yards across, floored 
with scoriaceous and ropy lava, which looks not unlike a great spring 
suddenly congealed. The evidence of flow and of viscosity in the 
now hard material is everywhere conspicuous. The surface is entirely 
bare of vegetation and without lapilli or other deposits. It is appar- 
ent at a glance that this is the hardened surface of the conduit from 
which the lava stream leading from it was poured forth. One is 
impressed with the smallness of the source in comparison with the 
vast volume of material that came from it. The lava stream at first 
went westward, then, curving, took a northern course about the base of 
the portion of its crater still remaining, and at a distance of a mile is 
about 500 yards wide; it continued to expand and reached a distance, 
by estimate, of 6 or 8 miles, and has a width in its expanded distal 
extremity of some 4 or 5 miles. These are rough estimates and are 
"The Geology of the Extinct Volcanoes of ( lentral France, by G. P. Scrope. London, 1858. 
