Russell.] LAVA STREAMS THAT ENTERED WATER. 115 
partially embedded in the black glass. In many instances, also, the 
interiors of cavities are lined with similar material, and at times masses 
of sand or clay are completely inclosed in the cellular, glassy lava. 
All of the facts enumerated unite in indicating that the lava while yet 
molten entered water and was quickly cooled, hardening to a glass, 
and that the steam generated blew the still plastic material into shreds 
and excessively irregular branching forms. As the upper surfaces 
of the lava sheets are not glassy and do not show the other character- 
istics referred to, it appears that the waters which the still hot lava 
entered were shallow. The still plastic material in numerous instances 
was rolled or folded into spindle-shaped and pillow-like forms, and 
even into spherical and oval balls, the surfaces of which in contact with 
water cooled quickly to a glass and the part within developed large 
crystals of olivine and other minerals. All stages in this rolling-up 
process, from folded masses to complete spheres, are illustrated. The 
still plastic lava, coming in contact with the sand at the lake bottom, 
adhered to it, thus giving the black glass a white surface, and in many 
instances masses of sand and cla}^ became involved in the basal portion 
of the flow. The characteristics here accounted for pertain to the 
lava which was fluid or plastic on entering water and was distended 
and variously modified by the steam generated, but not shattered. In 
other instances the rock was broken into angular fragments and a 
breccia produced. When this occurred it appears that the lava, 
although still hot, had lost its plasticity on entering the water and was 
fractured instead of being blown into shreds and similar branching 
forms. 
When a lava stream enters water it is evident that a connected series 
of products should be expected, ranging from compact or what may 
be termed normal lava to lava that has been shredded, as it were, by 
steam explosions acting on it from the exterior, and from normal lava 
again to coarse and fine breccia, the controlling condition being the 
degree of plasticity or rigidity of the magma. Between basic and 
acid lava it is to be expected that the former would be the more apt to 
furnish cellular, torn, and ragged masses or be rolled into balls; while 
the latter, owing to its less fusibility and greater brittleness when 
solid, would be the more apt to become shattered. 
The coarser breccia observed in connection with the steam-blown 
basalt in the walls of Snake River Canyon is composed of angular frag- 
ments of normal granular basalt, but the smaller fragments are glassy. 
The finer breccia resembles closely the tuff formed of lapilli blown 
out by steam explosions from volcanoes, but there seems to be a dif- 
ference between the two, although I am not sure it can be readily 
recognized. Lapilli are characteristically and perhaps always scori- 
iceous, while the fragments produced by the contact of hot lava with 
^ater may not be scoriaceous, and in the case of lava streams that 
