and the lamination of acid lavas. 43 
than the result of the expansion of such vapors. For the field 
study of the rock at Obsidian Cliff shows that in. places it was 
so stiff and-viscous before coming to its final rest that some 
layers which were pulled apart in descending the underlying 
slope never closed together again. Hence it is likely that local 
inequalities of pressure might be brought about by the crump- 
ling of so viscous a lava. 
We may imagine the process of formation of a lthophysa to 
have been somewhat as follows: Jn the still plastic glass from 
a center of crystallization a multitude of incipient microlites 
of feldspar radiated through a sphere of glass. As these anhy- 
drous microlites increased in size the nature of the cementing 
paste was changed. Being impoverished of alumina and alka- 
lies, it became more siliceous and relatively more hydrated. 
A point was reached where the absorbed vapors could no longer 
be retained in combination, but were released in innumerable 
bubbles which were either uniformly disseminated through the 
paste or ageregated into larger bubbles. The gas thus libera- 
ted acted as superheated steam and eventually produced the 
separation and crystallization of all the elements of the original 
sphere of plastic glass. Before the final crystallization of the 
paste when the hydrated glass gave up its combined vapors 
and became anhydrous, it shrank in consequence of the reduc- 
tion of volume and produced in some cases the cracks so fre- 
quently observed. That these cracks were formed before the 
final crystallization of the cementing paste is shown by the 
deposition on their walls of crystals of quartz, tridymite and 
fayalite. The conditions which produce a concentrically banded 
structure in some solid spherulites are most likely the same as 
those which lead to the formation of concentric shells in many 
lithophyse. It is evident that the development of lithophysze 
in a lava will depend upon a number of conditions both chem- 
ical and physical, the discussion of which is omitted from this 
paper. 
Laminated structure.—The lithoidal portion of the obsidian 
flow which forms Obsidian Cliff is beautifully laminated in 
thin layers of light and dark shades of a purplish gray color, 
which differ in their relative denseness and degrees of crystalli- 
zation. The more crystalline layers are full of minute cavities 
and thus become planes of weakness in the rock, which splits 
into thin plates, often not more than ~; of an inch thick. So 
that the lithoidal rock is traversed by a multitude of nearly 
horizontal cracks which follow the planes of flow through all 
their contortions. It is this inherent lamination or layer struc- 
ture, so commonly observed in the more acid lavas and espe- 
cially the rhyolites of the Yellowstone National Park, which 
was alluded to in the paper on columnar structure published in 
