42. J. P. Lddings—Nature and origin of Lithophyse 
ment of the obsidian in the flame of an oxyhydrogen blowpipe, 
when it becomes a greatly inflated white glass filled with gas 
bubbles which arise from the expansion of the liberated vapors. 
This shows that the gases which escaped at the surface with 
the formation of pumice were imprisoned in the dense obsidian 
which solidified lower down in the mass. As already stated, 
the larger spherulites are filled with multitudes of gas cavities 
recognizable under the microscope. 
These vapors were, undoubtedly, absorbed by the molten 
lava before its eruption, and were the only gases which may have 
taken part in the production of lithophysee, for chemical analy- 
ses of the obsidian, spherulites and lithophysee show that they 
have essentially the same chemical composition. The analyses 
of spherulites and lithophysee which occurred isolated in dense, 
unfractured obsidian are enough alike to be duplicates. The 
composition of both being the same, the transformation of a 
spherulite into a lithophysa, can only be a modification of its 
structure or a rearrangement of its minerals without any chem- 
ical addition or loss. Moreover, it is evident that such a trans- 
formation was in many cases effected by agencies entirely 
within the area of the lithophyse, for they often occur isolated 
in the dense obsidian having no connection with other cavities 
or sources of vapor. 
The change took place before the surrounding matrix had 
solidified, for in some instances the lithophyse have been 
crushed, prob: ably from a change of equilibrium in the lava, and 
the plastic matrix has been forced part way into the cavity of 
the lithophysa, which indicates that the matrix was still viscous 
and at a high temperature when the modification of the spheru- 
lite took place. And finally, though the matrix was plastic at 
the time of the formation of the cavities of lithophysze the lib- 
eration of gases did not expand or distend the substance so as 
to form the concentric shells, for where a lamination is present 
in the matrix it is also seen traversing the shells of a lithophysa 
without change of direction. 
From the foregoing it seems reasonable to infer that these 
lithopnyse, composed of prismatic quartz, tridymite, soda- 
orthoclase, fayalite and magnetite, are of aqueo-igneous origin, and. 
have been produced by the action of the absorbed gases upon the 
molton glass from which they were liberated during the crystalliza- 
zon consequent upon cooling. 
Instances where the layers of thinly laminated rock arch 
over hollow lithophysee and seem to have been pushed back 
by expanding gases, to which von Richthofen calls attention, 
are more probably slight flexures in the layers of the rock 
which have occasioned local relief of pressure and the disen- 
gagement of vapors which might give rise to the lithophysa, 
