438 
CROSS. 
the development of the feldspar seems to be so slight that 
one can hardly avoid concluding that the amorphous base 
must contain the elements of feldspar in addition to hydrous 
silica. 
2d. The hollow spherulites illustrated by Figs. 1 and 2, 
Plate 5, bear evidence of another character. It is found that 
the arborescent feldspar groups spring from many points on 
the surface of the cavity, and all the indications are that the 
cavity was formed prior to the crystallization. One can 
imagine the cavity to be a result of contraction in a sphere 
of colloidal matter. They are not found in glass, but are 
common in spherulites; hence the contraction seems to have 
taken place in a substance, not glass, occupying the spheru- 
litic area. 
3d. The huge compound spherulites of Silver Cliff repre¬ 
sent several peculiarities which seem explainable only on the 
hypothesis that a colloidal mass having practically the chem¬ 
ical composition of the magma separated out in the spheru- 
litic areas. The vein-like spaces described above may then 
be looked upon as radial fissures of contraction in such a 
sphere; and the two feldspathic crystallizations which took 
place after the formation of these fissures may be thought of 
as separations of the colloidal mixture into its two parts 
under the directing influence of the tension existing in the 
large sphere. 
In considering the further application of this hypothesis, 
that the separation of a colloidal substance precedes the for¬ 
mation of branching feldspar needles, it will at once occur 
to many observers that a large number of spherulites are 
holocrystalline and present no evidence as to the former 
existence of an amorphous constituent of any kind ; but this 
undoubted fact may possibly be explained by the very simple 
idea that an individualization of the silica as tridymite or 
quartz will effectually destroy all evidence of this early stage 
in the history of the spherulite; and this change may take 
place at once, before the rock has completely solidified, or in 
a much later period as a decidedly secondary process. It 
