182 MINERALOGY AND LITHOLOGY. 
found in the larger cavities. The cube is supposed to be of chloride of 
sodium, of which the fluid in the cavity is a super-saturated solution. 
These cavities are mostly entirely irregular in form, but sometimes one 
is found with hexagonal outlines, which correspond to those of the crys- 
tal which encloses it, and the cavity is really an inverted crystal. In Fig. 
6 on Pl. 6 some of the largest of these cavities, as they appear when very 
highly magnified, are represented. Such cavities are found in the quartz 
grains of nearly all the siliceous rocks; and hence what is here said will 
apply to most of those rocks which remain to be described. 
The bubbles which float in the fluids that these cavities contain can 
usually be made to disappear by heating the section to such a point that 
the expansion of the fluid is sufficient to fill the space. Some cavities 
contain a bubble, which disappears when the temperature of the section 
reaches about 30° C. In such cases the determination of the coéfficient 
of expansion, and the optical and other properties of the enclosed fluid, 
prove that it is liquid carbonic acid. Such cavities are very abundantly 
found in certain specimens of sienite from our state; but the bubble in 
the cavities of all the porphyries and granites that I have examined is 
little affected by heat, and would only disappear at a high temperature,— 
and the fluid is therefore water. 
Sorby * was the first to point out the geological significance to be at- 
tached to the presence of these cavities in the minerals of rocks, and he 
concluded that the mode of origin of a rock might be indicated from the 
circumstance whether it contained no cavities, or whether the cavities in 
it were empty, or filled with glassy material, or filled with fluid. The 
uniform presence of cavities filled with fluid in the quartz of granites he 
considered to be evidence confirming the views of Scrope and others, 
that water played an important part in the formation of these rocks. He 
moreover suggested that the relative volume of the fluid and the bubble 
might give some data for the calculation of the temperatures and pres- 
sures under which they were formed, since at a certain temperature the 
bubbles, which are spaces filled only with vapor of the fluid, disappear, 
and the fluid expands and fills the space as it must originally have done. 
But a determination of the temperature at which a cavity containing a 
vacuity in a fluid becomes full, gives only the temperature at which the 
* Phil, Mag. [4] xv, p. 152. Quarterly Fournal Geol. Soc., vol. xiv, p. 453. 
