198 SIG IIIS, SEO SI QUO IIN TiS 
In applying these results to ascertain the depth at which cavi- 
ties would close, the following additional assumptions are made: 
Some of the cavities are supposed to have a form best adapted 
to resist closing. In most cases, as shown by Professor Hoskins, 
this is probably spherical. The cavities are supposed to be very 
small as compared with their depth below the surface, so that 
the pressure due to gravity is practically the same for all parts 
of a cavity. The rocks are assumed as being among the strong- 
est—that is, having a crushing strength of 1700 kilograms per 
square centimeter. This amount somewhat exceeds the average 
crushing strength of the ordinary crystalline rocks, such as granite 
and schist, but is surpassed by about one-fifth in some of the 
very strongest hornblende-granites and basic rocks. It is prob- 
able that the stronger, and perhaps the strongest, rocks should 
be chosen, for the minute cavities in the interstices are con- 
cerned, and these can only be closed by the crushing or flowing 
of the individual mineral particles. Upon the one hand it is 
natural to suppose that some of the minerals composing rocks 
are stronger than any rock. Upon the other hand the cavities 
may be largely closed by the flowage or fracture of the weaker 
minerals of a given rock. However this may be, for the pur- 
poses of the present discussion it is plain that sandstones are to 
be placed upon the same basis as quartzites, and possibly shales 
upon the same basis as slates and schists. 
The specific gravity of the outer crust of the earth is assumed 
to be 2.7.1 As openings in the earth are usually filled with water, 
in obtaining a maximum depth at which cavities can exist per- 
manently, it is probably necessary to suppose that cavities are 
supported by the hydrostatic pressure of a column of water 
reaching to the surface, and therefore that 1 should be sub- 
tracted from the specific gravity of the rock in determining the 
depth at which the closing of cavities occurs. Under the slowly 
acting orogenic forces it is probable that the water must be 
* This estimate of 2.7 was kindly furnished me by Mr. G. K. GILBERT as a close 
approximation to the specific gravity of the continental masses. It is the same as my 
own best guess of the specific gravity of the pre-Cambrian rocks. 
