1886.] certain geological phenomena, ete. 23 
would be according to Laplace’s formula about 071914. If the 
depth were 150 miles, X would be about 0:115, and n about 5. If 
we use the latter values, the terms in % will be less than 0:039, 
and for our purpose may be neglected, and we arrive at the 
result, 
z—C€=mlog.(1+n) x (k—c) nearly. 
If now we suppose a still further quantity of gas to be libe- 
rated owing to the surface pressure being further slightly reduced 
from goc to gcc’, and that the height of the column of magma 
then becomes 2’, we have, by substituting 2’ for z and subtracting, 
z’—z=mlog.(1+n)(c—Cc’). 
The upswelling of the column is consequently sensibly proportional 
to the diminution of the superincumbent pressure, whether there 
be already vesicles or not, so long as the rocky matter is saturated 
with gas. 
The doctrine of the entire solidity of the earth rests chiefly 
upon the fact that no appreciable tides can exist in the interior 
of the globe, such for instance as would be formed in a sub- 
stratum of liquid freely intercommunicating throughout a layer 
of the sphere, provided the lquid were as is usually assumed 
incompressible. But the constitution of the magma now suggested 
renders it possible to account for the absence of measureable tides 
therein. 
Let AK be the thickness of the solid crust, 
KB the depth of the substratum at high tide; , 
and suppose that the rocky matter is completely A 
saturated with gas whether there be free gas 
among it or not. Let KC be the small space 4%; C 
through which the tide might fall, if the 
column consisted of inexpansible liquid. 
This fall would be caused by the difference 
of an amount of liquid represented by KC 
within the column. 
Then if we neglect the weight of the small 
amount of gas in KC, the pressure upon the 
column CB is lessened by goKO (see fig. 2), 
so that on the magma upon that account ex- 
panding, we shall have by our formula, 
expansion of BC= mlog. (1 +n) KC. 
It is obvious that, if the space through which BC expands should 
be nearly equal to KC there would be no fall of tide observable. 
This would be the case if m log.(1 +) were nearly unity, that is, 
if the volume of gas absorbed according to Henry’s law measured 
fy 
B 
(1) (2) (3) 
jes) 
ies] 
