170 Prof. J. Milne—Distribution of Volcanos. 
In the case where land slopes gently beneath the sea, as along the 
eastern coasts of North and South America, the inclination of any 
isotherm would be gentle and the decrease in thickness of the layer 
of surface rock which covers up and retains the more expansive rock 
beneath would be also gentle. In the case, however, where the 120 
mile slope beneath the sea was steep, the thickness and consequent 
strength of the restraining cover would rapidly decrease as it ran up 
from beneath the sea towards the land. 
Fie. 2.—CC’ is an imaginary isotherm when it passes from the land beneath the sea. 
These conditions are shown in Fig. 2. BB’ represents sea-level, 
AA the surface of the land. The dotted line CC’ an imaginary 
isotherm. Beneath this line we will suppose the rocks to be ina 
slowly viscous state. Above this line in a state sufficiently rigid to 
resist alteration of form. The load over the point C’ equals the 
weight of the thick layer of superincumbent rigid rock plus the 
weight of a column of water. The load over Cis only equal to the 
weight of a comparatively small layer of rigid rock. The layer of 
solid rock is something like the chain of a suspension bridge which 
is heavily loaded in the middle. 
If conditions anything like this exist in nature, then we should 
expect to find lines of fracture and volcanic outbursts along any line 
of land which slopes down quickly beneath the sea. Where the high 
land from which the volcanos rise forms a ridge between two seas, 
as, for instance, is the case in Central America, we should have the 
chain of solid material, so to speak, hanging downwards on two sides, 
and thus it would be strained from two directions. A similar remark 
would apply to chains of islands like the Aleutians, the Kuriles, and 
those which form Japan. 
When considering the position which isothermal surfaces pro- 
bably occupy beneath the surface of the land, we must remember that 
whenever we get a contour which approximates to a reentrant angle, 
an isothermal surface will probably come nearer to the surface than 
it would at an upward bend in the earth’s crust corresponding to a 
salient angle. Thus, if we had a plain from which a range of 
mountains suddenly rose up, we should expect at the junction of the 
plain with the mountains that an isotherm would be nearer to the 
surface of the ground than it would be at the summit of the range. 
