296 Prof. C. Lloyd Morgan—Elevation and Subsidence. 
contraction and yet further subsidence. Conversely beneath an area 
of denudation the decreased pressure may allow some of the solidi- 
fied magma (kept solid by pressure) to liquefy, thus giving rise to 
expansion and uplift. It may also permit the expansion of the 
water gas contained in the magma, and thus give rise to further 
expansion and uplift. And if, as some geologists have contended, 
loading and unloading are in themselves sufficient directly to depress 
or to cause the uplift of a flexible crust, it is clear that subsidence will 
more readily take place in that area which is not only being loaded 
above, but is being also thickened below by the condensation of the 
magma into solid; and that uplift will be more readily effected 
where the crust is not only being denuded above, but being eaten 
into by the melting of the underlayers below. 
May we not perhaps account on somewhat similar principles for 
the existence of the underlying liquid or viscous substratum? Mr. 
Mellard Reade and Mr. Davison have lately independently pointed 
out that, owing to the cooling and contraction of the earth’s crust, 
there is at some depth beneath the surface a level of no stress, where 
there is neither lateral compression nor extension, though the rocks 
are of course subject to the vertical pressure of the overload. Above 
this level the rocks are subject to compressive stress, and below this 
level to tensile stress. Picture the earth as composed, onion-fashion, 
of a number of concentric shells, and fix the attention on one of 
these shells at a depth of say three miles from the surface. Suppose 
this shell to be cooling and undergoing contraction. The shrinkage 
thus brought about will throw the shell into a state of tensile strain 
(like the oft-washed flannel shirt which has become uncomfortably 
tight). Now transfer the attention to the fact that the shells in- 
terior to the one we have selected are contracting. The shell has 
to fit a continually diminishing nucleus (like the frock-coat of a man 
who is rapidly losing flesh). In accommodating itself to this 
shrinking nucleus, the shell is subject to compressive stress. The 
question is, then, with regard to any given shell, Which tendency 
predominates—the compressive stress due to the radial contraction 
of the sphere, or the tensile stress due to the circumferential con- 
traction of the zone in question? At the level of no stress these 
tendencies are equal and opposite: above that level compressive 
stress predominates: below that level tensile stress predominates. 
According to Mr. Davison (Phil. Trans. 1887) the level of no stress 
lies five miles deep from the surface; according to Mr. Mellard 
Reade (Origin of Mountain Ranges, p. 125), it may be taken to lie 
ata depth of one mile: Mr. Osmond Fisher would reduce this to 
less than a mile (Phil. Mag. Jan. 1888). 
Is it not possible, I would suggest, that throughout the zone of 
maximum tension, due to circumferential contraction, the rocks may 
be rendered fluid by relief of pressure ? 
I have now sketched out in briefest possible outline the suggestion 
or suggestions I have to offer with regard to elevation and sub- 
sidence. I have introduced no calculations of the amounts of 
upheaval or subsidence. Although it is easy to see that the accumu- 
