166 
MR. ROBERT MALLET OX VOLCANIC ENERGY. 
that the exterior becomes rapidly cooled to a rigid state, leaves a central cavity when 
cold, the last consolidating particles having parted company towards the centre under 
the constraint of the rigid shell, whose diameter does not diminish so as to accommodate 
the total contraction. 
All very large cylindrical forgings of wrought iron, say of 2 or 3 feet diameter, become 
“flawed” as they cool, tearing asunder in greater or less irregular diametral and axial 
planes, as in section, fig. 3. The whole mass was at first, say, 
at a yellow heat and plastic; the outer coats cool first and 
most rapidly, and form a rigid arch (whose thickness we may 
suppose — covering the still hot interior, and whose di- 
mensions are determined by the volume of the latter which is 
grasped by the rigid arched shell. The cooling of the interior c 
proceeds, the heat being lost through the rigid shell ; and as 
the shell cannot yield to the radial pull of the contracting 
nucleus y a 1 , the diminution in volume of the nucleus is met 
by its tearing asunder in some one diameter, y x, and by 
deformation of the cylinder itself in the orthogonal direction as soon as this rent has 
been formed. 
The exact play and interdependence of these forces, here imperfectly stated, have 
been fully developed by the writer in a paper “ On the Coefficients of Elasticity and 
of Rupture of Wrought Iron, &c.,” printed in the 18th volume of the Minutes of Pro- 
ceedings of the Institution of Civil Engineers, pp. 307-312, to which he would refer. 
The case of a cylinder is (as regards illustration) the same as a sphere, as the material 
of the nucleus is there too viscid to escape endways. 
65. In the case of a small cylinder or sphere with a relatively very thick and rigid shell, 
the final contraction of the nucleus is met, as we see, by tearing asunder of the nucleus 
by radial tension, and not by crushing of the material of the covering dome by tangen- 
tial compression. 
But it is obvious that the one or the other must result, as the sum of the tensile 
resistances of the nucleus or that of the compressive resistance of the covering dome is 
the greater. 
And in a very large globe such as ours, where the active force is gravitative in all its 
parts, and where transverse section in the plane of any great circle enormously exceeds 
that of the shell, even if we suppose that some hundreds of miles in thickness, it is 
obvious that it is the shell that must yield by crushing up, and not the still heated 
nucleus by pulling asunder, though it is quite possible that in some remote future 
stage of final refrigeration a diametral rent might occur, or some central cavity be left 
in the then cold and rigid nucleus itself of our globe. 
66. The contraction now going on in our globe by its secular refrigeration is met there- 
fore by the compression of the colder and more rigid covering shell and crust, and by the 
crushing of its material along lines or at places or planes of weakness by their mutual 
a 
b 
