The critical temperature of water--that temperature at which it changes into steam under any pressure however great--being 412°, its pressure in the state of steam will be somewhere about 7150 lb. per square inch, let us say 500 atmospheres; then, if 400 atmospheres are required to balance 1 mile in depth of average rock, as we have stated above, the pressure of steam just cited would balance only 1-1/4 miles of rock. We can, therefore, see how inadequate it would be to force a column of lava up from even the depth of 9 miles. At that depth 3600 atmospheres of pressure are required to balance a column of lava, and there are only 500 available. It has been said that the downward pressure of steam would force up the lava through the vent of a volcano, but an arrangement of that kind would require a downcast shaft as well as the upcast one of the vent like as there are in collieries; but the downcast would have to go very deep to compress the steam--a gas now--to the required number of atmospheres. Far more likely that the steam itself would put an end to any increase of water, by driving it back through the channels by which it was descending; for if they are supposed to exist under a solid crust of 800 miles thick the pressure required would be 320,000 atmospheres, and with a crust of only 100 miles thick 40,000 would be required. The only way, therefore, in which volcanic eruptions can be produced in the earth, if solid or liquid, or partially solid and partially liquid, to the centre--in other words, from magma layers--is by the shrinking of the crust squeezing out the lava. With a hollow earth and shell of more or less 2200 miles in thickness, liquid to some depth on the interior surface the difficulty becomes very much less. The communication between the vents of volcanoes would be complete and simple, without any lateral forcing of the lava through magma layers made expressly for the purpose; it would be an open and natural flow from one place to another. That there are such volcano vents connected with each other has been very generally believed, and even almost proved by observation of eruptions taking place in two or more almost simultaneously, or at the least showing signs of violent agitation, the motive forces for which would be the gases which we have concluded must be imprisoned in the hollow centre. When their pressure came to be sufficient to blow or force out the liquid, or semiliquid matter, bubbling and boiling in the vents in constant activity, there would be an eruption, during and after which the gases would escape till their pressure was greatly reduced, when the volcanoes would return to their semi-active state. The gases would naturally be those of the many kinds that are found in eruptions, by reason of their being generated in the earth, mixed with steam and water in the manner we have already shown.