200 ANNUAL OF SCIENTIFIC DISCOVERY. 



sion which is perhaps more in harmony with the feeble and local ac- 

 tion which the interior is known to exert on the surface. Since, then, 

 the data are wanting to fix the amount -of shrinking in the crystalli- 

 zation of rocks, we may find in an analogous phenomenon some terms 

 of comparison. The difference between the density of cast metals 

 and the same after hammering can only arise from a contraction sim- 

 ilar to that which takes place in igneous rocks. The surface becom- 

 ing solid while the interior is yet liquid, the natural contraction of 

 this portion is prevented, and from this necessarily result vacant 

 spaces in the mass, which are afterwards compressed by the action of 

 the hammer. In calculating from the differences in density the vol- 

 ume of the vacant spaces thus produced, we find for iron a contrac- 

 tion of 0.075; for nickel, 0.045; for aluminum, 0.041; for copper, 

 0.011 ; for gold, 0.005 ; while the contraction of the earth necessary 

 to absorb the whole atmosphere would be only 0.004. From this it 

 results that an ino-ot of gold, the most solid obtained bv the fusion of 



^j ^2 *. 



a metal, contains more vacant space in proportion to its volume than 

 would be required in the globe for the absorption of its gaseous en- 

 velope; it is scarcely possible that any crystalline rock should be 

 wanting in this slight degree of porosity. 



From the preceding considerations, the successive absorption of 

 the air and water by the solid portions of the globe becomes in the 

 highest degree probable, and we may conclude that our earth will 

 one day present that same total absence of ocean and atmosphere 

 which we now remark in the nioon. It is evident that this progress 

 of the waters towards the earth's centre must have long been in op- 

 eration, and it becomes interesting to consider the effect which this 

 must have had upon the level of the ocean. Let us suppose that the 

 rocks near to the surface of the earth contain one hundredth of 

 water, a proportion which, from the above calculation, will not be 

 regarded as excessive, and that the water moreover does not exist in 

 this proportion at a depth beyond that at which the terrestrial heat 

 equals 100 Centigrade. If we take the augmentation of heat in de- 

 scendincr to be one decree for thirty-three meters, this will give a 



t^ ^j * *3? 



depth of about 3,000 meters, while one part of water by weight in 

 one hundred parts of a rock whose density is equal to 2.5, will cor- 

 respond to a volume of one-fortieth. We shall now calculate the 

 volume of this external layer which we have supposed to be thus im- 

 pregnated with water, regarding it as a prism having for its base the 

 surface of the earth, with a height of 3,000 meters, which would give 

 a mass of 1,530,000 cubic myriameters, containing 38,000 cubic myr- 

 iameters of water. The total volume of the ocean being one forty- 

 eight thousandth that of the globe, or 225,000 cubic myriameters, it 

 follows that this layer of 3,000 meters of earth would contain a vol- 

 ume of water equal to one-sixth of the present ocean. Whatever 

 may be the real value of these figures which we have adopted to ren- 

 der the demonstration more clear, the interest and importance of this 

 inquiry is evident. 



I am convinced that the ultimate complete cooling of the interior 

 of the earth is inevitable. We may affirm on general principles that 

 between two media of different temperatures, separated by a layer of 

 rock which is a conductor of heat, an equilibrium will at length be 



