264 ANNUAL OF SCIENTIFIC DISCOVERY. 



We believe, then, that the water which covers the surface of the 

 earth, and the air which surrounds it, will one day disappear, as a 

 necessary consequence of the complete cooling of the interior of our 

 planet. Rocks, with few exceptions, readily absorb moisture, and the 

 more crystalline varieties are the most porous ; we need not, how- 

 ever, consider the quantity of water which rocks may imbibe in this 

 way, for the total amount of this element on the earth's surface is so 

 small, when compared with the whole mass of the globe, that the 

 ordinary processes of chemical analysis would riot detect its presence. 

 If we take the mean depth of the ocean at 600 meters, 1 =1968 feet, 

 its weight will be equal to one twenty-four-thousandth of the earth, 

 which, being reduced to decimals, would give for one hundred parts, 



Earth, 99.9958 



Water, .0042 



In the Bulletin of the Geological Society of France (2d series, 

 vol. x. p. 131), Durocher has published a series of experiments made 

 to determine the quantity of water in those minerals which enter 

 into the structure of rocks, such as the feldspars, micas, hornblende 

 and pyroxene, and which are regarded as anhydrous in composition. 

 These minerals were reduced to coarse powder and exposed to moist 

 air, the proportion of water being determined both before and after. 

 It will be sufficient for our purpose to give the amount of water found 

 after exposure. The orthoclase of Utoe absorbed in this way 0.41 

 for 100 parts, while the mean of seven other varieties of the same 

 species was 1.28, and that of thirty specimens of various substances 

 1.27. We have already seen that if the whole of the ocean were to 

 be equally distributed throughout the earth this would contain only 

 0.0042, or one hundred times less than the least hygrometric of the 

 feldspars. It is probable that the water of the ocean thus absorbed 

 would enter into chemical combination ; at all events, it would occupy 

 a space much less than the pores produced by the shrinking of the 

 rocks. 



If, now, we attempt a similar calculation for the atmosphere, we 

 find that in supposing a height of eight kilometers, the total volume 

 of the air which surrounds our globe, brought to the density which 

 it has at the surface, would be about four millions of cubic myriame- 

 ters, the volume of the earth being equal to 1083 millions, or 270 

 times that of the air, so that a contraction of the primitive volume 

 producing a vacuum of four thousandths (^To) wou kl ^> e m re than 

 sufficient to absorb the whole of the atmosphere. (In calculating the 

 volume of the atmosphere we have multiplied the surface of the globe, 

 in square inyriaineters, by 0.8, which gives a sufficiently accurate result, 

 the more so that the density of the air in the interior of the earth 

 will be everywhere greater than at the surface.) 



It now remains to be seen whether the assumption of a shrinking 

 of four thousandths can be justified by analogies. In the want of di- 



1 This depth is deduced from the comparison of the relative areas of land and 

 water, which are taken as 1 : 3, the elevation and depression of the surface being 

 assumed as proportional to the square roots of their surfaces. The depth of the 

 Pacific Ocean, as deduced by Bache from the earthquake wave of Dec. 1854, was 

 about 13,000 feet. 



