62 CHEMISTRY OF THE ROCKS. 



the supposition of the lifting or overcoming any considerable 

 weight in the diffusion of gases. Under the first theory, diffu- 

 sion to a limited extent would be accounted for by the small 

 residuum of chemical or cohesive attraction that would remain 

 between the atoms when separated as they are in the gases ; and 

 under the last theory, by the mechanical impulsion of the mole- 

 cules, through their hitting against each other. Evidently it is 

 a principle which operates only within narrow limits and in the 

 lower temperatures of the gases. The sun gives no indications 

 of such a commingling of its gaseous elements. Spectrum 

 analysis, when applied to its outer edges, shows first hydrogen, 

 then the vapors of sodium and magnesium, and lastly those of 

 calcium and iron. The same fact and order of position are found 

 to exist in the more condensed layers of the sun spots. 



We have also further assumed that the elements in their gaseous 

 states have specific gravities corresponding to their atomic 

 weights. It is well known that all gases, whether simple or 

 compound, at the same temperature and pressure, and not near 

 to a condensing point or other change of state, contain precisely 

 the same number of molecules in the same volume. Therefore 

 it necessarily results that equal measures of the different gases 

 should have weights corresponding to the weights of the molecules 

 of which they are composed. Thus the atom of oxygen is six- 

 teen times as heavy as that of hydrogen ; therefore a cubic foot 

 of oxygen gas will weigh sixteen times as much as a cubic foot of 

 hydrogen gas. This is found to be experimentally true of all 

 the gases that can be measured or weighed. The apparent but 

 not real exceptions are, that in arsenic and phosphorus two atoms 

 of the element unite to form one molecule of the gas, thus mak- 

 ing it twice as heavy as it would be according to the general 

 rule; while in the case of mercury and cadmium, the atom 

 divides into two in forming their vapors. Hence we are not ab- 

 solutely sure in regard to the vapor-molecule, and therefore 

 vapor-density, of such elements as carbon, silicon, and calcium, 

 which chemists have not been able to volatilize. But there is 

 every probability, both from analogy and the position in which 

 some of them are found in the photosphere of the sun, that the 



