156 WRITINGS OF JOSEPH HENRY. [1855- 



case of carbonic acid, the range of distance of atoms is much 

 less in which this law is found ; for by mechanical pressure 

 the gas is converted into a liquid, a sudden change taking 

 place in the intensity of the repulsion of the atoms at this 

 point. Vapor of water, separated from the liquid which 

 produced it, obeys the same law as that of air; but in this 

 instance the range of atoms is still more limited than in that 

 of carbonic acid, and with a slight pressure, and at the ordi- 

 nary temperature of the atmosphere, the vapor is converted 

 into a liquid. 



The atmosphere being subject to the law of Mariotte, we 

 shall now proceed to inquire what will be its condition of 

 equilibrium or rest. 



First. If we suppose the whole atmosphere surrounding 

 the earth to be divided into a series of strata of equal weight, 

 as thin as may be necessary, and separated by ideal surfaces 

 perpendicular to the plumb line, these surfaces will rest 

 upon each other, and be in a state of equilibrium when each 

 part of the same stratum is of the same density. 



Second. In order to a stable equilibrium, the densit}' of 

 each stratum must diminish from below upward. 



Third. The upper stratum must be below the point where 

 the centrifugal force, derived from the rotation of the earth, 

 becomes equal to the weight of tlie air at this point. 



If the first condition is not fulfilled, (that is if the equality 

 of the density of the strata be not the same at all points,) the 

 heavier parts will flow below those which are less dense, and 

 buoy them up in the same manner as the heavier liquid sinks 

 below the lighter one; and it is evident that if the upper 

 strata were heavier than the lower ones, an unstable equi- 

 librium would be produced which the slightest agitation 

 would overthrow. 



Lastly, if the atmosphere extended upward above the 

 point where the centrifugal force equalled the weight of the 

 gas, the whole atmosphere, strange as it may appear, would 

 fly off into void space. To explain this, it is necessary to 

 demonstrate the important though paradoxical fact which 

 results as a logical consequence of the law of Mariotte, 



