Mr. Ivory ofi the Constitution of the Atmosphere. 199 



azote with one atom of oxygen. According to the view of 

 the matter we have here taken, there is nonsuch thing as an 

 atom of an-: for a vohime of that fluid contains the atoms in 

 21 parts of oxygen and 79 parts of azote, the atoms of each 

 ot the two gases being separately arranged in symmetrical 

 order through the whole volume of air. The difficulty al- 

 luded to is completely solved when it is observed that the 

 action of any one of the gases at a point of the common sur- 

 face IS transmitted to all the other gases; by which means 

 the equilibrium of the mixture is established, not by the ac- 

 tion of the particles, but by the mutual action of the elasti- 

 cities of the fluids. 



The method of investigation followed above, has the ad- 

 vantage of proving that the resolution of air into its element- 

 ary gases IS a necessary consequence of the law of Mariotte : 

 for the three equations (1.) are applications of that law to the 

 different gases. But the elements of which air is composed 

 might have been deduced from the theory of mixed gases, as 

 explained in the last Number of this Journal: and it may be 

 no improper addition to show that two trains of reasoning 

 which appear so little connected, Ifead nevertheless to the same 

 result. The common temperature being 6, letp, «, o denote 

 the pressure, density and volume of a portion of oxygen, 

 and p, q', V the like quantities of a portion of azote : if the 

 two fluids be introduced into an envelop, the volume of which 

 IS V = o + u', It is shown that the elasticity of the mixture, 

 that is, the pressure at every point of the envelop and the 

 elastic force which keeps in its place every particle within the 

 envelop, is equal to 



On' 



P- Y'+P--\r-- 



and if we suppose p = p', the same elastic force will be simply 

 p. Let R denote the density of atmospheric air under the ' 

 pressure^ and temperature 0: then, as observed before, R, 

 g, § will be proportional to the numbers 1, A, A': so that 

 ? = A R, 

 §' = A' R, 

 gu + g'u' = (Au + A'o'). R. 

 Now JO + g'v' is the sum of the masses of the constituent 

 parts of the envelop : wherefore, V being the volume of tiie 

 envelop, il D represent the density of the mixture, we shall 



„ , VD= (Ar> ^ A<u')R. 



t urther, we may assume 



V = Ay + A't>': 



