ON THE ENERGY OF STORMS — MARGULES 577 



If then the volumes of the charribers i and 2 are equal, we have 



{V )+U^rt*-±*l^^£r.t*. .(in) 



C p 2 p t V C p 



If p x = -65 , "" 1 mercury, p 2 = 755""" mercury, T = 273 then this 

 equation gives (V) = 1.55 meters per second. 



We will compare this value of (V) with the V that was deduced 

 previously in §(3) for two masses that initially lay in a deep trough 

 alongside of each other (see fig. 1). If h is the altitude [of the two 

 chambers] and 7\* T 2 * are the average temperatures and if the 

 greatest entropy in mass (1) is smaller than that of the lowest layer of 

 mass (2), then for chambers of equal volume the second analysis 

 gave the velocity V = \ Vghx (seethe expression (I) §25). This 

 may be brought into a form similar to that given above in (III) 

 if we put 



* A =log^ ^=log^ 



RTf * p h RT* p h 



ghz = gh — - = approximately RT* — 



whence 



1 ^ Pot ~ Po, ^ RT * (I * } 



Poi 



If for p 0l p 02 and T* we assume the same values as those just given 

 Pi = 765, p 2 = 755, T = 273 there results V = 16 meters per 

 second, or a velocity ten times larger than from equation (III), and 

 therefore a hundred times the kinetic energy per unit mass. 



§(32) We can also add the following problem: 



Let the chambers 1 and 2 of fig. 1 be limited above by a rigid 

 partition and contain masses of air that have equal entropies but 

 different pressures at the same altitudes, consequently there must 

 be a higher temperature on the side of the higher pressure. The 

 difference of pressure at any level is in the same direction at all 

 altitudes and is nearly proportional to the average pressure at 

 that level. The initial stages of 1 and 2 are respectively that of 

 stable equilibrium. After removing the separating vertical parti- 

 tion the adjacent layers on the same level unite. The altitude of 



