THERMODYNAMICS OF ATMOSPHERE VON BEZOLD 283 



But since this is not allowable from a strictly theoretical point 

 of view, therefore I would prefer to substitute the following method 

 of enunciation : 



"An adiabatic change of condition of moist air leaves the poten- 

 tial temperature unchanged, but a pseudo-adiabatic change raises 

 the potential temperature. The rise increases with the amount of 

 water that is abstracted." 



Of course these theorems are only applicable so long as we exclude 

 cases of mixture with air having a different temperature and differ- 

 ent absolute humidity, as also cases where water is added from any 

 source whatever. 



The precipitation or abstraction of water during adiabatic changes 

 of condition is also excluded because the very definition of "adia- 

 batic change" implies that the mass under consideration, that is, the 

 mixture of air and water remains the same quantitatively notwith- 

 standing the change of condition as to aggregation. 



On the other hand, under "pseudo-adiabatic changes" are included 

 all those in which the condensations that are formed either wholly 

 or partially fall away so that the quantity of water mixed 

 w r ith the given quantity or unit mass of dry air is diminished by the 

 abstraction or precipitation of the condensation. An increase of 

 this mass of water by addition from outside is also excluded by 

 this definition. 



In consideration of adherence to this latter definition of our 

 terms, the following theorem may be expressed : 



"An adiabatic change of condition may be either an expansion or 

 compression. A pseudo-adiabatic change is only possible with 

 expansion." 



Since the ascension of a mass of air is always attended by expan- 

 sion, and is pesudo-adiabatic after the beginning of the formation of 

 precipitation, if this expansion takes place without addition or 

 abstraction of heat, therefore by this process the potential tempera- 

 ture of the upper layers of air is increased. 



Since the potential temperature remains constant in descending 

 currents so long as no heat is added or taken away, therefore the 

 vertical movements of the air without change of heat should be 

 alone sufficient to make the average diminution of temperature with 

 altitude smaller than would result if the air contained no aqueous 

 vapor. Therefore, independent of all processes of gain or loss 

 of heat the simple ascent and descent of currents of air suffice to 

 explain why the temperature diminishes with altitude and in fact 

 slower than i° (or more exactly than 0.99 C.) per 100 meters. 



