282 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 51 



sion to compression but only after the last of the water and ice is 

 completely evaporated. Therefore the pseudo-adiabatic change of 

 condition is in its minutest features always reversible but in its 

 entirety is not so. I have therefore designated such processes 

 as "partially reversible" or "pseudo-reversible." 



But this theorem that the so-called adiabatic changes of condi- 

 tion of moist air in the free atmosphere are not completely rever- 

 sible, is one of the most important for all theoretical meteorology 

 and climatology. By it there become explicable not only the 

 foehn phenomena, whose study, as is well known, formed the 

 starting point for all incisive investigations, but also the contrast 

 in the character of the weather in the regions of high and low atmos- 

 pheric pressure; the difference in the conditions on the windward 

 and leeward sides of mountain ranges; the distribution of cloudi- 

 ness and precipitation in general; and finally, as above mentioned, 

 the law of the average diminution of temperature with altitude, at 

 least in its principal features, together with the relatively slight 

 diminution of the average temperature of whole small circles of 

 latitudes from the equator up to the "horse" latitudes. 



It is therefore worth while to first examine most carefully the 

 verbal expression of this theorem and then to deduce more rigor- 

 ously and elucidate more thoroughly than I did in my first memoir 4 

 the conclusions that result from it. 



The shortest and most rigid statement of the theorem is given 

 when we introduce the idea of potential temperature. I have else- 

 where done this and expressed the theorem in two ways as follows: 



"In the adiabatic change of condition of moist air, the potential 

 temperature remains unchanged so long as the dry stage is not 

 passed, but it increases when condensation begins and in propor- 

 tion to the quantity of water that is abstracted." 



Or otherwise it is expressed as follows when specially applied to 

 atmospheric processes: 



"Adiabatic changes in the free atmosphere and when there is 

 no evaporation, leave the potential temperature either unchanged 

 or higher." 



In both of these equivalent methods of expression I have at that 

 time made it understood that I assumed the adiabatic and pseudo- 

 adiabatic processes as equivalent, as we may safely do both in the 

 computations and the graphic presentations. 



* See Sitzb. Berlin, 1888, pp. 1189-1206, or pp. 243-257 of the preceding 

 collection of translations. 



