284 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 51 



Consequently for the reason thus explained the mean gradient 

 of temperature with altitude is smaller in the actual atmosphere than 

 for the convective equilibrium of pure dry air, and therefore on the 

 average the earth's atmosphere is in stable equilibrium. 



This fact was already recognized by Lord Kelvin. 5 



In the second memoir I have expressed this theorem in the follow- 

 ing form: 



"In general the potential temperature of the atmosphere increases 

 with the altitude." 



Evidently it will now be of interest to investigate more accurately 

 what the diminution of temperature with altitude would be were 

 it dependent only on vertical circulation without addition or abstrac- 

 tion of heat. 



-^ When one has first clearly understood this question then he can 

 obtain from observational data an idea as to how far these processes 

 actually affect the atmosphere and what role other circumstances 

 play such as hitherto have been almost exclusively considered. 



[In the original memoir there follows an investigation into "the 

 average distribution of temperature in a vertical direction." In 

 order to avoid repetition this portion is not reprinted, since this 

 subject, is treated in fuller manner in the following memoir. 6 

 Moreover, my succeeding remarks on the "influence of complete 

 convection on the mean temperature of the circles of latitude" are 

 not reprinted, since these are contained partly in No. VI of this 

 collection 7 and partly in No. XVI of this collection. 8 Note added 

 1905, W. v. B.] 



5 On the convective equilibrium of the atmosphere, dated January 2 r , 

 1862, published in the Memoirs Manchester Phil. Soc, 1865, (3) II, pp. 125- 

 132. 



6 No. XIV of these translations. 



7 See Mechanics of the Earth's Atmosphere, 1891, p. 254. 



8 See No. XX of present collection of translations. 



