METEOROLOGY. 515 



picions against the derivation from the convection theory, for the deriva- 

 tion for quiescent air is certainly free from all objection.* In fact an 

 hypothesis vas made in the assumption of the convection theory that 

 certainly is not proper, namely, that the arriving air in the instant of 

 its passiug by gives up the whole quantity of heat wiS {t — t') that is to 

 say that it is cooled through the whole interval t—t'; with this also 



diffusion is represented (see Stefan, Sitzvngsbericht, Vienna Academy, 

 vol. lxviii, page 385) by the formula : 



where 



' 5 P \ / X 



Pi ?1 = ~ -~ — A 12 (>i p 2 (Mi — U\ 2 ) 



O V 



Px = the density of molecules of aqueous vapor; 



it, = the velocity of molecules of aqueous vapor; 



Pi and « 2 , the same data for the air; 



? is the acceleration, which = under the present assumed 



state of equilibrium ; so also in this case is the evaporation, 



or u 2 =0. Therefore, we have 



d px 



j^- + A 12 fil p. i u v =0 



Since, now, p 2 : S 2 = p 2 T : p T, where S 2 is the normal density, for^ = 

 TOO""", therefore, multiplying and dividing by S h we have 



<lpi_ _A 12 T 8 2 dj p 2 

 S> '" Tp 3 t 



Pi Ui 



But i> 2 = P — pi) furthermore, p x u x is the quantity of evaporation for 

 the unit of surface ; for the spherical bulb of the psychrometer, there- 

 fore, this quantity is Q = 4* r 2 p } u. Therefore, if we put 



we have 



and after integration 



or 



n T T 



whence, evidently, D = ■ ° ^ -. Since p = 760 and -=- can be put 



-a-12 d l d 2-Lo lo 



= 1 for the temperatures occurring in psychrometer observations, there- 

 fore D is nearly constant. 



•[Even this derivation, however, implies certain assumptions that need further in- 

 vestigation. — C. A.] 





