514 SCIENTIFIC RECORD FOR 1888. 



since the radiation (I) can be put = OK (t—f), where O = the surface 

 and R = the coefficient of radiation, we should have had 



OR (*-/') + mS (*-*) = (Pi -Po) ^- 



or, 



PST OR-1 _ 



^=^->L 1 + H5j (*-*') < D) 



[This is, therefore, the Pernter-Maxwell formula, in which convection 

 and radiation are both considered.] 



The deduction of psychrometer formulae under the assumption of 

 perfectly calm air (i. e., neglecting convection and considering only 

 radiation, conduction, and diffusion) has been completely given by 

 Maxwell,* and Stefan ,t and they have arrived at the following expres- 

 sions, respectively: 



Maxwell, . . . lh=Vl -™(* + ^ ( , -V) (E) 



Stefan, p. = i> 1 -^_(K+Er) (t-V) (F) 



which latter becomes the same as the former if we put K' = -^, i. e., 



Maxwell-Stefan, . . . #o=i>i-^(p4-^T) {t-V) (G) 



In these equations we have [assumed the thermometer bulbs to be 

 spherical and of radius r, and] put 



p = the normal weight of the unit of volume of the air; 



K = the coefficient of conduction of the air : 



D = the coefficient of diffusion of aqueous vapor in the air. 



Since, now, K' = -g = 0.18, according to the experiments and com- 

 putations of Stefan, and D is also = 0.18, according to the statement of 

 Stefan, therefore the formula deduced for quiescent air acquires thf* 

 same form, i. e., 



PS/- Rr\ , 

 2>o = *i- 17^ + ^ (*-*') (H) 



as that deduced from the convection theory.! This must arouse sus- 



# See Encyclopaedia Britannica, 9th edition, Vol. vn, Art. Diffusion, London, 1878, and 

 Zeit. Ost. Gesell. Met., xvi. 



tSee Zei schrift der Ont. Gesell. fiir Meteorologie, xvi, p. 177. 



! Since D, the coefficient of diffusion, may be an unfamiliar term, I 

 will here, by its deduction from Stefan's diffusion theory, briefly show 

 what meaning it has in our formula. Let A 12 be a constant depending 

 only upon the nature of aqueous vapor and the air, then the piocess of 



