METEOROLOGY. 511 



Assuming t to be the true air temperature and the p to be correctly 

 given by t , we have the ordinary psychrometer formula of August — 



Po=Pi — A (t — h)P, whence, A = — ,f°__ ^ p 



Thirty observations of the dew-point and psychrometer gave Pernter 

 A = 0.0010415, or considerably larger than as deduced from observa- 

 tions at low altitudes. Pernter further considered that a similar re- 

 sult, A = 0.001284, deduced from eight observations with the Schwack- 

 hofer hygrometer, justifies at least the general conclusion that the 

 factor A increases with diminution of pressure.] 



" The object that such comparative observations always have is to 

 construct empirically a formula whose application to psychrometric ob- 

 servations will give the tension of atmospheric vapor with the greatest 

 attainable accuracy. Eegnault, long ago, and since him others, have 

 shown that the theoretical deduction of the psychrometer formula gives 

 no satisfactory result, and thus it might appear that one would do best 

 to renounce the theory and simply seek an empirical formula that shall 

 correspond to the results of observations. I, however, believe that it is 

 precisely the theory of the psychrometer that gives the best starting- 

 points, in order, with help of comparative observations, to arrive at a 

 satisfactory formula, and I must therefore introduce some theoretical 

 views. 



"From theoretical considerations we possess two forms of the psychrom- 

 eter formula — the one deduced from the convection, the other from the 

 diffusion and conduction theory of this instrument. It was August who 

 first deduced his familiar-formula from the theory of convection. Max- 

 well repeats this concisely in the following manner [see the reference be- 

 fore given ; but it should be noticed that the following is merely Max- 

 well's exposition of Dr. Apjohn's reasoning. See Trans. Royal Irish 

 Academy, 1834.] 



Let m = the mass of a quantity of atmosphere. 

 t = temperature " " " 



p = true tension of the aqueous vapor. 

 P = total barometric pressure. 

 a = the density of aqueous vapor relative to air. 

 A = the latent heat of evaporation. 

 Then will 



-?? m (X = the mass of the vapor in this quantity of atmosphere. 



Let 



Pi = the vapor tension corresponding to the temperature If 

 to which the wet thermometer sinks ; then will 



(Pi — po) -p- = the quantity of aqueous vapor evaporated from the wet 



bulb, and 



{pi —Po) -p- A = the quantity of heat necessary to such evaporation. 



