OF ARTS AND SCIENCES. 147 



where a' p and a' v are the true coefficients of expansion at the given 

 temperature ; 



•- /= — ( 1 + 5 m — — ) — , 

 J=4- 



l' : V*/to \ "^ /«„ ' V* \ » ) 



According to Thomson and Joule's experiments ?n = 0°.33 C. for 

 air and about 2°.0 for CO.. Hence ,x = 272°.99. 



The equations should be applied to the observations directly at the 

 given temperature, but it will generally be sufficient to use them after 

 reduction to 0° C. Using A'= .2375 according to Regnault for air, 

 we have for the latitude of Baltimore, — 



From Rontgen's value y = 1.4053 — = 430.3* 



" Amagat's " 1.397 — = 436.G. 



9 



" velocity of sound 33 1.78 m - per sec. — = 429. G. 

 Using Wiedemann's value for K, .2389, these become 



^ = 427.8; ^ = 434.0; L = 427.1. 



9 9 9 



As Wiedemann, however, used the mercurial thermometer, and as 

 the reduction to the air thermometer would increase these figures 

 from .2 to .8 per cent., it is evident that Regnault's value for K is 

 the more nearly correct. I take the weights rather arbitrarily as 

 follows : — 



Weight. J. 



Rontgen 3 430.3 



Amajrat 1 436.6 



Velocity of sound 4 429.6 



Mean 430.7 



And this is of course the value referred to water at 14° C. and in the 

 latitude of Baltimore. My value at this point is 427.7. 



* Rontgen gives the value 428.1 for the latitude of Paris as calculated by a 

 formula of Shroder v. d. Kolk, and 427.3 from the formula for a perfect gas, 

 and tliese hoth agree more nearly with my result tlian that calculated from my 

 own formula. 



