M. E.Wiedemann on the Specific Heats of Gases. 91 



and bv substituting the values of « and k we get 



- Tl -rt 2 T |+ m + -T^TT - U 



7?i — ?tt x m — m l t 1 + t 2 T(m — wij) m — m l ft 



T^ 7 F T + m ~T7^T ~T^ + Tx-T" T Tj^T ' 



an equation in whicn the two terms containing r cancel one 

 another. The correction is therefore independent of the tem- 

 perature of the surroundings, t. 



The correction necessitated by the fact that the whole of the 

 mercury- of the thermometer was not surrounded by liquid 

 must, of course, be applied to the temperatures t and t x . 



The temperature indicated bv the thermometer ten seconds 

 after closing the stopcocks was taken to be t x : a part at least 

 of the increase of temperature noticed during this short period 

 must be traced to the fact that the thermometer and its sur- 

 roundings, as also the silver gauze, did not possess the same 

 temperature as the liquid of the calorimeter. Nevertheless the 

 result is unaffected, whether we choose for t 1 this temperature 

 or that indicated at the moment of closing the stopcocks. 



We must now estimate the cooling of the gases. If M re- 

 present the constant (mean) temperature of the heating-vessel 

 during the experiment, then the cooling is represented by 



But if W be the weight of water in the calorimeter added to 

 the water-value of the same, G the weight of the gas, 0\ the 

 corrected temperature-increase of the calorimeter, and C the 

 specific heat of the gas under examination, then 



W.0, = cG0, 

 also 



c ~ G0* 



By multiplying the specific heats by the densities of the gases, 

 referred to air, we obtain the specific heats of equal volumes 

 of the different gases, if that of air be taken, for unit volume, 

 as equal to 0*2389 ; that is, we obtain so-called relative heats. 

 In order, to determine the influence of temperature upon the 

 specific heats of the different gases, they were heated once to 

 100° and another time to 200°, and the mean specific heat 

 estimated between this temperature t and the temperature r 

 (20-25°) of the liquid in the calorimeter. If we represent by 

 Q the amount of heat which must be given up to the gas in 

 order to raise its tpmperature from t° to /°. Q may be repre- 



