1813.] Specific Heat of the different Gases. 21 S 



raise the temperature of the calorimeter as high as it could be 

 raised by the sole effect of a current of gas, and that the obser- 

 vation of the rate at which it rose during such a process would 

 have been of no utility. We thought it better to raise it artifi- 

 cially, by means of a lamp of spirit of wine placed under it, to 

 a point which, frem preliminary trials, we knew to be near that 

 when the temperature would become stationary. We then made 

 the current of hot gas pass through it, and observed its tempera- 

 ture every ten minutes. Notwithstanding this precaution, it 

 would have been still very tedious to have waited till the cylinder 

 had reached its true maximum, and even difficult to have deter- 

 mined that point by inspection. We found it more convenient 

 to stop the process when, from the slowness of the heating, we 

 judged that the calorimeter was within 0*5, or 0*7 of a degree 

 of its maximum. ITien raising the calorimeter till it passed the 

 maximum a little, and making the current of gas to pass, the 

 calorimeter began to cool. We observed the rate of cooling 

 every ten minutes, and stopped the process when we judged that 

 it had come as near the maximum, on the one hand, as we had 

 got by the heating process, on the other. Taking the mean 

 between these two points, we obtained exactly the maximum 

 point, at which the calorimeter would have remained stationary 

 if the current of hot gas had been long enough continued. 



We determined the temperature of the calorimeter by means 

 of a thermometer with a cylindrical bulb, almost equal in length 

 to the height of the calorimeter. An opening, C, made in the 

 cover allowed the stem of the thermometer to pass out. The 

 thermometer had a very fine bore. Every degree* was 0*89 

 inch in length, which was divided into 10 parts, each of which, 

 by a little practice, we were able to subdivide by the eye into 

 other 10 parts. We were certain of not committing an error 

 equivalent to 2 of these last, or to 0'02 of a degree. As it was 

 of equal importance to know with accuracy the temperature of 

 the air which surrounded the calorimeter, we suspended, at the 

 distance of 3 inches from it, a very sensible thermometer. 



Before entering into the details of our experiment, we must 

 describe the methods which we employed in order to obtain an 

 uniform current of gas, to give it a constant temperature, to 

 determine its temperature when it entered and left the calori- 

 meter, and to appreciate the causes which might, independent of 

 that current, raise the temperature of the calorimeter. 



• The decrees alluded to in the paper are fliose of the centigrade thermome- 

 ter, each degree of which ia to a degree of Fahrenheit, a* 9 to b ; so that had the 

 tlirmiometer uteu divided into Fahrenheit'^ degrees, each degree would have 

 been 0-216 inch in length.— T. 



