On the Steam Calorimeter. 



243 



closed air-tight, permanently, if desired, is placed in the other 

 carrier, but containing air only. The receptacles have been pre- 

 viously calorimetrically compared, and adjusted to have the same 

 thermal capacity as they have the same external volume. Such 

 receptacles are most conveniently constructed for dealing either 

 with liquids or solids in small fragments. In these cases narrow- 

 necked vessels, such as may be closed air-tight without difficulty, 

 may be used. The calorimetric adjustment may of course be 

 effected by inserting a calculated weight of copper or other sub- 

 stance of known specific heat in the vessel deficient in calorific 

 capacity, or the difference may be allowed to remain and the con- 

 stant recorded. 



The thermometer enters the calorimeter midway between the 

 carriers ; and, as the calorimeter is of good conducting metal enclosed 

 in an outer shield of wood, a uniform temperature may be assumed 

 to prevail after some considerable interval of quiet has elapsed. 

 Now, whatever this temperature, or the temperature of the steam, 

 the receptacles and carriers alone are without effect upon the balance. 

 Nor will there be any increment perceived due to radiation.* If 

 one receptacle, however, contain a substance the balance will in- 

 dicate a precipitation due solely to this substance. It is to be 

 supposed that any error of transference of media will affect each re- 

 ceptacle alike, subsequent radiation will also affect them alike, and, 

 as their volumes are identical, the varying buoyancy of the media is 

 without effect on the balance. 



I have said the precipitation will be due solely to the substance. 

 Evidently, however, in very accurate work this cannot quite be 

 assumed. In fact the thermal capacity of the air expelled from the 

 one receptacle on the introduction of the substance must be considered. 

 The precipitation due to this weight of air, considered as possessing 

 the specific heat of constant volume, must be added to w, the observed 

 effect upon the balance. As the air in these receptacles cannot be 

 considered as dry air, this specific heat may be taken as having 

 approximately the value 0*176. The volume of the substance must 

 then, as with the use of the single calorimeter, be estimated, and the 

 weight of air occupying this volume at the prevailing pressure and 

 temperature calculated. For the pressure 760 mm., and temperature 

 10° C, if the steam temperature be assumed as 100° C, the addition 

 to w is 0*000037 gram per c.c. of volume occupied by the substance. 



* To secure this result, I have found — what might be expected — that a similar 

 condition of surface is necessary. With platinum vessels there would probably be 

 little difficulty in attaining this ; but with copper vessels I have found it necessary 

 to keep both surfaces very free from grease and oxidised all over. Washing in 

 ammonium hydrate and heating over a spirit flame seems to bring the surfaces to 

 the desired uniform and permanent condition. 



