172 



On a Modification of B 



unsen s 



Calorimeter. 



the tube by the body, this mercury being in its turn sur- 

 rounded by melting ice not in contact with it. The instru- 

 ment figured was made by 

 Casella. T is the graduated 

 tube of a glass thermometer 

 whose large bulb is encased 

 in the copper chamber M, 

 there being only a small air 

 space between the two. In 

 the centre of this large bulb 

 a test-tube t, provided with 

 a cork, is tightly inserted, so 

 that its lower portion is com- 

 pletely surrounded by the 

 mercury in the bulb. • 



Prof. Stewart intrusted 

 the calorimeter to us, in order 

 that we might determine its 

 working conditions and as- 

 certain how far it was reli- 

 able. 



The following was the way 

 in which the apparatus was 

 tested. The test-tube was 

 first filled with water to the level of the mercury inside the 

 bulb. The instrument being placed in its wooden case, the 

 latter was then filled with pounded ice and then allowed to 

 cool down until the column of mercury in the stem of the 

 thermometer T was apparently stationary between two obser- 

 vations taken at an interval of about fifteen minutes. 



After preliminary trials with various quantities of mercury 

 dropped into the test-tube, it was found absolutely necessary 

 to make some allowance for the loss of heat from the surface 

 of the mercurial thermometer by radiation and convection ; 

 for although the rise produced by dropping in was almost im- 

 mediate, yet during a comparatively long period of apparent 

 maximum that followed, the heat received must have been 

 equivalent to the heat lost by radiation and convection. (This 

 prolonged stationary period, it was always observed, was fol- 

 lowed by a comparatively rapid fall.) A curve of cooling, 

 extending from o, 3 to o, C, was accordingly obtained, with 

 times as abscissas and temperatures as ordinates. By aid of 

 this curve the loss of heat was allowed for by a simple method 

 of compensation. 



Brass and mercury were selected for comparison. The 

 results, after applying the correction above indicated, for 



