36 Reynolds and Moorby, Equivalent of Heat. 



In July, 1894, on my applying to him for some 

 research work, Prof. Reynolds asked me to undertake the 

 experimental part of a research (the method and appli- 

 ances having been already devised) on the work expended 

 in heating water from the freezing to the boiling point. 



The method suggested was substantially that set 

 forth below. 



The stream of water was to be supplied to the brake 

 at a temperature approaching as nearly as possible 

 to 32*^F., and was to be raised to a temperature of 2I2*^F. 

 before being discharged from it. 



If these temperatures could be kept steady at the 

 points indicated, two thermometers, the one in the supply 

 pipe and the second in the discharge pipe, would be 

 required ; and their only function would be to supply a 

 means of comparing, in the first case, the temperature of 

 supply to that of melting ice, and, in the second case, 

 the temperature of discharge to that of steam at the 

 standard atmospheric pressure. Thus the calibration of 

 their scales would be of no consequence. 



In order to eliminate losses of heat by conduction 

 along the shaft and by radiation from the brake, the trials 

 were always to be made in pairs, the one carrying a 

 moment of 1,200 ft.-lbs. on the brake^ and the second a 

 moment of 600 ft.-lbs. The duration of each trial was to 

 be the same, viz., one hour. Consequently, since the same 

 difference of temperature would exist in each case between 

 the final temperature of the water and the surrounding 

 air, the loss of heat by radiation would be sensibly the 

 same in both, and the difference of work done in the 

 two trials should be exactly equivalent to the difference of 

 apparent heat generated ; and by dividing the first of 

 these quantities by the second, a value of the constant 

 required would be obtained. 



In addition to the very obvious advantages contained 



