L. Merritt — Light from Incandescent Lamps. 169 



must also remain constant, and the heat absorbed by the'water 

 is measured by the product of this difference in temperature 

 into the amount of water passing through the calorimeter. 



Let t = time occupied by the experiment. 

 Tj = temperature of water entering. 

 T 2 = " " leaving. 



F = rate of flow of the water. 

 R = heat lost by radiation. 



Then H, the heat absorbed by the water is : 

 H = [T 2 -TJF*+R. 



This method of determining the heat absorbed was adopted 

 for several reasons. In the first place the correction for radia- 

 tion was very small ; we were always able to make the temper- 

 ature of the air intermediate between T 1 and T 2 and of such 

 value that the absorption of heat by the lower part of the 

 calorimeter nearly balanced the radiation from the upper part. 

 As it was impossible in our experiments to use a water jacket, 

 or similar arrangement for reducing the radiation, this fact was 

 of great importance. Then, if the ordinary form of calorime- 

 ter had been used, it would have been necessary to use a stir- 

 rer, which would have absorbed a considerable amount of light. 

 Finally, by using this method we were enabled to take a num- 

 ber of readings of temperature and current instead of only 

 one, and could continue the experiment as long as we wished. 



The electrical energy supplied to the lamp was determined 

 by measurements of current and resistance. The dynamo 

 furnishing the current was allowed to run at its normal poten- 

 tial and to send a current of 10 or 15 Amperes through a 

 dead resistance of heavy German silver wire. The lamp was 

 connected so as to be in multiple with enough of this resistance 

 to give the desired difference of potential. The resistance in 

 multiple with the lamp, and the resistance of the lamp connec- 

 tions, were measured after the close of each experiment. The 

 total current, and the lamp current, having been measured dur- 

 ing the experiment, the resistance of the lamp was calculated 

 from these data. The large tangent galvanometer in the Mag- 

 netic Observatory of Cornell University was used in measur- 

 ing currents. For H the average value 0*1718 was used 

 throughout the experiments. 



The thermometers used were Baudin's " specific heat " ther- 

 mometers, Nos. 10214 and 10294, graduated to 0°'02. They 

 were read every minute by means of telescopes placed about 

 two metres away and at the same height as the mercury in the 

 tubes. After each experiment the two thermometers were 

 compared to determine the difference in zero point. Curves 



