36 1 Prof. E. Wiedemann on the Behaviour of Gases 



discharges the gas returns to its original condition — which 

 is the case in our experiments, us no phosphorescence of the 

 gas was observed. 



This method, since it is one of estimation only, cannot of 

 course give so accurate a result as the heliometric method; 

 but it is sufficient for such qualitative-quantitative experiments 

 as the present. 



The measurements were made and calculated in the following 

 way. The electrical machine was first set in action by turning 

 on the water, the temperature of the thermometer in the 

 calorimeter read off, and the galvanometer observed. The 

 electrodes of the machine were then connected by a brass rod. 

 After three to five minutes the temperature was read again, 

 the brass rod removed by means of a cord passing over 

 pulleys, and the current allowed to pass through the dis- 

 charge-tube ; at the same time the galvanometer was observed. 

 After an interval of from one to eight minutes, the machine 

 was again closed, and the thermometer observed from minute 

 to minute, and read off at the end of three minutes, when the 

 movement of the temperature in the calorimeter had become 

 uniform. The experiment was allowed to last so long that a 

 change of temperature of two or more degrees resulted ; only 

 in special cases, in which the heating in the whole tube, or 

 that at the positive pole at very low pressures, was to be de- 

 termined, was the change of temperature less. 



After the experiment, the machine was stopped, and the 

 calorimeter allowed to cool by radiation, so that its tempe- 

 rature might not alter too much. The measurements before 

 and after the passage of the electricity enabled corrections to 

 be applied to the observed rise of temperature, according to 

 well-known methods. The correction could not be applied 

 for the conduction of heat from the parts of the tube not sur- 

 rounded by the calorimeter. Each experiment occupied at 

 least half an hour. 



The galvanometer-deflections were not reduced to angles, 

 since for the degree of accuracy reached at present the error 

 so caused may be neglected. 



The heating is proportional, during varied action of the 

 machine, to the deflection of the galvanometer, as my father's 

 experiments and control-experiments of my own have shown. 

 In fact, since the discharges are entirely discontinuous, and 

 the intensity of each single discharge is the same, when n 

 times as much electricity passes through there will be n times 

 times as many discharges, and consequently n times as much 

 heat produced. 



