ob'8 Prut'. E. Wiedemann on the Behaviour of Gases 



That the amounts of heat here, especially at high pressures, 

 are so much smaller than in the first case, is to be explained 

 by the fact that the potential necessary for each discharge is 

 here so much smaller than before, as the electrodes stood so 

 much nearer to each other. The minimum appears here at a 

 higher pressure than in the first case, since the decrease of 

 heating which occurs in the spark and at the positive electrode 

 is earlier compensated by the increase which takes place at 

 the negative pole. 



A third series of experiments is occupied with the determi- 

 nation of the total heating in hydrogen when between two 

 ball-shaped electrodes is interposed a capillary tube about 

 1 millim. wide and 30 millims. long (Plate IX. fig. 8). In 

 the first series the apparatus broke, in consequence of the high 

 potential necessary when the pressure was very small, since 

 the electrodes were somewhat too close together. The two 

 series of observations are therefore not directly comparable. 

 The first series gave: — 



p 



+ 



- 



184 

 15 

 0-4 



3-10 

 0-61 

 0-43 



2-76 

 0-62 

 0-45 



The second series, in which the water-equivalent was per- 

 ceptibly higher, gave: — 



p 



+ 



- 



1-3 



X 



0-134 

 073 



0-15 

 0-48 



Here also we see that the total heating at first diminishes 

 with decreasing pressure {first series) and then increases again 

 {second series). 



It must remain for further experiments to decide in this 

 case how positive and negative electricity behave, and what 

 part is taken by the heating at the positive pole, and at the 

 negative pole, and in the tube, and how these conditions alter 

 when the length of the interposed capillary tube varies. 



It seemed to be of interest to vary the quantity of electri- 

 city in each discharge, not only by change of pressure, but 

 also by interpolation of sparks — and the more so since in the 

 later experiments and in the study of the spectral phenomena, 

 when the evolution of heat within the capillary tube was de- 

 termined, this method became of manifold application. 



With the apparatus I, in which only the quantity of turpen- 



