50 Prof. E. Wiedemann on the 



As with this tube the numbers run exactly as with the 

 others, no curve is given. 



All these experiments show that for the low pressures at 

 which the dark kathode space is distinctly formed the follow- 

 ing holds good : — 



The 'production of heat in the electric discharge, starting from 

 the positive electrode, rises at first sloivly, then rapidly , reaching 

 its maximum within the " glow -rays," from which point it 

 decreases, attaining a relative minimum within the dark space 

 about the kathode, to ivhich, however, corresponds a much higher 

 production of heat than in the positive light. At the kathode 

 itself the production of heat attains a very high value, as shown 

 by separate experiments. 



If the pressure, is increased, then with decrease of length 

 of the dark kathode-space and shortening of the glow-rays 

 the maximum and minimum of heating-effect move closer 

 together, until at last they can no longer be separated by 

 experimental means. 



In order further to prove the peculiar course of the heating- 

 effects obtained by the methods described and shown by means 

 of the curves, the following arrangement was employed. At 

 the upper end of the usual float carrying the electrode (PL II. 

 fig. 1) a thermometer was attached, having its scale below, as 

 shown in fig. 7. The mercury reservoir a is cylindrical, 

 5 millim. broad and 17 millim. long. It is protected against 

 the kathode-rays by a horizontal screen b of mica carried by 

 a pair of platinum wires c. But as these measurements are 

 concerned, not with the temperatures of the gas, but with the 

 quantities of heat produced in the unit time, the rise in 

 temperature in equal times must be observed when the ther- 

 mometer is placed at different points of the tube. We must, 

 of course, also apply corrections for radiation and conduction : 

 this was done by observations before and after breakino- the 

 current in the usual way. The measurements were carried 

 out with a tube of about 140 millim. width and 500 millim. 

 long. The dark space had a length of about 35 millim. 



The following table gives the numbers found in three series 

 (L, II., III.) of experiments. The different series are not 

 comparable with each other as to absolute value. 



E denotes the distance of the electrode ; W the observed rise 

 in temperature in arbitrary units (generally j^o o° P er minute 

 for a current producing 100 millim. deflection). 



