360 Prof. E. Wiedemann on t lie Behaviour of Gases 



The galvanometer was a Wiedemann's reflecting galvano- 

 meter of moderate strength, with coils of gutta-percha covered 

 copper wire. The distance between the telescope and scales 

 was 189 centims, 



•2. The discharge-tube includes the space containing the 

 two electrodes and the part of the tube uniting them. This 

 either was altogether absent so that the electrodes were imme- 

 diately opposed to each other, or consisted of tubes of various 

 shapes. But care was taken to avoid any shape in which two 

 portions of the tube would be parallel to each other, as, for ex- 

 ample, in a U-tube. In this case induction-phenomena of a 

 very disturbing character occur. 



The electrodes were made of aluminium in all definitive 

 experiments, since this metal alone is not subject to dissipation 

 by the discharge. They were either balls of 5 millims. diameter, 

 or points 3 millims. thick at the base and 12 millims. long. 



In conducting the electricity to the electrodes, in order to 

 avoid as much as possible the scattering by points, and to ensure 

 good contact, the following arrangement was adopted (Plate IX. 

 fig. 1) : — a is a glass-bulb, b the electrode, c a platinum wire 

 screwed into it and covered with glass, d a glass tube joined 

 to a by fusion, into which the wire c projects. The tube d can 

 be filled with mercury, into which the wire conveying the 

 electricity plunges. The discharge-tube is melted on at m. 

 A similar arrangement is to be recommended for Geissler's 

 tubes, which are much used, as they are easily broken by the 

 pulling of the connecting wires at the projecting platinum 

 wires. 



If the heating was to be determined in the whole space 

 between the two electrodes, the whole apparatus was plunged 

 in a trough-shaped glass calorimeter of about 50 cubic centims. 

 capacity. The arrangement shown in Plate IX. fig. 2 was 

 employed to determine the heating at one electrode, e is the 

 electrode with the portion of the tube surrounding it, t the 

 thermometer, c the calorimeter with a tube-shaped portion I 

 in which a glass tube s melted on to the discharge-tube was 

 cemented. The part of the tube s next the electrode was 

 constructed out of a capillary tube, in order as much as 

 possible to prevent convection of heat by currents. 



In order to determine the heating in the tube, the arrange- 

 ments figs. 3 & 4 were employed. In the first, as in the ex- 

 periment with the electrode, a portion e of the discharge-tube 

 is carried vertically through the tube-shaped calorimeter a, 

 and is bent horizontally above and below the calorimeter ; in 

 the second the discharge-tube e is horizontal, and the calori- 

 meter a is pushed over it. The last arrangement must be 



