Electric Discharge in Gases. 43 



wider tubes were chosen and the rod came into contact with 

 the sides of the tube, any regular motion was impossible. A 

 star-shaped piece of platinum foil was therefore fastened to 

 the lower part of the rod with a little wax ; and as a further 

 guide the upper end of d was covered with a plate having in 

 it a hole just large enough to allow the rod to pass through. 

 In order to introduce the electrode /3 into the space A, either 

 the tube cl was made larger than the plate, or else the tube 

 A was cut in two ; and after the plate had been introduced into 

 A and the rod carrying it passed through 0, the tube was 

 melted together again; the terminal wires of the machine 

 plunged in « and e, or in a and S, as the case might be. 



In order to follow the rhythm of the discharges in the dis- 

 charge-tube, which it is difficult to do in the luminous phe- 

 nomena of the tube itself, on account of its great width, and 

 especially when the discharges are irregular, or are confined 

 to special parts of the tube, it is advisable, besides the wide 

 tube, to include also a second tube in the circuit, which it is best 

 to take of the form shown in fig. 2. This is made out of a wide 

 tube by contracting the middle portion before the blowpipe ; 

 the tube is exhausted so far that it offers as small a resistance 

 as possible to the current traversing it : a and b are flat 

 electrodes which offer a smaller resistance to the currents than 

 points do ; they are placed as shown in the figure, in order 

 to afford a rapid indication of the direction of the current. 

 The middle portion of the tube is observed in the revolving 

 mirror; its relatively great luminosity and small section render 

 it easy to judge in the rotating mirror whether the discharges 

 follow each other quickly or slowly. 



In the following experiments it is always assumed that the 

 exhaustion has been carried very far, so that the kathode is 

 surrounded by a dark space : — 



If we allow a positive electrode a (/3 of the figure) to 

 approach a negative plate k (« of the figure), the positive 

 electrode consisting of a wire which is cut off together with 

 the glass surrounding it, and which dips into the mercury, 

 the following phenomena are observed : — 



If a is at a considerable distance from k we have a head of 

 positive light about a, and stratifications between a and the 

 glow-rays, of which even the first quite fills the section of the 

 tube and is much curved. If a rises, the layers do not 

 change their position, but one after the other disappear as 

 soon as they touch the positive electrode ; this is also the case 

 with the last layer nearest the kathode. 



As soon as the end of the wire has passed through this, 

 there is only a small bundle of positive light to be seen upon 



