MR. J. S. TOWNSEND ON THE DIFFUSION OF HNs INK) GASES. 



1 11 



\\indow, W, through wliidi the rays from the Crookes tube, B, could jmsa. A second 

 I trass tube, C, 17 centims. long, fitted tightly into A, and could l>e move<l into any 

 desired position. The rod, F, to which the electrode, E, was fixed, passed through 

 the ebonite plug, D, which insulated it from the tube C. The electrode, having no 

 other support except D, could thus be put into any position in the tube A by moving 

 C. A series of very fine wires ('1 millim. in diameter) were soldered parallel to one 

 another, at distances 2 millims. apart, across the end of the tube C; the purpose 

 served by this grating will be explained when we come to deal with recombination. 



The gas entered the apparatus through the glass tube G, and, before reaching the 

 electrode, passed through the tubes T,. These tubes were soldered into holes bored 

 in two brass discs, a and /8, which fitted exactly into the tube A, so that no gas could 

 pass between the discs and the tube. The holes in the discs were equidistant from 

 one another, and lay on a circle whose centre was the centre of the disc. Twelve 

 tubes, 10 centims. long and '3 centim. in diameter, were thus arranged parallel to 

 one another, two of which are shown in the figure. The symmetry of this arrange- 

 ment ensured that the velocity along each of the small tubes would be the same. 

 Another twelve tubes, 1 centim. long and '3 centim. in diameter, were soldered into 

 the disc y. 



Fig. 1. 



The bulb B, and the RuhmkorfF coil with which it was worked, were contained 

 insidf a IM.\- covered with lead, L. A rectangular hole was cut in the box and the 

 lead, through which the rays from the bulb coxild pass. The lead covering prevented 

 the rays from falling on any other part of the apjiaratus except the aluminium 

 window W, and also screened the wire connecting F to the electrometer from electro- 

 static influence. 



The tube A was supported by two ebonite rings, 11 and R', which rested on the 

 lead, L, and insulated the tube. The potential of the tube was raised to 80 volts by 



