produced from a Gas by High-Speed Cathode Kays. 205 



covered with willemite. The direction of the primary cathode 

 rays could be thus observed and adjustments made with a 

 small magnet. A horseshoe magnet NS is placed around T 

 to prevent the entering rays, and any rays excited inside, 

 from escaping. It was experienced in preliminary investi- 

 gations, that when the ebonite pieces were struck by the 

 rays, some charge accumulated on the ebonite pieces and dis- 

 turbed the electrometer. To get rid of this effect, short 

 protecting tubes B, B / are fitted at each end of the gauzes, 

 and any unnecessary surfaces of the ebonite pieces are 

 covered with earthed foils. The gauzes and the electrode E- 

 are covered with candle-soot (§7). 



To the ionization-chamber are connected two U-tubes, Ui 

 and U 2 . A Gaede pump, McLeod gauge, spectrum-tube, and 

 P 2 5 drying-tube are connected to Ui, and a gas reservoir 

 and drying- tubes to U 2 . The U- tubes are cooled with liquid 

 air or a mixture of solid carbonic acid and alcohol. This 

 cooling condenses any mercury vapour coming from the 

 pump and also any wax vapour from the joints. The whole 

 apparatus was arranged to get rid of wax-joints as much as 

 possible. 



§ 3. Experimental Procedure. 



The separate parts of the apparatus before being put 

 together were washed with water and alcohol, boiled with 

 water for about one hour, again washed with distilled water, 

 and finally dried. Gluss parts before being thus treated 

 were washed with nitric acid. The electrometer-electrode E, 

 the gauzes, inside surfaces of the ionization-chamber and. 

 Faraday cylinder were covered with candle-soot. 



The whole apparatus was exhausted by the Gaede pump 

 and charcoal tube cooled with liquid air. To drive out any 

 gas adhering to the electrodes of the discharge-bulb, the bulb 

 was kept discharging while the apparatus was exhausted. 

 A small quantity of air or hydrogen gas, previously dried 

 by means of P 2 5 and CaCl 2 , was then introduced from a 

 reservoir. The apparatus was exhausted to a desired pres- 

 sure. The pressure was as low as possible, since, otherwise^, 

 the secondary cathode rays and the positive ions produced 

 at the same time with the secondary corpuscles would ionize 

 the residual gas by their passage to the electrometer-elec- 

 trode. The same reason necessitated reducing the successive 

 distances between the gauzes and the electrode to a value 

 much smaller than the mean free paths of the corpuscles 

 and the positive ions. The value, as mentioned above, was 

 3 mm. However, owing to the hardening of the discharge- 

 bulb, the pressure had to be kept above a certain value. 



