means of obtaining Continuous Electrical Oscillations. 259 



image is expanded into a fan-shaped area of light. Ex- 

 amining this, we see it crossed vertically by narrow or broad 

 lines irregularly placed. This indicates that the tube is ex- 

 tinguished at intervals, and this is because the electric field 

 disappears and the oscillations are interrupted. Not only 

 Neon tubes, but all other vacuum-tubes show this effect pro- 

 vided they are narrow tubes. If a wide tube is used, the 

 effect is not seen because the intermittent images of the tube 

 overlap each other. The phenomenon is well shown by 

 rapidly rotating a Neon tube near the spiral. The tube must 

 be fastened by silk threads to a strip of ebonite held on an 

 ebonite axle, and this rapidly rotated by a turn-table (see 



fi g- 2 )-. 



The inter mittency is always most marked when the arc is 

 first started and then diminishes, but it is always present 

 more or less, and it shows that the oscillations, even if un- 

 damped, are not absolutely continuous, but are cut up into 

 irregular groups. It seems to be caused by the arc changing 

 its point of departure on the carbon electrode. It is increased 

 if the rotation of the carbon is irregular, but I have never 

 been able to abolish it altogether. Another proof of the same 

 fact is derived as follows : — 



The helix is removed and another square coil of 12 turns 

 or so of insulated wire wound on a frame 60 cms. inside is 

 placed parallel and at a distance of 2 or 3 metres from the 

 similar coil in the condenser circuit (see fig. o). This 



Fig. 3. 



L 

 -Wffljp — i 



secondary coil is tuned by a condenser to the same frequency 

 as the arc condenser circuit, Across the terminals of the 

 secondary circuit another circuit is joined which consists of 

 the oscillation valve or glow-lamp wireless telegraph detector 

 devised by me four years ago, and already described*. This 



* See J. A. Fleming, " On the Conversion of Electric Oscillations into 

 'Continuous Currents by means of a Vacuum Valve," Proc. Rov. Soc. 

 Lond. vol. lxxiv. p. 476 (1905), Also Pnil. Mag. May 1906. 



