692 Professor J. A. Fleming [May 24, 



to one terminal of the condenser. This helix is tuned to the 

 condenser circuit and may be taken to represent the antenna when 

 the apparatus is used in wireless telegraphy. If we start the arc, 

 then high-frequency oscillations are produced in the helix, and by 

 the action of resonance the potential at the free ends becomes large 

 enough to create an electric brush discharge. There is, of course, a 

 strong oscillatory electric field outside the helix, and vacuum tubes 

 held there, particularly neon tubes, glow brilliantly. It has been 

 contended that these oscillations are undamped and continuous, but I 

 can show you a simple experiment with a neon tube which proves 

 that they are not always uninterrupted. If I hold a neon tube near 

 the helix, and move it rapidly to and fro, you see a broad band of 

 light, due to persistence of vision, but this is cut up by dark lines 

 and spaces. In the same manner if a neon tube is rotated near the 

 helix it does not produce a uniform disc of light, but the disc 

 presents the appearance of radial dark bands and bright spaces. The 

 same effect is seen with a vacuum tube filled with any other gas, 

 provided the tube is sufficiently narrow in the bore. It appears to 

 me that this proves incontestably that the oscillations are not 

 uninterrapted, but are cut up irregularly into groups of various 

 lengths.* 



To obtain these high-frequency oscillations the various contri- 

 butory factors — strength of magnetic field, length of arc, supply of 

 coal gas — have to be carefully adjusted with reference to the capacity 

 and inductance used and the voltage on the arc. No one who has 

 practically worked with the apparatus can say that it is a simple and 

 easy one to use . A very little want of exact adjustment causes the arc to 

 be extinguished or else it fluctuates greatly in current, and compared 

 with the extremely simple appliances required for spark telegraphy, 

 the advantage in ease of working is largely on the side of the spark. 

 But we have to consider whether there are not counter-balancing 

 advantages as a generator of telegraphic electric waves which make 

 up for the increased difficulty of working and greater complexity of 

 apparatus. The claim made for it is that if the transmitter 

 produces undamped continuous oscillations these can be reduced 

 to such small amptitude that they will not affect other neighbouring 

 wireless non-syntonic receivers even if only a little out of tune, but 

 can by the cumulative effects of resonance actuate their own corre- 



* Previous experimentalists seem to have been satisfied with examining in 

 a revolving mirror the flaming arc or brush produced at the secondary 

 terminals of a transformer, the primary of which forms the inductance in the 

 condenser circuit, and finding the image drawn out into a band of light con- 

 cluded that the oscillations were continuous. The neon tube is a more 

 delicate test, and reveals the discontinuity mentioned above. This discon- 

 tinuity of the train of oscillations seems to depend to some degree upon a 

 want of perfect regularity in the rotation of the carbon terminal. It may also 

 be brought about by the energy transferred to the condenser circuit being 

 radiated or dissipated faster than it is supplied. 



