454- Prof. J. J. Thomson on the Discharge of Electricity 



defined at the edges ; touching the tube, though this is 

 already connected to earth, produces a very marked effect in 

 increasing the facility of the discharge. We can, I think, 

 understand the reason of this if we consider the behaviour of 

 the tubes of electrostatic induction. When the spark passes, 

 these tubes (see p. 327) rush out from the jars and make for 

 the primary ; in their journey to the prima^ they pass through 

 the bulb and produce the discharge. Let us suppose now 

 that there is a large conductor situated somewhere near the 

 bulb ; the tubes, as before, rush from the jar to the primary, 

 but in doing so some of them strike against the conductor; 

 the tubes which do so lose the portion inside the conductor, 

 acquire two ends each on the surface of the conductor, and 

 swing round until they are at right angles to its surface ; 

 they remain momentarily anchored, as it were, to the con- 

 ductor, and if the conductor is in the neighbourhood of the 

 bulb, they will in general help to increase the maximum 

 density of the tubes passing through the bulb. Though these 

 tubes may not approximate to closed curves, and so directly 

 produce a ring-discharge, they may readily facilitate this dis- 

 charge indirectly ; for even those tubes which go radially 

 through the bulb may produce a glow-discharge from the 

 glass into the bulb, and may thus furnish a supply of dis- 

 sociated molecules through which the ordinary ring-discharge 

 can pass with much greater readiness. For nothing is more 

 striking than the enormous difference produced in the electric 

 strength of these rarefied gases by the passage of a spark. 

 It is sometimes difficult to get the discharge to pass at first, 

 but when once a spark has passed through the gas, a spark- 

 length one quarter the length of that necessary to originate 

 the discharge will be found sufficient to maintain it. 



It is sometimes convenient, in cases where difficulty is found 

 in starting the discharge, to avail ourselves of this property 

 by connecting the mercury of the pump to which the tube is 

 attached with one terminal of an induction-coil, the other 

 terminal of which is put to earth. When the induction-coil 

 is in action, a glow-discharge fills the pump and tube, and 

 while this glow exists the electrodcless discharge can easily 

 be started ; once having been started, it will continue after 

 the induction-coil is stopped. An experiment of this kind, 

 which I had occasion to make, gave very clear evidence of 

 the way in which dissociated molecules are projected in all 

 directions from the negative electrode in an ordinary discharge- 

 tube, but not from the positive. The discharge-tube was 

 fused on to the pump, and at an elbow two terminals, c and 

 d, fig. 14, were fused into the glass ; these terminals were 



