248 DISCHARGE OF ELECTRICITY. 



trolyte, which produces an equal eifect in stopping tlie discharge, 

 depends upou the length of spark intlie primary current, and so upon 

 the electro-motive force acting upon the air. The longer the spark the 

 greater is the molecular conductivity of the air in coinparison with that 

 of the electrolyte. This indicates that the conduction through the air 

 does not follow Ohm's law. This is what we should expect, as under 

 large electro-motive forces more molecules are split up and take part in 

 the conduction of the electricity. This great conductivity of rarefied 

 gases in those cases where the electricity has not to pass from metal, 

 etc., into the gas are in striking contrast with the infinitesimally small 

 values for the same property which are deduced from experiments on 

 tubes with electrodes. 



I was first led to suspect this high conductivity for rarefied gases by 

 observing the appearance presented by the ring-discharge in bulbs; 

 the ring, unless the pressure is exceedingly low, ceases at a distance 

 of little more than 1 centim from the surface of the bulb, this thick- 

 ness of conducting gas being sufficient to screen off the electromotixe 

 intensity from the interior. From experiments which I had made on 

 the screening effect' of electrolytes {Proc. Roy. ^Soe. xlv. p. 269), I knew 

 that it would require a very strong solution of an electrolyte to produce 

 screening comparable with this. To compare the screening effects 

 more directly than by the method just described 1 tried the following 

 experiment. The discluirge-tube. Fig. 11, was pumi)ed until the dis- 

 charge passed through it very freely; an exhausted tube was then 

 pushed down the central opening, it remained quite free from any visi- 

 ble discharge; the primary was now wound round a cylinder of tlie 

 same diameter as the discharge-tube of Fig. 11, and this cylinder was 

 tilled with distilled water. When the tube, Avhich had previously 

 remained dark wlien placed in the exhausted discharge-tube, was im- 

 mersed in the water, a brilliant discharge took place in it; and it was 

 necessary to add about 25 per cent of sulphuric acid to the water before 

 the shielding effect of the mixture was sufficient to keep the tube dark. 

 This experiment shows perhaps even more directly than the other the 

 great conductivity of a rarefied gas under large electro-motive forces 

 when nothing but the gas is in the way of the passage of the current. 



An experiment made in this connection illustrates the remark made 

 before as to the large effects produced by discharges through the gas 

 which are not accompanied by luminosity. A bulb A was fused into a 

 tube B which was surrounded by the primary coil C, I). B was ex- 

 hausted and then sealed off', while A was left connected to the pump. 

 When A was at atmospheric pressure a bright discharge took place in 

 B outside A; on pumping A a stage was reached in which no dischargi^ 

 could be seen in either A or B. On letting air into A the discharge 

 appeared again in B; on i)umping A still further a discharge appeared 

 in A, but not in B. The appearance presented by the discharge round 

 the bulb A (filled in this case with air at high pressure) is very remark- 



! 



