Passage of Electricity through Gases. 181 



tension necessary for the discharge), the reaction of this charge 

 on the charge of the electrode itself will he relatively small, and 

 the halo round the electrode will occupy a larger space. In 

 consequence of this, the discharge can spread itself in the 

 direction in which the potential diminishes most rapidly, i. e. 

 in the direction of the electrode I. If, however, the electrode 

 II. is positive, the charge on the glass opposite to the free 

 surface will he relatively strong, and the negative discharge 

 will extend from the electrode I. to the neighbourhood of the 

 electrode II. ; so that the discharge proceeding from the 

 electrode II., in consequence of the repulsion by the first and 

 the attraction by the second, will be bent towards the elec- 

 trode I. At the same time the halo proceeding from the 

 electrode I. will in the latter case spread itself beyond the 

 positive electrode, because the positive discharge, confined in 

 a limited space and emitted with great velocity from a small 

 spot in the direction of the resultant force, cannot at once 

 neutralize the negative electricity with which every point of 

 the halo is charged. 



A similar explanation applies to the beautiful experiment 

 of Hittorf, w T here two straight wire terminals are inserted at 

 right angles to one another in a sphere of glass. Since when 

 the rarefaction is sufficient, and the influence exerted by the 

 positive electrode not too strong, the negative discharge 

 spreads from a great part of the surface of the negative elec- 

 trode, and, according to our theory, a little earlier and before 

 the positive discharge at the point of the positive electrode 

 begins, therefore the latter, in order to neutralize its electricity 

 with that of the negative electrode, cannot turn directly towards 

 the negative electrode, but also must be diverted round the 

 latter in the direction of the negatively electrified glowing gas. 

 The same is the case with electrodes parallel to one another; 

 for then the discharge of the positive electrode passes to the 

 spot at which the attraction of the positive and negative 

 electricity in the tube is at its maximum. 



If an induction-coil be used, the phenomena are more com- 

 plicated than with an electric machine. The quantities of elec- 

 tricity driven by induction to the ends of the coil accumulate 

 not only there, but also more or less in the interior. If, 

 therefore, those ends are connected with the electrodes of a 

 tube, and if their electricity has attained to the tension neces- 

 sary for a discharge, the gas stratum at the electrodes will be 

 thrown off and rarefied by the initial discharge, since the 

 movement of the gas from the sides is slow in comparison with 

 that of the electricity. Thus the electricity flowing to the 

 electrode from the interior of the coil may be partly transmitted 



