614 Miss P. M. Borthwick on Potential Measurements in 



Hence the potential of E would be higher in the latter case 

 than in the former, whether P is positive or negative. The 

 fact that the increase is greater for P negative is readily 

 explained. It was found that the current, supplied by N, 

 necessary to produce discharge from P, is greater for P — 

 than for P + . This, together with the fact that the external 

 ions in the former case are the slower moving positives, implies 

 that the increase in the potential drop close to P, due to these 

 ions, must be greater when P is negative. 



Now the pull readings appeared to indicate that, in the 

 case of P -f- , the field at the surface of the point was less 

 when external ions were supplied than with no external ions. 

 The potential drop in the distance of 2 mm. from P does not 

 show this effect. The superposed potential drop dne to the 

 supplied positive ions may, of course, be more than sufficient to 

 swamp the effect; but in any case it seems clear that without 

 further information as regards the variation of potential 

 throughout the 2 mm., we cannot, from the present values 

 oi this potential drop, obtain reliable evidence of the value 

 of the field at the surface of P. The sparking effect mentioned 

 above prevents the obtaining of this information. 



The most that can be said, therefore, is that the measure- 

 ments of potential drop near the point, when external ions 

 are supplied, are in agreement with the results of the pull 

 measurements of the field at the surface of the point, in those 

 cases in which the information obtained is complete. 



Potential readings near the plate. — With the same appa- 

 ratus the slope of potential along the central line between 

 the electrodes, in the case of a point discharging on to a 

 plate, was investigated. For this P was connected to the 

 wimshurst and the plate was earthed through the galvano- 

 meter. 



The potential slope was traced by means of E for dif- 

 ferent values of distance between the electrodes (a?) and of 

 current (Cj. 



Curves III. show some of the results for different values 

 of x (C = 3'l microamps.), the full curves representing 

 positive discharge from P, and the broken ones negative. 

 The intercept made by each curve on the horizontal axis 

 gives the corresponding value of x (the plate being at zero 

 potential). It appears that the potential slope is more or 

 less uniform near the plate, except within about 1 mm. of 

 the latter, when there is invariably a marked increase. Now 

 Warburg {Ann. der PJiys. iv. 2, p. 299) obtained some results 

 bv the same method, for the same values of #, and found a 



