Observations at the' Poles of the Electric Arc. 961 



taken to eliminate disturbances due to heating effects in the 

 tubes leading to the gauge. These had to be kept strictly 

 horizontal, close together, and screened with the gauge from 

 the arc by asbestos sheets. 



For the observation and measurement or! some of the 

 residual pressures at low currents, the gauge referred to 

 above was not sensitive enough, and a Fry diaphragm 

 gauge (Phil. Mag. vol. xxv. 6th ser. p. 494) was substituted 

 for it. This was used in two grades of sensitivity : in one 

 case the sensitivity was *0076 dyne per mm. scale deflexion, 

 and in the other '0205 dyne per mm. It may be noted 

 that in no case, except with the new undenuded carbons, 

 were pressures recorded so high as the value 1 to 2 mm. 

 of water which Dewar quotes for an anode in his paper. 

 Nevertheless, for the size of the hole used the signs of the 

 effects are in agreement with those of Dewar — namely, a 

 positive residual pressure at the anode and a pressure which 

 was usually negative at the cathode. 



Before considering theoretical explanations of these effects, 

 two tests were applied :— 



(a) A reading was taken with a denuded carbon. The 

 current was then greatly increased for a few minutes, after 

 which it was reduced again to its previous value and the 

 reading immediately repeated. The absence of any change 

 was taken to indicate that the reading was uninfluenced by 

 the evolution of gas. There was no evidence of such evolu- 

 tion in any of the results given below, though naturally 

 those taken with carbons of wide bore are less susceptible 

 to this effect. But it should be mentioned that, particularly 

 in the case of the anode, a fresh surface at the end of the hole 

 is continually being exposed by the eating aw r ay of the 

 electrode, and if this effect extends up the hole for even a 

 short distance the evolution of gas may never be negligible. 

 Whatever may be said, therefore, as to the true magnitude 

 of the residual positive pressures — and this is discussed 

 further below — there is every reason to suppose that the 

 negative pressure often observed at the cathode is a true 

 effect uninfluenced by evolution of occluded gases (see also 

 below). It may be noted that the above remarks do not 

 apply to the evolution of carbon vapour in the hole ; this can 

 produce no permanent pressure on the gauge. 



(b) The possible influence of convection currents in the 

 arc was studied. It was found that for a given current and 

 length of arc the pressure reading was not affected by placing 

 a plate of fused silica just over or just under the arc, though 



