Observations at the Poles of the Electric Arc. 969 



at both carbons at the same time made quantitative study 

 quite impossible. The readings, such as they were, were 

 consistent with the results obtained separately on each pole, 

 and, taken in conjunction with the experiments on con- 

 vection referred to above, support the view that this plays 

 no fundamental part in the phenomena. 



The fact that there is no excess of ions of one sign in the 

 body of the arc was confirmed during the progress of 

 the above investigations by some experiments which were 

 carried out to detect the lateral loss of momentum, which 

 will, in general, take place when an unbalanced ionic stream 

 crosses the arc. With short arcs and flat poles of relatively 

 large area this can, in any case, certainly be regarded as 

 very small in amount. But its presence might be detected 

 if it could be accentuated by increasing the arc-length and 

 decreasing the pole diameter. Accordingly, the arc was 

 struck inside a quartz tube of 1| inches diameter, the 

 carbons lying along the axis of the tube. The anode was 

 solid, and was filed down to a diameter of 0*7 cm, and was 

 made pointed. The anode end of the quartz tube was closed 

 except for an opening to the pressure-gauge ; the cathode 

 end was open to the air. 



Under these circumstances with the longer arcs that are 

 possible in such a narrow tube, if there were an electric 

 wind in the arc from cathode to anode, a large fraction of 

 its momentum would be transmitted laterally to the air in 

 the quartz tube. This should lead to an excess pressure 

 around the carbon at the anode end of the tube. Assuming 

 all the momentum to be so transmitted, this pressure would 

 be the Dewar pressure, reduced, however, in the ratio of the 

 current-receiving area of the anode to the area of the quartz 

 tube. This ratio was of the order of 0*03 for 20 amperes. 



Definite results were difficult to obtain, partly owing to 

 hissing and partly to the great generation of heat. In some 

 cases with a current of about 20 amperes a deflexion of from 

 2 to 4 mm. of scale was obtained with the diaphragm gauge 

 in its most sensitive form — i. e., '0076 dyne per mm. 

 Between large electrodes giving no sideway loss this is 

 equivalent to a Dewar pressure of not more than l'l dynes 

 cm. 2 for 20 amperes. But as it was independent of the 

 direction of the current, it is reasonable to attribute it not to 

 wind, but to a slight change in density of the air in the 

 tube leading to the gauge, owing to its proximity to the arc 

 and the hot quartz tube. 



This therefore indirectly confirms the view that the 

 Dewar pressure does not originate from wind-producing ions 

 moving in the main body of the arc. 



Phil Mag. S. 6. Vol. 42. No. 252. Dec. 1921. 3 S 



