976 Prof. A. M. Tyndall on the Forces acting 



and the other to the open air ; and curves giving the 

 variation of hydrostatic pressure at, say, the anode with 

 length of arc should therefore dip by that amount as the 

 arc-length increases. 



But this is a pressure which would have been readily 

 observed if it had been present ; whereas the pressure- 

 distance curves given in curve III. of the previous paper 

 show no sign of it, and in fact slope, if anything, in the 

 opposite direction. 



This result is, moreover, consistent with other evidence 

 militating against the view that electrons of speed of the 

 order of 10 8 cm. per second would shoot across the arc 

 without exerting an appreciable drag. This velocity is 

 considerably below that at which an electron is usually 

 regarded as acquiring an abnormal free path. Also, Campbell 

 Swinton failed to detect in the arc electrons sufficient energy 

 to penetrate an aluminium sheet of thickness 0'0026 mm., 

 which, taking TVhiddington's constants for /3-rays, is equi- 

 valent to 0*094 mm. of air. The author is indebted to 

 Mr. E. Gr. Hill, M.Sc, for carrying out a somewhat allied 

 experiment on ions from a very fine point and from a white- 

 hot loop in point-plane discharge. Any projected electrons 

 must under these conditions be subjected at the point to the 

 action of a very high accelerating field, which would pre- 

 sumably increase their range. He failed, however, with 

 these sources to find ions of sufficient energy to penetrate 

 ..oil of this thickness, even when the potential of the point or 

 loop was 10,000 volts and the discharge-gap only a few milli- 

 metres. Nor was any effect obtained down to a pressure of 



5 mm. of mercury. 



Lastly, Pollock's argument that the forward E.M.F. of 



6 volts which, according to Duddell *, exists at the cathode 

 face must lead to the projection of electrons, is not above 

 question. 



Pollock neglected the opposing potential step which this 

 E.M.F. would set up ; yet, if the two were equal, the electrons 

 would leave the cathode with no excess momentum. It is 

 true that after leaving the cathode the electrons have to pass 

 through what is presumably the cathode dark space, where 

 they encounter a further step of 11'7 voltsf , in which, if they 

 have a free fall, they will acquire a speed of about 2 X 10 8 

 or 1*1 times that calculated by Pollock. The following- 

 argument shows that this might also give rise to a Duffield 

 effect. 



* Duddell, Phil. Trans, vol. 203. A. p. 332. 

 t Duddell. he. cit. 



