Electron Theory of the Carbon Arc. 365 



die magnitude of the cathode fall of potential reaches a limit 

 when ionization by collision, at the end of the average range 

 of projection from the cathode, is fully established. In this 

 case the value found by Mr. Duddell may be taken as a 

 constant, characteristic of the carbon arc. 



6. The Vapour Column. — From the ease with which the 

 arc is displaced, by even small magnetic forces, one is led to 

 consider that a large proportion of the current through the 

 vapour column is due to electrons moving with high velocity. 

 As the electric force in the column, about 4 volts per milli- 

 metre in the case under consideration, is utterly inadequate to 

 guide unaltered a swarm of electrons through molecules and 

 positive ions, the mode of conduction in the vapour column 

 would seem to be different from that usually considered in 

 the case of gases. 



Possibly, with a critical value of the cathode fall, the 

 collision of the projected corpuscles with molecules of the 

 highly heated gas, may start a progressive exchange of 

 -electrons among the oriented atoms, throughout the vapour 

 column, somewhat like that suggested by Professor J. J. 

 Thomson * for the case of conduction through metals. 



In such a case it may be considered that the electrons 

 would be handed on, as it were, from atom to atom with 

 unimpaired energy, and would thus reach the anode surface 

 with the critical velocity of the corpuscles starting the 

 exchange, together with that acquired during the flight 

 through the anode fall of potential, or with a speed of 

 3*6 x 10 s centimetres per second. 



In view of Professor Townsend's f discovery of the expulsion 

 of two electrons from an atom in ionization by collision, the 

 above suggestion, as to the mode of conduction through the 

 vapour column, would lead to the consideration of an ever 

 increasing volume of ionization towards the anode, thus 

 accounting for the shape of the arc, and an area of crater 

 proportional to the arc length. Without modification the 

 idea would, however, be inconsistent with the fact of a zero 

 density of electrification throughout the vapour column, as 

 the positive ions made near the anode in moving towards 

 the negative electrode would not meet an equally dense 

 negative stream. 



7. Summary. — The application, to the case of the arc, of 

 •the idea of the projection of electrons from the surface of 



* Thomson, "The Corpuscular Theory of Matter" (1907). 

 t Towusend, Proc. R. S., A. 80, March 1908. 



Phil. Mag. S. 6. Vol. 17. No. 99. March 1909. 2 C 



