134 DR. W. G. DUFFIELD, MESSRS. THOS. H. BURNHAM AND A. H. DAVIS ON 
are liberated, and their expulsion involves the recoil which has been measured in the 
present experimental investigation. 
Under these conditions the mechanical effect would he least likely to be disturbed 
by electric forces within the arc, because the oxygen atom approaching with two 
positive charges would contribute to the attractive force upon the pole an amount 
not very different from the repulsive force occasioned by the two receding electronic 
charges. 
In a normal arc the effects at the anode are very complicated, there is electronic 
projection due to thermionic and photoelectric action, and probably access of electrons 
and negatively charged atoms which carry the current to it. Nevertheless, the 
values of ejm obtained by the method already described is of the right order of 
magnitude, though three times higher than it should be, if the recoil is in this case 
also to be explained by the projection and impact of electrons and if they bear half 
the current. If we could accept the view that the momentum of the electron derived 
from the cathode is handed on through the vapour from atom to atom until it 
reaches the anode, the discrepancy would be reduced. Elsewhere we have shown 
that it is possible to reduce the carbon consumed by the anode to almost negligible 
quantities, it would be interesting to determine the changes in the anode recoil under 
these circumstances, but the experiments would be of very gi’eat difficulty. 
The writer tenders the above account of the mechanism of the arc with due 
appreciation of the assumptions underlying it. As far as the details are concerned, 
a great deal depends upon the accuracy of Duddell’s results, but any reasonable 
assumption regarding the magnitude of the potential drop across the cathode pole 
face would lead to a value for ejm which is of the order of magnitude of that of the 
electron and far removed from that associated with atoms. If instead of assuming 
random projection, we assumed normal projecticyi from a small area on the cathode, 
the values of ejm would be four times those given in Table X., and still in accord 
with their electronic rather than their atomic nature. 
The view I have taken of the mechanism of the arc attributes the fall of 
potential across the negative pole face to electronic projection there, contrary to the 
theory which regards the electric force as responsible for the extraction of the 
electron. Pollock,"^ assuming electronic projection, took the same view, and from 
Duddell’s work calculated the velocities in different parts of the arc in an important 
contribution to this subject. The discharge of electrons has frequently been assumed, 
but I do not think that there has hitherto been any mechanical evidence in its 
favour. 
Such action, photoelectric or thermionic, as occasions in the arc a discharge of 
negative electrons from the poles is probably assisted by the chemical interactions 
between the poles and the surrounding gas. This point has already been discussed 
in the paper by the writer, to which reference has been made. 
* Pollock, ‘Phil. Mag.,’ vol. XVII., p. 361, 1909. 
