Vol. 8, 1922 
PHYSICS: E. H. HALL 
307 
of magnitude of the constants, yet they are seen to correspond well with 
the values calculated by the thermodynamic equation. 
The results here presented therefore afford a striking confirmation of the 
general validity of the calculation of the ionization of elements from their 
ionization-potentials by the thermodynamic equation employed by Saha. 
1 Eggert, Physik. Zs., 20, 570, 1919. 
2 Saha, Phil. Mag., 40, 478, 809, 1909; 41, 267, 1921. 
3 Russell, Astrophys. J., 55, 129, 1922. 
4 Sackur, Ann. Physik., 36, 598, 1911; 40, 67, 1913; Tetrode, Ibid., 38, 434; 39, 255, 
1912; Stern, Zs. Electrochem., 25, 66-80, 1919; Tolman, /. Amer. Chem. Soc, 42, 1185, 
1920; 43, 1593, 1921; Lewis, Physic Rev., 18, 121, 1921. 
5 Arrhenius, Wied. Ann. Phys., 43, 18, 1891. 
6 H. A. Wilson, Phil. Trans. Royal Soc, A216, 63-90, 1915. 
' H. A. Wilson, Phil. Mag., (6) 10, 475, 1905. 
8 Wilson and Gold, Phil. Mag., (6) 11, 484, 1906. 
AN ELECTRON THEORY OF ELECTRIC CONDUCTION IN METALS 
By Edwin H. Hall 
Jeffi^rson PHYSiCAiy Laboratory, Harvard UiNrivERSiTY 
Communicated, September 5, 1922 
This is a modification and extension of the dual conception of conduction 
which I set forth in a paper^ printed several years ago and to which I have 
referred in more recent publications. The theory in its present shape, 
like every other theory of metallic conduction with which I am acquainted, 
involves propositions that may well be questioned, but it contains certain 
features not present, so far as I know, in other theories, and I hope these 
will be found worthy of consideration, even if they are finally to be rejected. 
I shall endeavor to state this theory, without extended argument, in 
the form of certain propositions, some of which are or involve outright 
assumptions. Any helpful hypothesis, not in conflict with well estab- 
lished facts or principles, I have considered justifiable, in view of the 
difficulties of the problem with which I am dealing. 
Behavior of the Ions in Conduction 
1. Complete metal atoms are of such hardness and symmetry that 
they are not subject to appreciable polarization or orientation in impressed 
electric fields of ordinary strength. 
2. An ion is formed from an atom by loss of an electron from the outer 
shell and is, accordingly, unsymmetrical, essentially polarized, and there- 
for subject to orientation in a uniform electric field such as may be ap- 
