PHYSICS: E. H. HALL 
163 
characters enumerated in this paper can be employed to good advan- 
tage in the identification of species belonging to these genera. 
iSaccardo, P. A., Phytophthora cadorum (C. et L.), Schroet. Syll. fung., 7, 1888, (238). 
2 The complete paper is published in /. Agric. Res., 8, 1917, No. 7. 
A POSSIBLE FUNCTION OF THE IONS IN THE ELECTRIC CON- 
DUCTIVITY OF METALS 
By Edwin H. Hall 
JEFFERSON PHYSICAL LABORATORY. HARVARD UNIVERSITY 
Read before the Academy, November 14, 1916 
The following paper is a development, in certain particulars, of a 
suggestion concerning electric conduction in metals which I made in 
// Nuovo Cimento for January-February, 1915. It rejects the free 
electrons within the metal as the vehicle of the electric current and puts 
in their place for this function the metal ions, necessarily equally numer- 
ous with the free electrons. Briefly stated, the idea of conduction 
which I have had in mind is the following: 
At very low temperatures the atoms of a metal are packed so closely 
that electrons pass readily from one to another in a continuous proces- 
sion through the metal, if there is an applied electromotive force to 
maintain progression in one direction. Free electrons and metal ions, 
if indeed ionization exists at the lowest temperatures, need not be called 
into action here. 
With rising temperature the heat vibrations separate the atoms so 
that they are not always in conductive contact, and not very many 
degrees above the absolute zero they are, on the average, so far apart 
that conductive contact between them is exceptional. 
In this state of things an electron will not in fact pass from one metal 
particle to another, even when they are in the closest contact of a col- 
lision, unless one of these particles is an atom and the other an ion. 
Figures 1 and 2 illustrate the fundamental difference, from the point 
of view of conduction, between an atom-atom collision and an atom-ion 
collision. In figure 1 the electric force tending to carry the extra elec- 
tron of the left-hand particle back to the positively charged right-hand 
particle, just before the particles separate, must be enormously larger^ 
than the greatest conteracting force we can apply from without. But 
in figure 2 the extra electron may belong as much to one particle as to 
the other, when once the particles meet, and a comparatively small 
general gradient of potential in the metal, due to an electromotive force 
