Vol.. 7, 1921 
PHYSICS: E. H. HALL 
107 
total heat conduction of metals; but I trust that enough has been done to 
show that the possibilities of this mechanism are worthy of serious con- 
sideration. 
I propose next to consider how far the ideas set forth in this paper will 
go toward explaining the various "transverse" electro-magnetic and 
hermomagnetic effects in metals. 
1 These Proceedings for April, 1918, pp. 98-103. I there used in turn "hypothesis 
A" and "hypothesis _B" concerning the behavior of the free electrons as a gas. In 
nearly all that follows I have used "hypothesis A" which is that, aside from electric 
influences, the free electrons tend to equality of gas pressure throughout the bar, where- 
as "hypothesis B" assumes that they tend to follow the law of "thermal effusion/' 
2 These Proceedings for March, 1920, pp. 139-154. 
3 Strictly, X'o was used here instead of X' c , with the same meaning. 
4 These Proceedings, 6, 1920 (617). 
6 It seems probable that in dealing with the conditions at the junction of two metals 
we really need to use the conception of number-density of free electrons in the "co- 
volume" of each metal. In fact, n may without harm be interpreted as having this 
meaning in all my formulas except the one, taken from Drude, in which I express the 
free-electron conductivity in terms of mean free path, mean velocity, etc., of the free 
electrons. See the latter part of this paper, where y, the grade of ionization, is 
estimated. 
6 Estimated by use of the Wiedermann-Franz ratio (average for 12 metals) 
applied to k 0 and km- 
7 Mean value, according to Lorenz (1881) between 0° C. and 100° C. 
8 See note under Cobalt. 
9 0i8 is given as 0.35 on the authority of Coolidge. 
10 These Proceedings, 6, 1920 (617). 
11 Physical Review, May, 1919. 
