Vol. 8, 1922 
CHEMISTRY: NO YES AND WILSON 
305 
Now Saha introduces the following assumptions as to the quantities oc- 
curring in this equation : 
(1) That the energy increase attending the ionization of a single mole- 
cule is equal to the electrical work Ve that must be expended in order to 
give to an electron a kinetic energy which just suflfices to ionize an atom 
of the element with which it collides; e being the charge on the electron 
and V the ionization-potential, which last may be determined by direct 
measurement or computed from the spectral series of the element. 
(2) That the heat-capacities of the neutral atom M and its ion M+ are 
equal, and that the heat-capacity of electron gas is the same as that of any 
other monatomic gas, namely, 2.b R per mol of the gas. 
(3) That the constant I, which is the sum /^i^ + -f Ie- — Im, of constants 
characteristic of the three substances (the positive ion, the electron, and 
the neutral atom), can be evaluated by assuming Im+ = Im and by cal- 
culating Ie-, the constant for electron gas, by a theoretical principle that 
has been shown applicable to ordinary monatomic gases, ^ which need not 
be here described. 
These considerations lead to the following values of the constants oc- 
curring in equation (4) : 
AH„^iV^^965Wy^j^g2^ AC^ ^=-,5.10. 
" R R 8.316 R R 
Substituting these values in equation (4) and changing from natural to 
ordinary logarithms, we get 
log K = - -f 2.5 log T - 6.56. (5) 
By this equation the ionization-constant K of an element at any abso- 
lute temperature T can evidently be calculated from its ionization-poten- 
tial V (in volts) ; and from this ionization-constant its actual ionization can 
then be found by equation (1) or (2). 
The main purpose now of this paper (of which a much more complete 
form will soon appear in the Journal of the American Chemical Society 
and the Astro physical Journal) is to point out that the earlier investiga- 
tion of Arrhenius,^ and especially the more recent researches of H. A. Wil- 
son,^ on the electrical conductivity of flames into which salt solutions are 
sprayed clearly indicate that the conduction arises from an ionization 
process of the type here considered, and that these measurements make 
possible a computation of the relative magnitudes of the ionization- 
constants. Thus the experiments of these investigators have established 
the following facts. 
1. The conductance of the flame is independent of the acidic constituent 
