187 
On the Recombination of ions in air and 
OTHER Gases. 
By W. H. Bragg, M.A., Elder Professor of Mathematics and 
Physics in The University of Adelaide; and R. D. Klee- 
MAN, B.Sc, Demonstrator. 
[Read October 3, 1905.] 
It is well known that when jjositive and negative ions are 
distributed through a given space a process of combination 
goes on until ions of one sign only are left. Let there be 
y positive ions, and u negative ions in each cubic centimetre 
at any instant, and suppose that the relations of any ion to 
all those of opposite sign are of the same character. Then 
the chance that an ion, say, a positive one, will enter into 
combination before the end of a short time ^t is propor- 
tional to ii^t ] and generally the number of combinations 
taking place in that time may be denoted by (iqm?>t where a 
is the "co-efficient of recombination." This has been clearly 
established by the experiments of Rutherford, Townsend, 
McClung, Langevin, and others. 
As a consequence, the current passing between two elec- 
trodes in a gas in which ions are being formed by external 
agents depends on the magnitude of the potential gradient 
or electric force- The relations between current and force 
have been carefully studied by many workers, and the observed 
facts have been compared with the results of calculation based 
on theory. The comparison is partly, but not completely, 
satisfactory. 
Certain experimental results which we propose to describe 
in this paper seem to throw light on the reason of the dis- 
crepancy. They point to the existence of another cause, dis- 
tinct from that represented by the expression o:np, which 
prevents ions from reaching the electrodes in the gas in which 
they are formed. This cause appears to be a process of re- 
combination of newly-formed ions with the atoms from which 
they have just been separated. The effects of it are propor- 
tional to the number of ions formed in a c.cm. in unit time, 
not to the product of the existing numbers of posi- 
tives and negatives. They aie independent of the shape of 
the ionisation chamber, and in this they differ from those of 
general recombination- They depend directly on pressure, 
and vary greatly from gas to gas. 
In order to bring these effects into relief it i^ only neces- 
sary to reduce tne number of ions in a c.cm. until the number 
of those that are lost by general recombination is negligible 
