[EVE] IONS PRODUCED BY THE BETA RAYS 71 
it follows that q = 292 divisions/minutes, where q is the number of 
ions per em. per second in the electroscope corresponding to a given leaf 
measurement in divisions a minute. 
If a quantity of radium Q is placed at a distance r from the elec- 
troscope 
where q is the number of ions produced per em* per second by the + 
rays from the radium C in equilibrium with the radium. When r is 
only two or three metres, we may neglect #7, because 7 is so small. 
In order to find the constant K’, specimens of radium of strengths 
-23, -52, 1-6 and 8-5 mg. were used. A specimen was placed at 
distances varying from 75 to 300 em., closely surrounded by cylinders 
of zine or lead, between the poles of a large electromagnet. In this 
way the 6 rays were absorbed and deflected, and the correction for the 
absorption of the y rays was as small as possible. The strengths of 
the radium specimens were determined by comparison with a Ruther- 
ford & Boltwood standard at McGill University, and with another 
standard which was tested at Manchester University by the kindness 
of Professor Rutherford. 
The mean result thus obtained was K’ = 3-74 X 10°, with a pos- 
sible error of about 5 per cent. Thus the number of ions produced by 
the + rays from 1 curie of radium C at 1 metre from the source in air, 
at atmospheric pressure, is 3-74 X 10° per cm.’ per second.* 
In the Phil. Mag., Sept., 1906, is an account of a previous deter- 
mination of the constant, and the value for K’ then found with an 
aluminium testing vessel, for pure radium bromide, withe—3.4 X 10, 
was 3-1 X 10°. This gives for radium, with e — 4-9 X 10°”, K’ = 3-68 
X 10°. This is 2 per cent less than the present determination. I have 
placed aluminium cylinders about 1 mm. thick over the electroscope 
used in the experiments now described, and find that with y rays alone 
there is an increase of about 5 per cent in the ionization in the electro- 
scope due to the surrounding aluminium cylinder, provided B rays are not 
coming from the source. The agreement between the old and new 
determinations is not, therefore, so good as at first appears. The 
former work, however, involved the very uncertain correction for a 
considerable absorption by lead, an error which has been now obviated. 
* In a paper to the Phil. Mag. Jan., 1911, I took K’ — 3-1 x 10° instead o 
3-74x 10%. All the results referring to y rays in that paper require multiplication 
1-2. 
