18 
Messrs Campbell and Wood, 
and reprecipitated. The mother liquor of the first crystallisation 
was compared with the crystals of the last, but no difference 
in the activity could be detected. (It is important in making 
these experiments that the specimens of the salt to be compared 
should be brought to precisely the same physical condition : for 
small differences in the density may make a difference in the 
quantity of the rays which escape unabsorbed from a com- 
paratively thin layer of the salt.) 
It was also tried whether aluminium hydrate or barium 
sulphate precipitated in the solution of a potassium salt would 
possess any abnormal activity. No such effect could be observed 
in the most careful experiments. 
It must be concluded, then, that the activity is a property 
of potassium itself. But it is still possible that some disintegra- 
tion product may be found, for in all the experiments just 
described some hours or days elapsed between the preparation 
of the samples and their examination. Further experiments will 
be devoted to this question. 
§ 4. Measurements were made of the penetration of the rays 
from potassium sulphate. The layer of the salt was covered by 
sheets of tin foil and the decrease of the ionisation caused by 
the addition of each layer was noted. Allowance was made for 
the activity of the tin foil itself. The results of these experiments 
are given in Table 2. The first column gives the number of 
sheets of foil, the second the ionisation observed (corrected), the 
fourth the value of - , where X is the absorption coefficient and 
P 
Table 2. 
No. of sheets 
Ionisation 
Decrease 
X 
p 
0 
467 
1 
361 
106 
27-2 
2 
299 
62 
20-0 
3 
265 
34 
12-7 
4 
240 
25 
10-6 
The surface density of the tin foil is '00945 grm./cm. 2 
p the density of the tin. The specimen of tin foil used is that 
employed by Crowther* in his researches on the f3 rays of 
* Phil. Mag. Oct. 1906. 
