The Radioactivity of the Alkali Metals. 
17 
ashes from wood which had grown on the Sussex Weald was 
obtained. These wood ashes were mixed with water and the 
resulting solution evaporated to dryness with sulphuric acid : 
the remaining solid consisted almost entirely of potassium sulphate 
mixed with sufficient organic matter to turn it a light grey. The 
activity was found to be 474 on the same scale as that of Table 1. 
The penetration of the rays was also tested by the method de- 
scribed in § 4 below. It was found that the decrease of ionisation 
caused by two layers of tin foil was 36 °/ 0 in the case of the 
commercial salts and 38'5 °/ Q in the case of the salt from the 
wood ashes. It may be concluded that the activity of the potas- 
sium does not depend on its place of origin. 
A still more convincing experiment of the same nature was 
tried. A specimen of orthoclase said to contain about 16‘5 °/ G 
of potassium was procured. The operation of extracting the 
potassium from the mineral was judged to be too tedious: the 
activity of the mineral itself was tested, and found to be 201. 
If all the activity had been due to the potassium it should 
have been 171, a value somewhat smaller. The difference may 
be accounted for by the presence of a small quantity of radium 
in the felspar, the presence of which was shown readily by testing 
for the emanation. The decrease in the activity of the mineral 
when it was covered by one layer of tin foil was 40 °/ 0 : since 
the tin foil was sufficient to cut out all the a rays of radium it 
was to be expected that the decrease would be larger than when 
a pure potassium salt was used. This expectation was confirmed : 
for the tin foil cuts down the ionisation due to the potassium 
salt by 2 6'6°/ n . However the decrease caused by a second layer 
of tin foil was 16°/ 0 , and agrees well with the decrease found in 
the same conditions for pure potassium salts, 17 - 2°/ 0 . 
These observations afford conclusive evidence that the active 
agent is either potassium itself or some element associated with 
it universally. 
(5) An attempt was made to decide between these 
alternatives by subjecting a potassium salt to fractional crystal- 
lisation. The first experiments were made upon potassium 
sulphate. 600 grammes of the salt were taken and divided into 
two equal portions by crystallisation from a solution saturated 
at 100 C. A difference in activity of 2°/ 0 was found, the crystals 
being slightly more active than the mother liquor. But when 
each of the portions was recrystallised once more by the same 
method, there was only a difference of 1 °/ 0 (barely detectable) 
between the extreme portions. 
Similar experiments were tried with potassium nitrate. 
1 kilogram of the salt was recrystallised 18 times, the crystals 
being each time separated from the mother liquor, redissolved 
2 
VOL. xiv. PT. i. 
