i 
with special reference to solutions of its salts. 563 
Crowther and others on the rays from uranium show that this 
quantity generally increases with an increase in the density of the 
absorbing material: the solutions, both of uranium and potassium, 
were less dense than the corresponding solids. 
If the density of preparations of uranium is constant, it appears 
that the activity is dependent only on the content of uranium. 
A weak solution of uranium was made containing '00067 gramme 
U per c.c. 210 c.c. of this solution placed in the tray gave an 
activity of 82. The same volume of the solution was then 
evaporated to dryness with 210 grammes of ammonium chloride, 
which is approximately of the same density as water. The activity 
of the solid resulting was 82'5. Incidentally it was found that 
uranium is 390 times more active than potassium as measured by 
the ionisation produced by its /3 rays in the apparatus used. But 
the value of this ratio deduced would depend on the nature of the 
ionisation vessel used. 
§ 8. Experiments were then made on the absorption of the 
j3 rays of uranium in solutions of potassium carbonate. A full 
account of these experiments and those which have arisen out of 
them will be given in a later paper. Here it will suffice to say 
that it appeared at once that the value of X/p in solutions cannot 
be predicted simply from that of the solute and solvent : solutions 
show ^markable abnormalities which made it appear possible that 
t v ' 1 apparent activity of the potassium solutions might 
n < > expectation, be explained by examining their absorp- 
•ays that they admit. An attempt was made accordingly 
directly the coefficient of this absorption, 
this purpose the activity of layers of varying thick - 
asured and the results compared with the theoretical 
/ = /„( 1- (4). 
In the case of the solution the thickness x was estimated from the 
volume of the solution and the area of the tray a (new trays 
were used) : in the case of the solid the mass of the salt (m) 
spread evenly over the tray was measured and the equation used 
in the equivalent form 
\ m 
I = / 0 ( 1—e **■ <r ) (4a) 
leading to a direct determination of X/p. 
The experiments lead to an evaluation of I 0 and X/p. 
a (equation (3)) is also known and hence the values of — . X/p, 
a quantity which ought to be constant, can be found. 
VOL. XIV. PT. VI. 
37 
