566 Mr Campbell, The Radioactivity of Potassium, 
Now if the active layer is emitting rays of different penetration, 
the softer rays will produce more than their fair share of the 
ionisation : for the harder rays pass through the ionisation vessel 
without being completely absorbed in the air contained in it. On 
the other hand, it is easy to see that in the measurements of the 
absorption coefficient the harder rays produce a greater effect upon 
the result than the softer rays, for the harder rays continue to 
cause a variation of the activity with the thickness of the layer 
after the layer has become so thick that it is practically infinite 
for the softer rays. All experiments have indicated that the 
variations of X/p are greater for soft rays than for hard. Hence 
in measuring this quantity for rays which are, on the average, 
harder than those which produce the ionisation, the correction 
for the difference in the value of X/p has been underestimated. 
For this reason it has seemed of no use to continue the 
extremely tedious work of measuring the absorption coefficient 
of the various potassium preparations. It has been shown in one 
instance — and that instance the most abnormal — that the variations 
of the absorption are such as to account, at least in part, for the 
deviation from proportionality to the content of potassium of the 
apparent activity. 
It may be noted that the heterogeneity of the rays for potassium 
suggests that a separation of the activity should be possible. For 
recent work tends to show that the apparent heterogeneity of the 
radium rays is due to the presence of several different elements 
emitting rays of different quality. 
I conclude, therefore, that while it has not been proved 
indisputably that the activity of the potassium is the same in 
whatever state of chemical or physical combination it occurs, 
there is no evidence that the activity is not the same. It appears 
more probable that the differences in the apparent activity 
of different preparations of potassium are to be attributed to a 
difference in the value of X/p for such preparations than to 
a difference in the activity of the potassium. 
§11. The following conclusions may be drawn from the work 
described above : 
(1) All attempts to observe or to produce any difference in 
the activity of different samples of the same potassium compound 
in the same physical state have failed completely. No evidence 
has been obtained of any such separation of the activity as is 
to be expected whatever view may be taken of the source of the 
activity. 
(2) The activity of a thick layer of a potassium compound is 
not accurately proportional to the amount of potassium which it 
contains. It is probable that the variation from strict propor- 
tionality is to be attributed to a difference in different compounds 
