174 



= 262. The 7?7 for each mixture is then calculated. In the 

 table the calculated and observed values are put side b}^ side, 

 and it will be seen that there is a good agreement : — 



<L CU OJ i "^ 



cs c3 e: bji'-2 Oj a C S 



4. 100 760 45-6 195 — 



Nevertheless, in a number of cases in which I have at- 

 tempted to calculate the value of Ul of one gas from a know- 

 ledge of the values of 111 for air, and for a mixture in known 

 proportions, I have obtained an unexpectedly high result, and 

 when I began some direct experiments on the question 1 was 

 quite prepared to find that the ionisation of a mixture of air 

 and gas was more than the sum of the ionisation of air and 

 gas separately. Further experiment will, no doubt, make 

 everything quite clear. In the meantime it is sufficiently evi- 

 dent that the principle is at least nearly true. For the pur- 

 pose of this investigation it may be taken as quite true, since 

 the correction to be made for the presence of air is, at the 

 most, only small. 



The quantity 7, as measured, includciS a small propor- 

 tion of ^ ray ionisation. It must be shown that this does 

 not harm the result. 



In the form of apparatus which T use the ionisation in 

 the portion AFj of the curve (see Fig. 1) is nearly 6 per cent, 

 of the ionisation at P, and I have not found enough variation 

 from gas to gas to justify an attempt at correcting for it. 

 Of course, the quantity is only sm.all. 



The curve shows only this ^ ionisation above A ; that 



