82 
Transactions of the Royal Society of 8021th Africa. 
The corresponding ionisation-density curve is represented by the line 
CBA in Fig. 6, which is the summation curve of the two ionisation-density 
curves DBA and EF, corresponding to the ions produced by at least two 
distinct groups of corpuscles leaving the gold. 
The method by which the logarithmic curve for the group of ions, whose 
ionisation-density curve is represented by EF (Fig. 6), was found, was to 
continue the line AB (Fig. 5) backwards, and to reconstruct from the 
antilogarithms the part of the curve BD (Fig. 6) which has its equivalent in 
OD in Fig. 3. The logarithm of the difference between this deduced cathode 
ionisation curve for the faster group and the experimentally obtained 
summation cathode ionisation curve for the two groups was again plotted 
against p (or x), giving the more sloping logarithmic curves in Fig. 5. 
Readings taken directly from these lines yield the following table of absorp- 
tion coefficients : 
Table II. 
Shoiving that there are at least Two Distinct Groups of Particles ejected with 
Different Velocities from a Very Thin Gold Leaf ivheyi Silver X Bays 
fall upon it. 
Density". 
Temp. 
Px 
Fast rays. 
Px 
Slow rays. 
\ 
Fast rays 
at 0° C, 
76 cm. 
A 
Slow rays 
at()° C, 
76 cm. 
i A, , 
1 /« 
1 Fast rays. 
1 
Slow rays. 
Air 
•00129 
23° C. 
7-36 cm. 
1-77 cm. 
5^35 
22 25 
4150 
17250 
0.2 
•00143 
26° C. 
7-05 „ 
1-53 „ 
5-64 
26-00 
3950 
18190 
CO., 
•00198 
23° C. 
5 05 „ 
■90 „ 
7-80 
43^76 
3940 
22100 
so.. 
•00293 
22^^ C. 
■i-26 „ 
•91 „ 
9-21 
43^2 
3150 
14720 
Mean = 
3797 
18065 
Eatio = 4^76 
The mean value of /d for the fast rays approaches closely Lenard's 
probable value that I have calculated at the end of Part I, viz. 3120. As 
lending support to the conclusion that there are two distinct groups of 
particles present, " the experiments of C. T. E. Wilson* on the photography 
of the tracks of the emitted electrons in gases indicate that for mono- 
chromatic radiations they would consist of a limited number of definite 
lengths rather than a series extending from zero to an upper limit. "f 
On these grounds, I think that the present accepted values of the 
absorption coefficients in air and hydrogen for the fastest corpuscular rays 
ejected from a plate by silver characteristic rays will need modification. 
* 'Proc. Roy.|Soc./ a, vol. xxxvii. 
t O. W. Richardson, ' Proc. Roy. Soc.,' a, vol. xciv, p. 271 (1918). 
