M‘Crettanp—The Energy of Secondary Radiation. 19 
Similarly, if », denotes the number of secondary particles emitted when 
the plate is made of, say, copper, we have 
N—n = 40a, 
or 
[eee eae 
N WN 
and 
pa Be He 
TG I IS 
But y is the value of our ratio p for lead, and + is the value for copper. 
l-=p, 26 
Lop £0 
In Column III. of Table I. we have the relative secondary radiations from 
lead and copper as measured by their ionising powers; from this column 
Combining this equation with the previous one we get 
Pi = ‘66, 
If we calculate the value of p for lead in this way, by taking lead with each 
of the other substances in Righi’s table (omitting platinum, which is marked 
doubtful, and bismuth, which is too near to lead to give an accurate result), we 
get values for p lying between ‘60 and 67, and giving a mean value of 64. 
The agreement between the values of p for lead as calculated from Righi’s 
results and as found above is not good. However, when we consider the widely 
different methods used, one comparing the radiations by measuring the ionisations 
produced, the other by measuring the charges carried, and as the leakage due to 
the ionisation of the residual gas surrounding the insulated plate in Righi’s 
experiments must have been difficult to eliminate, we could scarcely expect 
a close agreement. 
Tue SEconDARY RADIATION FROM A PuaTE aT DIFFERENT INCLINATIONS TO THE 
NorMAL. 
Before leaving this part of the subject, we may very briefly refer to the 
way in which the intensity of the secondary radiation from the plate was 
found to vary with the inclination to the normal as shown in the previous 
D 2 
