178 Mr Richardson, On an attempt to detect radiation, etc. 
The experiments therefore show that there is no leak produced 
by alternating currents in the wires. It remains only to consider 
the sensitiveness of the apparatus and the order of smallness 
within which the effect, if there be any, must lie. 
Approximations to this were obtained in three ways : — 
1. From the weight and dimensions of the gold leaf the 
quantity of electricity corresponding to a given reading was 
calculated. Now a leak from the 8th to the 7th millimetre per 
hour could easily have been detected, and the corresponding rate 
of diminution of surface density is 1 11 x 10~ 5 electrostatic units 
per second. This gives the least detectable loss of charge as 
1-42 x 10 -4 electrostatic units per second, taking into account the 
area of the leaves and the capacity of the whole apparatus. Since 
the charge on an ion is 6 x 10 -10 electrostatic units, we get that 
an appreciable leak would have been obtained if 2 x 10 5 ions had 
discharged against the conducting part of the apparatus per 
second. This is about 1000 ions per discharge since there were 
several hundred discharges per second. 
2. A further test of the sensitiveness of the apparatus was 
afforded by the discharging action of ultra-violet light. A spark 
gap of 0*5 cm. was set up 15 cms. from the bulb l, and a disc was 
placed in front of the bulb so that all the light had to pass 
through a hole 1 cm. in diameter. Although most of the ultra- 
violet light must have been absorbed by the glass of the bulb l, 
nevertheless enough penetrated the interior to produce a leak 
of 9 divisions of the electroscope in 15 minutes. 
3. A third test was afforded by the effect of radio-active 
substances. A little of the powder, which was of medium strength, 
was covered by thin aluminium, and after the discharge wire had 
been removed was inserted through one of the holes / by which 
it entered. It did not project so far as the charged cylinder, yet 
a leak of 9‘5 divisions in 24 minutes was produced. On with- 
drawing the aluminium foil, three small grains of the substance 
dropped out and these produced a leak of 9 divisions per hour. 
The experiments prove, if not that there is no radiation from 
the surface of metals carrying very rapidly alternating currents 
of great intensity, at any rate that it is smaller than the very 
delicate methods employed are capable of detecting. If the rays 
had been sufficiently active to produce 40 ions per c.c. per discharge 
throughout the volume of the gas in the small bulb ( l ) a measur- 
able leak would have been obtained. This furnishes therefore an 
upper limit to the intensity of the radiation. 
In conclusion I wish to thank Professor Thomson for suggest- 
ing this work to me, and for his advice and sympathy throughout 
the investigation. 
