500 Mr. E. F. Burton on a Radioactive 
air, was about 16°5 scale-divisions per minute. After heating 
the water in the bath to the boiling-point, air was bubbled 
for fifteen minutes through the oil and drawn into the cylinder 
A by means of aw ater-pump. The cylinder was then dis- 
connected from the tube H and hermetically sealed, after 
which measurements were made, from time to time, on the 
conductivity of the gas which it contained. The density of 
this gas was determined in ever y case and found to be about 
1°05, air being taken as unity. 
A Radioactive Emanation.—On first introducing into the 
cylinder the air which had passed through the oil, it was 
found to have an initial conductivity very ” oreatly in excess 
of that of normal air. Its conductivity steadily increased, 
after the cylinder was closed, for about three hours, when it 
reached a maximum value, atbes which it slowly decreased 
approximately in a geometrical progression with the time. 
Fresh air passed through different samples of petroleum into 
the cylinder under exactly similar conditions was found to 
possess different initial conductivities, but, in every case, the 
conductivity of the confined air steadily rose in about three 
hours to a maximum about 40 per cent. in excess of the initial 
value. It then decayed according to an exponential law, 
always dropping to one-half value in about 3°125 days. A 
typical set of observations on the conductivity of air bubbled 
through one of the samples of oil is given in Table IL., the 
time being reckoned from the moment when the cylinder was 
closed. 
TABLE I. 
Time | Current. | Time | Current. 
: |Arbitrary Scale. ; ‘Arbitrary Scale. 
hrs.) mins: | hrs. min. | 
me Tres 92 SE | reat 92 
2 ee 95°6 PS ee RAS, 83°5 
edd A | 103 50 eet 178 
i: ae L157 67 . 71 
Ware S. | 116°5 73 = 330 67-7 
2 43 | g7 | 95 oe 60°35 
9° 30" | re | Tie 55°5 
20 Cai 101 128 be 50°8 
23 95°7 | 138) ,e a0 486 
These results are shown graphically in fig. 2, where the 
ordinates of the curve represent the conductivity of the gas 
and the abscisse the times in hours. 
As in the experiments of Professor Thomson with the 
Cambridge tap-water and those of Strutt with mercury, all 
