204 Profs. E. Rutherford and H. T. Barnes on the 
variation with time of the heating effect of both the emanation 
and the radium from which it was separated. 
Description of Apparatus. 
As only about 30 milligrams of pure radium bromide were 
available in the experiments, special methods were devised 
to measure with accuracy the small heating effects involved. 
The maximum heat emission from 30 milligrams of radium 
bromide is not more than 6 x 10-* gram-calories per second, 
and it was necessary to measure with certainty a heat emission 
of at least 55 of this amount. 
In the following experiments a simple form of differential 
air-calorimeter was employed. In fig. 1 is shown a simple 
Fig. 1.—Differential Air-Calorimeter. 

sketch of the apparatus. It consisted of two 1-litre flasks 
with rubber stoppers through which passed three-way glass 
cocks, and tubes into which the radium or the emanation-tube 
could be lowered. The glass cocks connected the air in the 
flasks either to the outside air or to a manometer-tube 
which registered the differences in pressure. Some little 
difficulty was encountered at first in selecting a suitable 
liquid with sufficient mobility and low vapour-pressure for 
the manometer-tube. We finally used xylene, which proved 
in every way satisfactory. The difference in level of the 
xylene standing in the two arms was observed with a micro- 
scope provided with a micrometer eyepiece mounted on a 
cathetometer-stand. The two flasks were immersed in a 

