802 Proceedings of the Royal Society 
air of the room, the readings of a dry and of a wet bulb thermometer 
were noted. 
The following table gives the results of our different trials. The 
order of procedure is given in column sixth. A number of readings 
were first taken for the air of the room, then the receiver was 
exhausted to a pressure of 5 mm. or so, and the air subsequently 
allowed to enter slowly through the drying tube or tubes. When 
the state of the air was again changed, it was done by allowing 
the air to enter without passing through the drying tubes. The 
number entered in the seventh column is the mean of ten or six 
readings. 
It will be seen that in the case of the great majority of the 
comparisons, the reading for the undried air is less than for the 
dried air, and that this is universally true in the case of the first 
comparison in each series. The diminution of the difference in the 
case of the subsequent comparisons of a series may be due to the 
fact, that to obtain undried air a second time, we allowed the air of 
the room to enter through the already dried passages • of the air- 
pump. 
The series observed on 26 th May is the one in which the drying ’ 
was most carefully performed ; the conclusions which it supports are 
those which might be deduced from the other series taken collect- 
ively. In the case of this series, the individual observations are as 
follow : — 
Difference of Potential. 
Reading. 
First State. 
Undried. 
Second State 
Dried. 
Third State 
Undried. 
1st. 
240 
258 
253 
2nd. 
242 
270 
• 
262 
3rd. 
241 
283 
258 ' 
4th. 
248 
258 
255 
5th. 
250 
257 
255 
6th. 
247 
261 
250 
Mean, 
245 
265 
255 
