1897-98.] The Passage of Water through Indiarubber Films. 259 
It was soon evident that the balloons all lost weight at a rate 
which was fairly constant for each under the same conditions. 
Some were exposed to the air of the room, some placed in a 
chamber kept saturated with water-vapour, and some in a chamber 
kept dry by sulphuric acid. The loss fell in the moist chamber to 
Ts or 2 V ordinary air, and was more than doubled in the 
dry chamber (see PI. figs. I. and II., and Tables A., B., F.). 
The thickness of the rubber films, in their distended condition, 
varied from 0*1 to 0*025 mm. The extremes of the observations 
made are as follows : the numbers represent for twenty-four hours 
the decrease in radius of the balloon in micromillimetres (/ x = metre 
x 10~ 6 ) ; or, which comes to the same thing, the amount passing 
through per square metre reckoned in cubic centimetres : 
In the air of the room . . .11 and 32 
In the dry chamber . . .34 and 74 
In the moist chamber . . . *8 and 2 
These are the extremes recorded. For each balloon the divergence 
is less. The loss was not proportional to the thinness of the rubber; 
though the thicker balloons, on the whole, lost less than those of 
thinner make. 
One of the balloons showed a small leak when it had been filled ; 
it occurred to Dr Knott to close it with a piece of gummed paper. 
Kext day it was found to have been leaking a little ; but the 
leakage ceased, and this one behaved afterwards in the air of the 
room just like the others. 
It seemed clear, then, that the loss observed was closely analogous 
to evaporation, for it depended on the pressure of water- vapour in 
the air surrounding the balloon, not upon the hydrostatic pressure 
within it, which was practically the same under all the conditions. 
These experiments were all performed between the 19th July 
and the 6th October. This month (December) I find that similar 
balloons of thickness 0*025 mm. lose weight much more slowly — 
at a rate representing 4 to 6 cub. cm. per square metre per day, or 
only about ^ of that which obtained with corresponding balloons 
during the autumn. This difference no doubt depends on the 
lower temperature and consequent lower vapour-pressure of water, 
under ordinary conditions, at this time of year. 
