262 Proceedings of Royal Society of Edinburgh. [sess. 
water in the jar showed a slight but decided and constantly- 
increasing reaction to silver nitrate solution, showing the presence 
of a chloride in the water surrounding the bladder. 
These balloons began to leak by minute apertures three or four 
weeks after they had been filled. But in spite of this, they con- 
tinued to gain weight when immersed in water, though more slowly 
than before. 
I am indebted to Professor Crum Brown for giving me references 
on the subject, especially to papers published in 1866 by Payen* 
on the question of the permeability of indiarubber, in which he 
describes experiments of the same kind with bladders filled with 
water. His papers deal mainly with the microscopic structure of 
indiarubber, which he found to be penetrated by fine pores ; and 
he maintains the theory that these explain the well-known results 
of Graham with regard to the diffusion of gases through india- 
rubber, as against Graham’s own theory that the gases pass through 
the membrane in a state of solution. 
Payen’s balloons were, in their distended condition, *5-1 mm. in 
thickness, and the results varied with the state of the rubber. 
Pure (unvulcanised) rubber allowed 23 c.cm., vulcanised rubber 
only 4 c.cm., to pass per square metre in twenty-four hours. 
The temperature was 15° C., not very different from the tem- 
perature at which my experiments were made in summer. The 
rubber with which I have worked has all been more or less 
vulcanised. 
With regard to the explanation of the phenomena observed, I 
would only remark : — 
(1) That though Payen finds pores in the indiarubber, he fails 
to show that they are continuous throughout it. 
(2) That the fact that carbonic acid penetrates a rubber film far 
faster than other lighter gases, even than hydrogen, seems to 
require some other explanation than is given by assuming the 
rubber to behave merely as a porous septum. 
To summarise my results, then, I think I have shown — 
(1) That indiarubber, at least in layers up to 0’5 mm. thick, is 
steadily, though slowly, penetrated by water. 
t Compt. rend., 63, 533 ; Jour. Pharm ., [4], 4, 357 ; Zeitschrift fur Anal. 
Chem., 6, 109 ; Chem. Centralblatt , 1867, 93. 
