]g PERCOLATION THROUGH INDIA-RUBBER. CONDITION O*F EQUILIBRIUM. 



used for the purpose of manufacturing copperas for commerce. Slow motions of the 

 same kind occur when alloys are buried under the ground, or placed in exposed situa- 

 tions ; a silver Roman coin has thus been known to part with much of its copper, which 

 formed a species of crystallization on its surface, the patina of antiquarians. It is in 

 this way that trinkets of gold, on which small quantities of mercury have fallen, gradu- 

 ally recover their original brilliancy and purity. A number of facts of this kind, show- 

 ing that even in the most solid of metallic textures motions may take place, might be 

 referred to : these have been well considered by Boyle, in his tract on the languid mo- 

 tions of bodies. 



44. Caoutchouc, or gum elastic, is the substance which, of all others, has furnished the 

 most unexceptionable results on studying the phenomena of endosmosis. It, however, 

 at times exerts a synthetic action, which, so far as I know, has not yet been noticed. 

 Having capped an open tube with a thin piece of this substance, and thrown into it 

 200 measures of hydrogen gas, it was exposed to an atmosphere of 100 measures of 

 oxygen contained in a wider tube, into which it was raised. In eleven days the level 

 in both tubes had considerably risen, and the barrier, which was at first of a blackish 

 colour, became quite white. In sixteen days, the united volume of both gases was only 

 215 measures; this, on analysis, contained only 14 per cent of oxygen. It may here 

 be stated that a like mixture of 100 of oxygen and 200 of hydrogen, enclosed together 

 in a tube by the side of the former, had undergone little or no diminution. Now a 

 rough calculation shows that about one thirteenth of the united volume of the gases had 

 been condensed by the membrane into water, the remaining 62'35 parts of oxygen hav- 

 ing combined, in some chemical manner, with the substance of the caoutchouc, in the 

 process of bleaching it. 



45. This result points out the condensing effect of a membrane, which often, in many 

 arrangements, will have no small influence. Thus, in Dr. Mitchell's experiment, where 

 two bent tubes are screwed together, with a piece of gum elastic between them, the one 

 tube containing oxygen, and the other a double volume of hydrogen, we should be led 

 to expect, from the common theory of endosmosis, that however much the levels in the 

 two tubes might vary relatively to each other, the united volume of the gases ought to 

 remain constant. If the level in the hydrogen tube rose an inch, ought not the level in 

 the oxygen tube to sink an inch 1 But an appeal to experiment shows that such is by 

 no means the fact ; to a certain extent, the volume of gas in the tubes is constantly di- 

 minishing ; it is not due to leakage into the free atmosphere, between the membrane and 

 the glass that presses it, at least not entirely so ; for a part of the gases is condensed by 

 the direct action of the barrier to form water, and the remainder unites chemically with 

 it. In some instances, the action is still more obvious. If a vessel of atmospheric air, 

 the mouth of which is covered by a piece of India-rubber, be immersed in an atmo- 

 sphere of deutoxide of nitrogen, it will be found that red fumes do not appear in the 

 vessel, nor any other obvious indication of the presence of the deutoxide, but the mem- 

 brane soon begins to change its colour, and from being diaphanous, becomes of a dirty 

 umber brown, the volumes of the gases on both sides of it diminishing. 



46. Into a tube which was covered with India-rubber, standing on the shelf of the 



