1887.] Capillary Action in some Chemical Decompositions. id! 
been observed. Charcoal filters doubtless act in the same kind of 
way, though the chemical changes which occur in the films on 
charcoal are not so easily analysed. 
Suppose now that we have a platinum plate immersed in water 
containing oxygen in solution, then if the surface tension between 
platinum and the solution will be diminished by increasing the 
quantity of oxygen in solution there will be an attraction of oxygen 
into the film in contact with the platinum from the rest of the 
solution, and the film may acquire so much oxygen as would super- 
saturate it if it were not close to the platinum. Further, if the water 
contain hydrogen peroxide in solution the oxygen to saturate the 
film may be drawn from the peroxide, and the decomposition of a 
part of it may be thereby determined. Moreover, so long as a 
supply of oxygen is forthcoming from more, as yet undecomposed, 
peroxide it seems to be possible that the movement of diffusion, 
even of the liquid forming the film, may still go on. Besides the 
heat due to the decomposition of the peroxide will produce con- 
vection currents which will help diffusion in carrying some of the 
supersaturated solution into the body of the liquid. As soon as 
the supersaturated solution is removed from the sphere of the 
molecular action of the platinum, which according to Quincke has 
a radius of not more than ;,/,; millimetre and probably much less, 
the excess of oxygen will tend to roll up into bubbles, since by this 
arrangement the surface tensions will on the whole be a minimum. 
These bubbles will be formed quite close to the platinum and 
will appear to rise from its surface. 
Again, we know that the surface tension of a liquid is some- 
times greatly modified by the admixture of a comparatively minute 
quantity of some other liquid, and it is very probable that the 
addition of an acid, or of an alkali, to water may seriously affect its 
surface tension in contact with solids, not only with metals but 
also with glass. If the presence of acid impedes the liberation 
of oxygen it will also impede the decomposition of hydrogen per- 
oxide, while if an alkali facilitates the liberation of oxygen it 
will also facilitate the decomposition of the peroxide. 
At present there are no data by which to solve the question 
whether acid and alkali do impede or facilitate the liberation 
of oxygen from its solution in water, but it is only some such 
explanation which seems adequate to account for the facts observed. 
A metal so unalterable as gold induces decomposition of hydrogen 
peroxide in a neutral solution, but the addition of only a few 
drops of acid instantly stops the evolution of gas, while the 
addition of alkali hastens the streams of bubbles. It has been 
asserted that acids form chemical compounds with the peroxide, 
but there is no indication that this is the case, and the effect of 
the acid in stopping the escape of gas is quite evident when the 
