20 PERCOLATION THROUGH EXCESSIVELY THIN FILMS OF WATER, AS SOAP BUBBLES. 



are still so close, that anything like mechanical straining or leakage cannot take place 

 through it. The first attempts to ascertain the laws of movement and equilibrium of 

 gases passing through liquid films were made by stretching those films over the mouths 

 of vials, as here described ; subsequently, for several considerations, this arrangement 

 was given up : the horizontal film is at first too thick, it exposes too small a surface to 

 the atmosphere to which it is subjected, and it is not until towards the close of the 

 experiment that the action becomes at all rapid. Bubbles of water made sufficiently 

 adhesive by a little soap were, therefore, substituted. One of these filled with any gas, 

 and immersed in an atmosphere of another gas, at once exposes a large surface, and, 

 by swelling or collapsing, allows a free action. There are, however, three circum- 

 stances which tend to destroy such bubbles, and against these provision should be care- 

 fully made. Mechanical agitations of the surrounding air may be met by covering the 

 whole arrangement with a glass bell. Evaporation from the surface of the bubble, which 

 reduces its substance unduly, may be avoided by keeping all the gases under trial in 

 jars over water, until they are loaded with moisture, and thoroughly wetting the inside 

 of the covering-bell ; but it is not so easy to prevent that slow motion of the parts of 

 the bubble, which, in virtue of the earth's attraction, tends gradually to bring them to 

 the lowest part, while the walls of it become too thin to bear the weight, and are liable 

 to burst by the expansion of the gases accumulating within. 



54. After a number of trials, the following has been found to be the most suitable ar- 

 rangement for prosecuting these inquiries ; it is simple, not easily deranged, and allows 

 of sufficient latitude and change to suit other experiments. In it a soap bubble may be 

 preserved with certainty, for a time considerably exceeding an hour, and sometimes much 

 longer. As here described, it was used to illustrate the relative passages of hydrogen, oxy- 

 gen, and nitrogen through a watery film into atmospheric air. It is represented in section ; 

 A A (Jig. 11, pi. 1) is a small tin saucer, about three inches in diameter and half an 

 inch deep ; into it water can be poured, and it also serves as a platform to support a 

 large cupping-glass, b. Through the centre of this tin saucer, at c, passes a glass pipe, 

 f, i inch in diameter, the upper extremity of which is cemented into a hole of the same 

 size in a round, thin piece of copper, d, which is about half an inch in diameter, the 

 other end of the pipe opening into another cupping-glass, k, through a perforation in its 

 top, the communication being capable of being cut off by means of a cock, g ; the 

 lower cupping-glass serves as a support to the whole arrangement when placed upon 

 the shelf of the pneumatic trough. This apparatus is used as follows : The upper 

 cupping-glass being taken off the platform, is filled with any gas under trial, as oxygen, 

 and placed aside on the shelf. The lower cupping-glass is then filled with water 

 by depressing it in the trough, and the cock being closed, five hundred measures of 

 hydrogen, for instance, are thrown into it. After seeing that the copper plate, d, is 

 free from moisture, a drop of water, rendered slightly viscid by soap, is placed upon it 

 exactly over where the orifice of the pipe,yj opens. The upper glass, containing the ox- 

 ygen, is now placed upon the little tin saucer platform, as in the figure. The lower 

 glass is next depressed in the trough, and as soon as the cock is opened, a bubble of 

 hydrogen containing five hundred measures expands, the spare oxygen escaping from the 



