PERCOLATION THROUGH MASSES OF WATER. j^g 



51. The doctrine laid down in sections 47 and 48, of the condition of equilibrium 

 of gases on each side of a membrane, being the foundation of an explanation of all the 

 phenomena which have as yet been noticed, requires farther consideration and fuller 

 proof. Some remarks have been offered on the incomplete results which are obtained 

 bv the use of barriers consisting of pores of large size, such as stucco plugs. It is said, 

 however, that in the hands of Mr. Graham these have given some curious results re- 

 specting the rate of diffusion of gases ; experiments at once satisfactory and singular. 



52. The objections above mentioned have, however, appeared to me so weighty, 

 that I have not made use of such barriers, but resorted to liquids, which, for closeness 

 of texture, uniformity of composition, and, above all, on account of our accurate knowl- 

 edge of their habitudes and structure, are much preferable. They, also, have given 

 results as curious, but far more satisfactory ; and though, in the management of them, 

 something of that dexterity of manipulation is required which practice alone can con- 

 fer, yet they are easy of repetition, never failing to give precise and comparable results. 

 They also afford the means of prolonging or hastening the close of an experiment, which 

 at times is invaluable ; their action, too, is very uniform ; for a film of water so thin as 

 to be coloured acts as well as a mass several inches in depth, but the gases passing 

 through it more rapidly, a state of equilibrium on both sides is obtained in a few min- 

 utes. The following facts will serve as an illustration : Into a tube b (Jig. 10, pi. 1), 

 which was conoidal at its upper end, a disk of paper, a, was fastened water-tight, and 

 then upon that was poured distilled water till it was about j inch deep; the tube was 

 next filled at the pneumatic trough with hydrogen gas, which passed into the atmosphere 

 through the paper roof, and the water reposing on it ; but, though the tube was only f 

 inch in diameter, twenty-four hours elapsed before a column of hydrogen half an inch long 

 had gone out, and in seven days only one inch more. A common glass tumbler was 

 filled with hydrogen gas at the pneumatic trough, and by the side of it stood a small 

 bottle, the height of which was about 1? inch, its diameter ly inch, and the diameter 

 of its neck ? of an inch. The atmospheric air in this bottle being of the same tem- 

 perature as the hydrogen in the tumbler, a finger dipped in water rendered slightly viscid 

 with soap was passed over the mouth of the bottle, so as to leave a thin film stretched 

 there, the tumbler of hydrogen being then placed over it {Jig. 12, pi. 1). In the course 

 of two minutes, the film, instead of being horizontal, became convex, and continued to 

 be so until it had swelled into a large spherical bubble, which capped the top of the 

 bottle ; in sixteen minutes this had increased so much in size and become so thin that 

 it was of a dark metallic lustre, and it burst at last by swelling, so as to touch the bot- 

 tom of the tumbler. During this experiment the barometer was at 28-8 ; thermometer 

 at 68-75, Fah. 



53. The rapidity of this action being proportional to the thinness of the film used 

 as a boundary, it is obvious that the duration of an experiment may be managed by de- 

 termining beforehand the thickness of the film through which the gases shall pass. If 

 very thick, the time may be indefinitely long, and if very thin, indefinitely short. Nor need 

 we be limited in reducing the thickness to the greatest extent, for it is found by experi- 

 ment that, however thin the film may be, it still possesses cohesion enough, and its parts 



