68 APPARENT DECOMPOSITION OF METALLIC SALTS BY MEMBRANES. 



221. In giving an explanation of these curious facts, it is to be borne in mind, that 

 all the phenomena treated of in this chapter are the results of two contemporaneous cur- 

 rents, endosmosis never existing without exosmose. If these currents are established 

 in fluids, which, by their union, give rise to solid matter, its deposite may occur under 

 all the forms designated (219), (220). If, for instance, the precipitate be a light mate- 

 rial, and one of the currents exceed the other in volume, the small particles, as they are 

 formed, are drifted by the current, which is acting under the greater advantage, and the 

 deposite will take place wholly on one side. This, I suppose, is the mode of deposite 

 of oxalate of lime, which always goes with the greater current. The chemical change, 

 or union, it is to be remembered, takes place at the point of contact of the two 

 fluids, which is necessarily in the membrane itself; there they neutralize one another, 

 and if the circumstances of the experiment permit, the more powerful current carries 

 before it the precipitating particles as fast as they form, and the excess of unneutralized 

 material in it produces a precipitate of the same kind as soon as it mingles with the 

 mass of fluid on the side of the membrane towards which it is going. But, if any dis- 

 turbing causes intervene, if the precipitated matter has any affinity for the fibre of the 

 membrane, in the manner of a dye, or if it be too bulky to pass through the pores, or 

 too ponderous for the current readily to move, it is detained on the spot where it was 

 generated, and in a very short time the tissue becomes choked ; the biniodide of mer- 

 cury is subject to these circumstances. A number of disturbing causes will often 

 change the results of these experiments: when, for example, the precipitating particles 

 have a high density, their weight may carry them in a direction even opposed to 

 the stronger current. The relative specific gravity of the two fluids may also deter- 

 mine the course in which the particles shall go. 



222. That these experiments do not prove that membranes have a predilection for 

 passage in certain directions through them, the results of the earlier writers are suffi- 

 cient to show. They have shown that when Prussian blue is deposited in this way, 

 sometimes the precipitate is towards the salt of iron, and sometimes towards the prus- 

 siate of potash, the direction it takes being often influenced by very slight causes. Nay, 

 even analogous actions may be exhibited without using any membrane, barrier, or ob- 

 struction ; if into a half-ounce vial a quantity of strong sulphuric acid is poured, and 

 upon that a solution of chloride of calcium, so that the two fluids may intermix as little 

 as possible, in the course of a few days it will be seen that, as the fluids slowly diffuse 

 into each other, the sulphate of lime is deposited entirely in the supernatant solution, 

 and none in the strong acid, a result unquestionably depending upon their relative spe- 

 cific gravity and cohesion. 



223. Sections (188), (189), contain examples of what has been termed in these papers 

 decompositions of a certain sort. In relation to the first of these, it is by no means 

 clearly proved that oxygen really does leave the nitrogen in the atmosphere to go 

 through the barrier. The two gases may respectively pass into each other, as atmo- 

 spheric air and nitrogen. It is true that the more probable mode of passage is that 

 assumed in the section quoted. If so, it does not even follow that a real chemical de- 

 composition happens, for there is much reason to doubt whether atmospheric air is 



