J4 OF KNIHJSMOSIS. 



this absorption ; it may take place independently of apertures, pores, or vessels ; it may 

 take place between gases and gases, gases and vapours, or liquids, or solids, or mutually 

 and indiscriminately among them all. 



36. When a 'liquid rises in a capillary tube, those portions only are under the direct 

 influence of the attractive force of the tube which are nearest to it, the central columns 

 being entirely unaffected. Also, when water jets out through a narrow pipe, it is only 

 those portions that are directly in contact with the sides of the pipe that are subject to 

 its resisting influences, any disturbance which the central particles feel arising indi- 

 rectly from their cohesion. The same applies in the passage of liquids among pores: 

 the diameter of these pores amounting to a certain size, they will admit a passage with- 

 out exerting any direct influence. Thus, a pore in a piece of charcoal may suffer a 

 column of water to go through it, without in any wise affecting the central portion of 

 that column, by reason of its size ; but should the diameter of the pore be made to de- 

 crease, it is obvious a limit might finally be reached, where every particle that passed 

 should come under the direct influence of the physical force exerted by the pore, and 

 none pass by mere leakage or oozing. 



37. This leads us to consider the different effects that may ensue when the same 

 liquid or gas passes through pores of various sizes in the same solid. An example may, 

 perhaps, illustrate the results : The walls of a pore are so constituted as to allow an 

 easy passage of one gas, as oxygen, along them, and afford more or less resistance to 

 another, as nitrogen gas. Now, if we suppose this pore to be of very large size, and 

 atmospheric air to be passing through it, little or no change will happen in the consti- 

 tution of the passing gas, all the internal parts of the current being ut of the reach of 

 the walls of the pore ; but should the diameter of the pore be reduced to the diameter 

 of an atom of the compound gas, or thereabout, the oxygen, finding little or no resist- 

 ance, would glide through, and the nitrogen be retained, a perfect decomposition happen^ 

 ing. This shows the importance in all investigations relative to ENDOSMOSIS, or transit 

 of bodies through pores, of bearing in mind that, when those pores have a certain di- 

 ameter, the results of experiments made on them are illusive, not representing alone the 

 nature and value of the force exerted by the walls of the pore, but showing effects de- 

 pending also on the cohesion and other properties of the passing body. 



38. These observations apply to those experiments which have been made to illus- 

 trate the phenomena of endosmosis by forcing gases through plugs of stucco, which are 

 systems of capillary tubes of large size. Experiments on charcoal, plaster, &c., are also 

 open to the same strictures. Had these only been resorted to, the simplest phenomena 

 of endosmosis could not have been discovered. The disturbance X)f hydrostatic level, 

 which is so well shown by a sheet of gum elastic, or an animal membrane, cannot be 

 produced by the use of plugs with large pores or systems of capillary tubes. 



39. It might at first be expected that, as the diameter of a pore decreased, its indis- 

 position to admit a foreign body would increase ; but it is not so : that foreign atom 

 does not insinuate itself in a passive manner, nor does it go through the pore merely be- 

 cause it meets with no resistance. There is an active and very energetic force in play, 

 a force that is even greater than the cohesion of the parts of the pore itself. Hence, 



