12 METABOLISM 



fully poured consecutively into a glass cylinder, they will remain at first 

 distinct from each other. Sooner or later, however, the sharp boundary be- 

 tween them vanishes, as the copper sulphate molecules begin to wander in 

 amongst the water molecules and vice versa. This phenomenon is known as 

 diffusion, and it continues until the intermixture of the two fluids is complete, 

 and until the whole becomes uniform in character and in concentration. Sup- 

 pose, however, that the two fluids have no free surfaces but are separated by 

 a porous wall, then the phenomenon takes place under altered conditions, and 

 is known as osmosis. Let us assume that the fluids are water and an aqueous 

 solution of copper sulphate, and, further, that these fluids are placed in the 

 legs of a U-tube, separated by an intervening wall of clay (s. Fig. 2), animal 

 bladder, or parchment ; it will be found that the two fluids do not pass through 

 the dividing wall with equal rapidity ; the water, as a matter of fact, passes 

 far more rapidly into the copper sulphate than the copper sulphate solution 

 into the water. The result is obviously an increase of fluid on the copper 

 sulphate side, the level of the fluid on that side rising in proportion as it falls 

 on the other. Similar results are obtained when copper sulphate is replaced 

 by selected salts or by alcohol. That the result is essentially dependent on 

 the nature of the separating wall is shown by the fact that when a thin 

 caoutchouc membrane is introduced between alcohol and water, more alcohol 

 passes through the membrane into the water than water into the alcohol. 

 In every case, however, so long as the separating membrane is permeable to 

 both bodies, a complete intermixture in the long run always takes place, 

 resulting in the same concentration of fluid on both sides of 

 the dividing wall. 



The cell-wall comports itself precisely like the parchment 

 or animal bladder in this experiment, presenting less resistance 

 to the passage of water than of salt solutions. In course of 

 time the salts also manage to pass through, and it is only to 

 chemical compounds with large molecules, like gum, proteid, 

 &c., that the wall is quite, or almost quite, impermeable. 



Protoplasm in its osmotic characters differs from the 

 p. 2 cell-wall and agrees with the so-called semipermeable mem- 



brane in one very important point, viz. that it is quite 

 impermeable to certain materials, such as many salts, sugar, &c., although 

 it is readily permeable to water. If one places such a semipermeable 

 membrane in the position of the dividing wall in the experiment just de- 

 scribed (Fig. 2) the water passes into the copper sulphate, but no trace of the 

 copper sulphate passes into the water. Under these conditions a uniform 

 mixture of the two fluids, resulting in similar concentration on either 

 side of the membrane, cannot take place ; one substance must always be in 

 excess on one side of the dividing wall. As an example of a semipermeable 

 membrane we may take the so-called ' precipitation membranes ' produced 

 by the contact of aqueous solutions of ferrocyanide of potassium and copper 

 sulphate or of tannin and gelatine. The precipitation membranes thus ob- 

 tained, gelatine-tannate or copper-ferrocyanide, do not form suitable experi- 

 mental materials, nor can they be introduced conveniently into the U-tube 

 shown at Fig. 2. Even were their introduction possible, these membranes 

 are so fragile that they would be quite unsuitable for the purpose. For those 

 reasons Pfeffer (1877), in his fundamental researches on this subject, precipi- 

 tated the copper-ferrocyanide membrane in or on the wall of a porous pot, 

 such as is used in a Daniell's element. The precipitation membrane is thus 

 supported, and the apparatus takes the form represented at Fig. 3. If the 

 porous pot be filled with a 10 per cent, solution of cane sugar to which the 

 copper-ferrocyanide membrane is impermeable, and if the pot be then placed 



