OSMOSIS. 275 
The attraction of the substance of the membrane for water, at any 
rate, may then be a factor in the case. Ludwig 1 demonstrated, indeed, 
that the concentration of (he solution imbibed by an animal membrane 
may be lower than that of the solution in which it is soaked. 
Fick 2 distinguished between two possibilities for diffusion through an 
animal membrane — a "pore diffusion" in Brticke's sense, and a 
diffusion occurring through the spaces between the molecular aggregates 
of which the membrane may be considered to be built. The latter idea 
is somewhat of the nature of that formed of the diffusion of a gas 
through a film of liquid in which it is soluble, or is perhaps better 
illustrated in the experiment of L'Hermite, 3 in which, when water 
separates chloroform from ether in a tube, the chloroform increases at 
the expense of the ether. Fick's " homogeneous " membranes were 
made of collodion ; but his results show that such a membrane is not 
unalterable, since the amount of salt passing through increases with 
time, and it is difficult to escape the conclusion that in many cases 
some interaction of chemical nature takes place between the membrane 
and the substances to which it is permeable. 4 
The property possessed by certain substances of imbibing certain liquids 
(apart from capillary action), must be borne in mind in all considerations of the 
essential nature of the processes involved in the passage of fluids through 
membranes. This property can only be ascribed to some " affinity " between 
the molecules of the imbibing substance and that imbibed; thus gelatin 
swells in water but not in ether, while the reverse is true of caoutchouc. The 
retention of a gas, or a colouring matter by charcoal, of water by the silica 
of the opal, or that of pepsin by fibrin, are instances of the class of phenomena 
to which attention is here called, and to which the name of adsorption is often 
applied. When a homogeneous substance imbibes a solution, compounds of 
the imbibed with the imbibing substance may be formed, which may have a 
greater affinity for the solvent than the original imbibing substance, hut at the 
same time the osmotic pressure of the solution tends to retard the imbibition 
of the solvent ; hence, with a given jjair of substances, the amount of the 
solution of one taken up by the other will reach a maximum at a certain con- 
centration, a maximum, however, which may be well above that for imbibition 
of the pure solvent. 
The "affinity" of the imbibing substance for the solvent and dissolved 
substance imbibed may be of very different order, for gelatin takes up a more 
concentrate 1 solution of methyl-violet than that in the dye-bath; while, on the 
other hand, a f errocyanide of copper membrane will take up water while almost 
absolutely indifferent to dissolved cane sugar. 
Such " affinities " are not purely mechanical, since they vary with the 
chemical nature of the substances, and yet are not of the nature of chemical 
affinity in the usual sense of the term, since the "compounds" do not obey the 
laws of cunstant and multiple proportion. Ostwald has introduced the term 
mechanical affinity to meet the case. 
In the complex known as protoplasm there may be imbibing substances of 
different nature, permeated by a solution of substances whose chemical nature 
may, directly and indirectly, affect the imbibition of a solution brought in 
contact with the mass ; and, furthermore, undissolved particles may themselves 
1 Ztschr. f. rat. Med., 1849, Bd. viii. S. 1 ; Ann. d. Phys. u. Chem., Leipzig, Bd. 
lxxviii. S. 307. 
2 Untersuch. z. Naturl. d. Mensch. u. d. Thicre, 1857, Bd. iii. S. 294. 
3 Ann. de chim., Paris, 1854. Ser. 3, tome xliii. p. 420. 
4 Tamilian, Ztschr. f. physikal. Chem., Leipzig, 1892, Bd. x. S. 255 ; Walden, ibid., S. 699. 
