58 COLLOIDS IN BIOLOGY AND MEDICINE 



work were filled with membrane-forming precipitate, for as soon as 

 the membrane was melted, it again became permeable. In the same 

 experiment, it was determined that diffusion was hindered only by 

 visible precipitate membranes. 



There is no difficulty in forming by diffusion similar precipitate 

 membranes from pure organic materials. We have already, on page 

 55, mentioned that a membrane may be formed by the diffusion of 

 goat serum into goat-rabbit serum and we shall later refer to the 

 fact that H. BECHHOLD* 2 obtained membranes by the diffusion of 

 metaphosphoric acid into gelatin containing albumin. There is no 

 doubt that they may be obtained in other ways if desired. But it 

 must by no means be assumed that a membrane is something rigid 

 and unchangeable; on the contrary, it is constantly affected by the 

 substances which flow through it and bathe it, making it more or 

 less permeable, and in this way under certain conditions may evoke 

 a self-regulation or a valve-like action. 



Hitherto there have been no investigations as to the manner in 

 which the permeability of the precipitated membranes described is 

 influenced by crystalloids diffusing through them. A priori it is to 

 be assumed that such an influence exists just as in the case of re- 

 versible jellies. That membranes may be more or less rapidly 

 occluded by colloids is an observation which has been frequently 

 made during the performance of ultrafiltration. 



A number of dried animal and vegetable membranes and parch- 

 ment paper resemble precipitate membranes, in that they possess 

 the same or but slightly superior swelling capacity. In dialysis they 

 are used for the separation of colloids from crystalloids. 



Most of the membranes occurring in the organism are more or less 

 swollen; on drying they lose this property to a great extent, as they 

 are inelastic gels. 



Ultrafilters (see p. 95) formed by impregnating irreversible jellies 

 are similar to natural membranes, since they must be kept in water 

 to preserve their swollen condition. 



Membranes may be powerfully adsorbent, like reversible gels, and 

 in this respect powerfully influence diffusion and filtration. Thus 

 dyes, especially the basic ones, as well as certain groups of enzyms, 

 e.g., arachnolysin, staphylolysin and rennin (H. BECHHOLD * 4 ) are 

 strongly adsorbed by many membranes. Such adsorbed substances 

 may enter into chemical combination with the membrane (causing 

 either shrinking or loss of swelling capacity) and thus diminish its 

 permeability. This is the effect, e.g., of tannic acid, formaldehyd 

 and chromates. 



Alcohol, ether, acetone and sugar increase the permeability in cer- 



