VITAL PHENOMENA 105 



sieve membrane determines its electric charge, and hence its 

 permeability to ions. 



If we try to classify membranes according to their permeabil- 

 ity, the boundary between sieves and phase membranes seems 

 to break down. Bigelow and Bartell (1909) showed that the 

 permeability and osmotic properties of porcelain membranes de- 

 pend on the size of the pores. Whereas the pores in these mem- 

 branes can be seen, those in gelatine or collodion membranes are 

 invisible. Bechhold (1907) showed, however, that the permeabil- 

 ity of collodion or gelatine gels varies directly with the water 

 content, and with high water content large particles could pass 

 through them by filtration. Thus it seems that these emulsoid 

 gels have pores, a fact that may be detected in suspensoid gels 

 with the ultramicroscope. Rubber is also a colloid, at least 

 under certain conditions, although a rubber membrane has been 

 regarded as a continuous phase. Prof. Ira H. Derby tells me 

 that the rate of flow of liquids into a rubber membrane is mathe- 

 matically the same as the rate of flow into capillary spaces exist- 

 ing in small spherical aggregates, indicating that the liquid passes 

 through capillary spaces in the rubber. This is not only true of 

 native rubber, in which traces of the emulsion structure might 

 supposedly be left, but also of dental rubber, in which the spaces 

 between the original droplets have probably been obliterated by 

 the process of manufacture. 



Negative Osmose 



It has long been known that absorption by the gut and secre- 

 tion by the kidney may take place against osmotic pressure and 

 concentration gradient, and many attempts have been made to 

 rob these phenomena of their peculiarly "vital" attributes. Ham- 

 burger (1908) claims that membranes of different composition 

 on the two faces show irreciprocal permeability to water and 

 solutes. 



These experiments by Hamburger (which have apparently 

 been discontinued) may be concerned with the phenomenon of 

 negative osmose observed by Dutrochet, Graham, Girard, Bern- 

 stein and Bartell. According to Flusin (1908 a and b) the nega- 

 tive osmose is due to the fact that the swelling of the membrane 

 is greater on one side than on the other. 



