Walter Stiles 
188 
When a solution is in question a membrane may be equally 
permeable to both solvent and solute, or it may be impermeable to 
both. More frequently the permeability to the two components 
differs, and generally, in the case of solutions of crystalloids in water, 
the membranes are permeable to both water and the solute, and 
more readily to water. 
The different rates of diffusion of solvent and solute through a 
membrane were shown in the first half of the nineteenth century by 
Dutrochet (1826,1828,1840,1842) and Vierordt (1848) who examined 
the diffusion of aqueous solutions of salt through membranes of pig’s 
bladder. The water passes through the membrane faster than the 
dissolved salt, this difference in rate depending on the nature of the 
salt, its concentration, and, as shown by Graham (1854), on the 
nature of the membrane. 
Graham’s discovery of the impermeability of certain membranes 
to colloids led Moritz Traube to search for a membrane which, while 
permeable to water, should be impermeable not only to colloids but 
to crystalloids. Such a membrane, permeable to the solvent but 
impermeable to any solute may be termed a semi-permeable mem¬ 
brane. Although a perfect semi-permeable membrane has never been 
manufactured, some of the precipitation membranes prepared by 
Traube (1867) are rather near approximations to one. The best 
known of these, and the one which has probably been most used, 
is that of copper ferrocyanide, which is obtained in the form of a 
gel when solutions of potassium ferrocyanide and copper sulphate 
come into contact. For most purposes these membranes are too 
delicate if unsupported, and so they are usually precipitated in the 
wall of a porous pot, a device due to Pfeffer (1877). Other precipi¬ 
tation membranes are those of glue-tannic acid, copper tannate, lead 
tannate, ferric ferrocyanide, copper silicate and tin silicate. These 
membranes, like the collodion membranes already mentioned, are 
graded in their permeability. 
Tammann (1892) investigated the penetration of a number of 
salts through precipitation membranes of copper ferrocyanide and 
zinc ferrocyanide. He found the permeability of the two membranes 
was the same as regards the simple salts he investigated though not 
in regard to dyes. Traube (1867) had found the copper ferrocyanide 
membrane permeable to potassium, sodium and ammonium chlorides, 
and impermeable to barium chloride and nitrate, calcium chloride, 
potassium sulphate and ammonium sulphate. Tammann, on the 
contrary, found the membrane permeable to all these salts except 
calcium chloride. His findings in regard to permeability of the 
