152 
Walter Stiles 
Roaf and Webster (1912) that if a cell is washed with a solution 
either free from, or very poor in, a particular ion, the amount of this 
ion washed out from the cell may be practically inappreciable, for 
it will only be washed out until the concentration of the ion in the 
external solution bears to the concentration of the ion in the cell the 
relation required by the adsorption equation or molecular affinities, 
and under these circumstances the equilibrium concentration of the 
ion in the external solution may be very low indeed. 
In supporting the membrane theory, Bayliss (1915) deals par¬ 
ticularly with the view that the apparent impermeability of cells to 
some salts can be sufficiently accounted for by the presence of a 
membrane semi-permeable as regards colloids, but permeable to 
crystalloids. This view does not differ greatly from that of Moore 
and his collaborators already considered. To such a view Bayliss 
raises the following objections. Firstly, as, if a membrane is permeable 
to one ion of an electrolytically dissociated salt and not to the other, 
it is impossible for the salt or either ion to pass through the membrane 
(see Chapter V), if we have a salt of a protein which is electrolytically 
(but not hydrolytically) dissociated, and the protein ion is incapable 
of passing through the membrane, the other ion of the compound 
will also be incapable of doing so. But if there are two colloidal salts 
present, in one of which the kation is diffusible and in the other the 
anion, as there must presumably be if a neutral salt is taken into the 
cell and both ions so held, then both ions will diffuse readily out from 
the cell, so that in such a case a hypothesis not involving a membrane 
exhibiting complete or partial semi-permeability to such salts will 
not explain the observed facts. Secondly, Bayliss thinks there is no 
satisfactory evidence for the presence of compounds between proteins 
and neutral salts. Thirdly, glucose does not form a compound with 
proteins of the form required by the hypothesis, but exists free in 
the blood according to Asher (1912), and is apparently indiffusible 
into the corpuscles under normal conditions. Fourthly, the high 
osmotic pressure of certain cells cannot be given by substances of 
such high molecular weight as proteins, and seems only accountable 
by the presence of substances of comparatively low molecular weight 
in an osmotic cell. Finally, difficulties arise with regard to the dis¬ 
tribution of different ions within and without the cell if a membrane 
more or less impermeable to crystalloids is dispensed with. It has 
been shown both by Bayliss (1911) and Donnan (1911) that it is 
possible to have different concentrations of a salt on the two sides 
of a membrane through which the salt can diffuse freely, but that in 
