ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 127 
ions restore the norma! state of semi-permeability. In this connec- 
tion, the experiments of Clowes (1916), referred to in the preceding 
section, are significant. We have seen that the constituents of the 
cell membrane are most probably fatty substances, since they lower 
surface tension, together with concentrated solutions of emulsoid 
colloids in water, perhaps in the gel state, but certainly forming a 
watery phase. In other words, we have a system similar to that of 
oil and water in the experiments of Clowes, but more complex. 
Under the influence of sodium salts, Clowes’ emulsion was one of oil 
drops in a continuous watery phase and therefore permeable to 
water and to solutes therein. On the other hand, under the influence 
of calcium salts, a phase reversal occurred, so that there was then an 
emulsion of water drops in a continuous oil phase. Such a system 
would be impermeable to water, but permeable to substances soluble 
in oil. A change of this latter kind, if complete, would not give us 
the properties which the normal cell membrane possesses. It would 
be impermeable to water, as well as to solutes in water. These 
solutes would not be able to manifest their osmotic pressure and the 
volume of the cell would not have any relation to the osmotic 
pressure of an external solution, as we find that it actually has. 
Incidentally, however, it appears from some experiments by Lillie 
(1917), on the changes of permeability in fertilized egg cells, that, 
under exceptional conditions, the production of a “ waterproof” 
membrane may be possible. But what we have to explain is the 
change from a membrane permeable to-salts, sugar, &c. to one im- 
permeable to them, while remaining permeable to water. It is clear 
that the “Clowes effect,” as we may call it, must be only partially 
effected. The way in which Bancroft explains the mode in which it 
‘takes place is summarized by Clowes as follows:—‘The soaps 
present in the system tend, as stated above, to concentrate at the 
interface between water and oil and to form a coherent film. Soaps 
of monovalent cations, being readily dispersed in water but not in 
oil, form a film or diaphragm which is wetted more readily by water 
than by oil, consequently the surface tension is lower on the water 
than on the oil side. Since the area of the inside face of a film 
surrounding a sphere is obviously smaller than that of the outside 
face, the film tends to curve so that it encloses globules of oil in 
water, in this manner reducing the area of the side of higher surface 
tension to a minimum as compared with that of lower surface tension. 
On the other hand, a film composed of soaps of divalent or trivalent 
cations, being freely dispersed in oil but not in water, is wetted more 
readily by the oil than by the water, the surface tension is lower on 
the oil than on the water side, and the film tends to curve in sucha 
manner as to enclose the globules of water in an outer or continuous 
oil phase.” The process is dependent, on the presence of a film of 
soap between the phases and it is interesting to note the powerful 
effect of sodium oleate in destroying the membrane of the red blood 
corpuscles, while pure olein has not this effect. 
During the process of phase reversal, as described above, there is 
a stage in which one of the two phases is drawn out into elongated 
drops with narrow films or channels of the other phase between 
them. This is shown in Figure 2 of the paper by Clowes. Now, if 
