Permeability 241 
to happen when a gelatin or agar sol sets to a gel. In this way a 
substance which could not diffuse through the continuous phase owing 
to insolubility in it might be able to diffuse through the discontinuous 
phase but would not be able to enter the cell because of the discon¬ 
tinuity of the disperse phase. With inversion of the phases, penetration 
would then take place at once. The hypothesis would be equally appli¬ 
cable to a solution or a chemical combination theory of permeability. 
Clowes produces support for his theory from experiments on the 
effects of sodium hydroxide and calcium chloride in producing phase 
inversion in emulsion systems of olive oil and water, which are com¬ 
pared with the effects of these substances on protoplasmic permeability. 
Free (1918) has subjected this hypothesis to some criticism. He 
points out very rightly that such a hypothesis suggests that per¬ 
meability changes would be sudden, and even if it were assumed that 
the protoplasm (or plasma-membrane) were in a state of mobile 
equilibrium so that parts of it were in one state and other parts in 
the inverse condition, with frequent alteration in the relative quantity 
of the two parts which is the state of affairs Clowes appears to suggest, 
the balance would in all probability be very easily upset so that the 
whole of the protoplasm would go to one or other of the conditions. 
A further weakness of the hypothesis is the absence of any direct 
experimental evidence in its support, for the bearing of experiments 
with olive oil-water emulsions on cell permeability are, as Free 
suggests, rather dubious, as such emulsions are not generally present 
in cells. 
Lloyd and Free’s Colloidal Hypothesis 
The hypothesis of protoplasmic permeability put forward by Free 
(1918) and earlier suggested by Lloyd (1915) bears certain points of 
resemblance to the two hypotheses just discussed. These writers also 
start out from the supposition that the protoplasm consists of an 
emulsoid colloidal system. It is further supposed that two (at least) 
of the liquid phases of protoplasm differ importantly only in the pro¬ 
portions of water which they contain. Changes in permeability are 
supposed to be due to changes in the distribution of water between 
the continuous phase (dispersion medium) and the discontinuous 
(disperse) phase. When the globules of the disperse phase are large 
the spaces between them will be small, and vice-versa, and a substance 
diffusing through the continuous phase will travel faster or slower 
according to the dimensions of the spaces. Solubility in the disperse 
phase alone, and not in the dispersion medium, would, of course, not 
permit the penetration of a substance, 
Phyt. xxii. 5. 16 
