J. Loeb, in a recent and stimulating work (1), has given a convincing, if somewhat over-emphatic, study of the colloidal behaviour of proteins in solution, based largely upon the theory of the Membrane Equilibrium first suggested by Donnan (4). In one important particular, however, his argument is incorrect. Loeb observed, by certain means (2) devised by himself, the potential difference (P. D.) between a protein solution on one side of a semipermeable membrane and a solution of acid, or of acid and salt, on the other side. He found this P. D. to vary as the concentration of hydrogen ions, or of salt, was varied, in the same manner as did a number of other factors (osmotic pressure, viscosity and swelling). He found also that this P. D. could be “calculated” from the observed difference of
ρ
-H (or of
ρ
-Cl) in the two solutions, on the basis of the theory of the Donnan Equilibrium, and he concludes that the excellent agreement between calculated and observed is a strong argument in favour of his explanation of other colloidal phenomena by that theory. This conclusion is not correct: the equality found by Loeb of the observed P. D., to that calculated from the difference of
ρ
-H is a necessary consequence of
any
mechanism which does not offend the Second Law of Thermodynamics, and in itself offers no support to the theory that the Donnan Equilibrium underlies the colloidal behaviour of protein solutions. That theory may rest on other and stronger ground; since, however, Loeb appears, throughout his book (and especially in Chapters VIII and IX) and in other places (2), (3), to lay great emphasis on this agreement of the observed P. D. with that “calculated” from the observed
ρ
-H’s it is necessary to point out that this agreement proves no more than that the system investigated was in equilibrium, and that the observations were accurately made.