1923] S EI FRIZ— SALTS 



397 



II. Antagonistic effect of salts on toxicity of alcohol 

 Experiments by Loeb and Osterhout have shown that salts, 

 such as sodium chloride and calcium chloride, which are toxic to 

 protoplasm when alone, are not toxic when in combination. These 

 two salts when in solution together represent a typical case of what 

 is known as salt antagonism. According to the experimental work 

 of Osterhout, sodium chloride produces an increase in permeability, 

 while calcium chloride produces a decrease. Consequently, when 

 the two salts are together in proper proportion, that is, in a phys- 

 iologically balanced solution, the normal diosmotic properties of 

 protoplasm are not disturbed. 



In some experiments on the reaction of protoplasm to alcohol 

 (8), the writer found that ethyl alcohol in low concentrations is 

 highly toxic to the protoplasm of the cells of the Elodea leaf, and 

 that the osmotic value of these cells is greatly lowered by treatment 

 in dilute ethyl alcohol, due to increased permeability and conse- 

 quent exosmosis of cell contents. If sodium also increases perme- 

 ability, as Osterhout states, then one would rather expect it to 

 add to the toxic effect of alcohol. Interpreted in terms of effect 

 on osmotic value, as determined in these experiments, one would, on 

 the contrary, expect sodium, which was found to raise osmotic 

 pressure, to oppose the toxic effect of alcohol, since alcohol lowers 

 the osmotic value of the cell. This latter assumption proved to 

 be true. On the other hand, calcium, if it decreases permeability, 

 should oppose the dispersing effect of alcohol. On the basis of 

 effect on osmotic pressure, one would also expect antagonism 

 between calcium and alcohol, but to a much less extent than in the 

 case of sodium. In reality, calcium in solution with ethyl alcohol 

 has no apparent effect whatever on the toxicity of alcohol. 



European investigators have objected from the outset to the theory that "all the 

 bivalent cations" are able to decrease permeability, while all monovalent cations 

 (except H) increase permeability. The objections have been based in part on 

 their own experiments and in part on lack of confidence in the electrical conductivity 

 method of determining permeability changes. It is impossible to tell conclusively that 

 the changes in conductivity observed by this method are actually due to changes in 

 permeability. The selective permeability hypothesis of the early investigators appar- 

 ently has come into its own again. The valency hypothesis, while receiving substantial 

 support in man}' instances both in the non-living and the living world (for example, 

 in "antagonism"), cannot be regarded as a theory of universal applicability. 



