1. It seems first of all clear from our results that the effect of electrolytes upon electrophoretic charge is essentially the same, whether one is dealing with silica dust, bacteria, or yeast cells, although certain quantitative differences appear which will later be discussed.
2. The normal negative charge on the suspended particles appears to be slightly increased by very low concentrations of electrolytes, markedly so in the case of yeast cells. Increase in charge due to minimal concentrations of electrolytes has been recorded by Loeb (1922) for collodion particles.
3. Higher concentrations of electrolytes cause a marked and progressive decrease in negative charge, sometimes leading to an isopotential condition and sometimes to a complete reversal of charge with active migration toward the cathode. This effect is apparently due to the cation alone and increases with the valency of the cation, except that the H ion shows specially marked activity, between that of bivalent and trivalent ions. Since NaOH behaves like an ordinary univalent salt, increased alkalinity of a solution does not further depress the charge already depressed by salts; but, since the H ion is much more active than other univalent or bivalent ions, increased acidity does cause a further progressive depression of charge, even in salt solutions. Certain electrolytes appear to show individual peculiarities due to something else than their valency. Thus KCl for example is distinctly more effective than NaCl. Sodium chloride in general appears to exert less influence upon electrophoretic charge, either in low or high dilution, than do other compounds of univalent ions studied.
This depressing effect of moderately high concentrations of electrolytes is much less marked with yeast cells than with Bacterium coli. Silica dust is still less affected by monovalent and bivalent ions than are the yeast cells but appears to be more affected than either yeast or Bacterium coli by AlCl3.
4. Very high concentrations of AlCl3 (above 10–2 M) show a third effect, a decrease of the positive charge produced by concentrations of moderate molar strength. This is analogous to phenomena observed for trivalent salts by Northrop and De Kruif (1921–22) and for acid by Winslow, Falk, and Caulfield (1923–24).
5. Organic substances, such as glucose, glycerol, and saponin produce no effect on electrophoretic velocity until they reach a concentration at which viscosity changes are involved.
6. The first two results observed,—(a) the increase in charge as a result of slight additions of electrolytes, and (b) the marked decrease in charge with further concentration of electrolytes, depending on the valency of the cation, so far as vegetable cells are concerned, are entirely in accord with the theory of the Donnan equilibrium as worked out by Loeb (1922).
We might assume in explaining such phenomena that the plant cell contains a certain proportion of unbound protein material and that the first modicum of cation which enters the cell is bound by the protein, leading to an increase in the relative negative charge of the cell as compared with its menstruum, while subsequent increments of cation remain unbound in the cell and thus lower its charge. When we find, however, that the same phenomena are apparent with collodion particles, as shown by Loeb, and with silica dust, it seems difficult to apply such a theory, involving the conceptions of a permeable membrane and unbound organic compounds. Loeb (1923–24) suggests that the primary increase may be due to an aggregation of anions in the part of the electrical double layer adjacent to the suspended particles; but why there should be first an aggregation of anions and later (with increasing concentration) an aggregation of cations, is not easy to conceive.
The third result,—the reversion to a more negative charge in the presence of a marked excess of trivalent ions,—is again difficult to explain. Loeb, in this connection, postulates the existence of complex ion-protein compounds, which can scarcely be assumed in the case of the silica particles.