From the results of these tests it is clear that both Halicystis and Valonia have a high degree of tolerance for animal peptone, and a very high degree of tolerance for animal proteose and for egg albumen. The products of bacterial growths fostered by these proteins have a deleterious effect upon both species of algae; but, if it were possible to prevent bacterial growth entirely and at the same time supply proper food, it is probable that Halicystis and Valonia would show normal growth indefinitely in the presence of these three colloids. This is not true where exposure is made to yeast nucleic acid dissolved in sea water containing 0.00093 gm. per cc. of NaOH. Valonia is markedly less tolerant of this medium (perhaps of NaOH rather than the colloid used) than Halicystis. Such differential effects, however, reach a high point in the case of the solutions of diphtheria toxin and of edestin. Halicystis has a very high tolerance for diphtheria toxin, and Valonia a very low tolerance. In the case of edestin, the relationship is reversed. Here Halicystis has a very low tolerance, and Valonia a very high tolerance. In fact, it may be said that diphtheria toxin has no appreciable effect upon Halicystis, and edestin a very slight effect upon Valonia; while diphtheria toxin is extremely toxic to Valonia, and edestin is extremely toxic to Halicystis. We can offer no suggestions, at present, as to the way in which these effects are produced.
It is probable that the very thin protoplasmic layer of these species, which is certainly no thicker than 8µ, is sufficient to obstruct the passage of proteins having large molecules, like egg albumen, with a degree of efficiency that is extraordinary. In the tests we have reported, areas of from 20 sq. cm. to 40 sq. cm. have been submitted to the action of a relatively high concentration of egg albumen for several days without permitting the passage of sufficient amounts to give definable tests either with Spiegler's or with Tanret's method,— presumably less than 1 part in 250,000.
In the tests of the proteins having much smaller molecules (though the size may not be the explanation), there is some probability that the membranes exhibit a little permeability. The peptone and the proteose of animal origin, or biuret-positive substances derived from them, apparently pass the protoplasmic membranes occasionally in quantities sufficient to give biuret tests. The most probable case of protein passage, however, was that of the proteose of the scarlet runner bean, where specific detection of less than 1 part per 80,000 was possible. In this instance the proteose appeared to pass membranes that were healthy and were functioning normally. But since the cells of the algae had to be destroyed in making the tests, one cannot maintain this point. All one can say is that protein passage was indicated in carefully examined cells of both species, where no breaks in the protoplasmic membrane were discernible, and where samples of the treated cells behaved normally after treatment.