We therefore see that what at first sight appeared in the experiments of Herbst a necessity, namely, the presence of each constituent of the sea water, turns out as a special case of a more general law; the salts with univalent ions are injurious if their concentration exceeds a certain limit and this injurious action is diminished by a trace of a salt with a bivalent cation. Why was it not possible to prove this fact for the eggs of the sea urchin? Before we answer this question, we wish to enter upon the discussion of the nature of the injurious action of a pure NaCl solution of a certain concentration and of the annihilation of this action by the addition of a small quantity of Ca. The writer suggested in 1905 that the injurious action of a pure NaCl solution consisted in rendering the membrane of the egg permeable for NaCl, whereby the germ inside the membrane is killed; while the addition of a small amount of Ca (or any other bivalent metal) prevents the diffusion of Na into the egg,[265] possibly, as T. B. Robertson[266] suggested, by forming a precipitate with some constituent of the membrane, whereby the latter becomes more impermeable. The correctness of this idea can be demonstrated in the following way. When eggs of _Fundulus_, which are three or four days old and contain an embryo, are put into a test-tube containing 3 m NaCl they will float on this solution for about three or four hours; after that they will sink to the bottom. Before this happens the egg will shrink and when it ceases to float the embryo is usually dead. This is intelligible on the assumption that the NaCl solution entered the egg, increased its specific gravity so that it could not float any longer and killed the embryo. When we add, however, 1 c.c. 10/8 m CaCl₂ to 50 c.c. 3 m NaCl the eggs will float, the heart will continue to beat normally and the embryo will continue to develop for three days or more, because the calcium prevents the NaCl from entering into the egg.[267] For if we put a newly hatched embryo into 50 c.c. 3 m NaCl+1 c.c. 10/8 m CaCl₂ it will die almost instantly; hence the membrane must have acted for three or more days as a shield which prevented the NaCl from diffusing into the egg in the presence of CaCl₂. [265] Loeb, J., _Arch. f. d. ges. Physiol._, 1905, cvii., 252. [266] Robertson, T. B., _Ergeb. d. Physiol._, 1910, x., 216. [267] Loeb, J., _Biochem. Ztschr._, 1912, xlvii., 127. The same experiments cannot be demonstrated in the sea-urchin egg, first, because it can live neither in distilled water nor in very dilute nor very concentrated solutions; and second, because it is not separated as is the germ of the _Fundulus_ egg from the surrounding solution by a membrane which is under proper conditions practically impermeable for water and salts.