Similar reasoning applies to the bromides. We have long known of the hypnotic effects of the bromine salts and the specific effects of the bromides in epilepsy. These effects are due to the Br-ions, and one salt is as good as the other, provided it yields the same number of Br-ions, and its good qualities are not offset by a deleterious action of the kations. Experience with the bromides, moreover, brings to light the fact that it is indeed the Br-ions that determine the desirable effects of bromine compounds. Clinicians have long been acquainted with the fact that organic compounds containing bromine do not produce the effects given by the inorganic salts. This is because these organic compounds containing bromine do not yield any Br-ions at all, or because they yield only such small quantities as to be without effect in the doses administered. The same facts explain why manufacturers have been unsuccessful in producing an organic compound of bromine which could at all rival the ordinary inorganic salts.

The theory of electrolytic dissociation also explains why iron salts have certain general characteristics possessed by no other salts, and why the salts of Hg, Ag, Pb, Cu, etc., are classed in groups by themselves. In these instances, however, the characteristic activity of the salt is determined by the kations, for the effects of the Hg-ions, Pb-ions, etc., evidence themselves long before the effects of the anions spring into prominence. In the cyanides, again, the anions are the effective agents and determine the characteristics of their group. KCN and HCN show similar effects, perhaps, because they both yield CN-ions, and these manifest their effects in doses so small that sight is lost of the K- and H-ions.

Several years ago Dreser7 showed that the toxic effects of mercury salts are determined by the number of Hg-ions they yield upon solution. When mercuric chloride is added to albumin a precipitate is formed which can be readily dissolved in sodium thiosulphate, forming a so-called complex mercury salt. When the mercury exists in this complex form it loses its toxic properties, and even though equal weights of the metal be present, the complex salt is unable to inhibit fermentation; and frogs, fishes, etc., poisoned with it instead of the sublimate die more slowly. Dreser finds an explanation for these phenomena in the fact that the double salt is either not dissociated at all, or yields only a small number of Hg-ions. In cold-blooded animals the salt is slowly decomposed, and the toxic effects of Hg-ions formed poison the animal. In warm-blooded animals the decomposition occurs much more rapidly, and in consequence not much difference was found between the toxic effects of the mercuric salt and its more complex derivative. Yet all local irritative manifestations were lacking in the latter case.

There have recently appeared upon the market various organic compounds of silver (protargol, nargol, etc.) which have come into general use as substitutes for silver nitrate. It seems that these compounds exhibit all the beneficent and only a few of the deleterious qualities of silver nitrate. Undoubtedly an explanation similar to that given by Dreser holds here too. Silver nitrate owes its specific action to the Ag-ions it yields. The organic silver compounds probably yield none or only a small number of such ions. When, however, the organic compound is introduced into the body, it is decomposed, and the Ag exerts its specific effects to a degree dependent upon the number of Ag-ions liberated. These facts explain the differences in the behavior between the organic and the inorganic silver salts.

Paul and Kronig,8 and, more recently, Scheurlen and Spiro,9 have been able to show that the bactericidal power of solutions of electrolytes is dependent upon the ions contained in them. Equi-molecular solutions of mercury salts arrange themselves according to their degrees of electrolytic dissociation in the following order: HgCl2, HgBr2, Hg(CNS)2, Hgl2, HgCy2. When arranged according to their bactericidal powers, the order is the same. This power is then dependent upon the number of Hg-ions contained in the solution. HgCl2, which contains the largest number, has the strongest germicidal action, while HgCy2, which is least dissociated, has the feeblest. So weak is the action of the cyanide that at a concentration four times that of a bichloride solution capable of destroying all cocci and spores it permits the development of several thousand colonies of the staphylococcus and many colonies of the anthrax bacillus. If K-ions are substituted for the Hg-ions by the substitution of KC1 for HgCl2 in the antiseptic solution, the germicidal powers of the solution are decreased, another fact which proves that the Hg-ion is the specific germicide. These facts effectively dispose of the conception of Behring, still held by many, that the bactericidal power of a mercurial is dependent upon the amount of mercury contained in it, and is independent of the nature of the compound.

The germicidal effects of silver and gold salts are similarly found to be dependent upon the Ag- and Au-ions. That it is, indeed, the ions which are thus effective is proved by the fact that solutions of HgCl2 or AgNO3 in absolute alcohol or ether (solvents in which but slight dissociation occurs) have no deleterious effect upon anthrax spores.

Recently Loeb10, 11 has pointed out the influence of the valency and possibly the electrical charge of ions upon their toxic and antitoxic effects. Previous experiments had brought to light the poisonous character of a pure sodium chloride solution for the development of fish embryos, or on the beat of the heart. But these toxic effects are done away with when a small amount of calcium is added to the sodium chloride solution. Thinking that these were only special instances of a more general law, Loeb investigated the toxic and antitoxic effects of ions upon the development of the eggs of Fundulus, a marine fish. The eggs of this fish develop equally well in sea-water (their ordinary habitat), in distilled water, or in sea-water the concentration of which has been raised by the addition of NaCl. In a pure sodium chloride solu tion, however, of the same concentration as that of the sea-water, not a single embryo develops. If, now, a small though definite amount of a calcium salt be added, the poisonous effect of the NaCl solution is annihilated, and the eggs develop into embryos. Not only is calcium able to bring about this effect, but any bivalent kation serves the same purpose - Ca, Ba, Mg, Fe, Co, and even Zn and Pb. The nature of the anion is immaterial.