_Artificial_ immunisation generally produces the formation of so great a quantity of fixators that the phagocytes are unable to retain them and excrete them in part in the ambient humors, _i.e._ the blood plasma, or serum. When, afterwards, virulent morbid agents (microbes or figured elements) are introduced into an organism which has been immunised against them, they are at once faced, in the humors, with _fixators_, which immediately exert a biting action on them and render them sensitive to the action of the intracellular cytasis of the phagocytes. The same mechanism explains the specificity of the serums of vaccinated animals. The quantity of specific fixators in the humors depends on the surplus production of that ferment by the phagocytes and is not always the same. That is why different serums are preventive in different degrees. They are inactive if the phagocytes have not produced enough fixators to pass any out into the humors. For a serum is only preventive when it brings into the new organism into which it is injected a sufficient quantity of fixators ready to sensibilise the morbid agents afterwards introduced into the organism. The over-production of antibodies--fixators or antitoxins--corresponds up to a certain point with the frequency and quantity of vaccinal injections; that is why serums are usually preventive in artificial immunity and very rarely so in natural immunity. Through successive inoculations, the cells become accustomed to digesting the microbes, or figured elements, and manufacture, in consequence of that digestion, growing quantities of fixators. In natural conditions, on the other hand, morbid agents do not usually penetrate into the organism in massive or repeated doses; therefore digestion under natural conditions results in a less abundant production of fixators which can be contained in the interior of the phagocytes without leaking into the humors in sufficient quantities to render the latter preventive. It might be thought that immunity against pathogenic microbes is accompanied by immunity against their toxins. In reality that is not always the case, and very often the organism, now made refractory to certain microbes, remains sensitive to their toxic products. Thus antimicrobian immunity and antitoxic immunity constitute in most cases two distinct properties. In order to confer antitoxic immunity recourse must be had to vaccination by soluble poisons and toxins.