Faithful to his method of taking as a starting-point the simplest expression of the phenomenon to be studied, Metchnikoff began by lower beings. Unicellular organisms, such as myxomycetes, amoebæ, and infusoria, sometimes manifest a natural immunity to certain poisons. It is also possible to endow them with artificial immunity by accustoming them gradually to substances which, ingested straight away, would infallibly have killed them. Such phenomena, seen in unicellular beings, could only be ascribed to the reaction of the cell itself. Therefore Metchnikoff supposed _a priori_ that the phagocytes, being similar primitive cells of multicellular beings, would also react against poisons. And, in fact, he ascertained that the number of phagocytes in a rabbit's blood diminished considerably under the influence of a fatal dose of arsenic, whilst it increased under the influence of small doses of the poison, to which it was possible to accustom the animal. Dr. Besredka, a disciple of Metchnikoff, made some very interesting researches, which entirely confirmed the share of the phagocytes in the reaction against sulphides of arsenic. He had chosen the trisulphide, a very slightly soluble salt of an orange colour, in order to find it again easily within the organism. After having injected non-fatal doses of it into the peritoneal cavity, he obtained an exudate in which all the orange granules of the salt were to be found included within those leucocytes which have a large, non-lobed nucleus--the _macrophages_. These cells gradually digested the salt they had englobed, which ended by disappearing entirely within them, and the rabbit remained safe and sound. On the other hand, it died if the same doses of the same salt had been protected from the leucocytes by an elderberry bag, or when the leucocytes had been attracted elsewhere by a previous injection of carmine for instance. Those experiments removed all doubts as to the share of the phagocytes in the destruction of mineral poisons. Certain experiments on _microbian_ poisons spoke in the same sense. Thus MM. Roux and Borrel had observed that the diphtheritic toxin, which is inoffensive to rats even in large doses, kills that animal if a small quantity of it is introduced into the brain, the probable explanation being that, in cases of subcutaneous injections, the poison, "phagocyted" on the way, was destroyed before it reached the nerve cells. Thus experiments seemed to plead in favour of the view that the part played by phagocytosis is not limited to the struggle against microbes, but also extends to the defence against poisons and toxins. * * * * * After having studied the mode of destruction of these, Metchnikoff wished to elucidate the origin of the counter-poisons, the specific antitoxins discovered by Behring in the humors of immunised organisms, a question of which the study was even more difficult.