In 1895, Denys and Leclef1 first called attention to the effect of blood serum on phagocytosis by leucocytes. They were the first to produce reliable experimental evidence that substances which alter microbes in such a way as to permit their being ingested by the leucocytes exist in the blood serum of immunized rabbits. From careful experiments in vitro they concluded that in an immunized rabbit the leucocytes obtain their power of engulfing the bacteria from some other property of the serum. The immunized animal fights the bacteria, first, by the direct action of its serum; second, by its leucocytes.2 The latter always owe to the serum the commencement of their power.
In 1897, Bordet3 was unable to confirm their results. In the same year Mennes4 showed that the immunity of guinea-pigs inoculated with toxins or cultures of pneumococci depends on a modification of their serum whereby