Notwithstanding these untoward results in other hands, I am still confident that we have in this a measure of no small value, and worthy of application in suitable cases. A certain percentage of failures in immunization are to be looked for. Even cowpox vaccination is not always protective against itself; I knew one man who was successfully vaccinated every three years in a comparatively long series. Many habitually self-limiting diseases relapse in particular individuals. Even in the case of anthrax excess of glucose in the system, and the lack of some unknown influence of the spleen are respectively destructive of immunity. Even after the Pasteurian inoculation a certain number of inoculated animals are lost, it may be between the first and second injection, or it may be “two or three months” after the latter (Galtier). We must also bear in mind that in an infected herd or flock, there are always a certain number already infected at the time of the protective inoculation, and as the protective conditions are slowly established, through the action of the leucocytes, it is unreasonable to expect that serious illness and death can be obviated in such animals. In inoculation with the Pasteur lymph, the bacillus is held not to enter the blood, a position supported by the researches of Bitter, Perroncito, Wissokovicz, Lubarsch, Metchnikoff, Chamberland and Roux, so that the resulting immunization must come from the toxins. Add to this that Chauveau (1885), conferred immunity on a sheep by injecting intravenously, anthrax blood, defibrinated and sterilized by heat: Arloing obtained immunity in the sheep by injecting, subcutem, the clear supernatant liquid from old bouillon cultures of anthrax, from which all bacilli had been precipitated. Roux and Chamberland obtained the same result by using the pulp of an anthrax spleen treated with essential oil of mustard, so as to destroy the life of the bacillus, and then evaporated in vacuo to remove the essence. Smaller doses proved effective, than when the splenic pulp had been filtered or sterilized by heating to 58° C. The advantages of using sterilized toxins are numerous:— 1st. As the material can be derived from a case of the outbreak in hand, there is no risk of using the anthrax protective inoculation for black quarter, hæmorrhagic septicæmia or other disease which is so often confounded with it. 2d. There is no danger of the sudden enhancing of the potency of a weakened microbe on account of some condition of the animal inoculated, as no living microbe is employed. 3d. There is no possibility of planting the anthrax bacillus, on new soil, as is so liable to take place in using the weakened microbe.