the work of Toussaint, Chauveau, Wooldridge, Chamberland, and Roux on anthrax; of Charrin on the pyocyanic bacillus disease, of Chamberland and Roux on acute septicæmia, etc.; of Brieger, Chantemesse, and Vidal on typhoid fever; of Roux on symptomatic anthrax, and of Roux and Yersin on diphtheria. In most of these experiments the material used for inoculation was the cultivation medium modified by the growth of the organism, and sterilized either by heat, by filtration, or by both methods. The work of Charrin, Woodhead, Cartwright, and Wood has also shown that protection may sometimes be obtained not only by injection of the products of the growth of the pathogenic organism itself, but also of some quite different ones (Bacillus anthracis and bacillus of blue pus).
The products used were therefore of a very complex nature, and it was not known to what kind of compound they owed their property of conferring immunity. Roux and Yersin had, in 1888, tried to prove that their chemical vaccine for diphtheria owed its properties to an albuminoid body allied to unorganized ferment, but this last supposition is not generally accepted, although not disproved. (In order to understand the origin of the following improvements it is important to remember that the work of Panum (1856), Gautier and Selmi (1873) had revealed the production of very poisonous alkaloidal substances during putrefaction. The more accurate researches of Nencki, and still more of Brieger, demonstrated clearly the existence of an important class of poisonous alkaloids produced by the micro-organisms of putrefaction. Gautier (1881), on the other hand, was trying to prove that animal tissues are also capable of producing by their metabolism poisonous substances of allied nature. The experiments of Lauder Brunton and Sir Joseph Fayrer on cobra poison (1873) should be kept in mind in relation with this subject. It was soon found that, besides these poisonous albuminoids, other more or less poisonous products might be manufactured either by animal or vegetable cells; these products were found to belong to the ill-defined class of albumoses. I need only refer to the work of Weir Mitchell (1860) and Reichert on the albumoses of snake poison; of Sydney Martin on phytalbumoses—i. e., albumoses produced by vegetable cells, whether bacterial or others an important work, which led him to infer later on that albumoses were products intermediate between the non-poisonous albuminous substances of the culture media and the most poisonous alkaloids. Büchner, Wooldridge, Hankin, and others were also discovering toxic albuminous substances in various fluids or tissues of the body, some of which were deadly to bacteria.) Returning now to preventive inoculation, we find that in 1889 Sydney Martin in London, and Hankin of Cambridge, working in Koch's and Brieger's laboratories, had
