In view of these pronounced physiological properties of the proteoses, it is interesting to recall the now well-known fact that many of the chemical poisons produced by bacteria are proteose-like bodies, chemically, at least, closely akin to the proteoses resulting from pepsin-proteolysis. Thus, Wooldridge[158] as early as 1888 pointed out that an alkaline solution of tissue-fibrinogen exposed to the action of anthrax-bacilli suffered some change, so that when introduced into the blood it possessed the power of producing immunity to anthrax. This observation was verified by Hankin,[159] who further showed that the substance formed by the anthrax-bacilli was a veritable albumose, and that it truly possessed the power of producing immunity. Sidney Martin[160] carried the matter still further, and by growing the anthrax-bacilli in a pure solution of alkali-albuminate prepared from blood-serum, proved the formation of both primary and secondary albumoses, as well as of peptone, leucin, tyrosin, and a peculiar alkaloidal substance of pronounced toxic properties. Martin finds that the albumoses are not as poisonous as the alkaloid, and surmises that the alkaloid is contained in the albumose molecule in the nascent state; further, he suggests that the albumoses in small doses may exert some protective influence, while in larger doses they act as vigorous poisons. How true this may be I cannot say, but my own experience convinces me that the anthrax-bacilli grown in a culture medium composed of alkali-albuminate, prepared from egg-albumin, to which the necessary inorganic salts and some glycerin have been added, do give rise to albumoses and peptones which are truly endowed with toxic properties. [158] Versuche über Schutzimpfung auf chemischem Wege. Du Bois-Reymond’s Archiv f. Physiol., 1888, p. 527. [159] British Med. Journal, October, 1889. [160] Proceed. Royal Society, 1890, vol. 48, p. 78. Albumose-like bodies have also been obtained by Brieger and Fränkel[161] with the bacillus of diphtheria. These, too, were endowed with powerful poisonous properties, and when introduced into the tissues of the body gave rise to reactions resembling those produced by the Löffler bacillus. In my own laboratory, recent experiments made with the bacillus of glanders have shown that when grown in a slightly acid medium containing alkali-albuminate, albumoses, peptones, and crystalline bodies such as leucin and tyrosin are formed in considerable quantities. Kresling[162] has reported similar results. With the tubercle-bacilli, many like results have been recorded. Thus, among others, Crookshank and Herroun[163] have reported the finding of albumoses, peptone, and a ptomaine when the bacilli have been grown in glycerin agar-agar, and also in fluid media. [161] Untersuchungen über Bacteriengifte. Berlin, klin. Wochenschrift, 1890, p. 241 and 268.