Besides B. coli, there are several species of bacteria which liquefy gelatin, and a number of facultative organisms, whose presence is more or less accidental. By changing the food, and introducing with it quantities of foreign organisms, the composition of the intestinal flora may be changed. By introducing for a considerable period B. coli anindolicum, Lembke succeeded in entirely suppressing B. coli commune. On returning to normal feeding, the foreign organisms in some cases entirely disappeared. The researches of Dr. H. Becker[80] were applied more directly to the use of bacterial cultures for the bating of skins, and to the elucidation of the bacterial action of dog-dung infusions. He isolated 54 varieties of bacteria from dog-dung, and tried the action of pure cultures of many of them on a skin. [80] Zeit. f. Offt. Chem., Heft. xxiii. Jahr. X. p. 447. A list of the various bacteria isolated by Becker is given in tabular form on pp. 98–101. [Illustration: Fig. 16.--B. Erodiens (Becker).] [Illustration: Fig. 17.--Plate Culture from Fresh Puer.] Professor Becker’s Bacterium No. 12, which he has named Bacillus erodiens (Fig. 16), is undoubtedly a variety of B. coli, but has a more rapid motion, and does not coagulate milk, although it renders it somewhat thick. Cultivated in broth it gives off much gas, consisting of 12 per cent. carbon dioxide, 85 per cent. hydrogen, 3 per cent. oxygen. If glucose be added, the quantity of carbon dioxide rises to 40 per cent., and acid is produced. The most rapid growth is at 37° C., and at this temperature a broth culture has a distinct reducing action on skin. According to the medium in which it is grown, it produces acid or alkali, and thus comes under the heading of mixed bacteria. In sugar[81] solutions acid is produced, and in proteid solutions ammonia compounds, indol, and evil-smelling gases are given off. Thus, by varying the medium, the effect produced may also be varied. [81] Dr. A. C. Houston has been kind enough to make an examination of B. erodiens, and to ascertain its action on various sugars. It produces fluorescence in neutral-red broth cultures, acid and gas in lactose cultures, indol, and acid, and clot in milk cultures. It ferments dulcite with production of acid, but _not_ cane sugar, adonite, inulin, inosite, salicin, or raffinose. B. erodiens does not secrete any tryptic enzymes, hence its action on the skin is to be attributed either to an intracellular enzyme, or to its chemical products, which, being secreted _in situ_, have a more favourable and powerful action than if merely added to the bating liquid. It was for this reason that I proposed to use a mixed culture of bacteria, especially bacteria from the sweating process (see p. 105), which secrete a mild form of proteolytic ferment, capable of dissolving the more easily soluble portion of the skin fibres (or certain constituents), but not capable of attacking the hyaline layer.