A fairly pure culture may be obtained by inoculating a tube of bran infusion, previously sterilized, by means of a freshly made capillary pipette, which is pushed into the tube without removing the cotton-wool plug. The tube is placed in the incubator at 30° C., and the organisms causing the fermentation develop rapidly; as soon as the tube is observed to become cloudy, indicating the development of the bacteria, another tube is inoculated in the same manner, using a very minute quantity of liquid; this second “attenuation” is then carried on to five or six more tubes. In the fifth tube will usually be found a pure culture of the bran fermenting organisms, which I have called Bacterium furfuris. The bacteria (see Fig. 30) are mostly in the form of dumb-bells or pairs, each cell 0·75 µ × 1·25 µ: they vary slightly in size, some forming chains; those on the surface become surrounded by a kind of jelly, and form an iridescent pellicle on the surface of the liquid, the zoogloae form. When the nutriment of the liquid is exhausted or the acidity becomes too great,[154] this film sinks to the bottom and the liquid above becomes comparatively clear. There is no spore formation, and, consequently, the culture soon dies unless inoculated into a fresh nutrient medium. The liquor, after the fermentation both in the drenches and tubes of pure cultivations, is always slightly ropy. [154] Claflin (92) gives the limits of acidity for lactic bacteria from 0·02 to 0·5 per cent. Reynolds Green (97, p. 346) gives the upper limit for B. acidi lactici as 8 per cent. lactic acid. B. furfuris probably belongs to the group of coli bacteria, some of which ferment milk and other sugars with production of lactic acid and gases. This explains the ease with which the fermentation takes place, since the skins bring in immense numbers of bacteria from the bate, among which are the organisms described.