When the wort had cooled down in the flask and matters were arranged as represented in Fig. 75, I raised the two-necked flask so as to cause a little of the liquid and yeast to flow into the wort. Thereupon fermentation was set up, and the resulting carbonic acid gas made its escape by the drawn-out end of the doubled-necked flask. The entire arrangement with its supporting stand remained in this connection for eighteen months, sometimes on a stove, sometimes in the laboratory, exposed to all the variations of external temperature. At the end of that time I tasted the beer in the flask; it was perfectly sound, and the ferment, submitted to the microscope, showed not the slightest trace of any foreign ferments: and, doubtless, the experiment might have been protracted over any number of years with the same result. The only change that occurs in course of time is the appearance in the neck of the flask at the surface of the beer of a deposit of small prominences resembling a crystallization, but which really consists of those forms of ferment to which in Chapter V. I attached the name of _aërobian ferment_. The beer, after being transferred to a bottle that had been washed with hot water, was kept for several months in the heat of summer, without exhibiting the slightest trace of deterioration. [Illustration: Fig. 76.] [Illustration: Fig. 77.] The essential conditions of the preceding experiment can readily be realized on the large scale. For this purpose we may employ the apparatus in the above sketch (Figs. 76 and 77) constructed of tin or tinned copper. As appears from the sketch, this consists of a cylindrical tub resting on a support, and closed at the top by a cover, whose lower edge fits into a gutter containing water. The wort prepared in the copper is led into the cylinder, a process which does not materially lower its temperature. Now we know that wort in breweries which has been cooled in contact with the air, and so got charged with disease-germs, will, nevertheless, recover its faculty of keeping for any length of time in pure air, if we again raise its temperature to 80° C. (176° F.) or even 70° or 75° C. (158°, 167° F.) Having filled the tub with the hot wort and put on the lid, we then connect, by means of a caoutchouc tube _c d_, the metal tube _a c_ (which opens into one of the tubulures projecting above the lid) with the system of tubes _d e_, _f g_, of which _d e_ is fixed to the cylinder; _e f_ is a caoutchouc junction connecting _e_ with the bent glass tube _g_. We then dash over the apparatus, lid, tubulures, and their corks a quantity of boiling water. This collects in the gutter in which the lid rests, and any excess overflows into a second gutter outside the first, where, however, it cannot remain, but passes away by means of a ring of small holes between the base of the outer trough _i i_ and the cylinder. The overflow is collected in a third trough at the bottom, whence it can be removed by a pipe M.