By these different means, which are employed separately or combined with one another, we generally manage to obtain the yeast which we wish to purify in a very pure state. We need scarcely add that it is always well, in the case of our purifications, to begin with specimens which are already as pure as it is possible to obtain them. In making our choice the microscope is our best guide, but it is not a sufficient one. We should be strangely deceived if we believed in the purity of a yeast for the sole reason that when examined under the microscope it appeared to contain nothing of a foreign nature. The best means of assuring ourselves of the purity of a yeast consists in making some beer in one of our two-necked flasks, and leaving this flask, after fermentation, in an oven at a temperature of 20° or 25° C. (68° to 77° F.). If the beer, in the course of a few weeks, does not thicken, or become covered with efflorescence, if its deposit is microscopically pure, if, in short, it only tastes flat, we may have every confidence in the purity of the yeast which produced it. After we have purified a yeast we are, unfortunately, never sure that it has not undergone some change in the course of the manipulations to which it has been subjected in purification. It is indispensable, therefore, that we should test it, and see if the flavour of the beer produced by it is really the one that we want—viz., that of the beer from which we took the yeast that we submitted to purification. In the course of a series of practical experiments that we were carrying out in the large brewery of Tourtel, at Tantonville, in 1875, in connection with the new process of brewing, which will be explained in Chap. VII., the following circumstance occurred. We had purified some of the yeast of the brewery, by means of a succession of growths and adding a few drops of phenol, and had obtained a yeast of irreproachable purity. It happened that this yeast, which was repeatedly cultivated in the brewery during the summer of 1875, from six to ten hectolitres (130 to 220 gallons) of wort being used on each occasion, always produced a beer that had a yeast-bitten flavour and defective clarifying powers, notwithstanding that it possessed remarkable keeping properties, which it owed to the pureness of the ferment employed. As a matter of fact, the beer suffered no injury from journeys of more than 300 miles, by slow trains, in ordinary casks, containing from 50 to 100 litres (10 to 20 gallons), during the great heats of June and July, or from being subsequently stored for two months in a cellar, the temperature of which rose during that time from 12° to 18° C. (54° to 65° F.). The temperature of fermentation had been 13° C. (55° F.). Beer from the same brewery, made with the same wort by the ordinary process, did not remain sound for three weeks in this same cellar.