These methods are only suitable for observations at considerable intervals of time. For the continuous observation of the course of fermentation Harden, Thompson and Young [1910] connect the fermentation flask with a Schiff's azotometer filled with mercury and measure the volume of gas evolved, the liquid having been previously saturated with carbon dioxide (Fig. 3). The level of the mercury in the reservoir is kept constant by a syphon overflow, as shown in the figure, or, according to a modification introduced by S. G. Paine, by a specially constructed bottle provided with two tubulures near the bottom. This ensures that no change in the pressure in the flask occurs, and the volume of gas observed is reduced to normal pressure by means of a table. Before making a reading it is necessary to shake the fermenting mixture thoroughly, as the albuminous liquid very readily becomes greatly supersaturated with carbon dioxide, so much so in fact that very little gas is evolved in the intervals between the shakings. The exact procedure in making an observation consists in shaking the flask [p029] thoroughly, replacing in the thermostat, allowing to remain for one minute, and then reading the level of the mercury in the azotometer. After the required time, say five minutes, has elapsed from the time at which the flask was first shaken, it is again removed from the bath, shaken as before, replaced, allowed to remain for one minute and the reading then taken. In this way readings can be conveniently made at intervals of three or five minutes or even less, and much more detailed information obtained about the course of the reaction than is possible by means of observations made at intervals of several hours. Another form of volumetric apparatus, designed by Walton [1904], has been used by Lebedeff [1909]. An apparatus on a different principle has been designed by Slator [1906] for use with living yeast, but is equally applicable to yeast-juice, and a very similar form has been more recently employed by Iwanoff [1909, 2]. In this apparatus the change of pressure produced by the evolution of carbon dioxide is measured at constant volume, and comparative rates of evolution can be obtained with considerable accuracy, although the method has the disadvantage that the absolute volume of gas evolved is not measured. The apparatus consists of a bottle or flask connected with a mercury manometer. The fermenting mixture is placed in the bottle along with glass beads to facilitate agitation, the pressure is reduced to a small amount by the water-pump, and the rise of pressure is then observed at intervals, this being proportional to the volume of gas produced. As in the preceding case, the liquid must be well shaken before a reading is made. THE ALCOHOLIC FERMENTATION OF THE SUGARS BY YEAST-JUICE.