It is a matter of some interest to notice here that, in the mode of procedure adopted, everything combined to prevent the interference of air.  A portion of the liquid expelled at the beginning of the experiment, partly because of the increased temperature in the oven and partly also by the force of the gas, as it began to be evolved from the fermentative action, reached the surface of the mercury, where, being the most suitable medium we know for the growth of bacteria, it speedily swarmed with these organisms. [Footnote:  The naturalist Cohn, of Breslau, who published an excellent work on bacteria in 1872, described, after Mayer, the composition of a liquid peculiarly adapted to the propagation of these organisms, which it would be well to compare for its utility in studies of this kind with our solution of lactate and phosphates.  The following is Cohn’s formula: 

     Distilled water. ... . ... . ... . ..20 cc. (0.7 fl. oz.)
     Phosphate of potassium. ... . ... ...0.1 gramme (1.5 grains)
     Sulphate of magnesium. ... . ... . 0.1 gramme (1.5 grains)
     Tribasic phosphate of lime. ... ... 0.01 gramme (0.15
grain)
     Tartrate of ammonia. ... . ... . ... 0.2 gramme (3 grains)

This liquid, the author says, has a feeble acid reaction and forms a perfectly clear solution.] In this way any passage of air, if such a thing were possible, between the mercury and the sides of the delivery-tube was altogether prevented, since the bacteria would consume every trace of oxygen which might be dissolved in the liquid lying on the surface of the mercury.  Hence it is impossible to imagine that the slightest trace of oxygen could have got into the liquid in the flask.

Before passing on we may remark that in this ready absorption of oxygen by bacteria we have a means of depriving fermentable liquids of every trace of that gas with a facility and success equal or even greater than by the preliminary method of boiling.  Such a solution as we have described, if kept at summer heat, without any previous boiling, becomes turbid in the course of twenty-four hours from a spontaneous development of bacteria; and it is easy to prove that they absorb all the oxygen held in solution. [Footnote:  On the rapid absorption of oxygen by bacteria, see also our Memoire of 1872, sur les Generations dites Spontanees, especially the note on page 78.] If we completely fill a flask of a few litres capacity (about a gallon) (Fig. 9) with the liquid described, taking care to have the delivery-tube also filled, and its opening plunged under mercury, and, forty-eight hours afterwards by means of a chloride of calcium bath, expel from the liquid on the surface of the mercury all the gas which it holds in solution, this gas, when analyzed, will be found to be composed of a mixture of nitrogen and carbonic acid gas, without the least trace of oxygen.  Here, then, we have an excellent means of depriving the