We then replaced the porcelain dish in which the curved escape tube of A had been plunged, by a vessel filled with mercury. [Illustration: Fig. 61.] The following is a description of two of these comparative fermentations and the results they gave. The fermentable liquid was composed of yeast-water sweetened with 5 per cent. of sugar-candy; the ferment employed was _saccharomyces pastorianus_. The impregnation took place on January 20th. The flasks were placed in an oven at 25° C. (77° F.). _Flask A, without air._ January 21st.—Fermentation commenced; a little frothy liquid issued from the escape-tube and covered the mercury. The following days, fermentation was active. Examining the yeast mixed with the froth that was expelled into the mercury by the evolution of carbonic acid gas, we found that it was very fine, young, and actively budding. February 3rd.—Fermentation still continued, showing itself by a number of little bubbles rising from the bottom of the liquid, which had settled bright. The yeast was at the bottom in the form of a deposit. February 7th.—Fermentation still continued, but very languidly. February 9th.—A very languid fermentation still went on, discernible in little bubbles rising from the bottom of the flask. _Flask B, with air._ January 21st.—A sensible development of yeast. The following days, fermentation was active, and there was an abundant froth on the surface of the liquid. February 1st.—All symptoms of fermentation had ceased. As the fermentation in A would have continued for a long time, being so very languid, and as that in B had been finished for several days, we brought to a close our two experiments on February 9th. To do this we poured off the liquids in A and B, collecting the yeasts on tared filters. Filtration was an easy matter, more especially in the case of A. Examining the yeasts under the microscope, immediately after decantation, we found that both of them remained very pure. The yeast in A was in little clusters, the globules of which were collected together, and appeared by their well defined borders to be ready for an easy revival in contact with air. As might have been expected, the liquid in the flask B did not contain the least trace of sugar; that in the flask A still contained some, as was evident from the non-completion of fermentation, but not more than 4·6 grammes (71 grains). Now, as each flask originally contained 3 litres of liquid, holding in solution 5 per cent. of sugar, it follows that 150 grammes (2,310 grains) of sugar had fermented in the flask B, and 145·4 grammes (2,239·2 grains) in the flask A. The weights of yeast after drying at 100° C. (212° F.) were— For the flask B, with air 1·970 grammes (30·4 grains). For the flask A, without air 1·368 grammes.[113]