158 STUDIES OX FERMENTATION. 



use in our subsequent studies the ferment-bearing dust found 

 in October on the woody part of the vine, and, more particularly, 

 on the clusters themselves, as already stated. After our return 

 to Paris, and during the course of our experiments in October 

 and November, it sufficed to wash a few scraps of the bunches 

 in a little pure water, in order to obtain the grape-ferment in 

 abundance ; but later on in the winter we were astonished to 

 find that the same procedure yielded no ferment, only some 

 moulds. The bunches which, when put into boiled and filtered 

 must, in October, very readily caused that must to ferment, at 

 the end of winter could no longer produce the same effect, 

 however favourable might be the temperature to which we 

 raised the must. The particles of dust on the bunches had, 

 therefore, become sterile, as sources of alcoholic ferments. 



II. On February 17th, 1875, we purchased of Chevet, a 

 dealer in provisions, two bunches of white grapes, which were 

 perfectly' sound, presenting not the slightest trace of injury or 

 bruise. We took an iron pot full of mercury, which had been 

 heated to 200° C. (392° F.), and then covered over its surface 

 with a sheet of paper that had also been subjected to flame. 

 When the mercury had cooled down we placed several of 

 Chevet's grapes, singly and in small bunches, on the surface 

 of the metal, and, after having enclosed them in a glass 

 cylinder that had been previously heated with and by means 

 of the mercury, we crushed them in this vessel, in contact 

 with air, by means of a strong, crooked iron wire that had 

 been passed through the flame of a spirit lamp. The object of 

 all these precautions was to prevent any cause of error, such as 

 might have resulted from the accession of particles of dust 

 associated with the mercury, or floating about our laboratory. 

 We then placed our cylindrical jar in an oven, at a temperature 

 of 25° C. (77° F.) ; but though the experiment was continued 

 for several days following, no fermentation manifested itself. 

 At last, to assure ourselves that the pulp and liquid were, 

 notwithstanding tliis, well adapted to fermentation, we intro- 

 duced into the test-flask an almost imperceptible quantity 



