THEORY OF FERMENTATION 335 



of a fine caoutchouc tube, and replaced by about I c. 

 (about 17 minims) of liquid and deposit from another 

 flask, similar to the one we have just described, but which 

 had been fermenting spontaneously for twelve days; we 

 lost no time in refilling completely the exit tube with 

 water which had been first boiled and then cooled down 

 in carbonic acid gas. This operation lasted only a few 

 minutes. The exit-tube was again plunged under mercury. 

 Subsequently the tube was not moved from under the mer- 

 cury, and as it formed part of the flask, and there was 

 neither cork nor india-rubber, any introduction of air was 

 consequently impossible. The small quantity of air in- 

 troduced during the impregnation was insignificant and it 

 might even be shown that it injured rather than assisted the 

 growth of the organisms, inasmuch as these consisted of 

 adult individuals which had lived without air and might 

 be liable to be damaged or even destroyed by it Be this 

 as it may, in a subsequent experiment we shall find the 

 possibility removed of any aeration taking place in this 

 way, however infinitesimal, so that no doubts may linger 

 on this subject. 



The following days the organisms multiplied, the deposit 

 of tartrate gradually disappeared, and a sensible ferment 

 action was manifest on the surface, and throughout the 

 bulk of the liquid. Thje deposit seemed lifted up in places, 

 and was covered with a layer of dark-grey colour, puffed 

 up, and having an organic and gelatinous appearance. For 

 several days, in spite of this action in the deposit, we 

 detected no disengagement of gas, except when the flask 

 was slightly shaken, in which case rather large bubbles 

 adhering to the deposit rose, carrying with them some solid 

 particles, which quickly fell back again, whilst the bubbles 

 diminished in size as they rose, from being partially taken 

 into solution, in consequence of the liquid not being satu- 

 rated. The smallest bubbles had even time to dissolve com- 

 pletely before they could reach the surface of the liquid. 

 In course of time the liquid was saturated, and the tartrate 

 was gradually displaced by mammillated crusts, or clear, 

 transparent crystals of carbonate of lime at the bottom 

 and on the sides of the vessel 



