STUDIES ON FERMENTATION. 381 



with oxygen in combination, by being allowed to remain for one 

 year in an open flask in contact with pure air. This wort was 

 deprived of air in solution by a protracted boiling over mer- 

 cury. It was then pitched, out of contact of air, with an old 

 yeast. The yeast underwent no development at all, a proof 

 that oxygen in combination cannot act like oxygen that is free, 

 or simply in solution, in effecting the revival of the yeast ; 

 nevertheless, after the revival has been once started by means 

 of a small quantity of air, fermentation declares itself with 

 much greater facility than in the case of copper wort, placed 

 under the same conditions, but deprived of oxygen in com- 

 bination. 



§ V. Ox THE Influence of Oxygen in Combination on the 

 Clarification of Wort. 



Oxygen in combination has another effect which it is 

 essentially important to point out, for it concerns the clarifi- 

 cation of beer. One of the most valued properties of this 

 beverage is its limpidity and brilliancy. We know from the 

 results of the fourth experiment in the preceding paragraph 

 that in the case of a wort shaken up when hot with air, and 

 examined as soon as cold, that is, after an interval of only three 

 hours, we find a notable volume of oxygen to have been absorbed 

 by combination ; in the experiment to which we allude, this 

 volume was not less than 20 c.c. of oxygen per litre of wort. 

 The shaking up of the wort when cold with air saturated it with 

 oxygen in solution, but the quantity of oxygen which under 

 these conditions entered into combination, in the course of three 

 hours, is insignificant, although saturation by solution may be 

 attained in the course of one minute's shaking. If two samples 

 of the same wort are shaken up with air, one of them being hot 

 and the other cold, and both filtered after having been left 

 xmdisturbed for twenty-four hours, or_ even immediately after 

 the agitation, we cannot fail to be struck with the great differ- 



