CH. XIIl] SUGARS. 279 



weight of an apparatus from which the dry gas only can 

 escape. 



The problem is the same as that of the estimation of CO2 

 in carbonates, and full directions for the process are given in 

 Fresenius, Quantitative Analysis, 7th ed. (Carbonic Acid). 



It is necessary to make a parallel experiment with the 

 same quantity of yeast in pure water and to subtract the 

 amount of CO2 obtained in this experiment from the 

 total ; the yeast itself always evolves some gas on standing 

 for so long a time as is generally necessary for complete 

 fermentation. 



A convenient way of performing the estimation is to 

 use two nitrometers, introducing the solution and yeast 

 into one (over mercury) and the same quantity of yeast 

 in water into the other (over mercury), at the end of the 

 experiment the volumes of gas are read off and the 

 weight deduced from the volume of gas obtained. 



Where the amount of sugars is small it is better to 

 estimate the alcohol produced rather than the CO2, and 

 this may be done in all cases as a control experiment. 

 If the amount of alcohol in the distillate is considerable 

 it is determined from specific gravity tables, but if small by 

 Dupre's method of oxidising to acetic acid and estimating 

 the acetic acid by standard alkali. 



(For Dupre's process see /. C. S., vol. XX.) 



The following numbers may be used in calculating 

 weights of glucose in fermentation experiments. 

 CO., X 2-16 = glucose. 

 Alcohol X 2*07 = glucose. 

 Acetic acid x 1*58 = glucose. 



