Carbon Assimilation 
2I 5 
becomes acid owing to the presence of free acetic acid which is 
removed by the addition of dilute sodium carbonate solution until 
the extract is faintly acid to litmus. 
Portions of the extract are then fermented with yeasts which 
do not contain the enzyme maltase, namely, Saccharomyces 
marxianus, S. anomalns, S. exiguus. Two other portions are 
fermented with bakers’ yeast. The yeast is allowed to incubate 
for 21 to 28 days at 25"C, by which time fermentation is complete. 
After addition of alumina cream and filtration, the cupric-reducing 
power is measured. The differences between the cupric-reducing 
power of the extracts fermented with maltase-free and maltase- 
containing yeasts, must be due to the glucose resulting from the 
hydrolysis of maltose, and so the proportion of maltose in the extracts 
may be calculated. 
(iv.) The pentoses are estimated by distillation with hydro¬ 
chloric acid and weighing the furfural produced as phloroglucide. 
(v.) The reducing power of the maltose and pentose is 
calculated, and from the reducing power of the original extract, the 
reducing power of the hexoses can be calculated. The optical 
activity of the sucrose and maltose is calculated, and that of the 
pentoses on the assumption that they consist of /-xylose and 
Z-arabinose in equal proportions. By comparison of these data and 
the optical activity of the original solution, the optical activity due 
to the hexoses can be calculated. From this and their cupric- 
reducing power, the quantities of glucose and fructose can be 
calculated on the assumption that these are the only hexoses 
present. 
It will be observed that the accuracy of the determinations of 
glucose and fructose depend upon the accuracy of the following 
assumptions:— 
(i.) That sugars are the only cupric-reducing and optically active 
substances in the purified extracts. 
(ii.) That the only sugars that can be present in the extracts 
are sucrose, maltose, rf-glucose, ^/-fructose and pentoses. 
(iii.) That the pentoses present are only Z-arabinose and Z-xylose, 
and that these are present in equal quantities. 
In regard to this last assumption, Davis, Daish and Sawyer 
show that the error involved is not very large (about 7%) if the 
whole of the pentose is either Z-arabinose or Z-xylose. 
The accuracy of the glucose and fructose estimations also 
depends upon the accuracy of the following operations 
(i.) The completeness of the extraction of sugars. 
