153 
however, was altered in such a manner that the preci- 
pitated cuprous oxide was weighed directly after having 
been dried in a desiccator during one hour; the methods 
commonly used are to reduce the cuprous oxide to me- 
tallic copper, or oxidize it to cupric oxide. The reduc- 
tion process, however, is a rather troublesome method. 
The oxidizing of the cuprous oxide is a little more handy; 
the investigations of Kristensen (1916), have shown, 
however, that this method is a very tedious one. It is 
necessary to ignite and to weigh several times to bo 
jsure that the oxidation is complete. When ignited up to 
one hour in a dry current of air the weight will some^ 
times increase' The direct weighing, however, has 
jauother advantage in addition to its greater convenience. 
If a glass-thread stopper is used to prevent the asbestos 
filter to drop down from the Soxhlet’s tubes, as is ge- 
nerally the case, the tubes will burst to pieces by the 
ignition already after a relatively short time, and the 
analysis is spoiled. 
The values of the copper is obtained if the values 
of the cuprous oxide is multiplied by 0 , 8 S 82 . The ap- 
proximate value of the invert-sugar is obtained by divi^ 
ding the copper value with 2. As care always has been 
taken to use the same volumes it became easy to cal- 
culate the sugar percentage of the roots. 
In order to get the exact value of the invert-sugar 
it was necessary to polarize another portion. This was 
made in -the 20 c. c. glass tubes of the polarization 
apparatus. 
The total percentage of sugar was determined from 
a third portion, which was measured with a pipette. The 
sugar solution was first inverted by heating it together 
with 10 % hydrocloric acid on a water bath during half 
an hour. The percentage of sugar was then determined 
as invert-sugar by the modified Allihn’s method. 
