152 
invert so large a quantity of . cane-sugar during the 
warming up of the solution, that the results of the ana^ 
lysis become incorrect. Therefore I preferred to extract 
in cold water, and the experiments made have shown 
that 3 /4 of an hour is quite sufficient for a practically 
complete extraction of the sugar, when the solution is 
repeatedly shaken. Even extraction during */ 2 hour only 
gives about the same result as if it is continued during 
1 1 / 2 hour. Further, the method of extracting with cold 
water has the advantage of being more handy. 
The extracting was made in the following manner. 
The weighed quantity of substance — about 20 gr. pulp 
— was brought in a 250 c. c. measuring cylinder (or the 
double quantity in a similar one of 500 c. c.), which 
was filled with distiled water. The extraction continued 
during 3 /4 hour, and during this time the solution was 
often shaken. Then it was filtered through a sucking- 
filter. The filtered solution was coloured and rather 
muddy. In order to precipitate the albuminous substan- 
ces, which has been dissolved, a fixed quantity of the 
filtered solution was measured with a pipette; 2 c. c. 
mercuric nitrate solution was then added. The cause of 
my giving up the commonly used lead extract method 
is the fact that Keistensen (1916) has found that the 
lead has a disturbing influence on the reducing of 
Fehling’s solution. As soon as the precipitate had sett- 
led — which usually takes place in a few minutes — the 
solution was filtered ; 5. c. c. salt-solution was then added 
to a measured part of it for the purpose of precipita- 
ting the excess of the mercuric nitrate. When this 
precipitate had settled, which takes a time of 12 — 24 
hours, the solution was filtered. The sugar solution, 
now quite clear, was divided in 3 portions. 
The one part was used directly for the analysis of 
reducing sugars by the reducing of Fehling’s solution, 
according to Allihn’s method (König 1911). This method, 
