IOWA ACADEMY OP SCIENCES. 
71 
the separation of dextrose is precipitated with hydrogen sul- 
phide. This leaves a solution of levulose which, when concen- 
trated, is supposed to be capable of estimation in the polari- 
scope, or by gravimetric process. The method does not, how- 
ever, appear to be applicable to honey analysis. The products 
sought are destroyed in the reactions, and the results obtained 
were far from correct. We think the method quite applicable, 
however, where little reducing sugar is present. 
Allen’s method^ for the determination of levulose by taking 
readings in the polariscope of a normal solution of honey at dif- 
ferent temperatures, and calculating the amount of levulose 
that must be present, gave figures quite reliable and apparently 
accurate. 
The results for levulose by range of temperature on three 
determinations gave a mean of 32.93 per cent. 
The results for dextrose by precipitation with slaked lime on 
three determinations gave a mean of 43.58 per cent. 
The sum of the two sugars estimated separately is 76.51 per 
cent. The total reducing sugars present by Fehling’s solution 
is 76.87 per cent, a difference of 36 per cent in favor of the 
copper reduction. 
For the determination of moisture a paper coil method was 
Substituted for the open pan method. A comparison of results 
gives the following figures : 
Dried in open dish 8 hours, 15.50 per cent moisture. 
Dried in open dish 13 hours, 17.09 per cent moisture. 
Dried in paper coils 2 hours, 17.53 per cent moisture. 
Dried in paper coils 3 hours, 18.39 per cent moisture. 
Dried in paper coils 4 hours, 18.59 per cent moisture. 
It is generally stated that honey contains no dextrin ; -this 
sample, however, contained .36 per cent of dextrin. The 
method employed was the separation with absolute alcohol. 
For the determination of organic matter not sugar, acetate of 
lead was used as proposed by Wiechmann. f The result gave 
.14 per cent non-sugar material. An analysis of sample No. II 
gave .12 per cent of material as non-crystallizable, non-difiusi- 
ble. 
Should a sample contain non-crystallizable sugar, and also 
organic matter not sugar, these two methods would serve for 
the determination of such products, and might thus be useful 
^Allen’s Com. Organic Anal., Vol. 1. p. 291. 
+ Sugar Analysis, Weichmann, p. 83. 
