The determination of the extract in wine presents the same difficulties as with beer on account of the content of glycerine, which is greater in the former than the latter. For this reason the indirect method would seem especially applicable, but the difficulty here is that there is a larger amount of volatile acid present, which is only partially driven off with the alcohol, and the solution density of the solids of wine have not been so well established as that of malt extract, so that different tables and factors vary widely.1

For the direct estimation the French method is to evaporate 25cc. in a flat-bottomed dish, with vertical sides, and dry the residue to a constant weight in a water bath. The Berlin method is as follows:

Extract

Fifty cubic centimeters of wine are measured (at 15° C.) into a platinum dish (85mm. iu diameter, 20mm. in height, and of 75cc. capacity, weight about 20 grams), evaporated on the water bath, and the residue heated 2 1/2 hours at 100° C. Of wines rich in sugar, i. e., containing over 5 grams sugar in 100cc., a smaller amount should be taken, so that the weight of extract shall not exceed 1, or at most 1.5 grams.

1 Baumert found in his work on California wines that the tables of Schulze (Zeit Anal. Chem., 1880, 104) intended for beer extracts gave results in the indirect estimation which approximated much more closely to the results by direct estimation than figures obtained from Hager's tables (Chem. Cent., 1878, 415), which were intended especially for wine extract.

The Bavarian Chemists' Union depart from this method by drying the residue to a constant weight.

It is of great importance that chemists should adopt a uniform method of analysis, as results differ considerably by different methods. This is shown by the following results obtained from the samples analyzed by me, most of which were subjected to analysis by both methods, that is, drying to a constant weight and drying a certain length of time. In the very sweet wines, of course, such as the Angelica, Muscat, etc., containing so large a percentage of sugar, it can only be satisfactorily determined by using a small quantity, allowing it to flow well over the bottom of the dish, and drying to a constant weight as nearly as possible.

Comparison of methods for the estimation of extract in wine.

Serial Number.

Drying

for 2 1/2

hours at

100° C.

Drying to a constant weight.

4995

2.69

2.00

4996

2.82

2.26

4997

2.08

1.44

4998

1.57

1.16

4999

2.18

1.75

5000

2.24

1.74

5001

4.13

3.82

5002

9.78

9.30

5004

9.62

9.53

5005

2.51

2.09

5081

1.77

1.16

5083

1.80

1.16

5084

2.12

1.39

5088

3.64

2.83

Serial Number.

Drying for 2 1/2

hours at 100° C.

Drying to a constant weight.

5089

2.01

1.22

5094

2.30

1.52

5095

2.96

2.18

5096

2.92

1.71

5097

1.82

1.18

5098

1.90

1.34

5099

2.20

1.43

5100

2.47

1.96

5101

2.40

1.82

5103

2.19

1.71

5104

2.42

1.82

average

2.98

2.38

I have adopted the results obtained from the estimation by drying to a constant weight, and would recommend the following procedure for the determination of extract in wine. Weigh out 10 to 50 grams of the wine (according to its content of sugar) into a flat-bottomed platinum dish, evaporate on the water bath and dry the residue at 100° C. in an air bath, until a constant weight is obtained. This may be known to be the case when by weighing the dish and contents at frequent intervals (15 minutes) the losses in weight are found to be slight and equal for the same length of time of drying.