It is in every way better, in drawing conclusions from the results of analysis, to consider, not the proportion of esters alone, but the whole of the "secondary constituents," and especially the higher alcohols.

Girard and Cuniasse state that the sum of the secondary constituents, which they term the "coefficient of impurity," is rarely lower than 300 in genuine alcohol from wine.

M. Lusson (Municipal Laboratory, La Rochelle) considers that the sum in question is never lower than 340 in pure eau-de-vie; and that in eaux-de-vie de Cognac free from neutral spirit the sum of the esters and the higher alcohols is always above 300. The same author remarks that the sum of the acidity plus aldehydes lies normally between 10 and 36 per cent. of the total secondary constituents in eaux-de-vie. The lower number applies to newer spirits, the higher to products forty years old, but without being exactly proportional to age.

1 Amended during 1.915 to 35° under proof for certain areas, under the Defence of the Realm Act.

M. Rocques has shown that for well.made eau-de-vie de vin prepared in the Cognac district the ratio of the higher alcohols to the esters approximates to unity.

(In French practice, it should be remembered, the higher alcohols are determined colorimetrically and expressed in terms of an isobutyl standard. The colour produced by amyl alcohol is less than that given by isobutyl alcohol in the proportion of 3: 5. Hence results obtained with an amyl alcohol standard require to be decreased in the same proportion for comparison with the French figures.)

If the esters in a sample of brandy fall substantially below 80, and at the same time the total secondary products are substantially below 300 when the higher alcohols are determined according to the French practice with isobutyl alcohol as standard, the inference is that the sample contains an admixture of neutral spirit. An approximate idea of the minimum amount of such admixture may be obtained by calculating from the foregoing numbers taken as standards, allowance being made for the corresponding quantities of products present in average specimens of neutral spirit.

The analysis of brandy is carried out by the methods already described for the examination of commercial spirits.

The following analyses of brandies and of neutral spirit (alcools d'industrie) are given by Girard and Cuniasse: - 1

I. Eaux-de-vie of known origin. Analysed in 1896.

Saintonge. 1880.

Saintonge. 1896.

D'Aunis. 1896.

He de Re. 1896.

Armagnac.

40 years old.

Density at 15°; ...........

0.9157

0.8947

0.9011

0.8903

-

Alcohol, by vol...........

59.0

68.2

65.5

70.0

49.8

Extract, per 100 c.c.

-

Nil.

Nil.

Nil.

0.18

Secondary constituents: -

Acidity .................

105.7

17.5

21.9

17.1

146.7

Aldehydes ............

27.9

23.9

4.7

23.9

31.4

Furfural..................

2.3

2.6

1.3

2.1

0.7

Esters ..................

167.0

61.9

193.4

133.4

125.5

Higher alcohols.........

159.8

259.8

190.5

219.8

203.5

Totals

462.7

365.7

411.8

396.3

507.8

1 "L' Analyse des Alcools," 1899, p. 287 et seq.

The "secondary constituents " are expressed as parts per 100,000 of absolute alcohol.