NA TURE 



[November 3, 1904 



ing this point, it is obvious tliat the aggregate amount 

 and relative proportion of these products must depend 

 very largely upon the means made use of, and hence 

 perfectly genuine brandies must necessarily show wide 

 differences in the coefficient non-alcohol. 



In addition, it must be remembered that in the manu- 

 facture of brandy from wines of repute, the elimination 

 of the substances constituting non-alcohol must be 

 made with the greatest circumspection, since it is 

 upon their bouquet that the value of these brandies 

 depends, and this bouquet resides wholly in the non- 

 alcohol. 



On the other hand, if the brandy is being made from 

 damaged vt-ine the rectification must be most carefully 

 conducted, and may have to be pushed to a point that 

 the alcohol is obtained almost pure, that is to say, 

 almost free from tton-alcohol. 



As regards the influence of age, it is observed that 

 in those brandies which are found to improve on keep- 

 ing there is an increase in non-alcohol due (i) to the 

 formation of products of oxidation (acids and alde- 

 hydes), and (2) to concentration due to a loss of alcohol 

 and water. 



Brandies may be classified in the following 

 manner : — 



(i_) The brandies of the two Charentes, which are 

 habitually designated by the name of Cognac. 



(2) The brandies of Armagnac. 



(3) The brandies de vin du Midi and of Algeria 

 (trois-six de Montpellier, &c.). 



(4) Marc brandies. 



The brandies of the Charentes are obtained by 

 distillation of the wines of the district, and as the 

 reputation of these brandies depends upon their 

 bouquet they are submitted to a slight rectification 

 only in order to preserve that bouquet. 



The same may be said of the .Armagnac brandies. 



As to brandies made in other viticultural regions, 

 and in particular in the middle of France, their nature 

 is rnuch more variable. These brandies require to be 

 rectified in a manner, more or less complete, depend- 

 ing upon the nature of the wine or of the marc from 

 which they are derived, and varying, too, with the 

 quality of the brandy it is desired to produce. Certain 

 wines require, in fact, to be most carefully rectified in 

 order to produce merchantable brandy. Marc brandy 

 is made in all viticultural regions, and that of 

 Burgundy enjoys a special reputation. 



As regards the value of the coefficient in different 

 brandies, it is found that in those of Charente and 

 Armagnac the coefficient is very high. Thus, as 

 minima, a brandy of Clunis (1879, good, but not 

 guaranteed) gave 259 (Girard and "Cuniasse). A 

 Cognac of iS92_gave 287 (Rocques). .\s maxima may 

 be cited a Bois brandy of 1817, which gave 1174 

 (Lusson). This last number is exceptionally high. It 

 may be said that, ordinarily, the value of the coefficient 

 in Cognacs and fine champagne ranges between 275 

 and 450. 



But little analytical evidence has been published re- 

 specting the Armagnac brandies, but, such as it is, 

 it indicates that the coefficients in their case are less 

 than are generally found in Cognacs. 



The brandies obtained from the wines of the Midi 

 and Algeria show much wider variations, ranging from 

 25 to 500. 



Marc brandies have almost invariably a high 

 coefficient. The numbers range from 555 to 1487, and 

 it is interesting to note that the aldehydes frequently 

 form a large proportion of the whole. Thus a 

 Burgundy marc brandy was found to contain as much 

 as 519 of aldehyde, and one from the Midi as high as 

 730 of aldehyde. 



NO. 1827, VOL. 71] 



The question whether it is possible to fix minimum 

 and maximum limits to this coefficient naturally re- 

 ceived much consideration from the committee. The 

 fi.xation of these presents a certain interest, and that 

 from two different points of view. The fixation of a 

 minimum limit has interest for the analyst, as guiding 

 him in his inference as to the genuineness of the brandy 

 or as to the amount of " silent " spirit with which it 

 may have been mixed. The fixation of a maximum 

 limit has an interest from the hygienic point of view, 

 since it may become necessary if regulations are to be 

 established in this sense. 



The committee, however, are unable to recommend 

 that any such limits should be fixed, owing mainly to 

 the extremely variable character of brandy. Even in 

 the case of brandies of a definite character, as, for 

 example, Cognac, the non-alcohol coefficient is not the 

 only element of value, and any conclusions as to 

 character cannot be based solely upon it. Regard must 

 be had to the proportions of the different volatile sub- 

 stances and their relations among themselves. Expert 

 tasting (dSgustation) must be considered as an indis- 

 pensable complement of chemical analysis. 



The hygienic point of view, involving the fixation 

 of a maximum value for the non-alcohol coefficient, 

 was brought to the notice of the International Congress 

 of Chemistry in Paris in igoo, but the problem, as then 

 stated, received no definite solution. To base con- 

 clusions on the value of the coefficient alone, with no" 

 regard to the factors which it comprises, seems 

 illogical. For example, the acids, and in particular 

 acetic acid, frequently make up a large proportion of 

 this value, but it cannot be contended that these sub- 

 stances, at least in the proportion in which they are 

 present in brandy, have any detrimental influence. 

 Far more important are the aldehydes, ethers, the 

 higher alcohols, and furfurol. 



As regards the higher alcohols, the attempt has been 

 made to establish a higher limit. Thus in Belgium, by 

 a Royal decree of December 31, 1902, the sale is pro- 

 hibited of spirituous liquors containing more than 

 I gram of the higher alcohols and essences per litre 

 of absolute alcohol when these liquors have an alcoholic 

 content higher than 90°, and 3 grams when the 

 alcoholic richness does not exceed 90°. 



The committee remark that the effect of this regula- 

 tion would be to exclude some of the most famous, and 

 notably the oldest, brandies of the Charente, many of 

 which exceed the maximum Belgian limit, which, ex- 

 pressed as a non-alcohol coefficient, is 300. Thus : — 



Bois Brandy, 1817 (Lusson)... 

 Saintonge, Cazes, 1896 (Lusson) 

 Clemozac, or de Fesson, 1S93 (Lusson) 

 Clunis, 1875 (Lusson) 

 Cognac, 1873 (Rocques) 



Higtjer alcohol: 

 per tiectol 



.lbs. alcohol 

 ... 612 

 ... 372 



■■ 345 

 ••■ 34S 

 • 304 



of 



From the hygienic point of view the ethers, furfurol, 

 and especially the aldehydes, are undoubtedly of much 

 greater importance than the higher alcohols, since 

 admittedly the action of these substances on the 

 organism is far more deleterious than that of the higher 

 alcohols. From this point of view the attention of 

 hygienists .should be directed to the Marc brandies, 

 which, as already stated, frequently contain consider- 

 able quantities of aldehydes. 



Interesting and, no doubt, valuable as the report is, 

 it is hardly calculated to facilitate the work of the un- 

 fortunate public analysts who may be called upon to 

 express an opinion as to the genuineness of a sample 

 of brandy. The question, What is brandy? analytically 

 speaking, still awaits solution. 



