It would seem advisable to call attention to the very variable character of these wines as shown by the analyses. Considering the extensive use that is made of such wines for medicinal purposes, it is greatly to be desired that some standard should be required for their composition, or that their relative content of alcohol and sugar, at least, should be stated on the label, as is required by the Bavarian authorities. Among Mr. Parsons' samples will be found a "Sweet Muscatel" which contains as high as 31 per cent of sugar, and a "California port" which contains nearly 21 per cent. of alcohol by weight. The analyses of the sweet wines made by myself furnish a still poorer showing, for the low figures obtained for glycerine show that very little pure grape juice enters into their composition. Take the different samples of Angelica and Muscatel wines, for instance; these varieties are almost peculiar to California; they are made from a very sweet grape, of strong flavor. Comparatively few analyses have been made of them, but Baumert had among his samples two Muscatel wines and one Angelica. These contained the following percentages of glycerine : Muscat (H), .883; muscat (W), 1.424; Angelica, .698.

Compare these numbers with the percentage of glycerine contained in the following: 5003, Muscatel, .102; 5092, Muscatel, .104; 4994, Angelica, .140; 5093, Angelica, .052.

These results are so disproportionately low as to give strong ground to the suspicion that but very little of the pure juice of these strong flavored grapes entered into the composition of the samples I examined, but that they were chiefly composed of alcohol, sugar, and water. It would be an easy matter to imitate the strongly marked flavor of the grapes by means of artificial essences.

California wine-growers claim that they have in their very pure grape brandy an excellent and unobjectionable source of alcohol for the fortification of sweet wines, but certainly the samples above partake more of the nature of a liqueur, than of a natural wine.

The following table gives a classification of the wines analyzed in the Paris Municipal Laboratory during the years 1881 and 1882, showing the proportion which was declared adulterated, and the relative amount of the different varieties of adulteration as shown by the samples analyzed. It must be remembered that these analyses were made on suspected samples, and do not by any means represent an average of the quality of the wines sold in Paris.

A wine may be counted several times in this table; for instance, if it is at the same time watered, fortified, too much plastered, and artificially colored. The total of the percentages, therefore, adds up to over 100.

The samples analyzed by me may be tabulated as follows:

The percentages are of the total number analyzed; not of the number adulterated, as in the French wines.