It would seem natural that in American wines, which can be produced so cheaply and in such great abundance, this adulteration, which is such a favorite one with the manufacturers of the costly wines of Bordeaux, Burgundy, etc, would be very rare. The fraud is so simple, however, so easy of execution, and so difficult of detection, that it will probably always be a favorite one with unscrupulous dealers. It must be remembered, also, that with many American producers, whatever article they produce, more attention is paid to its quantity than its quality. Wine-growers are not the only persons who practice this method, as it can be done also by merchants and retailers, although in the latter case it is much more easy of detection. That which might be called scientific dilution, by means of the processes already described (petioti-zation, &c), is much more difficult of detection than the simple attenuation of the wine by the retailer. So little official supervision has been exercised over the wines sold in this country that certainly the fear of detection has not operated very largely as a preventive of this, or in fact any other adulteration.

In Dr. Baumert's work, which has already been alluded to,1 and to which I shall have occasion to refer frequently as constituting, small as it is, the only published investigation of American wines for adulteration, none of the samples fell below the German standard in percentage of extract (1.5 grams per 100cc). On the other hand, nearly all the white wines which I submitted to a complete analysis fell below this standard, and two of the red wines. A large number of the samples analyzed by Mr. Parsons also fell below it. That this limit is not placed at too high a figure, for California wines at least, seems evident from a study of the table I have prepared of Professor Hilgard's analyses of pure wines, from which it appears that only one series of analyses gave a minimum below it, while the averages are far above it. It might possibly be too low for Virginia wines, but the majority of those that fell below it were of California origin. The New York law specifies (§2) that "such pure wines shall contain at least 75 per centum of pure grape or other undried fruit juice." Just how a chemist, in the absence of legal definitions of what shall constitute a "pure grape or other undried fruit juice," is to decide upon the question of such adulteration by the above law is difficult to indicate.2

The samples which would be considered as watered according to the German standard are as follows: Serial Nos. 5084, 5099, 4997, 4998, 5081, 5083, 5089, 5097, and 5098.