In most cases the determination of the total milk-solids and the fat (the difference being the solids not fat) furnishes all the data required for determining the amount of watering which a sample of milk has undergone. The Society of Public Analysts use 9 as the average percentage of solids not fat in pure milk (which is generally considered as too low) and adopt the formula:-- 100/9 S = _x_, in which _x_ represents the percentage of genuine milk, and S the solids not fat. PLATE IV. [Illustration: Cream × 420.] [Illustration: Cows Milk × 420.] ARTOTYPE. E. BIERSTADT, N. Y. In skimmed milk the percentage of fat removed (_x_) can be ascertained by the formula:-- (2·5) / (9·0) S - _f_ = _x_, in which S = solids not fat, and _f_ = the fat found. In case the sample has been subjected to both skimming and watering, the water added (_x_) can be calculated from the formula[24]:-- 100 - (100 + 2·5) / 9 S - _f_ = _x_. The addition of mineral salts to milk is detected by the increased proportion of ash found; the presence of an abnormal amount of common salt by the high proportion of chlorine present in the ash, which in pure milk should never exceed 0·14 per cent. The use of sodium bicarbonate, borax, etc., is also detected by the analysis of the ash. Glycerine, salicylic acid, flour, and starch, if added, can be extracted from the milk-solids and their identity established by the usual characteristic reactions. The microscope is of great service in the determination of the quality of milk, and especially in the detection of the presence of abnormal bodies, such as pus, colostrum cells, and blood. In pure cow’s milk the globules are in constant motion; their usual size is 1/5000 of an inch, but this depends upon the nature of the food used. Plates IV. and V., which represent cream, pure milk, skimmed milk, and milk containing colostrum cells, were taken from photo-micrographic negatives furnished through the kindness of Mr. Martin. Numerous cases of severe illness have from time to time been developed by the use of milk which was apparently free from any of the usual adulterants. In a recent issue of the ‘Philadelphia Medical News’ (Sept. 1886) an instance of wholesale milk poisoning at Long Branch is described, and the results reached by a careful study of the epidemic are given. It was demonstrated that warm milk, fresh from the cow, if placed in closed cans under conditions which retarded the dissipation of its heat, may suffer fermentation resulting within a few hours in the genesis of a sufficient quantity of a poisonous ptomaine (termed _tyrotoxicon_) to produce dangerous toxic effects in those drinking it.