The water supply of a factory is also a question of prime importance. When taken from a shallow well, especially if surface drainage from the factory is possible, the water may be contaminated to such an extent as to introduce undesirable bacteria in such numbers that the normal course of fermentation may be changed. The quality of the water, aside from flavor, can be best determined by making a curd test (p. 76) which is done by adding some of the water to boiled milk and incubating the same. If "gassy" fermentations occur, it signifies an abnormal condition. In deep wells, pumped as thoroughly as is generally the case with factory wells, the germ content should be very low, ranging from a few score to a few hundred bacteria per cc. at most. Harrison[50] has recently traced an off-flavor in cheese in a Canadian factory to an infection arising from the water-supply. He found the same germ in both water and cheese and by inoculating a culture into pasteurized milk succeeded in producing the undesirable flavor. The danger from ice is much less, for the reason that good dairy practice does not sanction using ice directly in contact with milk or cream. Then, too, ice is largely purified in the process of freezing, although if secured from a polluted source, reliance should not be placed in the method of purification; for even freezing does not destroy all vegetating bacteria. FOOTNOTES: [1] Olson. 24 Rept. Wis. Expt. Stat., 1907. [2] Erf and Melick Bull. 131, Kan. Expt. Stat., Apr. 1905. [3] Storch (40 Rept. Danish Expt. Stat., Copenhagen, 1898) has devised a test whereby it can be determined whether this treatment has been carried out or not: Milk contains a soluble enzym known as galactase which has the property of decomposing hydrogen peroxid. If milk is heated to 176 deg. F. (80 deg. C.) or above, this enzym is destroyed so that the above reaction no longer takes place. If potassium iodid and starch are added to unheated milk and the same treated with hydrogen peroxid, the decomposition of the latter agent releases oxygen which acts on the potassium salt, which in turn gives off free iodine that turns the starch blue. [4] McKay, N. Y. Prod. Rev., Mch. 22, 1899. [5] Doane, Bull. 79, Md. Expt. Stat., Jan. 1902. [6] Harrison, 22 Rept. Ont. Agr'l Coll., 1896, p. 113. [7] Moore and Ward, Bull. 158, Cornell Expt. Stat., Jan. 1899; Ward, Bull. 178, Cornell Expt. Stat., Jan. 1900. [8] Harrison, 22 Rept. Ont. Agr. Coll., 1896, p. 108; Moore, 12 Rept. Bur. Animal Ind., U. S. Dept. Ag., 1895-6, p. 261. [9] Moore, Bacteria in Milk, N. Y. Dept. Ag., 1902. [10] Freudenreich, Cent. f. Bakt., II Abt., 10: 418, 1903. [11] Harrison, 22 Rept. Ont. Agr. Coll., 1896, p. 108. [12] Marshall, Bull. 147, Mich. Expt. Stat., p. 42. [13] Moore and Ward, Bull. 158, Cornell Expt. Stat., Jan. 1899.