The chemical examination of a suspected water is by no means a simple process, and in most cases had better be referred to an expert in such matters. It is highly desirable, however, that the physician should have sufficient technical knowledge to be able to make a rough analysis at least, if for no other reason than that he may be able to appreciate the results reported by the chemist. As a rule, when a water is so polluted with decomposing organic matter as to be positively dangerous it will have an unpleasant odor, which is best developed by half filling a quart bottle with the water to be examined and shaking it thoroughly. The so-called simple and ready methods which are from time to time advocated in the newspapers, such as the addition of sugar to the suspected water and allowing fermentation to take place, the use of tannin as a precipitant, or the decolorization of a solution of potassium permanganate, are really of very little value and should not be relied upon. In the hands of an expert the best simple method of determining the quality of a water is by evaporation of a known quantity and the ignition of the solid residue. From the amount of the total residue, the quantity left after ignition, the amount of blackening produced, and the odor, a very fair opinion can be formed as to the amount of organic matter present, and whether it is of animal or vegetable origin. It is not within the province of this paper to describe the methods used by chemists in water analysis, of which the principal are known as the Franklin and Armstrong, the Wanklyn, and the permanganate methods. A careful examination of these methods has recently been made under the direction of the National Board of Health, and a preliminary note of the results, prepared by Professor Mallet, has been published in the _Bulletin_. From this it appears that the chief value of chemical analysis is, first, the verification of gross pollution, which will usually be detected by the appearance and smell of the water; and, second, in periodical examination of a water-supply to detect changes from the normal or usual character of the water, which may be taken to have a certain local standard of purity. Special importance is attached to the careful determination of {184} nitrates and nitrites in water to be used for drinking, these being the results of oxidation of organic matters, and therefore giving evidence of previous contamination.