In a joint paper read before the annual meeting of the American Public Health association at Washington, D. C., H. W. Clark and Stephen D. Gage, of Boston, treated of the value of tests for bacteria of specific types as an index of the pollution of water. They said that the principal types of bacteria used as indices of pollution were the colon type, sporogenes type, sewage streptococcus type, and occasionally the two so-called paracolon types—i. e., the aerogenes or enteritidis Gartner type and the chologenes type. The colon type was the most numerous of these in normal sewage, and was accordingly of the most value as a specific indication of pollution. Bacteria of the colon type could be detected in ordinary city sewage in dilutions of 1-10,000 to 1-100,000. The presumptive test as applied to the colon type, reacting as it did with some fifty-eight well-described species of bacteria, covered too broad a fold unless followed by confirmatory tests. Tests for bacillus coli at the experimental station at Lawrenge had agreed well with results obtained by chemical analysis and by inspection of the sources of the samples. Surface waters having the greatest relative population on their watershed had shown bacteria of the colon type more frequently than others with less population. Ground waters, on the other hand, unless grossly polluted, never showed bacteria of colon type Samples of shellfish from sources polluted and unpolluted had almost invariably shown the presence of bacillus coli in the polluted samples and their entire absence in the non-polluted samples. Examinations for bacillus coli in different volumes of water showed that tests in both one and 100 cubic centimetres should be made, the tests in larger volumes being confirmatory of the tests in the smaller volumes; and in cases in which a considerable numher of samples were taken from the same source, tests in the large volumes usually gave more information day by day as to the quality of the water than did the tests in one cubic centimetre. Evidences of typhoid fever coincident with the finding of bacillus coli in normally pure waters—i. e., in waters known to be safe under ordinary conditions, but becoming dangerous by the accidental entrance of polluted water, had occurred within the observation ot the writers, in which, although the entrance of this pollution was known, the danger was not indicated by chemical tests or by an increase in the number of bacteria beyond the usual limits of variation for waters of this class. Comparative tests for the relative viability of bacillus typhosus and bacillus coli showed that there was a very great similarity between the length of life of the two germs under a great variety of conditions.