Unfortunately, there are no such bacteria which are a positive indicator of human pollution. The nearest thing to it are the coli bacilli. Thus B. coli are always found in any discharges from the human intestines. They are also found, however, in the discharges from the intestines of animals, birds, fishes, and on cereals, grains and many other places. They are even found in the air, the dust, and, because of their widespread occurrence, have often been termed “ubiquitous." Because of these conditions it is not a simple matter to say that any water supply which shows the presence of B. coli necessarily possesses any specific danger of human pollution or of infection which might at any time have been present. It merely suggests a potential danger in such a water. This difficulty is made more pronounced by the large number of forms or types in which the B. coli appear. The colon bacillus is not a simply definite form of unvarying properties. It represents, rather, a whole class or series of classes of bacteria resembling each other in a certain number of properties and differing in other properties. There is a graded series of bacteria, some of which have in common many qualities, but the extremes of the series have only a small number of things in common. Attempts have been made to divide these coli bacilli into two general classes, one called the typical coli and the other the atypical coli. The typical coli is the one which is most likely to occur in the human intestines, and the atypical coli is that form which is most likely to originate elsewhere. In 1905 the Committee on Standard Methods of Water Analyses of the American Public Health Association, of which the speaker was a member, drew up a set of standard tests for B. coli, designed to distinguish the typical from the atypical form, and to rate only the typical form as B. coli. More modern methods have shown the general tendency to make less difficult and more approximate tests of the occurrence of B. coli. They involve the use of the lactose peptone bile "presumptive test" as sufficient to show B. coli and to rate coli or coli gas-forming bacteria by this test. There is a good deal to be said for these simpler, speedier methods now used, even if they are not a positive guide for separating the percentage of B. coli which comes from the intestines of man, as distin guished from the bacteria reacting similarly in the laboratory and coming from the intestines of other domestic animals, grains, soils, etc. The percentage of organisms of the coli group, rated as positive by the presumptive test that would be considered typical B. coli by confirmatory tests, varies from at least thirty to ninety per cent., according to various workers testing different waters. Quite possibly in some waters the percentage of B. coli of typical forms would be much less than this. It is to be understood that even with this small percentage of typical B. coli there is no evidence to show what portion, if any, of the B. coli originated in the intestines of man. In Table III the relations of the B. coli result by different methods of testing are given. These figures by Professor Gärtner are taken from Dr. Houston's tests of London water. Three methods of testing are given: the first, the so-called presumptive test, is noted from the gas formation; the second column involves a so-called confirmatory test, as applied by Dr. Houston; and in the third column are given the results of so-called typical tests. Comparing the results given by the presumptive test and by the typical test for one cubic centimetre, we see that for the Thames raw water before storage the typical test showed almost as much coli as the presumptive test, or, in figures, about eighty per cent. as much. In the Lambeth Reservoir, after storage, the typical coli are only about fifty per cent. of those shown by the presumptive test. In the Chelsea Reservoir, after storage, the typical coli are shown to be only about forty per cent. as much as the presumptive coli. In the LaValle Reservoir, after storage, the typical coli are less than twenty per cent. of the presumptive coli. Generally speaking, the raw water shows a higher percentage of typical coli, and storage waters, which are largely purified by long storage, show a smaller percentage of typical coli. This may be because some, even a large part, of the original coli lose their typical properties in age, as is often claimed to be the case. It may be because the typical coli are less long-lived than the atypical coli. It might also be inferred from this table that the typical coli are better indicators of the water quality than the atypical coli. TABLE III. RELATION OF B. COLI RESULTS BY DIFferent MethODS. Dr. Houston, in a recent report on the value of storage, points out that storage results in a devitalization of the undesirable bacteria, on account of the water being an unsuitable medium for sustaining their activities. He speaks of this as the Glagesten" theory in the storage of impure water. Coli tests should be used most cautiously in forming an opinion as to the quality of an untreated raw water. Such data should be weighed carefully in connection with a sanitary survey in the light of existing conditions as to the opportunity for pollution. Realizing the inability to record the source of bacteria producing positive results by the coli tests, particularly as to whether or not they originated in the intestines of man, it is desirable to record what is available by way of information as to the relative prevalence in water of B. coli and B. typhoid. These results, so far as they have been studied, are quite erratic and unsatisfactory. We will make some mention of several comments, but before doing so it is desired to point out some of the irregularities known to the speaker due to changes in methods of determining B. coli, as follows: In 1898 the speaker had charge of the testing station at the Cincinnati water works. The water was drawn from the intake at the old Front Street pumping station, adjoining the Pennsylvania Railroad train shed. Into the Ohio River above this intake there were emptied the contents of thirty-two city sewers, and six miles above the intake the Ohio also received the flow of the Little Miami River, draining a populous watershed. On the basis of the determinative tests made for B. coli then in vogue it was found that sixty per cent. of the samples showed the presence of this organism when one cubic centimetre of the water was tested. (See the speaker's report on Water Purification at Cincinnati, 1899, p. 40.) In 1907 the intake was moved to the present site at California, seven miles above the old intake and one mile above the mouth of the Little Miami River. A marked falling off in the typhoid occurred during the summer of that year, and before the filtration plant was put in service in October of the same year. Since October, 1907, the city of Cincinnati has had one of the very lowest typhoid fever death-rates of any American city. By the present method of making the presumptive tests on the untreated Ohio River water at the upper intake now in use a positive result in one cubic centimetre samples is reported in eighty-five or ninety per cent. of the daily samples tested. No increase in population in Pittsburgh or other cities above can possibly explain the change in analytical data as compared with those obtained in 1898. At New Orleans, during the operation of the water testing station, Mr. R. S. Weston found in 1901-1902 that out of one hundred tests for B. coli he obtained positive results only three times when using from one to three hundred cubic centimetres of water per sample. (See Report on Water Purification Investigation, New Orleans, 1903, p. 47.) By the presumptive method as now practised positive B. coli results are obtained in eighty per cent. or more of the one cubic centimetre samples. The present intake is above the city limits, and there has been no striking increase in the population of the cities of the Mississippi valley sewering into this river of great size. The cities of Baton Rouge and Vicksburg have more extensive sewerage than a dozen years ago, but their effect upon pollution at New Orleans can scarcely be measurable. INTERNATIONAL JOINT COMMISSION STANDARD. During the course of this last year an International Joint Commission, consisting of members appointed by the United States and by Canada, have been considering the question of the allowable pollution in international boundary waters. The particular features they were studying were to what extent sewage and other polluting discharges into these boundary waters should be reduced in amount and purified, and to what point these boundary waters might be allowed to be polluted and yet at the same time not occasion, in connection with water purification plants, any menace to public health. In the course of this investigation this International Joint Commission obtained testimony from a number of sanitary engineers, and formulated a tentative standard of purity for these international boundary waters with respect to the proposition of not overloading water filters. The problem here was not the one of determining what coli count was permissible for a drinking water. That problem, indeed, could not have readily permitted a general solution in the light of our present knowledge. The conditions of such surface waters as form the international boundary waters are such that safety demands that all of these waters be properly filtered before being used as domestic supplies. The problem, |