the whole of the several individual constituents of the organic matter, taken together-is not subject to any such considerable range of variation, and that it may, without risk of serious error, be valued at about 40 per cent. Disregarding then for the moment the consideration of the nature of the dissolved organic matter present in natural water, and confining attention only to its quantity, it may be taken as admitted that the determination of the organic carbon in water is an absolute, and more than fairly accurate determination; that the organic matter of water is in a general way proportionate to the amount of its organic carbon; that the amount of organic matter may accordingly be represented by some multiple of the organic carbon; and that for purposes of comparison, at any rate, this multiple may be taken provisionally at 2'5, corresponding, of course, to the occurrence of 40 per cent. of carbon in the organic matter. The inquiry next presents itself as to what, on the basis of these propositions, is the quantity of organic carbon, and consequently of organic matter present in the three classes of water derived respectively from each of our typical varieties of source. Speaking generally, it may be said that the proportion of organic matter is decidedly least in spring or well-water-that is to say, if we limit our attention to such waters only as would be taken for town supply, for there are, of course, foul well-waters, in which the amount of organic matter is in excess of anything ordinarily met with in the water of lakes and natural rivers. As between lake sources in general, and river sources in general, it is not easy to assign the order of seriation. The proportion of organic matter found in different lake-waters is, on the whole, more uniform than the proportion found in the waters of different rivers. Accordingly, while in some river-waters the proportion of organic matter is somewhat higher, there are other river-waters in which it is very considerably lower than the proportion commonly present in lake-water. On the whole, it would seem that riverwater in general must take precedence of lake-water in respect to the smallness of its proportion of dissolved organic matter, subject, however, to the observation that the proportion of this constituent is more variable in riverwater-that it is liable to a greater range of variation, both as between the water of one river, and the water of another, and as between the water of the same river at different seasons. In order to give some idea of the quantities of organic carbon, and consequently of organic matter, present in the water, furnished from different varieties of source for town supply, I have made an abstract of the results set forth in the Registrar-General's monthly reports for the last year and a half, beginning, that is to say, with the report for January, 1883, and ending with the report last made, for June, 1884, in respect to the supplies of London, Birmingham, and Glasgow. I have resorted to the reports of the Registrar-General in part, because they alone furnish the results of a regular series of analyses of the water supplied to other towns than London; and, in part, because, with regard to London, I have, for the purpose of this address, preferred to bring before you the results of analyses for which I am not in any way responsible. There is not, however, any appreciable difference between the mean results for the period obtained by Dr. Frankland, on the one hand, and by my colleagues and myself on the other, in respect to the composition of the London waters examined by both of us. The five Thames waterworks companies, as is well known, take their supply from the river at Hampton, Ditton, and Molesey. They furnish just under 50 per cent. of the total supply of the metropolis. The East London Waterworks Company take their supply from the River Lee, some distance below Ware. Their subsiding reservoir at Walthamstow has the enormous area of 222 acres, equal to that of a good-sized lake. The company have power to take in addition ten million gallons of water daily from the Thames at Sudbury. It is but seldom, however, that they resort to the Thames at all; and very rarely indeed to the extent of more than one quarter of the daily amount they are privileged to obtain. Substantially, the water of MEANS OF EIGHTEEN MONTHLY ANALYSES, 1883-84. It will be seen that the organic matter of the Kent Company's water, which is a spring-water, is under onetenth of a grain per gallon; that the organic matter of the New River Company's water, which is mainly a river-water, is considerably under two-tenths of a grain per gallon; that the organic matter of the East London Company's water, which is a river-water, that of the Birmingham Corporation's water, which is a mixed water, and that of the Glasgow Corporation's water, which is a lake-water, are alike about two and a-half tenths, i.e., a quarter, of a grain per gallon; while the organic matter of the Thames Companies' supply of river-water is under three-tenths of a grain per gallon. It is to be noted, however, that, although the average proportion of organic matter in the Thamesderived water supplied to London, is a little in excess of that in the Birmingham and Glasgow Corporations' supplies, the excess is entirely due to the effect of the winter floods. Comparing the results in the summer nine months, March to September, 1883, and March to June, 1884, the proportion of organic matter in the Glasgow Corporation's supply is somewhat in excess of the proportion found in the Thames supply, the number of grains of organic matter, being for the Thames derived water 22 and for the Glasgow water 25 grain per gallon. Similarly, the organic matter of the East London Company's water, during the summer months, falls appreciably below that of the Glasgow Corporation's supply at the same season of the year, the season, that is to say, during which a high character of water is considered to be more especially demanded. The above statements, as to the particular fractions of a grain of organic matter ordinarily present in a gallon of different kinds of water, serve to convey some idea of the always exceeding smallness of the quantity. A better notion, however, of the minuteness of even the highest proportions of organic matter found, say, in any London water, is afforded by stating the results in another way. Thus, if we suppose for an instant that the Thames companies' water, instead of containing under three-tenths of a grain, contained seven-tenths of a grain of organic matter per gallon-a maximum which has been occasionally approached in the supply of one or other of the Thames companies at a period of flood-even this exceptional proportion would but correspond to the presence in the water of exactly the thousandth part of one Whether or not, variations per cent. of organic water. within the limit of such a small proportion of dissolved organic matter present in potable water, ranging from about the one-thousandth part of one per cent. exceptionally met with in a Thames-derived water, down to the eight-thousandth part of one per cent. habitually met with in a chalk-spring water, are matters of any real significance, must obviously depend on the character of the dissolved organic matter present in the different waters. It will suffice for the present to observe that, so far as mere quantity of organic matter is concerned, the water supplied to London from the Thames and Lee takes, on the whole, precedence of the highly reputed, and deservedly reputed, water furnished to Glasgow; as, doubtless, it will take precedence of the water about to be furnished to Manchester. As regards range of variation, it is noticeable that, while the mean proportion of organic matter present in the summer supply of the Thames Companies, as calculated from the average of forty-five analyses reported to the Registrar-General, is 22 grain per gallon; and the mean proportion present in the yearly supply, as calculated from the average of ninety analyses, is 28 grain per gallon; the proportion in one particular sample of water out of the ninety samples analysed, fell to 17 grain, in two samples it rose to 50 grain, in one sample to 53 grain, and in yet another sample to 60 grain; but that in no one sample out of the ninety did it reach the maximum of 70 grain per gallon, so as to constitute the thousandth part of 1 per cent. of the water. It happens similarly in nearly all natural waters, that while the absolute variation in the proportion of organic matter present at different times is almost infinitesimally small, the relative variation is, on the other hand, strikingly large. Thus, the proportion of organic matter present in the Glasgow Corporation's lake-water at different times varies as 1 to 2; that in the Kent Company's spring water, in the East London Company's water, and in the Thames Companies' water, varies as 1 to 3 or 3; that |