Now if the minute free-swimming animals thus increase when a certain amount of sewage supplies them with ample food, it is quite obvious that they must have a most important influence in removing objectionable impurities. The number of the excrements of entomostraca in the recent mud of such rivers as the Thames is most surprising. In one specimen, from Hammersmith, I found that there were more than 20,000 per grain; and the average number at Erith, in August, 1882, was above 7000, which is equivalent to about 200,000 per gallon of water at half-ebb, from the surface to the bottom. This enormous number must represent a very large amount of sewage material consumed as food; and though, as in the case of larger animals, a considerable part of their excrements no doubt consists of organic matter capable of putrefaction, yet there can be no less doubt that the amount entirely consumed in the life processes of the animals is also great. As named above, I kept cyclops alive for many months by feeding them on human excrement. It is thus easy to understand why, when they abound in the Thames, the relative amount of human excrement is very considerably less than in the winter, when their number must be much smaller. We thus appear to be led to the conclusion that when the amount of sewage discharged into a river is not too great, it furnishes food for a vast number of animals, which perform a most important part in removing it. On the contrary, if the discharge be too great, it may be injurious to them, and this process of purification may cease. Possibly this explains why in certain cases a river which is usually unobjectionable may occasionally become offensive. It also seems to make it clear that the discharge of rather too much sewage may produce relatively very great and objectionable results. Though such comparatively large animals as entomostraca may remove much putrefiable matter from a river, we cannot suppose that, except incidentally, they remove such very minute objects as disease germs, but it would be a subject well worthy of investigation to ascertain whether the more minute infusoria can, and do, consume such germs as a portion of their food. If so, we should be able to understand how living bodies, which could resist any purely chemical action likely to be met with in a river, could be destroyed by the digestive process of minute animals. Hitherto I have had no opportunity for examining this question critically, but have been able to learn certain facts which, at all events, show that it is well worthy of further examination. It is only during the last month that I have paid special attention to the number of the larger infusoria, and various other animals of similar type, met with per gallon in the water of rivers and the sca, which can be seen and counted by means of a low magnifying power. At low water in the Medway above Chatham, in the first half of June, the average number per gallon has been about 7000, but sometimes as many as 16,000. Their average size was about th of an inch. Possibly the number of still more minute forms may be equally great; but, even if we confine our attention to those observed, we cannot but conclude that their effect in removing organic matter must be very considerable; and judging from what occurs in the case of larger animals, those th of an inch in diameter may well be supposed to consume as food, particles of the size of germs. Up to the present time, I have however collected so few facts bearing on this question, that it must be regarded merely as a suggestion for future inquiry. 000 So far, I have referred exclusively to the effect of animal life. Minute plants play an important part in another way. The number per gallon of suspended diatoms, desmids, and confervoid algæ is, in some cases, most astonishing, and they must often produce much more effect than the larger plants. As far as I have been able to ascertain, their number is, to some extent, related to the amount of material in the water suitable for their assimilation and growth. In the mud deposited from pure rivers their number is relatively small, but in the district of the Thames, where the sewage is discharged, I found that in summer their number per grain of mud at half-ebb tide was about 400,000, which is equivalent to above 5,000,000 per gallon of water. This is two or three times as many as higher up or lower down the river, and, out of all proportion, more than in the case of fairly pure rivers like the Medway. Their effect in oxygenating the water must be very important, since, when exposed to the light, they would decompose carbonic acid, and give off oxygen, under circumstances most favourable for supplying the needs of animal life, and counteracting the putrefactive decomposition so soon set up by minute fungi when oxygen is absent. Taking, then, all the above facts into consideration, it appears to me that the removal of impurities from rivers is more a biological than a chemical question; and that in all discussions of the subject, it is most important to consider the action of minute animals and plants, which may be looked upon as being indirectly most powerful chemical reagents. ON THE CHEMISTRY OF POTABLE By PROFESSOR ODLING, M.B., F.R.S., F.R.C.P. I. SOURCES OF WATER SUPPLY. THE water of most towns is derived, as is well known, from one or other of three typical sources, constituted respectively by springs, or streams, or lakes, this last named source being taken to include not only natural lakes, but also those huge artificially formed reservoirs, in which the surface water collected from more or less extensive gathering grounds, is retained by means of embankments, and stored up for distribution to consumers. Although some streams, indeed, are fed largely by springs, and some lakes are little else than local expansions of streams or rivers, still a general distinction in character between supplies derived from the three just spoken of typical sources, is, on the whole, broadly recognisable; and this, notwithstanding the fact of the considerable differences in composition manifested by individual spring-waters, derived from different springs or wells, by individual stream-waters, derived from different streams or rivers, and by individual lakewaters, derived from different lakes or reservoirs. Despite the marked preference of a few authorities for some one variety of source to the exclusion of the others, the majority of engineers have come to recognise each variety as having its own characteristic excellencies and concomitant defects, and are accordingly ready to avail themselves, in different cases, of whatever kind of source is rendered by local circumstances most available and suitable for the particular town to be supplied. From all three varieties of source alike, numerous large populations have been furnished for generations past, with an abundant, and, as the result has shown, with a satisfactory and wholesome supply of water; although, indeed, in a few special instances, supplies from all three varieties of sources alike have, in different ways and under exceptional circumstances, proved detrimental to the health, and even, in certain cases, fatal to the lives of some among the population supplied. All three varieties. of source have, under the best advice of the time, been continually resorted to from the earliest period of water supply undertakings, nearly three centuries ago, down to the present day. Thus, while the youngest of the companies supplying London has obtained its supply, since 1862, wholly from deep wells sunk in the chalk, the new works for the supply of Liverpool and Manchester respectively, which may both take rank among the greatest hydraulic undertakings of the century, are intended to introduce and distribute river-water in the one town and lake-water in the other. All experience, indeed, goes to show that the supply of excellent water is not confined to its supply from any single variety of source; and that for the purpose of water supply, there are alike good wells and bad wells, good lakes and bad lakes, good rivers and bad rivers. Just, moreover, as the best of river-waters or lake-waters will fall short in regard to an excellency characteristic of well-water, so will the best of well-waters fall short in regard to an excellency characteristic of lake-water or riverwater. The comparison of the one sort with the other sorts of water must be made as a whole. To select arbitrarily the special good point of one sort of water as a standard, to the neglect of the countervailing good points of other sorts, and to measure the quality of the other sorts solely in reference to this selected standard, is clearly the conduct rather of an advocate than of a judge. 2. PURITY OF WATER. Water, as conceived of by the chemist, is a definite conpound of 100 parts by weight of oxygen, united with 12.5 parts by weight of hydrogen. So exceedingly difficult is it of production, even if it ever has been produced in an absolutely pure state, that it may be regarded rather as an ideal than a real chemical substance. All natural water, besides the matter, never entirely absent, which it holds in suspension, is a solution of various mineral matters, of various organic matters, and of various gases in the ideal water, or protoxide of hydrogen, of the chemist. Of the different kinds of foreign substance habitually or exceptionally present in natural water, some kinds are beyond question prejudicial, and in particular cases highly prejudicial; most kinds are simply innocuous; while not a few kinds, as saline matter in moderate proportion, and more particularly dissolved aerial matter, are positively beneficial. Looked at, however, from a strictly chemical point of view, any matter whatever foreign to the ideal chemical compound, water, constitutes an impurity of the natural water in which it occurs, whether this foreign matter be a prejudicial or a beneficial constituent, whether it consist of sewage matter, |