ing both sources of supply, the latter for drinking purposes, and the former for domestic and manufacturing uses. the opinion of the writer, a system of dual supply is not practicable for domestic purposes, but only for municipal or manufacturing requirements. In cases where water is used for power, a dual supply may also be considered indispensable, as the great pressure necessary for its economical use for this purpose is undesirable for a domestic supply. Again, a separate supply for municipal purposes can be used for road watering, sewer flushing, fires, &c., without affecting the domestic supply. In most manufacturing towns the factories are grouped together, and require large quantities of water at a very cheap rate, which, supposing the water is not used in any way for alimentary purposes, can often be supplied from sources sufficiently pure for purposes of manufacture, although the water may be unfitted for household consumption. Such a supply exists at Roubaix, in France, where the cheapness of the water contributes to the prosperity of the town.

Paris also is supplied in the same manner, there being a double set of mains in the streets, one for domestic purposes, and the other for street watering, &c. The supply of potable water, however, is limited, and in time of drought has to be supplemented by water drawn from the Seine.

A dual supply for domestic purposes would entail a considerable extra outlay on the construction of works, as a double line of mains would be necessary in every street, and the cost of most of the house fittings would be doubled, as they would have to be in duplicate. Another objection is the probability that the excellent quality of the water supplied for potable purposes would cause the quality of the other to be little considered, and the latter might deteriorate, or become subject to great contamination. Consumers, bearing this in mind, would probably use the purer water almost entirely. In ordinary times, this might not be of much consequence, but in times of drought it might be serious. This difficulty, it is reported, has just

occurred in Paris, which, as just stated, is supplied under this system, and where now, in the heat of summer, with the cholera possibly approaching the town, the drinking water is said to be failing, so that nearly all the supply will have to be taken from the other source, the Seine. Unfortunately, this water is not as carefully filtered and aërated as it would be if it were usually used for domestic purposes, and it may consequently prove a source of danger and disease.

When only one supply exists for domestic purposes, means are generally taken to render it as pure as possible, either by efficient filtration, or by such other means as may be found necessary. Should such water not be considered of a sufficiently high standard of quality, a pure source is selected, and the old one abandoned; when this is impossible, the use of household filters becomes universal.

When the Chelsea Water Company first moved their intake to Ditton, in carrying out the works an extra reservoir was constructed, and extra mains were laid for supplying unfiltered water for road watering and other purposes, but in practice this arrangement was found to be no saving, and to possess no advantage over supplying filtered water for these purposes, so it was discontinued, and now none but filtered water is supplied.

It may be contended that if water were sold by meter, the difficulty of supplying the two kinds of water for domestic purposes would be overcome, inasmuch as the extra cost of the purer water would prevent its excessive use, and so conduce to its conservation and the greater use of the inferior water. This argument, probably, would prove fallacious, as the servants who use the water are unaffected by any such considerations.

It might, further, be urged that the inferior water should only be laid on to the closets, and that the supply to them should be independent of the general domestic supply. This arrangement would leave the remainder of the domestic supply to be furnished by the better water. Baths even would have to be supplied from it, for when it

is remembered that they are very frequently supplied with hot water from an apparatus from which water is drawn for other purposes, it will be seen that it would be practicably impossible for them to be connected with an inferior supply. Apart from special cases, in which water is distributed under great pressure for power, and under smaller pressure for ordinary consumption, only one plan of dual supply seems feasible, which, however, from the necessity for duplicate mains, &c., must necessarily be costly, though it might prove advantageous in many respects if carefully worked out. The writer would hesitate to advise such a scheme, except under exceptional circumstances. It is that the domestic supply should be entrusted to a company selling water by meter, and the general supply (including supplies to closets) should be in the hands of the municipal authority, and paid for by a general rate levied on all property. In addition, the municipal authorities would receive payment for any water they might dispose of for manufacturing purposes. It will be seen that, under this arrangement, the municipal authority would have under its control water-works (supplying fairly good water it must always be presumed) for supplying closets, road and public garden watering, sewer flushing, fire extinguishing, and to sell for manufacturing purposes. The company which supplied water by meter for domestic use, and for those manufacturing purposes for which pure water is indispensable, should also be under the supervision of the municipal authorities.

Such a system as the above would probably only be found worthy of consideration in the case of a town which possessed a supply of good water, but of so limited a quantity that the rapid growth of the city rendered an increased supply indispensable, and where there was an inferior supply near, which could be made available at small cost. By these means the consumption of the pure water would probably be reduced to such an extent as to be available for a much larger population, and at the same time, from the fact of the municipal authorities controlling

the house drains, closets, and flushing supplies, an improved sanitary condition might result. It is, however, not easy to imagine a town in such a position.

Rain water, as it falls, is in a comparatively pure state, and if proper measures are taken to preserve the surplus falling in wet seasons, and to keep it from contamination, it will generally be found that the rain falling in each watershed is amply sufficient to supply the whole of the inhabitants dwelling in their districts. If any scarcity exists, it is owing, generally speaking, to the want of proper care, or to the fact that the district is so subdivided under different authorities, that joint action for the proper conservation of the rainfall is impossible.

MODE OF DISTRIBUTION, WITH SOME REMARKS ON DUAL SUPPLY.

By HENRY J. MARTEN, M.Inst.C.E.

UNTIL comparatively recent times no attempts were made to afford what we now term "a domestic supply," that is to say, a supply of water distributed to and available within each house for general domestic purposes.

The large supplies of fine water introduced into the principal cities of ancient times, by means of the magnificent aqueducts of which there are many remains to this day, were directed, so far as distribution is concerned, to the supply of public baths and fountains, and possibly here and there to the supply, for similar purposes, of a few of the private residences of the wealthy classes.

A house-to-house supply, such as we have in modern times, was unknown and undreamt of. The supply for domestic purposes had to be obtained by the occupants of the houses either from the public fountains, or from the streams flowing from them, or was supplied in detail by VOL. VIII.-H. C.

2 N

water carriers, somewhat after the fashion in which milk is now supplied from house to house. Although, according to Pliny, the ancients were aware that water under pressure rises to the height of its source, little practical advantage was taken of the knowledge of this fact. Water under pressure was occasionally conveyed for some distance through stone or earthenware tubes or in leaden pipes. The two former were, however, difficult to be kept tight, and the latter were not only very expensive, but, at an early period of their history, were found to act deleteriously on the water flowing through them. Although, therefore, immense volumes of water were frequently brought into ancient cities, as into Rome, where, according to Strabo, "whole rivers flowed through the streets," and where the total quantity so delivered probably exceeded three hundred gallons per diem to each inhabitant, there was practically no distributed domestic supply.

The inhabitants of modern London, therefore, although only furnished with a supply per head equal to one-tenth of that which used to flow into old Rome, are nevertheless infinitely better off than the ancient Roman, because the thirty gallons supplied to the modern Londoner are delivered exactly where he requires it, "upstairs, downstairs, or in my lady's chamber," and the supply is at all times available by the mere turning of a tap. We have here a good illustration showing how various combinations of applied science, united to the practical skill of modern times, have resulted in a tenfold economy, coupled, at a moderate estimate, with a tenfold advantage to the water

consumer.

If modern London were supplied in the same wasteful and unscientific manner as ancient Rome, the quantity of water required for the purpose would exceed in volume the average flow of the River Thames at Kingston, and whilst this immense volume of water would be rushing from public fountains, and down open channels made for the purpose in the streets, the inhabitants would be totally without the comfort and advantage which they at present