over this long conduit must make an appreciable addition to the cost of water. Nothing, it may be safely contended, but dire necessity would lead any legislature to sanction so costly and so perilous a scheme.

Without, therefore, at this stage of the inquiry, proposing at all to enter into detail, we hope that we have shown reason for the view, that the physical limits put by Nature herself on the water supply of the Thames valley are such as to indicate that in a century hence we shall only be able to prevent want by preventing waste. On the 122 miles area of London, after deducting more than one-fifth of the surface for roads, streets, and squares, 150,000,000 of metric tons of rain fall within the year.' This large quantity is now absolutely wasted; and at times is allowed to do incalculable damage. It adds to the cost of the main drainage, both by the increased size which it has rendered necessary in the sewers, and by the greatly increased pumping at the outfalls. It would supply, if economised, the need of 3,000,000 souls. This regards the metropolitan area alone. We have seen to what the water supply from rain over the whole Thames district amounts. Is it too much to ask of the chemists and the engineers of the day to arrest, before it flies off through the air, as much of the 4 milliards of tons of evaporated water, of which we have spoken, as may be needed for human wants?

The question of store for so large a quantity of water is, however, one of extraordinary gravity. To hold a day's supply for London would require a reservoir 1,000 feet square and 24 feet deep. All the reservoirs and all the filtering-beds of all the companies united would only cover rather more than four times the space-their aggregate being about 548 acres. These reservoirs and filter-beds, however, are recipients of a daily supply from pumping. In the case of depending on the rainfall, it would not be safe to provide for less than three or four months' consumption. The smaller period would require 18 square miles of reservoir, of the depth of 24 feet; or 9 miles of double that depth. The imagination flies at once to the lakes. But Windermere is only ten miles long, and about half a mile broad. Ullswater, with nine miles of length, and three-quarters of a mile of width, has a depth of 210 feet. But its feeders are not very important, and its height of 477 feet above Trinity highwater mark is such as to prevent it from receiving a very ample supply of water. If Ullswater were of the great depth of 210 feet over its entire area, its contents (independent of rainfall during the time) would be exhausted in supplying London for 200 days, and the problem of its capacity would be thus reduced to that of the area and rainfall of its watershed. The depth of Windermere is not known, or at least is not stated by Mr. Ansted; nor does it appear that the little river Eamont, which conveys the outflow of Ullswater into the Eden, has been gauged. These facts are cited, not with a view of entering into the question of the capabilities of the lake system of the North of England, but as bearing on the subject of

water storing. If the largest English lake, regarded as a reservoir, would be inadequate for the source of an inexhaustible water supply for the metropolis, no one familiar with the levels of the Thames watershed basin will conclude that it is practicable to construct within its area reservoirs that should be adequate to contain a three months' provision of storm water. The great depth of Ullswater would not be attainable by artificial means. The level of any reservoirs constructed to receive the flood overflow of the affluents of the Thames must be such as to render any great depth in them impossible. The area at present occupied by the reservoirs and filter-beds of the water companies of London is, as just remarked, 548 acres, as nearly as the published returns allow of an estimate being formed. The total contents of the whole of these basins is equal to about ten days supply of water to the entire metropolis. To provide for four months storage would thus require about twelve times that area, or more than ten square miles area of reservoirs. At 250l. per acre the cost of land for this purpose would amount to 16,000,000l., independent of excavation, embankment, masonry, and works of all kinds. It is evident that the construction of store reservoirs, to preserve the floods of the Thames Valley for the consumption of the metropolis, would be a work of impracticable magnitude.

This appreciation, however, is but one of a numerous class of considerations which point to the fact, that when density of population has reached a certain limit, its requirements cannot be adequately met by the method of concentration. On the contrary, it is only by division and distribution that problems of this magnitude can be solved. We find writers and speakers objecting to one source of water supply, proposing another, or even imagining a third. Meantime we are practically drawing upon them all; and we must look forward, if we do not shut our eyes to the future, to a time when we may so far exhaust all, that we shall be no longer able to rely on the wells, springs, brooks, and rivers of the district to quench the thirst of its inhabitants and to supply their domestic needs.

We shall then have no resource but to economise some of that water that is now wasted. As to that portion of it which is wasted in floods, which we may roughly take at some 12 per cent. of the rainfall, the means of saving it are, as we have just seen, so costly and so cumbrous, that nothing but necessity could drive us to make use of them. In France, where the character of the country in many parts is such as to allow of the construction of reservoirs in the mountain valleys, at moderate cost, and of considerable capacity, the plan has been tried, and has been abandoned.. In India, no doubt, it has been long since, introduced in some places with great advantage; but that has been under conditions wholly different from those which prevail in the United Kingdom.

It remains, then, that we turn our attention to that vast supply of rainfall which is now dissipated by evaporation. It may be mentioned that the evaporation from the surface of water is greater than

that from the surface of land. On the other hand, the proportion of land to water surface is such as to effect an immense economy in the evaporation of water collected in pools or streams. In France, where statistics are frequently more accessible than is the case in the United Kingdom, the water surface within the shore line of the country is rather less that 1 per cent. of the whole area. Thus a square foot of lake or river represents the drainage of sixty-six square feet of land, so that even if twice the amount of evaporation went on from the water surface, as compared with that from an equal surface of the land, there would be an economy of 300 per cent. effected by the collection of the water into pools.

If, instead of open and exposed pools, the water is collected into closed reservoirs, the evaporation may be still further brought under control. But this points towards the reconsideration of that very convenient source of domestic supply of water which half a century ago was almost universally employed in England, as three thousand years ago and more it was employed in many parts of the East. The water-butt has been so poisoned by the pollution of the rainfall that washes the atmosphere of our great cities, that it has been generally abandoned. But everything concurs to show that it is rather to the construction of numerous reservoirs of small size, and to the purification of the rainfall, whether in them or as it enters them, than to any other source, that we shall hereafter have to look for the security of our domestic water supply.

The water sources of the valley of the Thames are now drawn upon by the water companies to the extent of upwards of 120 millions of gallons per diem. The original sources, the wells and springs over the city area, and the brooks and burns that fed the Thames within the precincts of the suburbs, have been for the most part abandoned, the water being poisoned by the percolation of sewage. The chalk hills of Kent yield 7 million gallons of water daily. The wells and springs that feed the New River are drawn upon to the extent of above 26 millions of gallons daily. The River Lea yields about an equal volume. The remaining 60 millions of gallons are pumped, by five companies, from the Thames.

Each of these sources of supply is capable of further development. In addition, the chalk springs of the north of London, and the pervious beds of the valleys of the Wey and of the Mole, are to a certain extent available. The five Thames companies have power to take nearly double their present rate of supply from the river. But the general total derivable from all these sources combined has above been indicated.

These, then, are the physical limits of the water supply of London. Within forty, or, at the utmost, fifty years, if matters go on at their present rate, the population of the Valley of the Thames will demand a daily supply of 240,000,000 gallons of water throughout the year. That means 300,000,000 gallons daily in July or in August. But in a year as dry as was the year 1864 it is stated that this is a larger

volume than the whole of that which would daily run down the channel of the river, including the contributions of all its affluents, whether turned through the London sewers or not, as far east as the mouth of the Medway! It will therefore be necessary, by that time, either to seek a further supply, at great cost, risk, and peril, from some other watershed, or to economise directly an adequate quantity of that large proportion of the precious fluid which we now allow to run to waste in storm floods, or to elude our grasp by evaporation.

The problem of the water supply of London has an economical as well as an hydraulic aspect. It has to be regarded as it affects the ratepayer, no less than as it affects the consumer. Equality of charge, permanence of the principle of rating, payment in proportion to consumption-all these are important details of the economical question. But before attempting to solve any of these details, it is desirable to inquire what is the present cost of London water.

As to this, there is a considerable difference in different parts of London. First, however, it is well to take the question in block. Taking the whole revenue of the eight water companies, and comparing it with the population that they supply, it results that the average annual cost of water is 68. 6d. per unit of the population. Of this 6s. 6d., 48. goes to pay interest and dividends on capital, 10d. is the cost of pumping and filtering, and 18. 8d. covers all other expenses.

The quantity supplied at that cost is rather more than 32 gallons per head, for every unit of the population, for every day in the year. If we take the round number of 30 gallons per head, which is nearly equivalent to fifty metric tons of water per head per annum, the cost might be reduced to something like 68. per head of the population.

This price, however, is susceptible of reduction. If, instead of bearing a rate of dividend which rises in some cases to 10 per cent., only 5 per cent. were charged on the capital laid out, the charge for interest, which we have taken at 48. per head, might be kept down to 38. per head. If the economy of all the water companies were as great as that now attained by the cheapest of them, the halfcrown for working expenses might be reduced to 28. And if, these economies being effected, the consumption of water was paid for by measure, and the consumer allowed to regulate the quantity that he consumed, the 30 or 32 gallons of the present daily supply would in many cases be reduced by one half, and an annual charge of 28. 6d. per head of population might then pay for the water supply of London.

It might be practicable, under certain arrangements, to devote a surplus income from the water rates to the extinction of the capital stock. In that case, in a quarter or a third of a century hence, it is conceivable that the water rate might be reduced to the charge of one shilling per head of population per annum, which we must regard as the last result of the most cheerful optimism to anticipate. Such a result, however, could only be expected to follow a thorough, competent, and exhaustive inquiry into the whole subject, No. 608 (No. CXXVIII. N. S.)

hydraulic, sanitary, and financial, of a character of which we have no example in our annals to the present day.

The entire question of the amount which it might be fair and equitable to ask us to pay for the property of the London water companies depends on a perfectly simple principle, with which men of business are familiar. It is that the value of a business depends on the net profit which it earns, compared with the capital invested in the business. If a known profit-be it seven per cent., or any other proportion is secured out of the sum annually turned over, with the increase of that annual business the annual net profit will increase. But to arrive at a true valuation, we must either deduct the ordinary interest received on money from the estimated profit, in both cases, before making the comparison; or we must deduct from the increased profit the interest due to the increase of capital.

In the whole argument for the purchase contemplated by the Bill of Sir Richard Cross, the first of these considerations received due attention, but not a word was said as to the second. To that omission was due the magnitude of the sum which it was proposed to pay to the companies, for a simple transfer of ownership, without any provision for the increased wants of the metropolis being included in the measure.

If we examine the annual reports of the Metropolitan water companies from the year 1873-4 to the year 1878-9, we shall find what has been the annual increment in capital, in revenue, and in working expenditure during that period. The annual increase of capital during the five years has been 12.6 per cent., or close upon the average rate of the increase in the population of London, that is to say, 21 per cent. per annum.

The income has increased during the same period at the rate of 21.1 per cent., or by 44 per cent. per annum. The expenditure has increased by 17.8 per cent., or by a little over 3 per cent. per annum. Thus the profit has increased by of 1 per cent. per annum, but this has been distributed over a capital increased as above stated.

Thus the growth of the property has been two-fold. It has increased in magnitude, that is to say, in capital invested, or in the number of partners in the concern. And it has also increased in net profit, and that at a more rapid rate than the increase of capital. Thus, not only have there been so many shares added to the concern, but the value of each share, averaging old shares and new together, has been raised. Or, in round numbers, instead of ten and three quarter millions of capital, earning 63 per cent. per annum, the property was represented at the end of last year by twelve millions. of capital, earning 7 per cent. per annum.

Such being, on the published data, the outcome of the history of the companies for the last five years, it requires no very great stretch of the imagination to allow for a corresponding growth of the undertakings in the future. As to this, indeed, the companies urge that