But if organic matter be buried too deep, or passed too rapidly into the dead earth, or if the living earth be taxed beyond its power, water charged with putrid matters may percolate for hundreds or even thousands of feet in any direction without undergoing any real purification: the coarser particles, may be arrested, but the dissolved matters remain unchanged. In this lies the danger of cess-pools and grave yards, by which towage and putrid matters gain access to the ground-water at such a depth as to escape the action of the nitrifying bacteria of the upper soil; whereas sewage and excreta dug into highly cultivated surface-soil are so completely and rapidly "mineralized", as the French say, that no trace of organic matter, as such, can be found in the ground at a depth of from four to six feet This has been conclusively proved by Dr. Poore in his own garden at Andover.

Potable Waters. - Potable waters may be divided into rain. moor, river, lake, pond, marsh, spring, surface-well, and deep-well waters, those of each class differing much in their composition and characters.

Rain-water collected in the open country is very pure and aerated, though from the absence of CO2 not sparkling; but if collected on roofs it may contain much mineral and organic matter, and in or near towns soot, sulphuric and hydrochloric acids, etc. Close to the sea it may even be salt from the spray carried by strong winds.

Moorland-waters are collected, by impounding, for the supply of town- in the northern counties of England, in Devonshire, and Wales. They are as a rule pure, though containing more inorganic and organic matter than rain-water; but when they have percolated through peat, they take up certain organic acids which exert a solvent action on lead pipes, and, as at Sheffield, Leeds, Bacup, Batley, and elsewhere, this may give rise to lead-poisoning. The best preventive measure in such cases is to treat the water with lime, in fact to harden it. The lime neutralizes the peaty acids, and by forming on the surface of the lead an insoluble coating, protects it from the solvent action that all pure soft aerated waters possess in a greater or less degree.

Lakes in mountainous countries, as Wales, Cumberland, and the Scottish Highlands, yield the purest and most abundant supplies, provided they be not stained with peat. They are purer than moorland-waters is, since the collecting areas are largely composed of bare rock, suspended matters of all kinds are deposited in the bed of the lake, and the small amount of organic impurity is removed by the animal and vegetable life in the water. Contrary to what might have been expected, many lake-waters, as those of Bala, Thirlmere, and Loch Lomond, appear to have little or no action on lead.

Lowland lakes and ponds, in the midst of cultivated and populous districts, are wholly unfit for potable purposes, being fed by storm-waters and surfacedrainage charged with organic impurities. This condemnation does not apply to such as are not exposed to these dangers. The Tegel and Muggel See, from which Berlin now derives its supply, situated in sandy plains, surrounded by pine forests and remote from human habitations, and those supplying Dantzig and other towns in northern Prussia, are fully equal to the mountain-lakes already mentioned.

Springs rising on hillsides, and upland-streams impounded near their sources, may be considered as moorland-waters that have been subjected to a process of natural filtration and purification by the bacteria of the upper earth. They may be slightly hard, but furnish waters of exceptional purity.

Rivers for the most part owe their origin to such springs and streams, or to the overflow of mountain-lakes, but as they proceed seawards they receive an enormous accession to their volume from springs, affluents, surface-drainage, and storm-waters, together with the sewage of towns, river-side houses, and barges. The purity of rivers is to a great extent dependent on their length and volume. The Rhine at Bonn, t'<»r example, though it has traversed populous regions for several hundred miles, exhibits a freedom from pollution that could not have been expected from the size and number of the towns on its banks, and on those of its tributary We are thus brought face to face with the vexed question of the so-called "self-purification" of rivers. Of the reality of the phenomenon itself there can be no doubt, but its extent has been much overestimated and its nature misapprehended. It is not simply the effect of oxidation by continued exposure to the air, but it is the result of a complex process including not only aeration and subsidence, and dilution with affluent streams, ground-water, and springs, but also the action of animal and vegetable organisms - fish, entomo-straca, plants, and bacteria. It is in many respects analogous to what has been described as taking place in the living earth.

Rivers flowing through alluvial deposits always contain a certain amount of suspended matters, chiefly clay, which may even render filtration difficult, and after floods they are often turbid with mud brought down by storm-waters, or lifted from their beds by the violence of the current. Chlorides are usually more abundant than in spring-waters or lakes, the excess being due to the presence of sewage, which is indicated also by ammonia and the so-called albuminoid ammonia, or organic matter in a crude state. As a general rule rivers are not to be recommended as sources for the public supply of towns, a fact which is yearly becoming better recognized in all civilized countries, though filtration may do much to render such waters fairly pure and safe. At the same time, it must be admitted that no ill results have been traced to the use of the Thames

WELLS water in London, since the intakes were removed to above Teddington weir, where the tidal currents cease to be Celt

Surface-wells - which need not be shallow as measured in feet - are such as are sunk, in a pervious surface soil, to the ground-water resting on the first impermeable stratum. They are always to be viewed with suspicion, for though the surroundings may be faultless, and all organic matter nitrified before reaching the level of the water, this may be contaminated at some point in its course thither, as by cess-pits leaking at a depth where soil is dead, or by a straw-yard or dunghill, the earth beneath which is so saturated with impurity as to be powerless to nitrify the organic matter. Those wells in which the water rises nearly to the surface, arc specially dangerous. In short, the conditions under which alone a surface-well may be considered unobjectionable, are that the soil shall be of a kind to purify the water sinking through it. that the level of the groundwater be at no time less than six feet from the surface, that no source of pollution exist within some hundreds of yards (especially in the direction from which the water flows), that the sides of the well be absolutely impervious for a depth of at least eight to twelve feet, that surface-water and dirt be excluded by a parapet about eighteen inches high and a stone slab covering the mouth of the well, and that the water be raised by a pump, and not by buckets, which are frequently the means of fouling it. The plan first suggested, I believe, by Dr. Woodforde, Medical Officer of Health of Berkshin - of lining surface-wells by a series of large glazed stoneware sewer-pipes placed one over the other with the flanges downwards, well jointed with cement, and backed by a layer of concrete on the outside to give additional strength, is much to be preferred to the porous brick-and-mortar steining commonly used.