inlet as the fall of sewage level when the tank is emptied, or else that pumping be resorted to for filling or emptying the tank. The sewage is not perfectly quiet in most cases, but continues, with constantly diminishing velocity, a circulating motion or eddying caused by the comparatively rapid filling. The structural difficulties and details of liquid-withdrawing appliances, combined with the loss of head, cause this form of tank to be but little used.

When the velocity is so great that only the heaviest, mineral matters are deposited it is called a grit chamber. These are sometimes desirable where the combined system is used; but in the separate system so little sand or grit is carried that they are considered by experts to be unnecessary. They are generally objectionable because of the organic matter which is apt to deposit in them and putrefy, a velocity of even 135 feet per hour being insufficient to prevent this in Columbus. Such grit as finds its way to a tank might better settle with the remaining sludge; or a bottom baffle wall in the tank near the inlet end may serve to collect the grit and its accompanying organic matter.

In a plain sedimentation tank there are to be considered, besides the inlet and outlet, the length, width, depth and general form. Except for special forms to be described, tanks are generally made rectangular. Experiments at Columbus indicated that a velocity of 50 feet per hour would permit an amount of precipitation which could be increased very little by reducing the rate. Also that prolonging the stay in the tank beyond four hours did not materially increase the deposit. These figures might vary somewhat with differences in the nature of the sewage, but agree almost exactly with the ideas of some English authorities. Accepting them, we would have a tank 200 feet long, and with an area of cross-section obtained by dividing the flow in cubic feet per hour by 50. In addition to this, allowance should be made for 1 or 2 feet of quiescent sludge in the bottom of the tank. From 6 to 8 feet depth, allowing 3 to 6 feet for depth of actual cross-section of moving sewage, is generally considered most desirable.

The width obtained by such a calculation might be taken as

that of the tank, and in a small plant probably would be. But to permit of putting a tank out of service when cleaning without intermitting the treatment, several tanks may be provided; and this is also made desirable by the tendency to the formation of cross-currents and other causes of non-uniform flow in a wide tank. Several tanks placed side by side are therefore desirable if the volume of flow exceeds say 1,000,000 gallons a day.

Such a tank will remove from 40 to 60 per cent of the suspended organic matter, and a higher per cent of suspended inorganic matter. The sludge must be removed at intervals of three to six days in summer and two to four weeks in winter, if active putrefaction is to be avoided. The sludge deposited will be 80 to 95 per cent water, and disposing of it presents serious difficulties in many cases. This will be considered in another article. While large numbers of bacteria are removed by sedimentation, the number leaving in the effluent is still so large that subsequent treatment to remove them is necessary if bacterial purification is considered.

This outlet

The super

The removal of the sludge may be facilitated by special construction or apparatus, the simplest of which is the sloping of the bottom of each tank toward a central gutter, which itself slopes toward an outlet at one end or in the middle. may lead by a pipe to a sludge pit or sludge bed. natant liquid may be drawn off with the sludge, may be pumped to an adjacent tank, or may remain in the tank after the sludge has flowed out. A scraper, of the nature of a squeegee, is sometimes used to force the sludge to the outlet, but this is generally unnecessary. In intermittent-flow tanks the effluent is generally drawn off by a hinged pipe, its free end being maintained, by a float, about 3 to 6 inches below the surface, the lower end being connected to an outlet pipe. When the effluent begins to run cloudy the remaining contents of the tank, or sludge, is drawn off into a sludge well.

Sedimentation tanks should be practically watertight to prevent pollution of the soil and undermining of the foundation. They should have a hard and smooth surface to facilitate removal of sludge. Steel plates might be used to met these requirements

but would be unnecessarily expensive and subject to rapid deterioration by rust. Brick or concrete is ordinarily employed, the latter being more common at the present time. The interior of a concrete tank should ordinarily be floated down to a smooth sidewalk finish. Brick-work should be smooth, with the joints pointed. Enameled brick are sometimes used because adhering matter can so easily be removed from them. The tanks are underground in the majority of cases, since the surface of sewage in them is practically at the level of the flow line of the sewer, except when it is necessary to pump the sewage.

Sedimentation tanks are sometimes roofed over, in other cases they are left uncovered. Roofing is somewhat expensive, especially where the tank is large. It offers the advantage of protecting the tank from winds, which would create eddies and currents in a large tank, which would interfere with sedimentation; they maintain a more uniform temperature, preventing the surface of the sewage from freezing (although this is likely to occur only in very cold climates); and conceal the tanks from view and prevent the diffusion of odors, thus palliating imaginary or real offenses to sight and smell. For small tanks an ordinary frame roof, with the gables closed in, will ordinarily serve the purpose, although a more durable and ornamental structure may be obtained by the use of masonry and a slate roof. For larger tanks a more common construction is a concrete or brick roof of groined arches supported by masonry pillars resting at regular intervals upon the floor of the tank. If, as is desirable, a large tank is divided by longitudinal walls into a series of narrow tanks, pillars are unnecessary and either frame or masonry roofs may be supported on the partition walls. Given the dimensions calculated as indicated above, the remainder of the design and construction of a sedimentation tank would be similar to that of any like structure for containing water; except for the special inlet and outlet constructions, as already described.

On account of popular prejudice, as well as to reduce the cost of the considerable area occupied by horizontal tanks, they will generally be placed as far as possible from built-up sections.

Where this cannot be done, the area required can be reduced by use of a vertical tank.

In a vertical tank the sewage flows vertically downward to nearly the bottom of the tank, then outward under a suspended wall and upward to the outlet, the area of the upward flow being much greater, and consequently the velocity much less, than of the downward. This assists the precipitation of suspended matter. The "Dortmund" tank is of this type. The upward velocity is at the rate of .005 to .01 foot per second. Experience shows that these remove less organic matter than do horizontal-flow tanks, and few are used.

ART. 74. TANK TREATMENT. PRECIPITATION

To hasten the sedimentation and render it more thorough, as well as to remove a part of the matters in solution, chemicals are sometimes added to the sewage. It was at first thought that by chemical treatment a large part of the organic matter in solution could be rendered insoluble and precipitated, and Slater cites over 450 patents granted in England for chemicals to be so used. It is now generally recognized, however, that practically only the solids in suspension and 5 per cent to 15 per cent of those in solution can be removed by this method. As only about one-fourth of the total solids are in suspension, it is evident that but a small percentage of them is removed, although these may include half of the organic matter.

Precipitation is largely or entirely a physical process. When lime, for instance, is added to sewage it unites with the carbonic acids to form carbonate of lime, and with sulphuric acid, if any be present, to form sulphate of lime or gypsum; both of which are insoluble in water and settle to the bottom of the tank, entangling and carrying down with them flocculent matters in suspension. If a large amount of lime be used, calcium hydrate instead of carbonate is formed, clarifying the sewage. Sufficient lime generally remains in solution in the carbonic and other acids to render the sewage alkaline. If iron sulphate or aluminum sulphate be added to sewage thus made alkaline, a flocculent

precipitant of hydrate of iron or hydrate of aluminum is formed which seems to precipitate slightly more of the soluble matter than does lime. Ferrous sulphate seems to be useless without the addition of lime to combine with the excess of carbonic acid and with the sulphuric acid of the ferrous sulphate. Ferric sulphate is more readily precipitated and more completely insoluble than the ferrous salt, and the use of lime with it is not so necessary; as is also the case with aluminum sulphate or crude alum, ordinary sewage containing enough alkali to decompose these salts. It is found that if more lime is used than will combine with the carbonic acid in the sewage, no benefits result from the additional lime; and the free lime is objectionable because of the danger that it will kill fish in the water reached by the effluent, and that it will cause a secondary precipitation in the effluent or stream which receives it. With ferric and alum salts, however, the precipitation increases with the amount used, though at a less rate after a certain point is reached. Some sewage containing industrial wastes, such as that of Worcester, Mass., contains so much ferric sulphate that it is useless to add more.

Of the great number of materials (not necessarily "chemicals" in the popular use of the word) proposed, only a few have been found practicable, many being too expensive. Lime, ferrous and ferric sulphate and sulphate of alumina are believed to be the only ones used in this country. A few patented preparations are used in England. From tests made by the Massachusetts State Board of Health certain conclusions were reached as to relative effectiveness and cost, which are given in Table No. 24. In each case, the time allowed for sedimentation was one hour. The calculations of cost were based upon the following unit costs in the year 1908: Lime (70 per cent available CaO), $6 per ton. Copperas (55 per cent available FeSO4), $10 per ton. (Sugar sulphate of iron, containing 64 per cent available FeSO4, can be used, reducing the cost about 15 per cent). Crude alum (58 per cent available Al2(SO4)3), $20 per ton. Ferric sulphate, $27 per ton.

Worcester, Mass., and Providence, R. I., are believed to be the only cities in this country now using precipitation. Owing