being conducted on 250,000 gallons per day of septic effluent and those at Baltimore on the effluent from a sewage previously treated by a septic tank and trickling filter. It was found that the Baltimore effluent could be satisfactorily disinfected by the use of about 75 pounds of bleaching powder per million gallons, the bacterial efficiency of this being 95 per cent, and the combined bacterial efficiency of sprinkling filter and bleaching powder being between 98 and 99 per cent. This amount of bleaching powder represents three parts per million of available chlorine. The cost of such treatment is estimated at $1.co to $1.50 per million gallons. To remove 98 per cent of total bacteria from crude sewage, it was determined, would require from five to ten parts per million of available chlorine and cost from $1.50 to $3.50 per million gallons. It was also found from these experiments that the disinfection of septic sewage required from ten to fifteen parts of available chlorine. However, it would appear that it would be very advantageous to disinfect the sewage before septic treatment rather than after; this requiring less chlorine and being equally effective in destroying the pathogenic bacteria, but leaving in the effluent the liquefying and nitrifying bacteria to continue the purification after discharge into the stream.

Mr. Phelps has prepared a table of estimated costs of treating sewage and effluents of various kinds with bleaching powder, these figures being based upon a plant having a capacity of five million gallons per day, the cost given being that per million gallons. The treatment is classified according to amounts of available chlorine used, and these are considered to apply as above stated, namely, from two to five parts for filter effluents of varying quality; from five to ten parts for sewages, and from ten to fifteen parts or more for septic sewages.

It is seen that the costs of disinfecting by bleaching powder given by Phelps are only about one-third to one-fifth of those given by the Ohio Board of Health. A considerable part of this difference is probably due to the difference in size of the plantsfive million gallons per day in one case as compared to from 40,000 to 160.000 gallons per day in the Ohio plants. Moreover,

the bleaching powder is assumed in the Ohio report to contain but 25 per cent available chlorine, and to cost 2 cents per pound. An estimate of the cost of a plant for one of these towns, with an assumed flow of 600,000 gallons per day, is given as $151 exclusive of arrangements for supplying water to dissolve the chloride of lime.

In applying chloride of lime or other disinfectant, time and thorough mixing are necessary. Heat also plays some part in the effectiveness of action. Certain experiments seem to indicate that it is better that the mixing of the disinfectant with the sewage should take place somewhat slowly and should continue throughout a period of an hour or more before the effluent is diluted by discharge into a stream.

The action of the chlorine is probably through the free nascent oxygen which it liberates from the water in which it is dissolved, the oxygen destroying the bacteria in the same manner as ozone would. It is for this reason that organic impurities in the sewage increase the amount of chlorine required, since a considerable part of the oxygen would be taken up by these rather than be used in destroying the bacteria. Chlorine is also obtainable as chlorine gas, and as oxychlorides, these existing in three forms, Cl2O, Cl2O3, and CIO2. Any of these could be used as a dis

infectant, but the cost is greater than that of bleaching powder. Potassium and sodium permanganate have been used for the oxidation of organic matter in streams and in sewage as laboratory experiments. Apparently the only objection to their use is that either is more expensive than the chlorine compounds, without any greater efficiency.

Bleaching powder is manufactured chiefly by the electrolytic process at Niagara Falls and can be purchased at about one cent per pound guaranteed 40 per cent available chlorine.* This cost of 2 cents per pound of available chlorine is equivalent to 21 cents per million gallons for each part of available chlorine. It has been suggested, and indeed some small plants have been operated upon the principle, that the chlorine might be manufactured from caustic lime by the application of electric current. As this is practically the method employed in the manufacture of bleaching powder at Niagara Falls, where current is unusually cheap, the process conducted on an enormous scale, and the bleaching powder merely a by-product, it does not seem at all probable that it will be possible to create chlorine electrolyticallyin the comparatively small quantities required in sewage disposal plants as cheaply as it can be purchased. There is an advantage in the chlorine gas generated in the sewage itself, however, in that it is more powerful and more fully available than the chlorine in bleaching powder, and it is possible that some method may be devised for the use of electricity for generating chlorine in the sewage itself.

During 1913 and 1914 apparatus was perfected for applying chlorine gas directly to water and sewage, the gas being purchased compressed into liquid form in steel containers similar to those used for charging soda water. The cost of material for a given amount of chlorine is somewhat greater than by the use of bleaching powder, but there is less waste and less storage room required. It seems probable that, to secure equal disinfection, the suspended matters in the sewage must be more thoroughly removed or comminuted when gas is used.

The use of electricity to decompose sea water, or a solution of

*The European war has caused an increase of this price.

magnesium and sodium chlorides, has been used in this country under the name of the Woolf process at Brewsters, N. Y., and at Danbury, Conn., but in 1895 the latter place was enjoined from discharging the effluent from this treatment into the Still river, and adopted filtration in its place. At Brewsters ICCO gallons of water containing 160 pounds of salt was subjected to an electric current of about 700 amperes and five volts, the positive electrode being of copper plated with platinum and the negative of carbon, a 4-H.P. dynamo being used. One part of this solution was used in 100 parts of sewage, or $3.20 of salt to each one million gallons. Practically the same process was used in Bombay in 1897, but abandoned after four months' trial, it being found that the same amount of free chlorine could be obtained with chloride of lime at one-half the cost. (Recent use of electricity in sewage treatment will be referred to in the next article.)

It is noticed that Phelps refers to the destruction of only 98 to 99 per cent of the bacteria. It is found that, whatever method of disinfection or sterilization be employed, the use of comparatively small amounts will effect bacterial removal up to, say, 95 per cent; that double this amount would be required to increase this to 98 or 99 per cent; and that two or three times this latter quantity would be required for complete sterilization, if indeed this last would be possible with any practicable amount. This phenomenon was termed by George C. Whipple as that of a "resistant minority," there apparently being a certain very small percentage of bacteria in all sewages which are destroyed very much less readily than any of the others. Further investigation is necessary to determine whether or not this small resistant minority contains any pathogenic bacteria. There are some reasons for thinking that it does not. At any rate, the great addition to the cost required for destroying this last one per cent would not ordinarily be justified by the results obtained.

ART. 83. MISCELLANEOUS METHODS

Wave Filters

Special conditions or special ideas concerning sewage purification have naturally led to the designing and in some cases constructing of a number of variations on the methods and devices described in the previous articles. One of these is the "wave filter," which has been used in at least three or four plants in this country. At Kenton, Ohio, are three wave filters, each 10 feet wide and 100 feet long, filled with broken stone and pea coke, the stone being from 1 to 3 inches in diameter; the depth of the filtering material decreasing gradually from 2 feet at the upper end of the filter to 6 inches at the toe. Dosing devices discharge the sewage at the upper end of these filters into each, in rotation, and the sewage passes in waves or sudden flushes through the filter to the toe, where it flows through a number of 2X8-inch openings into an effluent channel. The dosing intervals were approximately five minutes. The filters are supposed to serve to aërate the sewage passing through them and to remove by straining and surface adhesion a large part of the suspended matter, which is oxidized after the draining of the filter at the termination of each dose. It was believed that the wave action would displace the carbonic acid and nitrogen gases which would be formed in the pores during the periods of rest. The material was removed from each filter twice a year and spread on adjacent land in thin layers where exposure to sun, wind, and rain sufficed to restore it to such condition that, after screening, it was suitable for use again. The general experience here and in other plants appears to have been that these filters have not developed the efficiency expected, and their use has been quite limited.

Cultivation Filters

In the Scott-Moncrieff "cultivation filter" the sewage passes upward through the gravel or broken stone, leaving the solid matter behind, but carrying with it all matter liquefied from