EXPERIMENTS UPON THE PURIFICATION OF SEWAGE AND The experimental work of the Lawrence Experiment Station is still under the active direction of Hiram F. Mills, C. E., member of the Board, who originally designed the station and planned its work. Mr. Allen Hazen is chemist in charge of the station, and Mr. George W. Fuller is in charge of the biological department. Mr. Harry W. Clark is assistant chemist. The compiling of the records and the preparation of the tables and diagrams have been the work of Mr. F. L. Fales. Professors T. M. Drown and W. T. Sedgwick of the Massachusetts Institute of Technology are respectively consulting chemist and biologist with a general oversight of the chemical and biological investigations. The following detailed account of the work of the station has been prepared by Mr. Allen Hazen, chemist in charge: — The special report of the board upon the Purification of Sewage and Water, 1890, contains a full account of the work at the Lawrence Experiment Station during the first two years up to Nov. 1, 1889, and also a summary of the results for 1890. During the past two years investigations have been continued upon purification by filtration along the same general lines, but in such a way as to throw additional light upon numerous points of great practical importance. Most of the filters described in the previous report have been continued in use to obtain information as to their permanence, while many others have been started to investigate special points connected with sewage purification and the filtration of water. Filter Tank No. 1, which is one of the original filters and has been in use four years, has filtered sewage during the past two years, 1890 and 1891, at an average rate of 85,920 gallons per acre daily for every day in that time, and with a removal of 94 per cent. of the organic matters, as shown by the albuminoid ammonia, and 98 per cent. of the bacteria. Filter Tank No. 2, also one of the original filters, has filtered for the same period, at an average rate of 49,360 gallons per acre daily, and has removed 97.5 per cent. of the organic matters and at least 99.99 per cent. of the bacteria of the applied sewage. With many other filters correspondingly good results have been obtained. The purity of the effluents obtained by intermittent filtration was thoroughly discussed in the report on the Purification of Sewage and Water, and it will suffice to say in this connection that there is no other method of sewage purification which yields results at all comparable to those obtained by intermittent filtration under favorable conditions. One of the most important results of the past two years' work is the fact that, by systematically breaking the scum which forms on the surface of filters, a very much larger quantity of sewage can be purified without any deterioration in the quality of the effluent. We had found that different materials require different treatments for the best results; that a system of applying sewage best adapted to a fine material fails to yield the best results when applied to a coarse sand, and vice versa. By studying the performance of our different filters under various conditions, in connection with the mechanical composition of their materials, we have discovered many of the causes of these differences, and have secured data which will enable us to determine, in advance, the general line of treatment required by different sands. In a similar way we can form an approximate estimate of the quantity of sewage which can be successfully applied to various materials under known conditions.* We have continued the study of the effect of winter weather upon filtration, and have found that even with unprotected filters good results are possible in our climate under some conditions to be discussed in the following pages. The effluents obtained in winter, while less perfect than those of the warmer months, were still good, -much better than could be obtained by other processes, as, for instance, chemical precipitation. An excessive quantity of acid in the sewage prevents satisfactory purification, but our experiments have shown that if, for the treatment of such sewage, a filter containing limestone is employed, the acid is neutralized, and a good result is obtained. This is a most important demonstration, as it assures the successful use of sewage The reader should bear in mind that the results given in this paper were obtained at the Experiment Station where all of the work is under scientific supervision, and the sewage is applied to small areas with uniformity and in definitely measured doses. In applying these results to actual work on the large scale some allowance may have to be made for less favorable conditions. filters, regardless of possible acidity of sewage. If the sewage is only occasionally acid it will cause no trouble with ordinary filters, but if it upon trial proves to be so strongly and continuously acid as to prevent satisfactory purification, the addition of limestone to the upper layers can be depended upon to correct the acidity, and to insure as good a result as could otherwise be obtained from normal sewage. The permanency of filters has also been made the subject of special study. The leading fact of intermittent filtration is that the organic matters of the sewage are destroyed instead of being stored in the filter, as is largely the case with other methods of filtration ; and it seemed probable from the available data at the time of writing the special report upon the Purification of Sewage and Water, that there would be no continued accumulation of matter in the filter. Further study has shown, however, that the conditions which allow the purification of the maximum volume of sewage with the best results are such that a small percentage of the more stable organic matters of the sewage resist the oxidizing action of the filter, and accumulate in its upper layers, until after a time the surface may become choked to such an extent that it will remain saturated with water, thereby excluding the air. This can be remedied temporarily by turning the surface under, but it does not seem probable that continued inversion of the upper layers will allow the indefinite use of the same material while still maintaining the application of the maximum quantity of sewage. With lower rates of filtration the storage may be less in proportion to the amount of sewage, and, in any case, the replacing with fresh material of a moderate amount of the upper layers of sand from time to time will ensure the permanency of the filter. It is quite possible that the removed sand, when taken entirely away from the filter, will in time so far regain its original properties as to again allow its advantageous use as filtering material. We are giving much attention to methods for hastening this renovation. The filtration of water also is receiving increased attention. The tables showing the work of the water filters for the time covered by this report are appended, together with a brief summary of the results obtained during 1892, up to the time of going to press. Owing to the important data bearing on this subject now being obtained, and as yet incomplete, a detailed discussion of these results is for the present postponed. In the following pages are given tables representing the work of the various filters in continuation of those given in the special report upon the Purification of Sewage and Water, with other analytical data, and a brief description of the leading facts concerning each. As many of the points of greatest interest are involved equally in the history of several filters, it will facilitate discussion to consider many subjects by themselves, rather than under the heads of the various tanks. A discussion of some of the results of the work on intermittent filtration will now be presented, after which will be given in full the data in regard to the various filters upon which the conclusions are based. SEWAGE PURIFICATION AS AFFECTED BY THE MECHANICAL CONDITION OF THE FILTERING MATERIAL. use. The purification of sewage by intermittent filtration depends upon oxygen and time; all other conditions are secondary. Temperature has only a minor influence; the organisms necessary for purification are sure to establish themselves in a filter before it has been long in Imperfect purification for any considerable period can invariably be traced either to a lack of oxygen in the pores of the filter, or to the sewage passing so quickly through that there is not sufficient time for the oxidation processes to take place. Any treatment which keeps all particles of sewage distributed over the surface of sand particles, in contact with an excess of air for a sufficient time, is sure to give a well-oxidized effluent, and the power of any material to purify sewage depends almost entirely upon its ability to hold the sewage in contact with air. It must hold both sewage and air in sufficient amounts. Both of these qualities depend upon the physical characteristics of the material. The ability of a sand to purify sewage, and also the treatment required for the best results, bear a very close relation to its mechanical composition. It is our present purpose to more definitely formulate the results obtained with a view to predict the action of any material from its mechanical composition. Mechanical Composition of Materials Used. In making a mechanical analysis the sand is first sifted through a series of sieves, each, in a general way, twice as fine as the one next coarser. The sand passing the finest sieve is divided into several portions by beaker elutriation. Each portion is weighed and the range in the sizes of its particles is determined by micrometer measurement in the case of the smaller particles, but the diameters of the larger particles can be more conveniently and accurately calcu lated from their weight. The diameters of all particles are taken at, as nearly as possible, the diameter of a sphere of equal volume. Of course the sand grains are irregular. With the Lawrence sands the average lengths of their axes, selecting the longest and taking the other two at right angles to it, are to each other as 4:3: 2, and the mean diameter, taken as the cube root of the product of the three axes, 2.88. The longest diameter thus averages to be nearly 40 per cent. longer than the mean diameter, while the middle diameter, which is the width as seen by a microscope, is an approximation to the mean diameter. The analyses of some of the materials which have been most carefully studied in their relations to sewage purification are shown in the following table. The figures given show the per cent. by weight of the different materials having smaller diameters than the sizes given in the first column. These results have been revised by improved methods of analysis, so that in some cases the following figures differ slightly from those given in the report upon the Purification of Sewage and Water. For study and comparison the results have been plotted, and are shown on the accompanying diagram, the height of curve at any point showing the per cent. of material finer than the size indicated at the bottom of the diagram. The lines representing the diameters are spaced according to the logarithms of the diameters of the particles, as in this way materials of corresponding uniformity in the range of |