The results have also shown that the subterranean supply at the lower cross-section of the river, at San Geronimo, is abundant, and can probably be increased to an indefinite degree by continuing the filtration gallery; while at Santa Catarina the same type of development can be made for a high-source supply, although requiring a long and expensive tunnel and conduit. DAVID T. PITKETHLY, ASSOC. M. AM. SOC. C. E. (by letter).--Having been engaged on the design of sewerage systems for some years, the writer finds this paper of peculiar interest, particularly the sewerage portion. There are some points in the design, however, which do not appear to be clear. The system is described as "strictly separate," and yet the sewers are designed to run half-full, providing a capacity of 200%, the 100% basis, or 380 liters per capita, being 90%, or 180 liters, in excess of the calculated water supply of 200 liters per capita. It has been the writer's practice to design sanitary sewer systems on the basis of the water consumption, and to assume the whole daily amount to reach the sewer in 16 hours, thus providing capacity sufficient to care for the maximum or wash-day flow without causing the sewers to run above the calculated hydraulic gradient, which should be placed within the pipe so as to provide air space for ventilation under all circumstances. The practice of calculating sanitary sewers to run half-full is a good one when ground-water is expected in sufficient amount to fill the remaining portion of the sewer, but when no ground-water, or roof-, or surface-water is allowed to enter the system, or all precautions are taken to exclude such, then the system may be designed so that the expected maximum, or wash-day flow, will fill the sewer to the desired hydraulic gradient. The method of ventilating the sewers does not seem practicable. The houses are principally of one story, and yet the stand-pipes on the sewers have openings 25 ft. 9 in. above the sidewalk. Are the ventilating or vent pipes of the house plumbing carried to a height to balance this, or will these chimneys draw the air from the house drains and fresh-air pipes, breaking the seal in the so-called disconnecting traps, thus causing the circulation of air in the house piping to be downward through the sewers instead of upward through the fresh-air inlets and vents, as designed? It is interesting to note that crude sewage, as well as the liquefying (septic) tank effluent, is to be applied to land for irrigation purposes, but the application of crude sewage without any attempt at removing the suspended matter, or the effluent from the septic tanks where only a partial removal occurs, seems to be bad practice. The author states that: "The degree of purification in the tanks was relatively unimportant; the object to be obtained consisted chiefly in distributing on the land an effluent which would be innocuous and clear."