One of the most important points in connection with filtration as above described is the fact that bacteria which themselves possess definite size are too large to pass through the fine pores of the mud cover of the sand filter. The result therefore is that when the sand filter is perfectly used and the effluent water perfectly clear the bacteria which the water first contained are left almost entirely upon the mud film. Examinations of the effluent water from the sand filters in Washington have shown as low as 8 or 9 organisms only per cubic centimeter, and rarely do they go above 20. The experience in other cities is doubtless comparable to that in Washington. Thus the sand filtration serves a double purpose; it not only removes the suspended matter and makes the water clear, but it also retains the principal part of the bacteria which the water originally contained. It is evident that a water which is never muddy would not be greatly benefited by passing through a sand filter unless the sand was fine and compact because it would be impossible for it to form the mud cover necessary to its perfect function. If the sand is fine enough the bacteria may be largely removed. There is always an element of danger in the use of sand filters. Imperfections or breaks may appear and through these the filtered water may be reinfected; also in cleaning the filters, and immediately after cleaning pollution may take place. In spite of all of these faults, the sand filters have generally resulted in lowering the death rate from typhoid. The sand filters have also been used with previous sedimentation or precipitation. In these cases rapid and effective work has been done. In Cincinnati, notably, the death rate from typhoid has been greatly lessened. It does not exceed 4 to 6 for each 100,000 of population. In Washington, on the contrary, the deaths from typhoid have not been materially decreased by the installation of the same filtration plant. Possibly the presence of the disease in Washington may be attributed to some other cause. Since writing the above the death rate at this time (June, 1918) has materially diminished and the number of cases of typhoid fever in the city has fallen below 30.