CHURNING AND WASHING BUTTER. 253 
depth of the fine-sand layer. In a general way, the slower the 
rate of filtration is, the more thorough it is; and, vice versa, the 
more rapid the rate of filtration, the more incomplete is the 
removal of the bacteria. If the filter-bed is constructed as 
described above, the rate of filtration will be about 18 gallons 
per hour, and about 96% of all the germs present will be removed, 
together with the impurities present in suspension. 
(4) The filter used at the Iowa Experiment Station was in 
constant use for about three months, without having been 
changed. At the end of this time it did as efficient work as 
at any previous time. The length of time a filter-bed can be 
used without being changed depends upon the purity of the 
water to be filtered, and also upon which kind of filtration is 
used, the continuous or the intermittent. The more impure 
the water which has to be filtered, the oftener the filter-bed 
should be changed. Whenever the rate of filtration is decreased 
to such an extent as to make the process impracticable, the 
filter-bed should be taken out and cleaned. If the water 
to be filtered is of average purity, a change of the filtering-mate- 
rial once every four months is ordinarily sufficient, no matter 
whether continuous or intermittent filtration is used. A 
filter-bed may do efficient work even a longer time than this. 
The same filtering-material can be used again providing it is 
thoroughly washed previous to replacing it in the filtering-can. 
Kinds of Filtration.—The two kinds of filtration in use are 
(1) Continuous, and (2) Intermittent. 
By the continuous method of filtration the inflow of water 
into the can is constant during night and day. The stream of 
water admitted into the filter-can is sufficient to cause the 
surface of the filter-bed to be covered with water all the time. 
This method excludes all oxygen from the filter-bed, except 
that which is in solution in the water. 
During the process of filtration a slimy coat is deposited 
on the fine sand. This seems to be the real agent absolutely 
necessary in order to eliminate bacteria by’a process of filtra- 
tion. A filter-bed without this slimy deposit on it simply takes 
