124 
EIGHTH REPORT. 
is reduced in thickness to a degree that is regarded no longer safe, 
when the bed is rebuilt. The practice of the English for years was to 
dig up their filters once a year, but the Americans beginning along in 
the 80’s have shown that those ideas were wrong, and that the older 
a bed becomes the better it is; so that now we start these filters with a 
thickness of four feet of sand and run down to sixteen or twelve inches 
before renewing; or more frequently we run them down about 
eight inches at a time and then replace the sand, thus keeping the lower 
part of the bed undisturbed. After scraping, the water from these 
filters has to be wasted for a short time to give these surface layers an 
opportunity to form anew, and the filter should always be refilled ivith 
filtered water, which must be permitted to rise very slowly from the 
bottom. Care must, of course, be taken to prevent disturbing the surface, 
and the water must be brought upon it at all times without agitation. 
The first attention to filtration in this country was when the city 
of St. Louis, in 1866, sent Mr. James I*. Kirkwood, its engineer, abroad 
to study methods of water purification. At that time he found slow 
sand filtration in use at Berlin, Altona, Nantes, and Marseilles on the 
continent, and at Leicester, York, Liverpool, Edinburgh and Dublin, 
as well as London. He returned and presented a report which was 
published in 1869, but St. Louis went no farther. Great Britain had 
meanwhile accepted purification as a matter of course, and they thought 
no more of it. We over here did not feel the necessity for it generally, 
but as the result of Mr. Kirkwood’s trip abroad there were two or 
three filters established in the United States. The first at Poughkeepsie 
in 1872, designed by Mr. Kirkwood, and the second a smaller plant, 
also from his designs, at Hudson, N. Y., in 1874. About that time 
St. Johnsbury, Vt., put in one, the designer of which is unknown. 
.Those remained all the filters of this type in the United States until 
1893. During that year the city of Lawrence, as the result of agita¬ 
tion after an epidemic of typhoid, built its famous filter, which was 
the first to be employed in a large city in this country, Lawrence having 
at this time 40,000 people. In 1894 Mt. Vernon, New York, put in a 
filter to. supply a community of 11,000 people; in 1895 Ashland, Wis., 
as the result of a typhoid epidemic there, added another 10,000 to the 
users of filtered water, and the city of Milford, Mass., put in a filter, 
making another 10,000 so supplied. So while at the end of the period 
from 1872 to 1893, only about 36,000 people were supplied with filtered 
water in the United States, that number was more than doubled within 
the next five years. In 1899 the city of Albany put into service a slow 
sand filter, which added approximately 250,000 to the number of per¬ 
sons using filtered water. Since that time Philadelphia has put into 
service a portion of its filter plant, it being designed ultimately to 
supply the entire city, the full capacity of the plant to be 350,000,000 
gallons per day. Washington has filters under construction which are 
nearly ready to put into service. Pittsburgh has filters under con¬ 
struction. New Haven has a plant which is going into service, and 
Yonkers, N. Y., has one now in service. The general cost of slow sand 
filters is about five dollars per person, and the cost of operation ranges 
from $1.25 to $4.00 per million gallons filtered. We find that 98 to 99 
per cent of the bacteria are removed by this process. The more impure 
the water is when it goes through the filter the larger the percentage 
