IOWA ACADEMY OF SCIENCES. 
149 
wells located on the inside of this bend near the river may 
be sufficiently free from iron, it is probable that water from 
beneath and from the upland side may carry iron in solu- 
tion into such wells. This can only be determined by driv- 
ing a few wells and analyzing the water; but, as even a 
small quantity of iron in solution is so very undesirable, 
the chances of securing a satisfactory supply are not very 
attractive. 
The last possibility is to filter the water from the river 
itself. Here the difficulty is to filter out the bacteria and 
the very fine clay which is so slow in settling, and after a 
rain, to dispose of the mud which the river carries in abun- 
dance. If the '^English method” of filtration were used, 
17,424 square feet of filtration surface would be needed per 
million gallons. This would require a trench twenty feet 
wide and 871 feet long, or two trenches each at least half 
as long, that the surface of one filter may be cleaned while 
the surface of the other is flooded. A filter placed in the 
bed of the river would answer for a while, but it would 
not be conveniently placed for care. If 500 feet were 
placed in the river bed and 500 feet in a new trench, a 
total excavation of at least 46,611 cubic yards would be re- 
quired, 12,666 cubic yards of sand and gravel for the filter, 
and 16,000 feet of tile for the bottom of the filter, besides 
controlling gates and a receiving well. 
If the ‘‘American method” of quick filtration is employed 
a centrifugal pump is needed, settling basins, and tanks 
for sand filters. At one grain of alum per gallon of water 
it would require 143 pounds of alum per million gallons as 
a coagulant of the fine clay. The bacteria will be caught 
in the coagulated clay and filtered out at the same time. 
This method can with reasonable care give the city the 
water supply which it needs. It seems the only method 
that is feasible. 
