160 On the Composition and Properties of Drinking-Water, 
be removed by heating as by the lime process, and which renders 
water permanently hard. 
The composition of spring or well-water from the chalk-strata 
varies but little in different localities. Its hardness rarely 
exceeds 18 degrees, and pretty uniformly amounts to from 16 to 
17^ degrees. Water of that degree of hardness contains in 400 
gallons about 1 lb. of carbonate of lime, held in solution by 7 
ounces of carbonic acid gas. 1 lb. of carbonate of lime, in round 
numbers, consists of 9 ounces of caustic lime and 7 ounces of 
carbonic acid. It is evident, therefore, that the addition of 9 
ounces of caustic or quicklime to 400 gallons of such water 
will have the effect of depriving it of the 7 ounces of carbonic 
acid gas, which holds 1 lb. of carbonate of lime in solution ; and 
that both the lime added and that originally present must be pre- 
cipitated together as neutral insoluble carbonate of lime, minus 
a small quantity, amounting to about 2 grains in the gallon, 
which pure Avater is capable of dissolving. 
The original hardness of chalk springs may thus readily be 
reduced by Clark's lime-process from 16 to 18 degrees to from 
2 to 4 degrees. 
This process is sufficiently simple to be left to the execution of 
a workman of ordinary intelligence. All that is required for him 
to do is to stir lime-water or milk of lime, made by mixing 
quicklime with water — about 1 lb. to 40 gallons of water — into 
the water intended to be softened, until the carbonic acid which 
holds the carbonate of lime in solution is completely neutral- 
ized by the addition of quicklime. The spring-water, on the 
addition of the lime-water, at first has the appearance of thin 
milk, but the preciptation of the carbonate of lime proceeds 
with rapidity, and in the course of 24 hours the water may be 
syphoned off from the precipitate and received in a perfectly 
clear condition into the supply cistern or tank. The only pre- 
caution necessary to be taken is to insure the absence of an 
excess of lime. To this end the water in the settling tank has 
to be tested from time to time with a few drops of a weak solu- 
tion of nitrate of silver. This test gives a white precipitate in 
the original hard spring-water, and shows whether the» quantity 
of lime required has been exceeded, by the brown colour of the 
precipitate then formed. In practice the addition of lime-water is 
stopped as soon as a sample of tlie filtered water from the settling 
tank gives a brownish-coloured precipitate with the nitrate of 
silver test. In that case more of the hard water is added, and 
well mixed with the contents of the settling tank. After subsi- 
<lence, a sample of the softened water is again tested with a drop 
or two of a nitrate of silver solution, and the addition of more 
hard water if necessary is repeated a third time, or until the 
