the same extent. Instead of removing the carbonicacid by boiling, it is rendered insoluble by the addi-tion of lime, forming with the latter insoluble car-bonate of lime, which is precipitated together withthe carbonate of lime or magnesia resulting, asin the preceding process, from the decompositionof the soluble bicarbonate. Thus, while the bicar-bonates are removed, nothing is added to the water,and it is suitable for all purposes. This methodwas devised by the late Ur. Clark, Professor ofChemistry in the University of Aberdeen , and pub-lished by him about the year 1841.
To apply it on a small scale, about 1 oz. of quick-lime is required to be added to each 700 gallons foreach part of temporary hardness in 100,000 parts ofwater. The quicklime is slaked in a pailful of water,well stirred to form milk of lime, and then pouredinto the cistern, which should already contain atleast 50 gallons of the water to be softened. Anycoarse particles from the lime may be left in thepail. Fill the pail again with water, and stir andpour this also into the cistern. Then add theremainder of the 700 gallons. If the running ofthe water does not thoroughly stir up and uniformlymix the contents of the cistern, this must be doneby means of a wooden or iron paddle. The waterwill appear very milky, and should be left at rest todeposit the precipitate. In about three hours it willbe lit for washing, and in about twelve hours fordrinking.
This process is now in use on a large scale innumerous places, being applied to the entire watersupply of several towns. In such cases clear limewater is used in place of milk of lime, and thoroughmixture is attained by running the hard water andthe lime water into depositing tanks simultaneouslyin streams suitably proportioned. The time re-quired for subsidence is reduced by the use oftelescopic discharge pipes, by means of which theclear water is constantly taken from the surface,the overflow pipe being lowered as the water runsoff. The quantity of lime water employed is con-trolled by occasionally testing in a porcelain cupwith a solution of nitrate of silver. If the lime bein excess, a brown colour is produced; but whenthe proper proportion only is present, a very faintyellow is seen on adding the nitrate of silver. Thecost of softening by lime is very small; 1,000,000gallons of water, with a temporary hardness of 20parts per 100,000, could be softened at a cost forlime and labour of about 23s. The carbonatedeposited for the most part carries down with ita large part of the organic colouring matter in thewater, so that after subsidence the clear watergenerally exhibits, when seen in a large bulk, amagnificent blue tint, a very important improvementbeing thus produced in the character of the water,as is shown in the accompanying table. Theresult in the case of the Thames water wouldprobably have been still better, had not the softenedwater been necessarily exposed in a cistern on theroof of a house to a London atmosphere for twenty-four hours.
VOL. II.
985
Description.
Total
Solid
Matters.
Organic
Carbon.
Organic
Nitrogen.
Hardneas
Caterham water supply,before softening, Febru-ary 14, 1873,.
27-68
•028
■009
21-2
Ditto, after softening, ....
8-80
•015
•003
4-4
Thames water, supplied byGrand Junction Co., be-fore softening, average of9 experiments in 1870-71,
27-85
■lei
•026
20-8
Ditto, after softening, ....
12-85
•122
•022
5-1
Lea water, supplied by NewItiver Co., February 14,1871,.
30-60
•135
•018
22-4
Ditto, after softening, ....
13-76
•100
•Oil
60
The system is now in active operation, on thelarge scale, at the Chiltern Ilills Water Works, atCaterham, at Canterbury, and at Bushey.
On the Filtration of Water. —The most obviousresult of the filtration of water is the removal ofsuspended matter, so that a turbid water becomesbright and clear. Its efficiency in this respect dependschiefly on the closeness of texture of the filteringmedium, i.e., the smaller the interstices throughwhich the water has to pass the more efficient willbe the purification. Very finely divided suspendedmatter, such as clay, requires a very fine filter, andalso that the passage of the water should be slow.The efficiency of a filter is often judged solelyby its clarifying action, and when the water ispassed rapidly through a thin filtering medium,e.g., a paper filter, there is probably little if anyother result. When, however, the stratum offiltering material is of considerable thickness, andthe rate of filtration slow, an important reductionin the quantity of organic matter present in solu-tion may be effected in addition to the removal ofsuspended matter. The organic matter is oxidisedby the oxygen dissolved in the water, supplementedperhaps in some cases by oxygen supplied by thenitrates present. It is desirable, therefore, thatthe water to be filtered should be well aerated, and,if practicable, that the filter itself should from timeto time also be aerated. The importance of thiswas seen in some experiments made on the fil-tration of sewage through various materials. Thesewage itself probably contained no oxygen in solu-tion, and certainly did not contain sufficient for thedestruction of the organic matter in solution. Itwas found that when it was passed unintermittentlythrough a thick sand filter, at first considerableoxidation took place, but this soon ceased, and thefilter speedily became choked; but when the sewagewas supplied intermittently, so that the filter waswell aerated after each charge, the action could becontinued indefinitely if the amount of organicmatter in each charge was not too great for theamount of atmospheric oxygen introduced at eachinterval. When the charge was increased beyondthis, the purification ceased, and the filter speedilybecame choked. If the quantity of oxygen is suffi-cient, the degree of purification will depend on thenature of the filtering medium, its state of division,124