Sterilisation Experiments on the Cambi'idge Water. 578 



about most efficient sterilisation, especially if the following con- 

 ditions are observed. The chlorine should remain in contact with 

 the water as long as possible ; the ideal method of treatment, 

 therefore, would be by means of an automatic "injector" to throw 

 " bleach " solution into the mains leading from the pumping wells 

 to the reservoir. During dry weather, the necessary amount of 

 chlorine once determined, the work once started and the chlorine 

 flow regulated, the process may be allowed to go on almost with- 

 out attention. During rainy weather rough tests should be made 

 as to the amount of chlorine absorbed from time to time. When 

 no colour reaction is obtained with iodide of potassium and starch 

 it would be necessary to increase the strength of the chlorine 

 solution until a reaction is obtained. This of course may be done 

 either by increasing the amount of " chloride of lime " added to 

 the solution already in the chlorine supply tanks, or by increasing 

 the speed of the pumping engines injecting the fluid. The test is 

 made in three or four minutes and any increase of organic matter 

 is determined at once and, in our experience, is so small that it 

 may always be neutralised even after the heaviest rainfall observed 

 during the time our experiments were being made by an increase 

 of about 20 7o of chlorine, this addition invariably ensuring com- 

 plete sterilisation. The chlorine and water should be thoroughly 

 mixed in the rising main and should be conducted to a reservoir 

 of sufficient capacity to contain a couple of days supply. Equalisa- 

 tion of the chlorine would thus be obtained and the conversion 

 of any slight excess of chlorine into hydrochloric acid would be 

 ensured. Water so treated is absolutely free from organisms of 

 the Bacillus coli type and therefore from bacilli of the typhoid 

 and paratyphoid groups. 



