570 Prof. Woodhead, The residts of 



of the samples of water taken from the weir. Sterilisation, as 

 regards the coliform organisms, was absolute. 



At this time however I was satisfied that we were not ob- 

 taining the greatest possible amount of work from our chlorinating 

 cylinder, and that a small proportion of the water was passing 

 through the cylinder unnecessarily quickly and was therefore in 

 contact with the chlorine for too short a time. Using per- 

 manganate of potash in place of chlorine it was found that 

 a trace of colour made its appearance at the weir in about 

 18 minutes. In order to prolong the period of contact and to 

 prevent this somewhat rapid and irregular passage of a portion 

 of the water a number of simple baffle plates were inserted. By 

 these the passage of the water was so far delayed and regularised 

 that the first trace of an introduced colouring matter made its 

 appearance in about 20 minvites, so that none of the water could 

 pass through the cylinder without being in contact with chlorine 

 for at least that period before it was neutrahsed; the bulk re- 

 mained in the cylinder for about an hour. 



After these alterations had been made another series of experi- 

 ments extending over 12 days (12 hours each day) was carried 

 out. Here again we had a run, first for three days with one part 

 of chlorine per million parts of water, the chlorine afterwards being 

 neutralised by a calculated amount of bisulphite of soda. The 

 results obtained were exceedingly unsatisfactory on the first two 

 days and, even on the third, the results were not good. On 

 searching for the cause of this it was found that the bisulphite 

 solution was not sterile and that the organisms contained in it 

 were passing into the water taken at the weir. The chlorine had 

 not time before it was neutralised to kill the organism present 

 in the bisulphite of soda solution. This of course afforded an 

 explanation of the irregular results obtained in the preceding 

 series. Even on the third day, when excellent results were ob- 

 tained during the greater part of the day, acid-forming organisms, 

 not true Bacillus coli, were present at 12 noon and at 3.30 p.m. 

 The results obtained by the other bacteriologists coincided exactly 

 with our own — sometimes the treated water was free from Bacillus 

 coli in 500 c.c, at others the presence of this organism could 

 be demonstrated. On the next two days the amount of chlorine 

 was halved. During the early part of one day acid-forming 

 organisms were present in 500 c.c. but, from this time onwards, 

 sterilisation was complete. During the next three days one part 

 of chlorine in four million parts of water was used and only once 

 during the whole of this time was the Bacillus coli found in 

 500 c.c. of water. On the last four days of the experiment one 

 part of chlorine in seven to eight million parts of water was used 

 but no bisulphite of soda was added. Of 13 samples of 500 c.c. 



