80 The Philippine Journal of Science i9i8 



count was lowered. This was reduced still further during week 

 III, when 1 part of chlorine per million was used, and during 

 week IV, when 2 parts of chlorine were applied. The results 

 of the determinations for B. coli are not shown in Tables III to 

 XI, but are summarized in Table XIII. Weeks V and VI show 

 the efficiency of copper sulphate in killing B. coli. Although the 

 1 cubic centimeter portions showed no difference, 10 cubic 

 centimeter inoculations showed (Table XIII) that the average 

 of 2.33 positive B. coli tests per week, obtained when 1 part 

 of copper sulphate per million parts of water was used, was 

 reduced to 0.66 times when 2 parts were applied. During week 

 VII, when no disinfectant was used, the B. coli count again 

 increased, but during the following week it was reduced by the 

 daily application of 0.5 part of chlorine. Continuing the same 

 application for another week (Table IX), the increase in the 

 B. coli count, as well as that of other bacteria, showed that it 

 was a disadvantage to use the water in a tank longer than one 

 week. When fresh water was again used, the daily addition 

 of antiformin, equivalent to one-seventh part of available chlorine 

 per million parts of water (Table X), reduced the B. coli count 

 to 2, while the daily addition of 0.5 part of chlorine (as chloride 

 of lime) brought it down to zero. In an attempt further to 

 reduce the amount of chloride of lime, 0.25 part of available 

 chlorine was added daily during week XI. However, B. coli 

 again made its appearance and the total bacterial counts 

 gradually increased, showing that this amount of disinfectant 

 was insufficient. 



The total bacterial counts, with daily variations and the effect 

 of some of the disinfectants, are shown in fig. 1, where the 

 Roman numerals refer to the tables from which the averages 

 are taken. The 37°C. counts are higher than the 25°C. counts, 

 and in both cases the curve obtained when no disinfectant was 

 added (VII) is highest. When 0.5 part of chlorine per million 

 parts of water was used once a week, the bacteria were kept 

 down somewhat, but at the end of the week they were nearly 

 as numerous as in the former case. However, III (1 part of 

 chlorine weekly) shows a considerable bacterial reduction at 

 the end of the week. This reduction is more pronounced in the 

 case of IV (2 parts of chlorine weekly). The most efficient 

 disinfection was obtained when 0.5 part of chlorine (from chlor- 

 ide of lime) was added daily. This is represented by the 

 dotted line VIII. In spite of the fact that curves based upon 

 bacterial count often show surprisingly great jumps, which can- 

 not be accounted for, these curves illustrate the conditions very 



