122 EDWARDS. 



Fifty-three per cent of all the 700 waters have been condemned for one 

 cause or another. (See Chart I.) Of the total number examined, 

 amoebae were present in 39.4 per cent, but there were 46 per cent positive 

 in those reports in which it was specifically stated that these organisms 

 were searched for; 3 flagellata were encountered in 20 per cent of all 

 waters, and the percentage of positive results on the reports in which 

 these organisms are entered is 23.5; ciliata were mentioned as being 

 present in 7.5 per cent and of the number in which protozoa were stated 

 as being present or not, in 9.5 per cent; protozoa, without specification 

 of identity, were mentioned in 1 per cent of all the waters and in 1.16 per 

 cent of the number of completed examinations in which mention is made 

 of such animal forms. High bacterial count was given as a reason for 

 condemning a water in 9.3 per cent. However in 43 per cent of those 

 cases where the fact of making such a count was mentioned, pathogenic 

 bacteria were present in 0.83 per cent of the total, and in 28 per cent of all 

 the cases in which a sjDecific examination for these organisms was men- 

 tioned. Numerous special examinations for the presence of the cholera 

 vibrio were undertaken, but up to January 1, 1908, no such specific 

 organism was found, excepting in one estero where a railroad company 

 had been emptying its sewage pails. 



Chart II shows the percentage of condemned waters from various 

 sources. The waters from public places included those from the court- 

 house, the public offices, the prison, restaurants and eating houses, 

 excepting in instances where it was specifically stated that the sample 

 examined was distilled water direct from one of the distilling plants. It 

 is probable that if the number of samples of the latter class were to be 

 subtracted from the total, the percentage of contaminated water would be 

 much higher; as it is, it reaches 55 per cent. Fifty per cent of the spring 

 water was condemned. The Manila supply, including samples from the 

 intake, furnished the highest percentage — namely, 95.7. The study 

 of the examinations from the latter source furnishes one of the strong 

 arguments in favor of sand filtration. The colony count of the water 

 secured from Manila hydrants varied from 40 to 600 bacteria in 1 cubic 

 centimeter water. The low figure might be kept constant with an 

 efficient sand filtration plant. 



The samples of water passed through various household filters of 

 divers construction gave 95 per cent to be condemned, the second highest 

 number, but it is well to add that in nearly every instance representatives 

 of this class were reported upon unfavorably because of the presence of 

 animal forms. This result demonstrates the hopelessness of the expecta- 

 tion that a supply secured by filtration should be free from animal forms, 

 but we know that water which has been subjected to slow sand filtration 



3 That is, reports in which amoebae are especially mentioned as ''positive" or 

 "negative." 



