BACTERIA IN WATER 1019 



supply for a large proportion of farms in the eastern United States 

 are usually very rich in bacteria and are by no means to be regarded 

 as safe sources, except in cases where great care is observed as 

 to cleanliness of the surroundings. In such wells the filtration of 

 the water entering the well may be subject to great variation accord- 

 ing to the geological conditions of the surrounding ground. The 

 proximity of barns and sinks may lead to dangerous contamination 

 of such waters. 



Examinations by various bacteriologists have shown that such 

 wells frequently contain as many as five hundred bacteria to the 

 cubic centimeter. 



Perennial spring waters are usually pure. Examinations by the 

 Massachusetts State Board of Health 4 in 1901 showed an average 

 of about forty bacteria per cubic centimeter. As sources of water 

 supply for general consumption, however, springs can hardly be 

 very important because of the insignificant quantities usually derived 

 from them. 



Of much greater practical importance are deep artesian wells, 

 which, under ordinary conditions, are largely free from bacterial 

 contamination. 



Quantitative Estimations of Bacteria. The quantitative estima- 

 tion of bacteria in water is of necessity inexact, because of the 

 difficulty of always securing fair average samples from any large 

 body of water, and because of the large variations in cultural re- 

 quirements of the flora present in them. All these methods depend 

 upon colony enumeration in plates of agar or gelatin, preferably 

 of both. For the sake of gaining some basis of comparison for 

 results which, at best, can never be entirely accurate, an attempt 

 has been made by the American Public Health Association 5 to 

 standardize the methods of analysis. 



Water for analysis should always be collected in clean, sterile 

 bottles, preferably holding more than 100 c.c. If w^ater is to be 

 taken from a running faucet or a well supplied with piping, it is 

 important that it should be allowed to run for some time before 

 the sample is taken, in order that any change in bacterial content 

 occurring inside of the pipes may be excluded. It is obvious that 

 in water pipes through which the flow is not constant, bacteria may 



4 Mass. State Bd. of Health, 33d Annual Eeport for 1901. 



5 Fuller, Trans. Amer. Public Health Assn., xxvii, 1902. Eeport of Com. on 

 Standard Methods of Water Analysis, Jour. Inf. Dis., Suppl. 1, 1905. 



