Januaky 12, 1906.] 



SCIENCE. 



63 



sidered in connection with the chemical 

 data. 



A certain added amount of information 

 may be gained as to the weight to be placed 

 on the total bacterial content of a water, by 

 also determining the number of colonies 

 that develop on agar plates at blood tem- 

 perature and the number that decompose 

 lactose with the formation of acid. Ac- 

 cording to Rideal and also to Winslow, in 

 an unpolluted water the proportion be- 

 tween the total number of colonies ob- 

 tained on gelatine plates at 70° Fahrenheit 

 and the number obtained on agar plates at 

 98 degrees Fahrenheit should not be greater 

 than 12 to 1, and Winslow and Prescott 

 state that in normal Massachusetts waters 

 the total number of organisms growing at 

 the body temperature rarely exceeds 50 per 

 cubic centimeter, and that acid producers 

 are generally absent. 



The information that can be obtained by 

 the examination of a water for the colon 

 bacillus is much more positive and impor- 

 tant than that which can be obtained from 

 a bacterial count, for though undoubtedly 

 the colon bacillus is not confined to the 

 secretions from mammals, the intestines of 

 the higher vertebrates form a better en- 

 vironment for its growth and multiplica- 

 tion than any other which occurs in nature, 

 consequently drainage from domestic and 

 agricultural wastes of human life must be 

 considered as the method by which large 

 numbers gain access to surface waters. It 

 follows, therefore, that a water containing 

 large numbers of colon bacilli must be 

 looked upon as a polluted water, and the 

 generally accepted statement of to-day is 

 that if the colon bacilli occur in sufficient 

 numbers in a surface water to be detected 

 in the majority of one cubic centimeter 

 samples tested (at least six samples be- 

 ing taken), it is almost positive indica- 

 tion of recent sewage pollution. Failure 

 to detect colon bacilli when water is thus 



tested is not, however, a proof that the 

 water is a normal water, though it is usu- 

 ally a proof- that the pollution, if it exists, 

 is not recent nor continuous. 



Having attempted to give what I believe 

 can be learned from the sanitary analysis 

 of a surface water, ground and artesian 

 waters remain to be considered, and with 

 these waters the analyses assume far greater 

 importance than with surface waters, for 

 the area of the source of the water is often 

 indefinite and rarely determinable with ac- 

 curacy, and a careful and complete survey 

 of what may be called the watershed is 

 very difficult and generally impossible. 



In determining the character of a ground 

 water we make use of the same data that 

 we have considered in speaking of surface 

 waters, but the deductions drawn from the 

 data are very diiJerent. This is due to the 

 fact that the two waters differ so greatly 

 in their chemical and biological character- 

 istics that they can not be judged by the 

 same standards. 



Ground waters are surface waters which 

 have percolated through the soil, and the 

 changes which they have thus undergone 

 are very similar to the changes which take 

 place in the process known as slow sand 

 filtration and it is from a study of this 

 process that we are able to follow the 

 changes that take place when surface 

 waters pass into the soil. 



In slow sand filtration of water we 

 find that odor, color and turbidity are re- 

 moved, that about 90 per cent, of the 

 nitrogenous organic matter is oxidized, and 

 that a part, varying from 50 to 75 per cent, 

 of the nitrogen of the organic matter, is 

 found in the filtrate as nitrates, that the 

 amount of chlorine remains unchanged and 

 that the bacteria are to a very large extent 

 removed. 



To apply the information thus obtained 

 to ground waters it can be stated that an 

 unpolluted ground water should be free 



