41 



invc'stifjalioiis wt-rr coiitiiuicd diiriiiL; 'Ih' Idllowini;- tivr years on ihv Hudson and 

 Moliawk rivtM-s, and reports of tlu* work aiv to In' found in the annual reports of 

 tlie New York Slate I>oard of llerillli sinee ISSS. Tlie inspe<'tion of sources of 

 pollution showed what went into the rivers, hut the ciieuiical analyses failed to 

 i<how with suflieieut (h'tlniteness the eflect of this pcdlution on the water. \\'e 

 tlien tried liie nietliod of determining the numbers of bacteria in tlu' water, and 

 wliile that was fairly satisfactory when the conditions were siirij)le, we found it to 

 be absolutely necessary that there he no disturbing conditions whatever, so that it 

 was diftii'ult in most cases to find a time when the method could be applied. 



We are told by the l)acteriologists that the bacteria which are objectionalde 

 in drinking water are the bacteria introduced by sewage. We therefore concluded 

 that we should determine the i)roportion of such bacteria in the water. At this 

 juncture Dr. Theobald Smith, of the Bureau of Animal Industry, Department of 

 Agriculture, at Washington, suggested a method of making this determination, 

 which he was developing, and we ai)plied the method to our study of the risers 

 with results that are so far quite satisfactory. 



The method rt-sts ujion the assumplions that Ikicillua coli communis is a species 

 ■which is very common in sewage, which does not proliferate under ordinary con- 

 ditions of running water, and who.se numbers may therefore be assumed to bear 

 a fairly definite relation to the amount of sewage pollution. Numerous check 

 ob-iervations i!j)hold these assumjjtions. 



The method of determining the numbers of "coli" in a given sample uses the 

 fermentation tube, now frequently called the Smith tube in bacteriological labo- 

 ratories, as the use of the tube in bacteriology ha-i been developed by Dr. Smith. 



The tube, as shown by the pictures, is a bent tube with one end closed and a 

 bulb at the other. It is filled with a clear bouillon of beef with peptone, salt and 

 •2 |)er cent, of glucose, properly neutralized, or made slightly alkaline. The tube 

 and filling are sterilized by boiling on three successive days, the air driven off into 

 the closed end of tiie tube, being decanted so as to leave the liijuid in the tube 

 sterile, and that in the closed end of the tube without oxygen. The liquid in 

 the bulb is now inoculated with bacteria and the tube placed in an incubator 

 kept at 98° F. for 86 to 48 hours. Classifying the l)acteria likely to be found in 

 water as motile and non-motile, and as aerobic and facultative anaerobic, it is 

 readily seen that non-motile and aerobic germs will develop in the bulb only, and 

 will leave the liquid in the tube clear. Motile bacteria that can develop without 

 oxygen will reach the tube and change the character of the liquid. The tempe- 

 rature of 98° at which the tube is kept will prevent the development of nearly all 

 the common water bacteria. Certain l)acteria produce gas from media containing 



