SIGNIFICANCE OF COLON GROUP IN WATER 169 



and Pusch, 1903) examined a considerable series of 

 waters from different sources by incubating measured 

 samples with equal amounts of nutrient broth and iso- 

 lating upon agar. In 45 samples of well-waters they 

 found B. coli 7 times in .01 c.c, 9 times in .1 c.c, and 

 7 times in i c.c. In the other 22 cases it could not be 

 found in i c.c. and in 4 cases not in 100 c.c. One sample 

 showed it only in 600 c.c. and i not in 750 c.c. Of 

 29 river-waters, only 2 failed to give positive results 

 in .1 c.c. and 14 showed B. coli in .001 of a c.c. or less. 

 In sewage the number varied from i to 1,000,000 per 

 c.c. The authors conclude that a quantitative estima- 

 tion of the B. coli content furnishes a good measure of 

 the faecal pollution of water. There is still a school of 

 bacteriologists in Germany, however, who are inclined 

 to place little value on the colon-test. We have pointed 

 out how Knise and his pupils at Bonn led in the attack 

 on it in 1894. Fourteen years later Kruse (1908) 

 concluded after a full smnmarj^ of the literature that 

 the colon test was on the whole less valuable than the 

 gelatin count, although he admitted that when the 

 test is made quantitative it is valuable as a supple- 

 ment to the plate count. Konrich (1910), working in 

 Gartner's laboratory at Jena, after perhaps the most 

 exhaustive study ever made of the whole subject, 

 concludes, that to include the colon test in forming 

 judgment on the sanitar}- quality of a water is to 

 complicate the procedure without impro\-ing it and that 

 one would do well to omit the test except in certain 

 special cases. 



The German criticisms of the colon test are based 



