BAOTERIOLOQICAL STUDY OF WATER. 497 



— and employs in the subsequent analysis the same 

 amount of water that was used in making this plate. 



If the original water contained so many organisms 

 that there developed on a plate or tube made with one 

 drop too many colonies to be easily counted, then the 

 sample must be diluted with one, ten, twenty-five, fifty, 

 or one hundred volumes, as the case may require, of 

 sterilized distilled water. This dilution must be accu- 

 rate, and its exact extent noted, so that subsequently the 

 number of organisms per volume in the original water 

 may be calculated. 



The use of a drop is not sufficiently accurate. The 

 dilution should therefore always be to a degree that will 

 admit of the employment of a volume of water that 

 may be exactly measured, 0.25 and 0.5 c.c. being the 

 amounts most convenient for use. 



Duplicate plates should always be made and the 

 mean of the number of colonies that develop upon 

 them taken as the basis from which to calculate the 

 number of organisms per volume in the original water. 



For example: from a sample of water 0.25 c.c. is 

 added to a tube of liquefied gelatin, carefully mixed and 

 poured out as a plate. When development occurs the 

 number of colonies are too numerous to be accurately 

 counted. One cubic centimetre of the original water 

 is then to have added to it, under precautions that pre- 

 vent contamination from without, 99 c.c. of sterilized 

 distilled water — that is, we have now a dilution of 

 1 : 100. Again, 0.25 c.c. of this dilution is plated, 

 and we find 180 colonies on the plate. Assuming that 

 each colony develops from an individual bacterium, 

 though this is perhaps not strictly true, we had 180 

 organisms in 0.25 c.c. of our 1 : 100 dilution, therefore 



