EXAMINATION OF WATER 389 



tained in a Petri dish, provided due care is taken to 

 thoroughly mix the sample with the molten nutrient 

 medium. 



As attempting to measure a volume of less than O'l c.c. 

 is not a satisfactory operation, it is best, in the case of a 

 water suspected to contain a large number of bacteria, to 

 dilute the water 50 or 100 or more times, as follows, before 

 proceeding to the examination. Small sterile flasks con- 

 taining about 49 c.c. of sterile distilled or, better, sterile 

 natural water, receive 1 c.c. of the water under examina- 

 tion. This is well mixed, and again 1 c.c. of this first 

 attenuation is taken and introduced into another flask until 

 the degree of dilution is considered sufficient. Plates are 

 then made from 1 c.c. of the various attenuations. The 

 plates are then allowed to stand at a temperature of about 

 22° C, and examined daily. It is usual to count the 

 colonies on the third or fourth day from starting the plate. 

 The accuracy of the results obtained depends to a very 

 large extent upon the care with which the organisms are 

 distributed through the nutrient medium. Care should 

 also be taken that the original sample of water is well and 

 thoroughly shaken, to evenly distribute the organisms con- 

 tained therein, before withdrawing the quantity for the 

 examination. A number of plates containing varying 

 quantities of the sample should always be taken. Great 

 care and practice are required, so as not to have too many 

 organisms on a plate. A good plan is to aim at having 

 about 100 colonies on each plate. 



Counting the Colonies. — This is done by means of Wolff- 

 hiigel's apparatus. This consists essentially of a glass 

 plate divided into squares, each a centimetre square. Some 

 of these squares are subdivided. The plate or dish is laid 

 under this scale, and the number of organisms present is 

 found by counting the number of colonies in a few of the 



