v] METHODS OF INOCULATION 71 



successively with the water with which they are to be filled, 

 are quite satisfactory. 



(4) In order to determine the number of bacteria present 

 per I cc. in a given sample of water by means of gelatine 

 plates, a sufficient quantity of the water ought to be used to 

 yield a fair number of colonies, such as can be counted 

 fairly accurately.^ If the water has a large number of 

 bacteria — this can be easily determined in a few minutes by 

 making a film preparation — i.e. by depositing one or two 

 drops of the water, after shaking, in the centre of the cover- 

 glass, drying and heating, staining and mounting and sub- 

 jecting it to microscopic examination ; after some practice 

 it is quite possible to say from such examination whether the 

 water has comparatively few or many bacteria, and accord- 

 ingly to make the plate-cultivation with a small quantity, 

 say j^fj- or less or more up to i cc. 



It ought further to be remembered that the rapidity with 

 which the colonies appear in the plates depends in the first 

 place on the temperature at which the plates are kept. If 

 the plates are kept in a cool place — e.g. in a cupboard at the 

 temperature of the room in the cold months — the growth is 

 extremely slow, and the colonies appear only after many 

 days. I have made comparisons in this respect. I have 

 seen it stated by the water analysts of the London water 

 companies that during particular months of the cold weather 

 the number of bacteria in London waters as determined by 

 gelatine plates kept in a dark cupboard at the temperature 

 of the laboratory, which was certainly under i7°C., and the 

 counting being done forty-eight hours after, was between 7 

 and 70 per i cc. ; whereas in my experiments under similar 

 conditions the plates were counted not after forty-eight hours 



^ What is here stated of water refers to all other fluids of which the 

 number of microbes are to be determined by pla'e-cuUivatioi}. 



