290 A MANUAL OF BACTERIOLOGY 



boiling point for a short time; and that they are 

 more or less perfectly removed by some processes 

 of nitration and precipitation, whilst other pro- 

 cesses of the same nature are worthless, or even 

 worse' (Frankland). 



Before describing the methods for the bacterio- 

 logical examination of waters, we must allude to 

 (a) the collection of the samples, and (b) the trans- 

 port of the same. 



To collect the samples of water accurately stop- 

 pered bottles (70 cc. capacity) are used. These 

 must be perfectly clean, and rinsed out with dis- 

 tilled water. Each bottle is put into a small tin 

 canister, and the canisters (containing the bottles) 

 are heated in a steriliser to about 180 C. for at 

 least three hours. * The bottles thus sterilised can 

 be easily transported without suffering contamina- 

 tion by dust to the place where the sample is to be 

 collected. In collecting the sample of water the 

 outside of the bottle should be rinsed in the water 

 before removing the stopper, and when the bottle 

 is opened the water is at once allowed to enter and 

 fill the bottle to the extent of four-fifths, the stopper 

 being immediately replaced and tightly screwed in, 

 so that the exposure to the air is reduced to a 

 minimum. The bottle is replaced in the tin 

 canister, and the lid closed. In collecting samples 

 of water from rivers, reservoirs, lakes, or ponds, it 

 is better not to remove the stopper until the bottle 

 is completely immersed in the water, and to replace 

 it while still beneath the surface.' After collection 

 the sample of water should be examined as soon as 



