46 DISTRIBUTION AND ORIGIN 



in solution to allow some to develop. This is, however, not surprising when 

 we consider that thirty thousand bacteria only contain about T uth of 

 a milligram of solid matter. Rain-water is contaminated with bacteria which 

 it has taken up in its passage through the air ; one sample contained thirty- 

 five germs per litre. The water of rivers, lakes and springs contains veiy 

 varying numbers of micro-organisms, their abundance depending largely 

 upon the amount of organic matter present. As soon as water is rendered 

 impure by the influx of such substances as are carried into rivers by sewers 

 and drainage canals, it ceases to be merely a liquid in which bacteria can 

 retain their vitality for a long time unimpaired, and becomes a nutritive 

 medium in which they proliferate with great rapidity. The water of the 

 Spree for instance contained, above Berlin, on an average 6,140 bacteria 

 per cubic centimetre, that taken below the city no less than 243,000. 



The bacteriological analysis of water is a very simple process. If the 

 microscope shows that bacteria are not too numerous, a measured quantity 

 of water (generally i c.c.) is mixed with nutrient gelatine and poured out on 

 to a glass plate or shallow dish. The germs distributed through the gela- 

 tine are allowed to develop into colonies which can then be counted in the 

 usual way. The mere enumeration of the bacteria in water is of far less 

 importance than the determination of their nature and properties. The 

 majority of forms found in rivers and springs are of an entirely innocent 

 character ; they are bacteria whose natural habitat is water, and they per- 

 form a useful task in consuming the dissolved organic impurities. But when 

 by any means pathogenic species obtain access to such waters these may 

 easily become a source of infection, particularly when, as is the case with 

 cholera and typhoid organisms, the bacteria belong to species which can 

 live and multiply side by side with the ordinary saprophytic forms (see 

 Chaps. XIII and XIV)*. When the water to be analysed contains a very 

 large number of bacteria it must be diluted, before mixing, with an equal 

 quantity of sterilized water. For the detection of isolated pathogenic forms 

 in comparatively pure water, that is to say in water containing very few 

 bacteria of any sort whatever, a method of ' nursing ' or ' enrichment ' is 

 employed. A sterile nutritive solution (peptone-sugar) is added to the water, 

 so that the few germs present may multiply before the gelatine is added. 

 This method involves of course the drawback that the non-pathogenic species 

 multiply also, and the danger that they may outnumber and overpower the 

 less abundant bacteria we are looking for. 



Numerous aquatic bacteria are able to resist freezing for long periods, 

 and ice sometimes contains enormous quantities of them (e. g. 2,000 per c.c.). 

 [Even glacier ice is not free from them.] 



Of all the natural habitats of bacteria there is none in which they are 



* For marine bacteria, see Chap. VII. 



