PATHOGENIC BACTERIA IN DIFFERENT WATERS 295 



sterilised water, and on this head a great deal 1 of the 

 work done is unfortunately quite unreliable. It has 

 already been pointed out (see p. 267) how essential it 

 is to employ special methods for the detection of 

 typhoid bacilli in natural waters, owing to the impossi- 

 bility of recognising them with ease and certainty on 

 an ordinary plate-culture containing numerous other 

 colonies ; it is, of course, equally important to employ 

 these special methods in looking for typhoid bacilli in 

 purposely infected unsterilised waters, and this has, 

 unfortunately, only been done in very few of the re- 

 searches the results of which are tabulated above. 



There cannot be the slightest doubt that, if only the 

 ordinary method of plate-cultivation is employed in 

 such investigations on unsterilised water, the typhoid 

 bacilli will be generally overlooked unless they are 

 present in large numbers. Again, the attempts which 

 have been made by some experimenters to count the 

 typhoid colonies on such mixed plates, and the nume- 

 rical estimates given of the typhoid bacilli in such 

 unsterilised waters, must be wholly illusory, for the 

 number of typhoid colonies which develop what may 

 be called a typical appearance (i.e. one which enables 

 them to be readily diagnosed with reasonable certainty) 

 depends on a variety of different circumstances, amongst 

 which may be mentioned the age of the plate, the ex- 

 tent to which the colonies are crowded together on the 

 plate, very probably, also, the nature of the other 

 colonies on the plate, and certainly the degree of 

 vitality possessed by the typhoid bacilli themselves. 

 Thus most of the investigators in question rely for 

 their diagnosis of the typhoid colonies on mixed plates 

 on the characteristic surface expansion-colonies of the 

 typhoid bacilli, but nothing is commoner than to find 

 only a vanishing proportion of the total number of 



