184 BACTERIOLOGICAL EXAMINATION OF WATER. 



bacillus. Broth cultures also are not suitable for use in capillary 

 tubes ; at the end of twenty-four hours the growth is rarely 

 sufficient to give a marked reaction in the tubes, and if the 

 culture be incubated for a longer time false clumping is likely 

 to occur in the culture. P'or these reasons it is always best to use 

 a twenty-four hours growth on agar, and make an emulsion from 

 it bv rubbing up one or two loopfuls of growth in sterile broth. 



Pfeiffer's Test. This consists in injecting into the peritoneum 

 of a guinea-pig varying quantities of a twenty-four hours agar 

 growth of the microbe under investigation so as to determine 

 the fatal dose. Ten times the fatal dose, mixed with a little 

 serum (0*01 c.c. to O'OOl c.c. according to the strength of the 

 serum) from a highly immunised animal, is then injected into 

 the peritoneal cavity of a guinea-pig. If the bacillus under 

 examination be the B. typhosus the animal will remain well 

 whilst a control animal injected only with the bacillus will die. 

 In the peritoneal cavity of the animal which recovers the bacilli 

 are first converted into granular masses, and then disappear 

 under the lysogenic action of the specific serum. If the bacillus 

 be a variety of the B. coli, the animal will infallibly die ju^t 

 like the control animal. This test is of the highest importance ; 

 but, unfortunately, owing to the difficulties of the experiment it 

 is not applicable to general hygienic work. The test also has 

 this limitation, the typhoid bacillus may have lost its virulence, 

 so that a fatal result may not be obtained when the control 

 animal is injected with the agar culture. 



Results obtained by Immunising an Animal with the Micro- 

 organism under Examination. True races of the B. typhosus 

 when injected into an animal gradually produce a serum which 

 will agglutinate other races of B. typhosus. The extent to 

 which the serum so produced can be diluted, and still agglutinate 

 other races of B. typhosus, depends on the race of B. typhosus 

 used for, and the length of time occupied by, the process of 

 immunisation. Experiments at Netley showed that the least 

 pathogenic race of B. typhosus produced a serum which, when 

 diluted 1-100, completely agglutinated all the other races of 

 the typhoid bacillus in the laboratory collection. I have never 

 been able to isolate from water a micro-organism belonging to 

 the pseudo- typhoid group which would produce a serum capable 



