QUALITATIVE BACTERIOLOGICAL ANALYSIS. 201 



carefully examined. Each colony whose appearance is suspicious 

 must be inoculated on to a tube of litmus-agar . . . It is 

 not enough to inoculate two or three tubes from suspicious 

 colonies . . . five to ten tubes may be sufficient if the 

 water is comparatively pure. But if the water has been exposed 

 to very obvious contamination, the resulting colonies on the 

 agar will be very varied in aspect, and it will be necessary to 

 inoculate from them ten to twenty or an even larger number of 

 litmus-agar tubes." The litmus-agar contains twenty-five 

 grammes of litmus and thirty grammes of milk sugar to the 

 litre of nutrient agar. " On the day after their inoculation a 

 number of the litmus-agar tubes will be found to have turned 

 red, and may be at once discarded. After a further twenty- 

 four hours others may turn red, and may also be discarded. 

 . . . The remaining tubes that are still blue, and that have 

 the naked eye appearance of the growth of enteric, must now 

 be subjected to microscopical examination." The cultures 

 resembling B. typhosus are finally sub-cultured in milk, potato, 

 &c., and agglutination tested with anti -typhoid serum. 



Using this method, Hankin states that he has frequently 

 isolated the B. typhosus from suspected waters in India ; and, 

 as most modern bacteriologists working with other methods have 

 failed to detect the B. typhosus in suspected waters, the pro- 

 cedure appears deserving of critical study. The tube which 

 Hankin selects from his first series is usually that containing 

 two to three drops of Parietti's fluid, which corresponds to 

 0*05 per cent, carbolic acid (1 c.c. = 30 drops), the amount 

 generally used by bacteriologists at the present day. The 

 second series, commencing with three drops, would run up to 

 six drops of Parietti, or in other words up to O'l per cent, of 

 carbolic acid, in which B. typhosus will usually grow, if it has 

 not been debilitated by prolonged immersion in sewage. So that 

 if the B. typhosus were not accompanied by B. coli it should 

 be detected by this method, providing of course it was originally 

 present in considerable quantity, so that typhoid bacilli might 

 be present in the small quantity of water used for the experi- 

 ment. If, however, B. coli is present with the B. typhosus (as 

 it usually is in most polluted waters) the result will be different. 

 B. coli will grow in 0'2 per cent, carbolic ( = 12 drops of 



