THE TYPHOID BACILLUS 287 



characteristically in glucose bouillon and nutrient gelatin, besides 

 showing the specific serum reaction. Probably not one time in ten 

 thousand would such bacilli prove on further investigation not to be 

 typhoid bacilli. A still further test is to inoculate animals with several 

 doses of the dead bacilli whose identification is sought, and note whether 

 there is produced a serum which specifically agglutinates undoubted 

 typhoid bacilli. 



