Elimination of Bacteria from the Blood. 83 



experiments was the common bile duct or pancreatic duct ligated, 

 and entrance to the gut by this route excluded. 



For almost all our experiments rabbits were employed. The 

 test bacterium was Bacillus prodigiosus, which unless otherwise 

 stated, was suspended in 0.8 per cent, salt solution and inoculated 

 into the ear vein. After two experiments had shown that the 

 injected bacteria were to be found in the small intestine, further 

 experiments were undertaken to discover their portal of entry. 

 Four routes seemed possible : (1) the bacteria may enter the lung 

 and then gain access to the gastro-intestinal tract ; (2) they may 

 enter the intestine through the common bile duct or (3) through 

 the pancreatic duct, or (4) traverse the wall of the intestine. Ex- 

 periments were undertaken to discover whether this last route was 

 possible, and to this end the other paths were one by one excluded. 



In two experiments the pylorus was ligated in order to exclude 

 the possibility of bacteria descending from the respiratory tract. 

 As in all the experiments the inoculation of test bacteria was not 

 made until the animal had recovered from the shock of the opera- 

 tion, which interval was generally one to two hours. After the 

 inoculation the animal was killed in from one to three hours. 

 Cultures were then made from different levels of the gut, under 

 the precautions as to sterility which I have mentioned in my 

 previous paper. Varying amounts of the contents of the intestine 

 were used for culture, from one platinum loop of 4 mm. diameter 

 to 1 c.c. ; this was transferred to large tubes of broth, from which 

 in turn agar plates were made. In some instances forty to fifty 

 broth tubes were inoculated. In these preliminary experiments, 

 in which the pylorus was ligated, it was proved that if three loops 

 of agar culture of Bacillus prodigiosus were inoculated, this bacillus 

 could be recovered from the small intestine two and one half 

 hours later. 



Of three experiments in which the common bile duct was 

 ligated, two gave positive results. The smallest amount of culture 

 employed was two platinum loops. Bacillus prodigiosus could not 

 be recovered from the stomach, but was present in large numbers 

 in the bile. In one of these experiments rabbits' serum was used 

 as a menstruum instead of salt solution. 



In five experiments the pancreatic duct as well as the com- 



