1907.] Bacteria may be present in Ventilating Pipes, etc. 265 



4 hours and incubated, when typical members of the coli group were again 

 readily isolated. As a final test, the possibility of the B. typhosus being 

 ejected from typhoid stools was ascertained by using the apparatus shown in 

 fig. 7. The trap was filled with sewage, and litmus-lactose-nutrose-agar plates, 

 the media facing upwards, were suspended in the vertical pipe, which was 

 afterwards covered with a glass plate. A typical stool, obtained from a case 

 of enteric fever under treatment in the military hospital, was then mixed 

 with 2 gallons of water placed in the can connected by indiarubber piping, 

 1*5 inches in diameter, with the house side of the trap. The taps were next 

 opened, and the contents of the can having passed slowly through the trap, 



Ftg. 7. 



'//ss//s//SSJ/SSS/SW/SJ/sASS/St 



were collected in the can connected to the P outgo of the trap. The taps 

 were then turned off and, the cans having been changed, the infected sewage 

 was again passed through the trap. This procedure was followed until the 

 sewage had passed 12 times through the trap. The plates were then removed 

 and incubated. Next day several transparent, blue colonies were observed 

 these were fished and planted on agar slopes. The growths resulting were 

 tested with an anti-typhoid horse serum, and one was found to be completely 

 agglutinated by the serum diluted 1 — 500. The growth was then submitted 

 to the usual tests, which showed that under the conditions mentioned a true 

 B. typhosus had been carried up the vertical pipe. 



Two days later the experiment was repeated with another stool from the 

 same patient, with the result that three colonies of B. typhosus were isolated, 

 one being in a plate 2 feet above the trap and the other two in a plate 

 3 feet 6 inches above the trap. 



The experiments were so conducted that no splashing could possibly occur, 

 and on looking through the glass plate on the top of the vertical pipe, when 

 sewage was flowing through the trap, a few bubbles were seen bursting at the 

 surface of the fluid. The pipes and trap employed were quite new, and had 

 not been used in any of the previous experiments. 



