1913.] Action of Electric Discharge upon B. coli communis. 353 



acid were mixed, after all air had been replaced by hydrogen, by means of 

 repeated exhaustings with a Geryk pump and refilling. "When the potash 

 and pyrogallol were allowed to mix they constituted both a test and an 

 absorptive agent for oxygen, very little of which was present, judging from 

 the very faint coloration of the solution. 



After this treatment the hydrogen was led to the bell-jar, which for these 

 trials was fitted with an exit delivery tube. This exit tube was connected 

 through two more wash-bottles, the first being another test of pyrogallate, 

 and the second merely to prevent diffusion of air back into the first. The bell- 

 jar was rendered air-tight by sealing it to the metal plate with a stiff wax. 

 The apparatus was then exhausted by means of a Geryk pump and slowly 

 re-filled with purified hydrogen, this being done three times, after which the 

 bell-jar was found to be free from oxygen. The discharge was now switched 

 on, a stream of hydrogen being kept continually passing through the 

 apparatus during any exposure. 



Continuous discharge upon infected agar, for periods varying from 

 30 minutes to 2 hours, failed to produce any toxic effect, the colonies 

 developing as quickly and in as great a number after exposure as normally. 

 This result is the reverse of that obtained by previous investigators, Thornton; 

 and also Fouler ton and Kellas, having stated that the discharge proved fatal 

 in hydrogen as well as in air, though the latter give no indication that any 

 attempt was made to exclude oxygen completely. They attribute their 

 result to the formation of hydrogen peroxide, which, by quantitative tests and 

 subsequent trials with definite concentrations, they show to be produced in 

 quantity sufficient to destroy the bacteria. 



Thornton, on the other hand, assumes at the outset that no hydrogen 

 peroxide was formed in his experiments, but makes no statement as to any 

 test employed to detect it. It may have been that the compound was indeed 

 formed, and that it was responsible for the sterilisation. Such a state of affairs 

 is probable if the hydrogen atmosphere contain small quantities of oxygen, as 

 was shown by some experiments of ours with such mixed atmospheres. 

 Infected plates exposed to the discharge for 40 minutes, under such 

 conditions, show after incubation a small clear space immediately beneath 

 the discharging point, but the effect never approaches that obtained in air. 

 Quantitative determinations, made by a series of comparative colour tests 

 with the above-mentioned titanium solution, disclosed the fact that the 

 presence of oxygen induced the formation of hydrogen peroxide in varying 

 quantities. 



