VITALITY OF PUTREFACTIVE ORGANISMS. 233 



contact with infectious matter every one of the flasks 

 •showed its power of sustaining and multiplying life. 



§ 29. Remarks on Hermetic Sealing. 



A few brief remarks on this subject may, I think, 

 be fitly interpolated here. Hermetic sealing during 

 ebullition is an operation requiring some apprenticeship 

 to perform it aright. The neck of the flask ought to 

 be so narrow that the pressure of the steam within shall 

 be always sensibly greater than that of the atmosphere 

 without. This condition would be readily fulfilled if 

 the liberation of the steam were absolutely uniform, and 

 not by fits and starts. But it never is uniform, and if 

 the channel througli which the steam issues be wide, it 

 is scarcely possible to avoid regurgitation. Sometimes 

 the pressure within is above that of the atmosphere, and 

 steam freely issues ; but at the next moment, through 

 liquid adhesion to the flask, and partial condensation 

 above, the internal pressure may be below that of the 

 atmosphere and permit air to enter. This alternate 

 triumph of the inner and the outer pressure may be 

 rendered plainly evident by the motions of the water 

 condensed in the neck of the flask. The liquid acts as 

 an index which moves to and fro, sometimes forward, 

 sometimes backward, as the pressure varies. It is quite 

 evident that contamination may be, and it is quite cer- 

 tain that contamination has been, thus introduced into 

 flasks reputed to be free from air. 



Even with considerable care and fairly disciplined 

 manipulatory skill success is not invariable. Ten per 

 cent, is not at all a large allowance to set down as 

 defective in ordinary hermetically-sealed flasks. The 

 recent opening of about two hundred flasks employed in 

 my earlier experiments, under water and under caustic- 



