382 L. D. BUSHNELL 



were used but in many cases we failed to obtain growth. Care 

 was always used not to add bits of potato. 



Just before inoculation, the bottles of media were heated in 

 the steamer and then cooled as rapidly as possible to drive out 

 the air. Care was always exercised not to shake up the medium 

 during the inoculation and sealing process. After inoculation, 

 the cotton stoppers were removed and the bottles were plugged 

 with a No. 8, two-holed rubber stopper, fitted with two glass 

 tubes. One of these tubes was constricted near the upper end 

 and served as an attachment for the vacuum pump, and for 

 sealing after the container had been exhausted. The second 

 tube was bent to serve as a manometer tube (P) as indicated in 

 figure 1. The rubber topper was fitted with the tubes, wrapped 

 in gauze and then covered with a thick layer of cotton. The 

 lower end of the manometer tube and the upper end of the straight 

 tube were plugged with cotton and all autoclaved at twenty 

 pounds pressure for an hour. It is necessary to wrap the cotton 

 stopper in the mouth of the bottle, and the rubber stopper with 

 gauze to prevent bits of cotton adhering to them, since this would 

 prevent a perfect seal. 



After inoculation, the cotton stopper was discarded, and the 

 rubber stopper carrying the tubes, was unwrapped and well 

 flamed with a Bunsen flame. The rubber stopper was then 

 pushed firmly into the neck of the bottle and covered with a 

 thick layer of rubber cement. Thin blocks of wood were placed 

 on the top of the stopper, and all wired in place with No. 16 u'on 

 wire. 



After the bottle had been fitted, it was placed in a special incu- 

 bator at 37°C. and immersed in a copper tank (AI) containing 

 heavy mineral oil. The lower end of the manometer tube was 

 placed in a dish containing mercurj', after removal of the cotton 

 plug. The bottle was then exhausted as much as possible with 

 a Geryk pump. When the mercury column ceased to rise in the 

 manometer tube the tube attached to the pump was sealed at 

 the constriction with a blast lamp. Bj^ this method we could 

 obtain practically a complete vacuum as determined by the 

 temperature and barometric pressure at the time. Figure 1 

 shows the details of this apparatus. 



