128 Messrs. S. G. Shattock and L. S. Dudgeon. [Feb. 20, 



B 



which is submerged and kept in a Dewar vacuum flask of liquid air. The 

 use of mercury was avoided in producing the initial vacuum, in order to 

 exclude the presence of mercury vapour, which might in various ways 

 invalidate the results of the experiments. 



The apparatus was stored in a dark room to eliminate the action of light, 

 and for the same purpose metal foil was wrapped around the tubes during 

 the process of exhaustion with the pump. The charcoal at the temperature 

 of liquid air has a remarkable capacity for 

 gas absorption. The action of this substance 

 was maintained sometimes for three days, 

 sometimes for five, during which time the 

 vapour of water distilled continuously, and 

 was condensed on the walls of the condenser 

 at —195° C. whilst the gases were absorbed 

 by the charcoal. The way connecting the 

 tube with the bulb of charcoal was then 

 sealed off in the blowpipe flame, and the 

 tubes were stored in light-tight photographic 

 bags in a metal box, which was, furthermore, 

 kept in a closed cupboard. 



The possible ways in which the vacuum 

 may suffer are (1) in sealing off the side 

 outlet, by the evolution of gas from the 

 heated glass ; (2) owing to the presence of 

 microscopic fissures at the junctions of the 

 apparatus, especially if many chambers are 

 connected up with the same charcoal re- 

 ceptacle; (3) by reason of the fact that 

 glass is hygroscopic, and that even after a 

 prolonged period of high exhaustion, water 

 might fail to be removed, at the ordinary 

 temperature, from the inner surface of the 

 vacuum tube or from the glass slip inocu- 

 lated with the organic film of peptone containing the micro-organism. 



The vacuum tubes employed were stout test-tubes of soft glass, furnished 

 with a side channel by means of which they were connected (1) with the 

 exhaust pump, and (2) with the bulb of charcoal. (See figure.) 



The inoculations were made upon rectangular slips of thin glass (No. 2 

 microscopic cover) by means of a large platinum loop, the slips having been 

 previously heated for one hour at 150° C, in the Petri dish in which they were 



Fig. 1. — A. Open end of test-tube, 

 sealed off after the insertion of 

 the inoculated slip. B. Side tube, 

 connected with exhaust pump, 

 and also with glass bulb contain- 

 ing charcoal, etc. This is eventu- 

 ally sealed off in the blowpipe 

 flame. C. Slip of thin glass with 

 dried bacterial film. 



