380 L. D. BUSHNELL 



is to obtain an apparatus by means of which the various changes 

 may be noted and recorded over a series of days, since the fermen- 

 tations change with the age of the culture to a marked extent. 

 We hoped to obtain more accurate information by this method 

 although the amount of work necessary is multiphed in proportion 

 to the number of determinations made. 



Of course, it is impossible under all conditions to make the 

 medium absolutely the same. However, we have attempted to 

 do this as far as possible, by obtaining enough material in one lot 

 for an entire series of tests. The same asparagus was used 

 throughout; and for the carbohydrate fermentation tests, we 

 used a 2 per cent peptone-water to which, after sterilization, was 

 added 1 per cent of the carbohydrate in question, except in case 

 of glucose of which 0.5 per cent was used. We used peptone- 

 water rather than broth because this could be made more uni- 

 formly, and did not require an adjustment of the reaction. Un- 

 fortunately these organisms would not grow in any of the mineral 

 solutions usually recommended for the culture of nitrogen fixing 

 anaerobes, so their action in a medium of known composition 

 could not be determined. 



In their experimental work Wolf and Harris used a large 

 2-liter bottle for culturing the anaerobes. Tliis bottle was fitted 

 with a two-holed rubber stopper through which passed two glass 

 tubes, one of the tubes extending into the space above the liquid 

 and one into the liquid itself and nearly to the bottom of the 

 bottle. The container was exhausted and the culture allowed to 

 develop in vacuum until the space above the liquid was filled with 

 gas. Samples were removed at intervals by means of the tubes. 

 The presence of the gases was utiUzed to force out enough hquid 

 for an analysis. 



The apparatus described by these investigators could not be 

 used by us, since our cultures did not grow very rapidly, and we 

 invariably found that a leak had occurred and air had entered. 

 Also, we experienced difficulty with contamination, in sampUng 

 according to their method. Aside from the difficulties mentioned, 

 the apparatus recommended by Wolf and his associates would 

 probably give more uniform results for a single series of examina- 



