ESTIMATION OF THE RESPIRATORY QUOTIENT 261 



rubber stopper with three holes. Insert through one of the aper- 

 tures a suitable glass tube which reaches almost to the bottom of 

 the flask, and has the outer short end bent at right angles. A 

 second tube of the same pattern but which does not extend so 

 near the bottom as the first by a cm. is put in place in a second 

 hole, and a short tube bent at 45 is put in the third. Invert 

 the flask and attach a funnel to the shortest tube and pour into 

 the flask 100 cc. of a nutritive solution which has been thoroughly 

 infected with spores of Aspcrgilius. The nutritive solution should 

 fill the inverted flask to within 2 cm. of the bottom, and should 

 not reach the ends of the long tubes. Allow the preparation to 

 stand for a day or two until a mycelial layer has been formed on 

 the solution, then open the outlet and allow the solution to flow 

 out slowly, leaving the mycelium attached to the walls of the 

 flask. Introduce distilled water and let it stand for a few min- 

 utes to take out the traces of remaining nutritive fluid. Remove 

 the water and introduce the substances the influence of which is 

 to be tested. First put in a 2 per-cent. solution of dextrose, fill- 

 ing the flask to the original level and, after the preparation has 

 stood an hour or two, draw fresh air through the longer tubes by 

 means of a filter pump, and then fit the flask with a manometer 

 attached to one of the tubes and an apparatus for withdrawing 

 air for analysis to the second (Fig. 128). The latter consists of a 

 three-way stopcock with two long arms, attached to a burette 

 containing mercury. Fit the short arm of the three-way tube to 

 the upper end of the burette B, and bend one arm so that it may 

 be immersed in mercury in the small dish b. Attach the burette 

 to a suitable support and raise the bowl at the end of the flexible 

 arm until the mercury rises above the stopcock at a. Now at- 

 tach the other arm of the tube to the free end of the tube pro- 

 jecting from the flask. To do this it should be suitably bent 

 three times at right angles. Fill the dish D with water and at- 

 tach the mercury manometer C, making all joints secure by wire 

 bindings. This may be tested by lowering the bowl E with the 

 stopcock open, drawing the mercury up in the manometer arm. 



