drawn off. This allows a certain amount of gas to enter, but the 

 exact percentage contained in the bell glass must be determined 

 by analysis. This is accomplished by passing a sample of the 

 gas in the Winkler-Hempel burette, which necessitates allowing 

 water to pass from / to // in order to counterbalance the air 

 pressure. After exposure to light for a required length of time, 

 other samples of air can be taken and analyzed as before. The 



Fig. 2. Ajiparalus for determining the amount of carl)on dioxide taken up by 



potted plants. 



number of samples of air which can be taken depends entirely 

 upon the capacity of the inner tube, //, and also the amount of 

 gas, or air, which is utilized each time as a sample for analysis. 

 We have found it better to supply the plants with a considerably 

 large percentage of carbon dio.xide, as this renders the results 



