10 



been prepared in the manner described above, the shoots 

 can be placed in the cyhndrical vessel A for the assi- 

 milation-experiments, 



Shoots having 700 à 750 leaflets (not counting the buds) 

 were always chosen for the experiments. 



The shoots were fastened with a pièce of raffia to a 

 glass rod, which, after removing the rubber stopper, was 

 put in the vessel À, which was closed again afterwards. 

 By means of the COj-apparatus the water was now pas- 

 sed from the réservoir B through the apparatus, but fîrst 

 the valve /c- was shut, whilst the valve in tube h was 

 open. In this way the air could escape from the tubes a 

 and b and also from the réservoir A. After this the tube 

 h was shut and the valve /r- was opened. 



Not before the air had been driven out of the other 

 tubes and the water had flowed through the apparatus for 

 some time (to take up ail the air still attached to the sprigs 

 and the glasswalls) was the water caught in the flasks. 



b. Collecting the water. We hâve already told, that 

 the water is caught in one of the flasks F^ and Fo. The 

 contents of thèse flasks being about 300 cM'\ was accurately 

 measured. Before the rubber stoppers were put on, the 

 flasks were filled up to the brim with an alcohol-solution 

 of 7 7o' tinged with methylenblue. The current of water was 

 lead to the bottom of the flask by the tube e and dis- 

 placed the blue alcohol which was floating on the surface. 

 In this way, mentioned by Blackman^), I succeeded 

 in collecting the water whilst avoiding its contact with 

 the air. The disappearance of the blue colour told the 

 moment when the whole was driven out of the flask and 

 its contents consisted only of the ordinary efïluent water. 



By opening first the valve ki and shutting /co and by 

 reversing the manipulation, F^ will be filled first and then 



') Proc. of the Roy Soc. of London. Vol. 83 B. 1911. p. 379. 



