Wo to a lower température than it had in the bath W^. 



The water after having been raised to the desired 

 température and having passed the shoots, is again con- 

 ducted through a glasstube d which has been bent several 

 times and which is also placed in a waterbath: hère it is cooled 

 to a low température (that of the tapwater) by means of 

 a water current. After this the water is caught in one 

 of the small flasks Fj and Fo. différent précautions being 

 taken, which will be described further on. The liquid, dis- 

 placed from the flasks by the overflowing water, is caught 

 in a measuring cyHnder, in order to détermine the velocity 

 of the watercurrent. 



It will be évident that the current is brought about by 

 siphonic action and that the velocity can be regulated by valves. 



Figure 1 shows that it is also possible to conduct the 

 water immediately to the flasks, without flowing along the 

 /fe/o<iea-shoots. Before and behind the cylindrical vessel 

 A a T-piece is put in the water-pipe and by opening and 

 shutting the valves k^, k^, /c^ and k^, the course of the 

 water may be regulated. 



Lastly it should be mentioned that a half-watt lamp L, 

 with a maximum power of 4000 candies was used as the 

 source of light. By linking in a rhéostat R (see Fig. 2) the 

 light'intensity could be regulated. The daylight was excluded 

 entirely and at the same time care had been taken to 

 reflect the light of the lamp as much as possible on 

 the shoots by enveloping the lamp with white asbestus, 

 covered outside with a black cloth and by placing the 

 vessel A in a small wooden box K. This box is furnished 

 at back and front with a pane of glass, the front one to 

 arrest the heat of the lamp, the back one for observation 

 of the shoots during the experiment. 



In the upper side of the box is an opening, to allow 

 the passage of the tube A. This box K is also covered 

 with a black cloth. 



