382 Dr. F. F. Blackman and Mr. A. M. Smith.  [Dee. 19, 
chamber being kept horizontal. It may take some minutes to make a 
completely air-tight joint all round, and the application of the gas-flame may 
warm the chamber up a few degrees, but as there is a thermometer inside, 
it can be seen that the temperature reached is not injurious. During this 
time there is no water inside and, before admitting water, the chamber is 
tested for leaks by a gauge. 
The chamber is then adjusted upon a tall brass stand (not shown in the 
figures, but similar to the one figured in “ Assim. and Resp.,” III, p. 57), and 
the whole lowered into the water-bath, previously regulated to the desired 
temperature. The position of the stand is shifted till the front of the 
chamber is 4 cm. behind the glass window of the bath, and the appropriate 
connections made at (¢), (7), and (9). 
The chamber is now to be filled up* with the proper CO:-containing 
liquid which has been previously made and transferred into the supply 
bottle A. This is done by disconnecting the outlet tube e just above the 
valve Y and sucking till the solution first fills the chamber and then the 
valve and tubes. As soon as the tap /# is turned to one of the pipettes D 
or E, the flow will commence. In this particular experiment this was done 
at 11.19 a.m., and pipette D (previously filled up with blue liquid) then 
began to overflow above at a steady rate, the overflow liquid dropping into 
the litre cylinder F. Illumination of the plant was begun at 11.25 by 
lighting the triple high-pressure burner (placed 13°4 cm. from the front of 
the chamber), and the circulation of water through the water-screen in front 
cf the bath was started at the same time. 
Everything has now been arranged, and the experiment starts its course 
at 11.25, see Table I on p. 383. It is obvious that, with a slow flow through 
the chamber, it will be some time before the contents of the pipette D repre- 
sent the full effect of assimilation in removing CQ» from the liquid. With 
rates about 300 c.c. per hour, the current must flow for more than an hour 
before it is worth while analysing the effluent. 
In this experiment, as the table shows, the first sample of the effluent was 
withdrawn at 12.45, the current being then switched through pipette E 
already full of blue liquid. The COs left in the 200 cc. of liquid from D 
was still equal to 22:18 ec. N/10 HCl, and comparison with the later 
numbers in the column shows that the full effect had not then been reached. 
The experiment was continued till 7 P.M., and six more analyses of the effluent 
were made, about one every hour, each representing the liquid flowing in 
40 minutes. About once an hour also an analysis was made of the afitwent 
* Before doing this, the dropping of acid on to the marble in the tower is started and 
adjusted to the appropriate rate known from previous experience. 
