Stomata , Leaf Water -content, and Witting of Plants. 367 
balance; on the right-hand pan of the balance were placed weights just 
insufficient to counterbalance the potometer. The beam was then released 
and its swing observed. With the loss of water by the plant the weight of 
the potometer eventually became equal to the weight on the right-hand pan, 
and the time when this occurred was noted with a stop-watch. From the 
right-hand pan of the balance were then removed weights totalling about 
50 to ico mg., the actual (most convenient) figure being found by 
experience. The beam was again set swinging and the time taken to 
restore the equilibrium by further water loss was re- 
corded. Thus, instead of determining the decrease in 
weight during a period of known length, the time 
during which the plant lost a known weight of water 
was measured, and from this the transpiration rate was 
calculated. For the short period observations necessary 
in the present experiments this method was found 
more suitable than the more usual one of weighing 
at regular intervals. Stomatal changes were recorded 
automatically by the use of a modification of the 
recording porometer previously described ( 16 ). The 
modified form of the recorder has been in use for some 
considerable time, and, owing to its portability and 
reliability, it is deemed worthy of detailed description 
here. Air is drawn from the porometer cup through 
a Mariotte bottle aspirator (a) (Fig. 2) in the usual 
manner, a constant pressure difference being maintained. 
In the bottom of the aspirator is a shallow layer of 
mercury (b), the surface of which just touches the end 
of the tube (C), from which the bubbles of air emerge 
into the aspirator. Down the centre of this bubble 
delivery tube passes a mercury platinum electrode (d), 
the platinum point of which just fails to make contact with the surface 
of the mercury in the aspirator. Leads (G) from this mercury layer and 
from the electrode above it connect the two in series with a battery 
and a magnetic pen marking on the paper on a time-drum rotated by 
clockwork. The air enters the bubble delivery tube and passes down 
the annulus between it and the electrode, eventually bubbling up through 
the water in the aspirator. As a bubble is formed it depresses the 
mercury surface immediately below it. On the release of the bubble, 
the mercury, before coming to rest again, oscillates slightly, and in 
doing so makes contact momentarily with the platinum point of the 
electrode, thus closing the electric circuit through the battery and the pen. 
The pen is deflected and makes a record on the drum which indicates the 
release of a bubble. Thus the passage of every bubble is recorded, and 
D d 2 
