94 Darwin and Pertz. — On the artificial 
be supposed that the result would be an absence of all 
curvature. But this is not so : what happens is that the 
plant curves first in one, then in the opposite direction. 
These to and fro movements occur with surprising but not 
exact regularity in a rhythm corresponding to that of the 
application of the stimuli. In our former experiments the 
reversal of the stimulus occurred at intervals of half an hour ; 
we have now succeeded in building up a periodic movement 
in a J5-minute rhythm 1 . 
That during the continuance of the alternate stimuli a plant 
should nutate in a given rhythm is sufficiently remarkable, but 
it is far more interesting that the rhythm should continue 
after all stimulation has ceased, and this we again find to be 
the case. 
Method. 
We have again used the intermittent klinostat, employing 
of course a horizontal axis for geotropic experiments, and 
a vertical axis in the case of heliotropism. 
A cord is wound round the rotating axle and supports 
a weight over a pulley ; the axle also bears a pair of arms 
projecting from it at right angles and separated from each 
other by i8o°. As long as one of these arms is fixed the 
weight cannot turn the axle, and the plant is geo- or helio- 
tropically stimulated. At regular intervals a clock escapement 
frees the arm, and the axis rotates through i8o°, when the 
plant is at once subjected to the opposite stimulus for another 
equal period of time. The act of rotation is rendered gentle 
by a fan-governor, so that the plant is not unduly jarred. 
Experiments. 
We find that heliotropic experiments succeed with much 
greater regularity than those in which the stimulus is gravita- 
tional. This was to some extent evident in our 1892 results, 
but we can now state the case more definitely. We have 
made twelve heliotropic experiments 2 on Phalaris canariensis ; 
1 We also tried an hourly period, but without success. 
3 Including one with an hourly period. 
