Francis Darwin. 
16 
purpose the intermittent klinostat was arranged so that the plant 
remained for unequal periods in the alternate positions. When the 
difference of the periods had approximately reached its limit the 
resulting curvature was very small, and required three to four hours 
for its completion. This was effected by the plant being alternately 
horizontal for 353" and, after a sudden rotation through 180°, 
horizontal in the other position for 367". Thus the plant curves 
when the alternating periods of exposure differ by 4 %, and this 
percentage was found to hold for all observable rates of rotation. 
Thus if the plant is alternately exposed for 25" and 26" a slight 
curvature results, and here again the exposures areas 100: 104. 
Other very interesting results are obtained by intermittent 
stimulation, namely a knowledge of the relaxation time, i.e., the 
time required for the practical extinction of the state of excitation. 
The methods were two, i.e. (i.) to expose the plant alternately 
in the horizontal and vertical positions : (ii.) periods of horizontally 
were alternated with periods of continuous rotation. Both methods 
give alternations of excitation and repose, and the two methods 
are mutually confirmatory. When the period of Rest is to the 
period of Stimulation as 5 : 1, the reaction occurs as quickly as 
when the plant is exposed to continuous stimulation. When Rest 
is to Stimulus as 11 : 1, slight variable curves occur after three 
to five hours. When Rest is to Stimulus as 16: 1, no curvature 
follows. These results were obtained with epicotyls of Vicia faba : 
with roots of V. faba there was no curve with 5:1, and hardly any, 
with 3:1. The results were found to depend strictly on the 
proportion between the times of stimulation and repose, and 
this proportion was the same whether the actual periods of 
exposure were short or long. It is clear that the above experiments 
give the relaxation time. When the period of Rest was to that of 
Stimulation as 16 : 1, no reaction followed because the excitation 
of the first stimulus had died out before the second occurred, and 
the second before the third, so that there was no possibility of 
adding together the excitations. 
Let us return to intermittent stimulation, when the proportion 
between Rest and Stimulus is such that the excitation from one 
stimulus does not die out before the next is added. Fitting tried 
the following experiment. A plant A is exposed to intermittent 
stimulation, i.e., to alternate periods of repose and excitation. At 
the same time a similar plant B is exposed to continuous stimulation. 
Then both A and B are placed on the continuous klinostat so as 
