66 
Bateson and Darwin. 
horizon as it curves upwards, is exposed to an increasing 
stimulus. In the same way the stem, starting from above 
the horizon, curves away from the optimum position, and 
thus encounters a stimulus diminishing in a similar ratio. 
It is clearly therefore extremely difficult to find out what 
were the initial geotropic tendencies corresponding to the 
two positions. The difficulties inherent in this experiment 
made it seem desirable to apply to the question a method 
differing from those hitherto employed. 
If a flower-stalk (or other apogeotropic organ) remains 
for an hour or two pinned down to a board in a horizontal 
position, so that no curvature can take place, a well-known 
result is seen on its being released : — the freed end springs 
up with a sudden geotropic curvature. Our method is based 
on this fact. Geotropic stems were immoveably fixed at 
various angles, and the amounts of curvature occurring on 
release were taken as representing the geotropic stimulus 
corresponding to each position. Whatever may be the faults 
of the method, it has one merit, namely, that the organ is 
exposed to a constant instead of to a varying stimulus, as 
must be the case if the stem is free to curve during the period 
of stimulation. 
Sachs 1 has compared shoots constrained in this manner 
in a horizontal position, with shoots fixed at one end only, 
and therefore free to move from the first. He shows by an 
analysis of the distribution of growth and tension, that the 
curvatures in the two cases are of an essentially similar nature. 
These results encourage us to believe that we are right in 
drawing conclusions as to normal geotropism from the be- 
haviour of constrained shoots ; for we do not consider the 
difference pointed out by Sachs between the two classes of 
curvature sufficient to vitiate our method. 
Our experiments were made in the following manner : — 
Young flower-stalks of plantain (Plantago lanceolatd) were 
gathered, and after the removal of the flower-heads were 
pinned on to boards. This was not done by transfixing 
1 Arbeiten, i. p. .204. 
