1900 | MECHANISM OF ROOT CURVATURE 55 
cells on the lower side. I see no other alternative. Greater 
extensibility of the membranes on the lower side could not 
enlarge the cells unless they took up more water. The evidence 
is against the greater osmotic pressure on the lower side; and 
the conclusion seems inevitable that the cells on the upper side, 
from the immediate effect of the stimulus, allowed water to 
escape into the intercellular spaces. This could be brought about 
either by increased elasticity of the walls or increased permea- 
bility of protoplasm to water. From what we know of effect of 
stimuli on other plant organs, the latter view seems more 
probable, though it is possible that both the elasticity of the 
walls and the permeability of the protoplasm are increased by 
the stimulus. If the stimulus can cause the cells on the upper 
side of a horizontal Hippuris stem to give out water into the. 
intercellular spaces as a direct effect, we are justified in thinking 
' that the shortening of the concave side in other stems and in roots 
also can be brought about in the same way. Noll has observed 
that when curving stems have been gashed on one side, water is 
gradually forced out of the gash, and shows in the form of 
drops. The water must have come from the intercellular spaces 
into which it had escaped from the cells on the concave side. 
MacDougal asserts (39, p. 309) that there are no intercellu- 
lar spaces in the motor zone of roots into which water could pass 
from the cells. He is certainly mistaken as regards Vicia faba 
roots in which well developed intercellular spaces are found in 
the cortex very near to the initial zone. In sections from fresh 
roots their presence is easily demonstrated by the air contained 
in them. MacDougal himself, however, says in another place 
(30, p. 309) of the roots of Zea mais: “Intercellular spaces (of 
the convex side) are larger and more abundant than in the con- 
cave side.” Evidently there are intercellular spaces in roots of 
Zea also. This makes the condition favorable for water to 
escape from the cell if the protoplasm becomes more permeable. 
That the stimulus produces this condition in roots has not been 
proved; but it seems to me to be the most reasonable explana- 
tion of the phenomena exhibited in the concave side of curving 
