ne es ey ee ee 
1900] MECHANISM OF ROOT CURVATURE 59 
further investigation. On the whole, the view that the proto- 
plasm in the cells of the concave side becomes more permeable 
seems to explain more phenomena than any other view yet 
advanced ; and while it is not absolutely demonstrated, I believe 
it to be the condition which lies at the basis of those changes 
which, as my experiments show, take place in that half of a 
stimulated root which becomes concave. 
As for the changes which take place in the half that becomes 
convex, there is certainly an increased growth on that side; and 
Noll has shown that there is in some cases a change in the qual- 
ity of the membrane. It is not proved, however, that this 
change in the membrane is a condition preceding growth, as 
Noll believed, since it may be merely the method of growth. 
This point can be decided only when our knowledge of the 
mechanics of growth is more complete. 
In the light of the foregoing experiments and arguments the 
mechanism of the curvature of roots is as follows: The stimulus 
is transmitted from the sensitive root tip to the curving parts, in 
the cortical parenchyma. The effect of the stimulus is to 
increase the normal tension between cortical parenchyma and 
axial cylinder on the side that becomes convex, and to decrease 
Or reverse the normal tension between the cortical parenchyma 
and the axial cylinder on the side that becomes concave. The 
change in tension also extends to the different layers of the cor- 
tical parenchyma on the concave side, the outer layers becoming 
negative with respect to the inner ones. So much has been dem- 
Onstrated. The evidence is in favor of the view that the ten- 
sions on the concave side are changed by the protoplasm becom- 
ing more permeable to water, some of which passes out into the 
intercellular spaces, possibly to be taken up by the convex cells, 
which later contain more water than the concave cells. The 
shortening of the concave side may be masked sometimes by a 
certain amount of growth. 
The same explanation will apply to curvatures of dicotyle- 
donous stems if allowance is made for the different normal ten- 
Sions found in them. 
UNIVERSITY OF MICHIGAN. 
