4 
: 
| 
’ 
; 
} 
j 
: 
} 
1897 ] THE CURVATURE OF ROOTS 361 
the concave side on release from the compression have under- 
gone an extension by which their length has been approximately 
doubled, and is in excess of the actual length of the cells of the 
convex side. 
The conclusion is warranted that the excitation of the root 
in a direction opposite to newly formed curvature does not result 
in a straightening of the curvature by the relaxation or contrac- 
tion of the extended convex cells, after a period of growth has 
ensued. The straightening of the curvature is due to the accel- 
erated elongation of the concave side in the same manner as in 
the formation of the original curvature. A compression or 
shortening of the convex side does not occur until the concave 
side has extended sufficiently to compress it mechanically. It is 
pertinent to state here that anything like an active contraction 
or relaxation of the cells on the side becoming concave either in 
curvations or recurvations is not to be found in roots. On this 
account the straightening of curvatures by recurvation is not to 
be adduced as evidence that curvature is due to elastic stretching 
in the manner in which it has previously been done by Sachs, 
Noll, and others. Furthermore, my preparations show that the 
walls of the originally convex side have lost their attenuated 
condition, and that the cells of the originally concave side have 
taken up this character. The straightening of curvatures by 
plasmolysis is an altogether different process, since in this 
manner the greater elastic stretching of the convex side would 
be directly released, and would allow the root to return to a 
position determined by the physical characters of the wall. The 
complications which attend the plasmolysis of tendrils (14) 
would be wanting, and the straightening of the curvature in this 
Manner, as well as the difference between the membranes of the 
convex and concave sides, would justify the conclusion that the 
curvature is due to the elastic stretching of the convex side of 
the root, and that this elastic extension was fixed or held in an 
elongated position by the loss of elasticity in any one of many 
Ways; by changes of the quality of the wall induced by the ecto- 
Plasm, or by the intussusception or apposition of new building 
