1900] MECHANISM OF ROOT CURVATURE 4 II 
of the organ is not rigid but flaccid, while a contracting muscle 
takes up water (36, pp. 446, 447), and is more rigid than in the 
unstimulated condition. Loeb’s theoretical explanation as to 
how gravitation causes a curvature presupposes that gravitation 
acts directly on the cells concerned in the curve (36, p. 448). 
This is not the case, however, in roots where the tip only is 
sensitive to the stimulus, and the changes in the cells where the 
curve is produced are brought about by an influence transmitted 
from the tip. The objections of Elfving and Noll to Wortmann’s 
theory as applied to unicellular plants seem to me to be equally 
valid against the theory of Loeb for these plants. 
An article on root curvature was published by D. T. 
MacDougal in the Boranicat Gazerre for May 1897. Among 
other conclusions he has the following : ‘The curvature of roots 
is due to the excessive active elongation of the internal layers of 
the cortex, of the side becoming convex, made feasible by the 
increased stretching capacity of the longitudinal membranes. 
The extension of the membranes is accompanied or preceded by 
changes in the quality of the membranes as indicated by their 
reaction to staining fluids.” MacDougal believes, then, that the 
convex side is alone active in producing the curvature, and he 
supports his view by measurements of cells on both the concave 
and convex side of curved rpots, which are compared with 
m€asurements of cells in straight roots. The figures given show 
that, in the roots of Zea Mais curved geotropically 105°, the 
cells of the convex side are 3.2 times as long as those in the 
normal root, while the cells of the concave side are 2.1 times as 
long as the normal ones. Not only does the convex side grow 
much faster than the normal root, but the concave side also is 
much accelerated in growth. These results contradict the obser- 
vations of several writers that the concave side often actually 
shortens, and indicate a rate of growth in stimulated roots which 
has not been observed by any other writer. I can only conclude 
that the cells measured in the normal root did not correspond 
to those measured in the curved root. The paper is obscure on 
the point as to how the author knew what cells to measure. 
