1897 | THE CURVATURE OF ROOTS 
It is important to note that this is the first 
establishment of the fact that the radial 
diameter of the convex side of any root 
becomes greater than that of the concave 
side. Cieselski affirmed the same fact con- 
cerning Zea, Phaseolus, and Pisum in 1871, but 
it was disproven by Sachs a year later, and 
recently by myself in the same plants. It 
had come to be regarded, therefore, as a well 
founded fact that the radial diameters of the 
convex sides of stems increase during curva- 
ture and those of roots decrease, and that 
while the longitudinal diameters of the cells 
of the convex side of roots increased the 
radial diameters did not change or decrease, 
while exactly the reverse conditions were to 
be found in the concave side. 
It is noteworthy in this connection that 
the roots offering similar conditions to stem 
curvatures exhibit similar reactions, and it 
seems reasonable to conclude, therefore, that 
since the morphological character of the tis- 
Sues involved is not always identical, this sim- 
ilarity in behavior is founded upon mechan- 
ical necessities. Furthermore, it is to be said 
that the roots of Phoenix offer unmistakable 
gaa of the shortening of the concave 
side. 
Fig. 7. Median 
longitudinal sec- 
tion of curved por- 
of excitation. C, 
tables XIII, XIV). 
XIV. INTERPRETATION OF EXPERIMENTAL RESULTS. 
The most important question involved in the solution of the 
Various problems connected with curvature is the determination 
of the nature of the changes involved in the extension of the 
cells of the convex side of the organ, to ascertain whether the 
elongation of the membranes is due to the actual intussusception 
of new material, or whether the membranes undergo induced 
