328 BOTANICAL GAZETTE [May 
saema triphyllum, and Phoenix dactylifera. On account of the fact 
that Zea has been used in so many researches of this character, 
and because so many of the minor features are well known, I 
have taken it through every phase of treatment. 
It is universally admitted on all hands that the forces actually 
productive of curvatures are manifested in the newly formed 
cortex of the convex side of the root, and the point upon which 
question is raised is whether the elongation of the cortical cells 
is due to actual growth of cells or is due to a sudden induced 
ductility and elasticity of the longitudinal membranes. As will 
be seen below my results give direct evidence upon this point. 
The first direct work upon this point was done by Cieselski 
(1), who concluded that the changes in the motor zones of 
curving roots consist chiefly in a greatly exaggerated increase 
in size in all directions of the cells (of the cortex) of the convex 
side, and not only a decreased growth of the cells of the concave 
side, but also a compression of ‘these cells. The cells of the 
convex side are enlarged in all three axes, and the cells of the 
concave side in every axis are below the average size, while the 
walls are wrinkled and folded. Since Cieselski’s work has been 
made the basis of so much recent work which must be corrected 
in the light of my own results, I quote his paragraph contain- 
ing this matter in full, and reproduce the figure showing the 
structure of a curved root. 
Schon die der Untersuchung des Langschnittes einer solchen stark ge- 
kriimmten Wurzel fallt es auf, dass die Zellen der Epidermis und des Rinden- 
parenchyms der unteren concaven Kante vielfach gegeneinander verschoben, 
keilformig zusammengedruckt sind, und nicht selten Falten in den atisseren 
Conturen des concaven Bogens erscheinen, wihrend die obere convexe Kante 
eine gleichmassige Spannung und stark ausgepragte, regelmassige Entwick- 
elung der entsprechenden Zellen zeigt. Das mikroskopische Bild iiberzeugt 
uns hiernach mit wee Eeneethen, dass die an der convexen Seite 
gelegenen Zellen e h allen Richtungen erlitten 
dadurch die Zellen der concaven Kantenicht nur an der entsprechenden Ver- 
grésserung gehindert, sondern sogar comprimirt haben, wie dies die vielfachen 
wir nun die Grosse der Zellen an den beiden Kanten genauer, so finden wir, 
dass die der convexen nicht blos der Lange nach, sondern auch nach den 
beiden anderen Dimensionen weit iiber das normale Mass ausgedehnt habe 
