? 
52 BOTANICAL GAZETTE [ JANUARY 
membranes on that side are really more extensible than those of 
normal stems. My own experiments show that the convex half of 
the stimulated roots which are split when the curve begins, curve 
much more rapidly that the similar half of unstimulated roots; 
and to that extent my experiments may be said to support the 
view that the convex side of stimulated organs is more exten- 
sible than similar cells of normal organs. But that this is the 
only factor, or even the main one, in the curvature of stimulated 
organs is disproved by the experiments of Sachs and Kohl for 
stems, and my own experiments on roots, which show that in 
both stems and roots changes also take place in the concave side 
of an opposite kind from those in the convex side. Enough has 
been said to show that MacDougal’s conclusions as to the 
mechanism of root curvature cannot be accepted as adequate. 
Since the theories of Wortmann, Kohl, and Loeb all regard 
the concave side as alone active in the formation of the curva- 
ture, they are inadequate at the best. They take no account of 
the decreased tension in the convex side of curving stems, and 
the increased tension in the convex side of curving roots. Wort- 
mann’s theory also fails to account for the contraction of the 
concave side. Moreover, objections have been made to it by 
Noll, Kohl, and others which remain unanswered, and probably 
are unanswerable. 
Aside from the inadequacy of Kohl’s theory as a complete 
explanation of the curvature, it is not even proven as the cause 
of contraction of the concave side. While theoretically possible, 
perhaps, it requires that the cells in the region of curvature have 
their longer diameter parallel to the longitudinal axis of the 
curving organ. Kohl’s own work shows that this condition does 
not exist in grass nodes, and my observations show that it does 
not exist in the tip half of the curved zone of Vicia faba roots. 
Moreover, Kohl’s theory would seem to require that at the 
beginning of plasmolysis the cells of the concave side should — 
lengthen, and the curve be somewhat straightened; but Kohl and 
Noll both found an increase of the curvature at the beginning of 
plasmolysis, when the curve was not too old. 
