14 BOTANICAL GAZETTE [JANUARY 
6. By increased formation of osmotic substances on the 
convex side, accompanied by extension. De Vries. 
3. Difference in the quality of cell membranes. 
a. Concave membranes thicker, more resistant. Wort- 
mann. 
6. Convex membranes thinner, more ductile. Wiesner, 
Noll, MacDougal. 
To be sure, all the theories must go back ultimately to the 
activity of protoplasm; but there is a question whether any fac- 
tors are concerned in the curvature which are not concerned in 
ordinary growth, and whether the activity of the protoplasm is 
manifested in a change in its own condition, a change in the cell 
contents, or a change in the condition of the cell membrane. 
The curvature of grass nodes can be explained only by an 
increased growth of tissue on the convex side. Some contrac- 
tion on the concave side is not excluded, but such contraction 
could not possibly develop energy enough to curve the organ. 
Plastic stretching on the convex side is also inadequate in grass 
nodes. Indeed, the changes in quality of the membranes on the 
convex side, which Noll has described, are similar to those which 
Strasburger has shown take place in cells of Cladophora when a 
new branch forms. Hence Noll’s plastic stretching may be only 
a phenomenon of growth, and not a condition peculiar to curving 
organs. Increased growth on the convex side, however, will not 
explain the shortening of the concave side which has been 
reported by several observers. De Vries’s view of increased 
turgor on the convex side is not tenable. 
Original experiments. 
_ MATERIAL AND METHODS. 
In the following experiments a study was made of the 
traumatropic curvature of roots, and the materials used were the 
roots of Vicia faba, the broad or Windsor bean. Some of the 
experiments involved a separation of different tissues, and this 
Species was chosen because of the large size of the roots, which 
were, therefore, easily manipulated, and in which the liability to 
