eee ee ee 
aa gears lll, tte es nde ae aE dea ea ei Cohan 
1900] MECHANISM OF ROOT CURVATURE 29 
The separation of the cortex from the axial cylinder was 
done by first cutting off the cortex with a sharp razor on two 
opposite sides, in to the axial cylinder, for a distance from the tip 
equal to the length of the growing zone. This left a lamella of 
the root consisting of the axial cylinder, with a narrow strip of 
cortex on two opposite sides. Then 1™ of the tip was cut off 
exactly transversely. With a very fine-pointed, thin scalpel the 
strips of cortex were separated from each other, but all left 
attached to the upper part of the root. The separation can be 
made quite accurately with roots as large as those of Vicia faba. 
The operation left the strips of tissue free to curve or straighten, 
lengthen or shorten. The time required to perform the opera- 
tion was two or three minutes, and the roots were dipped in water 
several times during the operation to prevent wilting. After the 
separation they were laid in water and were observed for changes 
during several minutes. 
TABLE VI. 
POSITIVE TENSION OF CORTEX IN STRAIGHT ROOTS AS SHOWN BY 
DIFFERENCE IN LENGTH ON FREEING FROM AXIAL CYLINDER. 
In water In KNO, 10% 
Cort Cortex Cortex Cortex Cortex ‘ortex 
Date 1897 — ae shorter | and axis — longer shorter | equal to 
ib than axis | than axis equal than axis | than axis axis 
January ? 8 7 I 
March 29 10 5 I 4 9 6 2 I 
March 30 10 5 2 3 7 6 I 
April 20 16 15 I 15 13 2 
44 32 3 9 31 25 2 4 
72.9% 6.8% 20.5% 80.6% 6.5% 13% 
In these experiments the roots were observed especially for 
the difference in length between the cortex and.axial cylinder 
after the strips were separated from each other. Forty-four roots 
were observed in water, and thirty-one of these were plasmolyzed 
in 10 per cent. KNO,. In almost three fourths of the roots the 
cortex was longer than the axial cylinder when they were placed 
