502 BOTANICAL GAZETTE [DECEMBER 
grew upon the solutions. The upper half of the root was in moist 
air and the lower half in the solutions. The roots grew straight, 
so that ¢urvature could not account for the difference in the forma- 
tion of the root hairs. 
TABLE VII 
LENGTH OF HAIR (IN MM.) 
PLANT pg nego 
In air In solution 
Pelianthus.3..50.... 8 8 1 (o. 25M sucrose), 
wg Pi eee fee a ae 8 5 1(0.200oM 
Be ea 5 5 pe Pe > San 
ee ie Oe a ee 5 5 t(o.osM * 
PAO i 5 8 2(0.10oM “ 
Discussion 
The measurement of the length of cell walls as an indication of 
differential growth seems of little value unless the number of the 
cells in any region is taken into account. Since there is not an 
epidermal cell corresponding to each individual cortical cell, it 
would seem that the sum total of the epidermal cells should be 
known in order to determine differential elongation as a factor. 
So far as known, the measurements taken were of individual 
adjoining cortical and epidermal cells. 
BARDELL’s own measurements show that for corresponding 
epidermal and cortical cells there is no definite relation in length 
which would cause the epidermal cell to produce a root hair, as 
neither do the measurements found here on corn. 
The factors which determine the initial swelling of the outer 
wall of the epidermal cell are of primary importance, while other 
factors come in which determine the later growth into a root hair. 
It has been proved in root hairs that there are portions of wall 1 
less resistant than others. The bursting of the root hair, the swell- 
ing, the branching, and the varying thickness of the membrane, 
all give evidence of this. The indications are that such weaker 
places exist in the membrane at the time of root hair formation, 
although this has not yet been proved definitely. These weaker 
regions would determine the region of the wall in which the locali- 
