46 BOTANICAL GAZETTE | JANUARY 
were separated into two lots, one for the control of the other. 
The average time was about six hours after branding, and the 
extreme limits were 3.5 and 11.5 hours after branding. In the 
control experiments the roots were taken from water and the | 
outside carefully wiped with a dry cloth to get rid of the water 
adhering to them, and just enough of the tip was cut off to take 
in all the curve. These tips were put into a weighing flask, 
which was closed air tight. In the other experiments the roots 
were split into two parts as nearly equal as possible, one half 
being the concave, the other the convex side of the curve. The 
two halves were cut off by acut transverse to the root at the point 
of cutting, and just enough taken to include the whole curve. 
The convex halves were put into one flask, the concave halves 
in another. The flasks had all been previously weighed, now 
the weight of the roots was ascertained. They were then dried 
quickly at a temperature of 100” C. until they ceased to lose 
weight, and were weighed again. The loss in weight, which 
equals the amount of water, was calculated on the basis of the 
fresh weight. 
A total of 143 roots were split and seventy-three were taken 
without splitting. The control and the split roots were weighed 
in fourteen different lots of each. With the exception of one lot 
they all showed a greater percentage of water in the convex 
halves than in the concave halves. That one exception was 
the one in which the time was the shortest, and the excess of 
water in the concave side was only 0.07 per cent. In the other 
lots the extreme limits of variation were 0.08 per cent. and 1.9 
per cent. in favor of the convex side. The total amount of water 
varied from 87.42 per cent. to 91.93 per cent. in the convex 
side and from 87.49 per cent. to 90.86 per cent. in the concave 
side. In the whole tips the variation was from 85.96 per cent. 
to 91.15 per cent. One third of the lots of the latter showed 4 
greater amount of water than the corresponding split roots. 
One third were between the convex and concave sides of the 
corresponding split ones, and one third fell below the concave 
side in the corresponding lots. 
