Osmotic Pressure , Root Pressure 5 and Exudation 111 
to assume such a gradient between the root-hair and the wood- 
elements. It has been pointed out in this Journal by Thoday 1 (1918), 
in an admirably clear and useful paper, that such a gradient is in 
no way necessary; all that is required in the chain of cells is a gradient 
of absorbing power with the highest absorbing power in the secreting 
cell; this might easily be consistent with a reversed osmotic gradient. 
Furthermore it may be pointed out that if the water supply is ample 
and there are no intermediate losses of water such a gradient will 
necessarily arise whatever the distribution of osmotic pressures. 
If in fig. 2 we imagine that the root-hair, A , exposed to water has 
the highest osmotic pressure and the cells B to E, and the xylem 
vessel F, progressively lower pressures, water will still pass in from 
Fig. 2 
A to F. For as A takes up water its absorbing power will tend to 
fall below that of B, when B will immediately take water from it, 
the absorbing power of B will then tend to fall below the osmotic 
pressure of C, and C will immediately take water from B , and so 
on; thus a gradient of absorption will arise. As the cells get turgid 
and cease to have any absorbing power, F will be able to draw water 
from outside A , however low its osmotic pressure, and the force with 
which water is drawn in will depend solely on the difference between 
the osmotic pressure of the contents of F and the osmotic pressure 
of the solution external to A. The osmotic pressure of the other cells 
can be neglected, though the resistance to water passage resulting 
from the interposition of these cells will reduce the rate at which 
water will pass from A to F. 
The existence of an osmotic solution in the vessel F could be 
explained, as in Pilobolus, as a result of diffusion from F through the 
partially permeable membrane between the two. The accumulation 
of fluid under pressure in F would depend, as already stated, solely 
on the osmotic pressure of the solution in the vessel, and would 
1 Hofler (1920) has recently dealt with the water relations of the cell in 
a manner very similar to that of Thoday but somewhat more fully; he makes 
however no reference to Thoday’s paper. 
