554 Dr. A. J. Ewart .
[ Nov. 2 .
" The Ascent of Water in Trees .
" By Alfred J. Ewart , D.Sc , Ph. D. , F.L.S. , Lecturer on Botany in the University of Birmingham .
Communicated by Francis Darwin , For .
Sec. R.S. Received November 2 , \#151 ; Read December 1,1904 .
( Abstract .
) As the result of a series of experimental observations bearing upon this problem , the author has been led to the conclusions stated in brief below .
The flow of water through open vessels filled with sap takes place in accordance with Poiseuille 's formula for the flow through rigid cylindrical tubes , divergences being due [ to the presence of irregular internal thickenings in the vessels , and to local constrictions or deviations from the circular outline .
Hence the velocity of flow is directly proportional to the pressure and to the square of the radius of the tube , inversely proportional to the length of the tube and to the viscosity of the liquid .
A small number of large vessels , therefore , offer very much less resistance to flow than a large number of narrow ones having the same length , and the same total internal area of cross-section .
Since viscosity is largely dependent upon temperature , the latter forms an important factor in regulating the flow , the viscosity and the resistance falling with a rise of temperature .
With an average rate of flow the total resistance due to the viscosity of the water flowing through the vessels is always less , and in climbing plants with large vessels is considerably less , than a head of water equal in height to the stem .
The adult vessels of actively transpiring Angiospermous trees always contain air-bubbles , and these introduce a resistance to flow which is inversely proportional to the radius of the tube , when the air-bubbles and water-columns move together .
When the air-bubbles are comparatively stationary , as in most vessels , the resistance is still further increased , and it becomes very great when the vessels are small and the air-bubbles numerous .
In intact vessels containing air the rates of flow under similar pressures are proportional to a power of the radius lying between 1 and 2 , the volume passing to a power of the radius lying between 2 and 4 .
Estimations of the amount of flow , made from the rate of flow and the diameters and number of the vessels , showed that the actual ow takes place in the wood of Dicotyledons almost entirely through the cavities of the vessels and hardly at all through the tracheides , In young stems saturated with water under pressure , a considerable flow takes place through the pith , but practically none in intact transpiring stems .
1904 ] The Ascent of Water in Trees .
In a cut stem , apart from the blocking at the cut surfaces , a gradual diminution of conductivity occurs along its entire length after water has been passed through for some time .
This appears , in part at least , to be due to the development of micro-organisms in the vessels , but may be aided by swelling , by lessened permeability , or by other changes in their walls .
The length of the vessels in the wood of the branches examined averages from 7 to 36 centimetres , the tracheides of the yew being from 02 to 05 of a centimetre in length .
Since , however , the vessels appear mainly to end at the nodes where branches arise , it is possible that they may be much longer in the young wood on old bare trunks .
The resistance to transverse flow through saturated wood is 800 to 45,000 times greater than to longitudinal flow , the resistance to filtration under pressure through a single partition wall being from 2 to 10 times greater than that to the flow through the entire length of a vessel filled with water in the wood of a crab apple .
The total resistance to flow in the erect stems of actively transpiring plants appears to correspond to a head of water of from 6 to 33 ( shrubs and small trees ) , or from 5 to 7 ( large trees ) times the height of the plant .
- Hence in the tallest trees the total pressure required to maintain active transpiration may be equivalent to as much as 100 atmospheres .
No leaf could produce or maintain an osmotic suction of this intensity , and in the presence of large air-bubbles in the vessels the stress transmitted in them from the leaves could never be as great as an atmosphere .
Vines* found , for instance , that the suction force of a transpiring branch was never greater than two-thirds of an atmosphere .
The supposition that these forces might summate is entirely erroneous .
On the contrary , the leaves at the base of a-tree would pull water down from the upper vessels and leaves , instead of up from the roots , in the absence of any pumping action in the stem , and of any root-pressure .
If the air-bubbles in the vessels were exceedingly minute , they might be under a small positive pressure , while the water outside was under a maximal strain of five atmospheres .
This would suffice to overcome the resistance offered during active transpiration by 30 to 80 feet of stem , hence the results obtained by Strasburger with dead stems .
The maximal osmotic suction exercised by the leaves , as determined by comparing the osmotic pressures during active transpiration of the leaves at the top and bottom of an elm 18 metres high , appears to be from 2 to 3 atmospheres , and is usually less than this .
At the same time the total resistance to flow in the trunk of this tree would be from 10 to 12 atmospheres .
Tt appears , therefore , that to maintain flow , a pumping action of * 6 Annals of Botany , * 1896 , vol. 10 , p. 438 .
556 The Ascent of Water Trees .
some kind or other must he exercised in the wood , for which the presence of active living cells is essential .
In support of this it has been shown that the production of wood in a slowly growing tree is greater than is necessitated by mechanical requirements .
In other words , the production of new wood is largely determined by the length of time during which the wood-parenchyma can remain active .
There is no known means by which these cells can directly pump water in a definite direction , although the existence of a power of absorbing and exuding water under pressure has been empirically determined to exist in the living wood of cut branches .
It is suggested that the wood-parenchyma cells by the excretion and re-absorption of dissolved materials may bring into play surface-tension forces within the vessels of sufficient aggregate intensity to maintain a steady upward flow , and to keep the water of the Jamin 's chains in the vessels in a mobile condition ready to flow to wherever suction is exercised upon it.* The rapid rates of diffusion required for such action do actually exist in the wood-parenchyma cells .
It appears that the terminal branches of trees at heights of from 22 to 44 feet above ground exhibit little or no power of bleeding in spring .
Possibly in such trees the pumping action is only used or developed in the wood of the older stems , or is only exercised when transpiration is active , and when the water-columns in the vessels attain a definite size relatively to the wood-parenchyma cells .
The importance of the Jamin 's chain in the vessels is that it renders a staircase pumping action possible , and enables the water to be maintained in the vessels in a labile condition , ready to flow to any point where moderate suction is exercised .
This pumping action being diffused and probably regulated , need not produce any high pressure of exudation at the terminal branches of tall trees , which , in fact , appears always to be absent at high levels .
* Surface-tension actions would be possible in the absence of air-bubbles wherever the wood-parenchyma cells contained oil or any other substance non-niiscible with wrater , as they often do .

