378 Knight. — Further Observations on the Transpiration , 
balance in the plant must be an extremely delicate one. As water is lost 
by transpiration it must be replaced by translocation of water from the 
petiole or the stem, and this simultaneous progression may even extend to 
root absorption, thus permitting the maintenance of almost perfect equilibrium. 
Otherwise even a small change of transpiration rate induced by slight altera- 
tion of the environmental conditions would immediately result in a change 
in the water-content ; a small increase in the wind velocity, for example, 
would produce visible wilting in a very short time. The leaves of a few 
plants, e. g. Helianthus , the large-leaved species of Saxifraga , and the root 
crops, certainly do lose some of their turgidity on very hot days, but it is the 
exception rather than the rule to find the ordinary mesophytes wilted even 
on the days of highest atmospheric evaporating power. In the absence, 
therefore, of any regular occurrence of temporary wilting, it must be con- 
cluded that changes of transpiration rate are accompanied by equal 
compensating changes of absorption rate. Such close correlation between 
these two processes is more easily conceivable if part at least of the flow of 
water through the plant is maintained by the pull, upon continuous tensile 
water columns, resulting from evaporation and capillarity, than if the eleva- 
tion of water is effected by directive c vital ’ action of leaves or stem. (See 
Dixon (io), pp. 24, 25.) 
Another aspect of the question of change of water-content is worthy of 
consideration. It appears paradoxical that whilst a change of 1 per cent, in 
leaf water-content may produce such profound results, yet two leaves on 
the same shoot, apparently quite similar as regards turgidity, may differ in 
water-content by as much as 4 per cent. This apparent discrepancy may 
be the result of structural differences — for example, difference of thickness 
of cell wall. A leaf is a composite structure made up of many different 
types of tissue, doubtless differing in the percentage of water which they 
contain. It is evident that a difference in the proportion of the various 
tissues in two leaves would result in a different percentage water-content of 
the leaf as a whole. 
The variation of leaf water-content during the day and night is rather 
a different problem and needs special consideration in relation to the results 
already recorded by other workers. Clark (5) has investigated changes of 
water-content of leaves of trees, and found that in addition to a seasonal 
change from a maximum in spring to a minimum at leaf fall there also 
occurred considerable variations through the day. It was nevertheless found 
impossible to determine the time of day at which the maximum or minimum 
water-content was reached, on account of the variation in the results. 
The seasonal change of water-content is attributed by Clark to seasonal 
structural changes similar to those postulated above, but no explanation is 
advanced to account for the large diurnal changes which, were recorded. 
The difference in leaf water-content determined at different times of the 
