337 
the Ascent of the Sap in Trees. 
leaf, where I believe they terminate. Plate XXVII. presents 
two sections of the leaf-stalk of the horse-chesnut ; the first be- 
ing taken from the middle of the stalk ; and the second from its 
base. Lying parallel with, and surrounding the abovementioned 
vessels, appear other vessels, which I conclude return the sap 
to the tree ; for, when a leaf was cut off which had imbibed a 
coloured infusion, I found that the native juices of the plant 
flowed from these vessels, apparently unaltered, as has been 
remarked by Dr. Darwin. These* vessels descend through the 
inner bark, (as delineated iq. Plates XXIV. XXV. and XXVI.) 
and appear to extend from the extremities of the leaves to the 
points of the roots. 
The whole of the fluid, which passes from the wood to the 
leaf, seems to me evidently to be conveyed through a single 
kind of vessel ; for the spiral tubes will neither carry coloured 
infusions, nor in the smallest degree retard the withering of the 
leaf, when the central vessels are divided. But the annexed 
figures appear to point out at least two kinds of returning ves- 
sels. And I think it by no means improbable that two kinds 
exist, with distinct offices; for there is a new layer of alburnum, 
and a new internal bark, to be formed. I have, however, seen 
it asserted somewhere, in the writings of Linn^us, and other 
naturalists, that the internal bark is annually converted into 
alburnum. But this is totally erroneous ; and a vigorous shoot 
of the apple-tree often presents in its transverse sections, when 
three or four years old, as many layers in its bark, each of 
which once formed its internal vascular lining. 
As the bark appeared to me now to receive its nutrition 
through the leaf, I wished to see what effect would be produced 
by gradually reducing the quantity of the leaves. I had a luxu- 
Xx 
MDCCCI. 
