on the Motion of the Sap in Trees. 185 
the buds, and numerous roots, of considerable length, were 
emitted, whilst no sensible growth took place between the base 
of the young shoots and the mould in the pots. 
It having been proved by Du Hamel, that inverted parts of 
trees readily emit roots, I expected to derive further information 
from cuttings of this kind : I therefore planted, in the autumn 
of 1802, forty cuttings of the gooseberry-tree, and an equal 
number of the common currant-tree; one half of each being 
inverted. Of the former, not one of the inverted cuttings suc- 
ceeded ; whereas few of the latter failed ; and in these I had an 
opportunity of observing the same accumulation of wood above 
the bases of the annual shoots, and the same mode of growth, 
in every respect, as in the inverted vines ; except that no roots 
were emitted at their upper ends. The same thing occurred, 
without any variation, in inverted grafts of the apple-tree. 
If it be admitted, according to the theory I have on a former 
occasion laid before you, that the sap descends from the leaves 
through the vessels of the bark, and that such vessels are, in 
their organization, better calculated to carry their contents to- 
wards the original roots than in the opposite direction, it will be 
extremely easy to explain the cause of the accumulation of 
wood, and the emission of roots, above, instead of below, the 
base of the annual shoots. The vessels of the bark (the vaisseaux 
propres of Du Hamel) commencing in the leaves, were formerly 
traced by M. Mariotte, and subsequently by myself, (being 
ignorant of his discovery,) to the extremities of the roots; and, 
when a cutting, or tree, is planted in its natural position, the sap 
passes downwards through these, to afford matter for new 
roots, and to increase the bulk of those already formed, having 
given proper nutriment to the branches and trunk in its descent. 
mdccciv. B b 
