on the Descent of the Sap in Trees. 283 
duced has been very extraordinary. More new wood has been 
generated in the old trunks, within the last two years, than in 
the preceding twenty years ; and I attribute this to the facility 
of communication which has been restored between the leaves 
and the roots, through the inner bark. I have had frequent 
occasion to observe, that wherever the bark has been most re- 
duced, the greatest quantity of wood has been deposited. 
Other causes of the descent of the sap towards the root, I 
have supposed to be, capillary attraction, and something in the 
conformation of the vessels of the bark. The alburnum also 
appears* in my former experiments, to expand and contract 
very freely, under changes of temperature and of moisture; 
and the motion thus produced must be in some degree commu- 
nicated to the bark, should the latter substance be in itself 
wholly inactive. I however consider gravitation as the most ex- 
tensive and active cause of motion, in the descending fluids of 
trees ; and I believe, that from this agent, vegetable bodies, like 
unorganized matter, generally derive, in a greater or less degree, 
the forms they assume; and probably it is necessary to the 
existence of trees that it should be so. For, if the sap passed 
and returned as freely in the horizontal and pendent, as in the 
perpendicular branch, the growth of each would be equally rapid, 
or nearly so : the horizontal branch would then soon extend too 
far from its point of suspension at the trunk of the tree, and 
thence must inevitably perish, by the compound ratio in which 
the powers of destruction, compared with those of preservation, 
would increase. 
The principal office of the horizontal branch, in the greatest 
number of trees, is to nourish and support the blossoms, and 
the fruit or seed ; and, as these give back little or nothing to the 
