65 



because the largest branches are on the 

 leeward side; yet if the upper part of 

 one half of the side of a stem is dead^ the 

 opposite living side will deposit round 

 the whole living part below. And I 

 imagine that it is thus that the windward 

 roots of an exposed tree are nourished 

 by the descending sap from its leeward 

 branches. 



As far as I have remarked, though 

 young roots are round, the older ones 

 greatly incline to the oval shape ; and 

 in all the transverse sections of roots 

 which I have examined the eccentricity 

 of the common point from which the 

 {medullary?) rays diverge, and which is 

 occasioned by the comparatively over- 

 growth of the upper sides of the new 

 annual rings, is very striking. May 

 this be from the descending sap having 

 a greater facility of depositing on the 

 upper sides, or from the roots meeting 

 with a less mechanical resistance from 



E 



