SECT. II. COMPOSITE ORGANS. 221 



by means of passing through the bark of a tree 

 several small threads of silver in a horizontal di- 

 rection, so as to penetrate the liber.* If the liber was 

 converted into wood, the threads, it was to be pre- 

 sumed, would be found ultimately imbedded in 

 the wood ; and if it was not converted into wood, 

 they would be found still in the bark. Accordingly 

 when a trunk which had been so treated was at the 

 end of several years opened up and inspected, the 

 threads were found to be deeply imbedded in wood : 

 it is plain, therefore, that the new layer of wood 

 forms originally a layer of liber 9 according to the 

 common acceptation of the term. But to try also 

 the value of Grew's conjecture with regard to the 

 separation of the liber into two parts, the one ex- 

 panding towards the circumference and forming 

 new bark, and the other condensing towards the 

 centre, and forming new wood, Du Hamel varied 

 the above experiment so as that some of the threads 

 were passed through the outer part of the bark, near 

 the epidermis; others through the inner part of it, 

 near the liber; others through the liber itself; and 

 others between the liber and wood. At the end of 

 several years, when the trunk was opened up and 

 inspected, the threads that were originally placed in 

 the outer bark and near the epidermis were now 

 covered only with a thin and decayed crust, which 

 broke readily _in to pieces ; those that were originally 

 placed in the outer bark, but near the liber, were 

 * Phys. des Arbres, liv. iv. chap. iii. 



7 



