FUNCTION.] ORIGIN OF WOOD. 193 



water of a steady but rapid current would be diverted from 

 its course by obstacles in its stream. Again, in Guaiacum 

 wood, the descending tubes of pleurenchym cross and inter- 

 lace each other, in a manner that is unintelligible upon the 

 supposition of wood being formed by the mere deposit of 

 secreted matter. If the new wood were a mere deposit of the 

 bark, the latter, as it is applied to every part of the old wood, 

 would deposit the new wood equally over the whole surface 

 of the latter, and the deviation of the fibres from obstacles in 

 their doAvnward course would scarcely occur. This, therefore, 

 in my mind, places the question as to the origin of the wood 

 beyond all further doubt. Or, if further evidence were 

 required, it would be furnished by a case adduced by Achille 

 Richard, who states that he saw, in the possession of Du 

 Petit Thouars, a branch of Robinia Pseudacacia on which 

 R. hispida had been grafted. The stock had died ; but the 

 scion had continued to grow, and had emitted from its base a 

 sort of plaster formed of very distinct fibres, which surrounded 

 the extremity of the stock to some distance, forming a kind 

 of sheath ; and thus demonstrating incontestably that wood 

 does descend from the base of the scion to overlay the stock. 

 The singular mode of growth in Pandanus is equally 

 instructive. In that plant, the stem, next the ground, is 

 extremely slender ; a little higher up it is thicker, and emits 

 aerial roots, which seek the soil and act as stays upon the 

 centre, As the stem increases in height, it also increases 

 notably in diameter, continuing to throw out aerial roots. 

 If the roots were pruned away, the stem would be an inverted 

 cone ; but, if we add to the actual thickness of the base of 

 the stem the capacity of the aerial roots at that part, the two 

 together will be about equal to the capacity of the stem at the 

 apex ; which suggests the idea that the woody matter that 

 descends from the leaves may really be their roots, passing 

 through the horizontal cellular system of the stem. An 

 analogous but much more remarkable case is the following 

 mentioned by me in the Penny Cyclop&dia, article Endogens, 

 vol. ix. p. 396. In an unpublished species of Barbacenia 

 from Rio Janeiro, allied to B. purpurea, the stems appear 

 externally like those of any other rough-barked plant, only 



VOL. II. O 



