92 PROCESS OF NUTRITION. CHAP. III. 



a faded and sickly appearance. Some of them were 

 then removed to damp situations, and others to dry 

 situations, to know what the effect of such removal 

 might be. The former gave indications of recovery, 

 the latter of continued decay.* It is plain, there- 

 fore, that in the former case moisture must have been 

 absorbed from the atmosphere by means of the 

 epidermis of the leaf, or at least of the branch. 

 Mariotte cut off from a tree a branch terminating in 

 two boughs, which he suspended upon the edge of a 

 vessel filled with water, so as that the one was 

 within and the other without the vessel. The 

 former preserved its verdure for several days, but the 

 latter began almost immediately to wither. 

 Experi- But the most complete set of experiments upon 

 Bonnet, the absorbent power of leaves which has hitherto 

 appeared is that of M. Bonnet, of Geneva. Satisfied 

 that leaves are furnished with absorbent organs for 

 the purpose of the intro-susception of moisture, 

 as deducible from the experiments of Hales and 

 Guettard, his object was that of ascertaining whether 

 the absorbent power of both surfaces was alike. 

 With this view he filled several vessels with water, 

 on the top of which he placed a number of leaves, 

 some having the upper, and others the under surface 

 applied to the water, so as that they only floated in 

 it but were not immersed. If the leaf retained its 

 verdure longest with its upper surface applied to the 

 water, the absorbing power of the upper surface was 

 * Phy. des Arb. liv. ii. chop. iii. 



