iv.] VARIOUS THEORIES, &c. 97 



elements have always some water in their cavities, 

 and (2) the water contained in the tracheides, &c., of 

 the splint increases as we ascend the tree, Dufour 

 simply seeks to show that this does not matter : it is 

 not urged that the membranes should be in contact 

 with much or little water, and in fact complete satura- 

 tion brings about the maximum mobility of the 

 imbibed water molecules. The water in the cavities 

 must be looked upon as so much reserve-water, upon 

 which the membranes can draw : how it comes there is 

 not explained, but it has nothing to do with the 

 imbibition-theory. 



As a demonstration of the accuracy of Sachs' 

 views, Dufour again brings forward experiments with 

 sharply bent shoots, showing that in the majority of 

 cases the bending of the vascular bundles on them- 

 selves closes the cavities, and water can no longer 

 be forced through by such pressures as Elfving and 

 others used : nevertheless, the plants did not cease to 

 transpire and conduct water. 



Dufour also employed the following method : 

 Branches were sawn into in such a way that at 

 two places on opposite sides of the stem, one being 

 a little higher than the other, the whole of the tissues 

 were severed as far as the pith: the argument was 

 that the continuity of all the vessels, tracheides, &c. 



H 



