23«S LECTURE XIV. 



several days' filtration now shows that the long wood-cells, which were opened by the 

 section, are completely filled with the fine cinnabar particles ; and that these have 

 penetrated even into the cavities of the bordered pits and completely filled them. 

 It may here be definitely observed that the very minute cinnabar particles have pene- 

 trated even through the narrow pore of the pit into its cavity, but are there held fast, 

 since a thin membrane prevents their passage over into the neighbouring cells of the 

 wood. Only the cut cells of the upper end of the piece of wood have taken up 

 particles of cinnabar, not the uninjured ones. A similar result is obtained if, instead 

 of cinnabar emulsion, mercury is poured into the apparatus described, and left in it 

 for some time. The mercury does not pass through the wood of Conifers ; but it 

 fills the cells opened at the upper transverse section and the cavities of their bor- 

 dered pits, without passing over into the neighbouring cells. 



After thus establishing that the wood-cells of the Coniferse are not in open 

 comrhunication (which is likewise true of the wood- cells and vessels of ordinary 

 foliage trees), the theory which has existed for two hundred years, according to which 

 the water ascending in the wood is considered to move as in capillary tubes, falls to 

 the ground of its own accord. The hypothesis established by Quincke^, that the 

 ascending water may be drawn up as an extremely thin molecular layer on the 

 surfaces of the walls of these capillaries, is likewise definitely refuted by means 

 of our discovery. Both theories, however, would be wrecked without that, on 

 the fact that even if continuous capillaries were present in the wood, they are 

 closed in the roots below and in the leaves above ; and that, further, the transverse 

 section of these capillaries — i. e. the wood-cells and vessels respectively — is far too 

 large to explain, according to the known laws of capillary tubes, an ascent of the 

 water in the wood to a height of more than a few metres. 



Now, however, there comes in another quite decisive fact, by which the capillary 

 theory of the ascent of the sap is likewise put aside : the fact, namely, that at the 

 time when transpiration is going on in the leaves — that is, at the time when a rapid 

 current of water is ascending in the wood — the cavities of the w^ood-cells and vessels 

 of foliage-trees are not filled with water, as should be the case according to the 

 capillary theory ; rather, the wood-cells and vessels are only to a small extent filled 

 with water, the vessels, however, being quite empty. This highly important fact 

 is already to be concluded from the simple observation that fresh wood, cut out 

 of the stem or branch of a transpiring tree in summer, and then thrown into 

 water, floats ; and since the cell-walls of the wood are without doubt specifically 

 heavier than water, if their cavities v/ere at the same time filled with water, 

 such a piece of wood must sink like a stone. Instead of this, however, it 

 floats, i. e. it is lighter than water ; and this fact is not otherwise explicable than 

 by the assumption that cavities exist in the wood which are not filled with water. 



' The possibility that the ascending water may pass up to the leaves as an extremely thin layer 

 on the inner side of the wood cell-walls, was not suggested by me, but by the physicist Quincke ; 

 though it is true it was made public by me in my text-book. This view was of course only tenable 

 so long as the bordered pits of the wood were supposed to be open, according to the older statements 

 of Schacht. Since the proof that the pits of the wood are closed was afforded, however, by Sanio's 

 and De Bary's anatomical investigations, and by my repetition of the elder Hartig's filtration experi- 

 ments, Quincke's theory can, of course, no longer be spoken of. 



