80 VEGETABLE PHYSIOLOGY 
cavities of the vessels, and subsequently solidify, these 
passages can be occluded for a distance of a few centi- 
metres. Gelatin or paraffin can be used for the experi- 
ment, being injected at a moderately low temperature such 
as will not injure the vitality of the tissue. If after it has 
solidified a fresh surface is made by a clean cut a very 
short distance from the end, and the branch immersed in 
water, the leaves very soon flag, even if some pressure 
is applied to the water in contact with the cut surface. If 
the path of the liquid were the cell-walls, no obstacle being 
offered to the transfer of water to them, the upper portions 
ought to remain turgid. The experiment shows that the 
normal channels are blocked by the paraffin or gelatin 
used, and flagging results. 
A similar demonstration that the water passes by the 
cavities or lumina of the cells is afforded by the experi- 
ment of compressing the stem in a vice; if the pressure 
is carried so far as partially or entirely to obliterate their 
cavities, the rate of flow is materially interfered with. 
The progress of a dye injected into the surface of a cut 
branch also points to the same conclusion. If such stains 
as fuchsin or eosin, which colour wood very rapidly, are 
forced up into a stem and sections made almost immedi- 
ately, the lignified walls will be found to be in process of 
staining, and the colour will be seen to be deepest on the 
side of the wall abutting on the lumen, often only penetrating 
partly through the thickness. If the wall itself were the 
path of the pigment solution, its thickness would be stained 
uniformly as far as the dye penetrated at all. 
The rate at which the transpiration current naturally 
flows varies a good deal, plants showing differences among 
themselves as to facilities of transport. In a fairly vigorous. 
tree it may be taken to be about 1-2 metres per hour, 
though in some plants it has been observed to be three 
times as rapid. In other cases as low a speed as ‘2 metre 
per hour has been found. It is a little difficult to measure 
in most cases; the plan generally adopted has been to 
