1 88 BOTANICAL GAZETTE [September: 



because the air in the tracheae expands until finally it closes then^ 

 to the passage of water. Removing a part of the air makes a 

 lower tension necessary for what is left to plug the tracheae, and 

 so permits more than the normally possible suction. I have 

 illustrated this with a small branch of a greenhouse grown Euca- 

 lyptus, the cortex removed, inverted so that when any bubbles 

 were drawn from the stem they could not stay in contact with it 

 and prevent the absorption of more water. It was in air-tight 

 connection with a tube i"""" in diameter, full of water and stand- 

 ing in mercury. The mercury rose in one hour and fifteen 

 minutes loi'^"', where it was stationary. This suction was- 

 increased to 19 1"^'", but drew air from the pith and wood, and 

 fell slowly to 106^^. A little air continued to be extracted, 

 but the suction increased, until 52 hours after the experiment 

 began it was 173""°* ; and after 79 hours 235°^"^. The rise of mer- 

 cury had become very slow, and at this point a leak stopped the 

 experiment. The young leaves, up to those more than half 

 grown, were already becoming dry, though the cut end was, and 

 had constantly been, in water. It is because the possible suction 

 is chiefly a function of the air in the tracheae that Vines (1896: 

 538) finds that a large part of the leaves may be removed from 

 a branch without much effect on the suction. For the same 

 reason — because a part of the tracheae are closed — a suction of 

 less than the greatest the transpiration stream can exert may 

 nearly stop the absorption of water (Vesque, 1884; Strasburger^ 

 1891 : 788). It seems clear to me that the evidence we have 

 warrants the conclusion that the tensions in the tracheae of liv- 

 ing plants^ — at least of dicotyledons — -in nature are never even 

 approximately as low as zero.^ 



What happens if the tension is artificially lowered? Bohm 

 (1864:546) put more than a hundred Salix cuttings in a vessel 

 where the pressure against the absorbing ends was reduced to 8 

 to lo"'"' Hg. The result was *'Dass die Blatter nach zwei oder 

 hochstens drei Tagen vertrocknet waren. Das Absterben 



*This was treated as an established fact by Naegeli and Schwendener (1877 ► 

 380). 



4- 

 5 



\ 



