i66 BOTANICAL GAZETTE [September 



as a fact that in wood there is no general upward movement of 

 free air. Therefore, if it is believed that the behavior of this 

 tube depends on the movement of the bubbles, the experiment 

 is not appropriate to the problem. I shall show that the cohe- 

 sion explanation is open to exactly the same objections in nature 

 which disqualify it in this tube. This leaves a true explanation 

 to be sought outside of the cohesion of the water, or the rise of 

 the bubbles ; by analogy this is probably true in my experiment. 

 The missing link in the problem is some play of surface ten- 

 sion. Assuming for the moment that the suction is transmitted 

 largely by the air and that the bubbles are carried upward, there 

 is still present continuous water in the plaster, and the upper 

 end of this column would then be subject to the pull of its own 

 12*" of vjdiier, plus the suction at the bottom. Its surface tension 

 — which I would rather call capillarity than cohesion here 

 might enable it to endure this strain. As a matter of fact, 

 though, the tension of the water in the plaster and that of the 

 air and the water around it at the same height must be practi- 

 cally the same. And since the surface tension opposes the 

 movement of the bubbles, but not that of the water, it is natural 

 that the latter should move when either must. 



Any further discussion here would be duplicated in sifting 

 the theories on the ascent of sap in nature. The positive result 

 of our experiment is that the water column being continuous, 

 but air being present, a suction of less than one atmosphere can 

 still operate as a suction more than 12"^ low^er down. 



There is an immense mass of published experimentation on 

 the rise of the transpiration stream, whose results are not doubted. 

 During the last four years I have repeated a large part of it for 

 the sake of first-hand familiarity, I shall not add to the litera- 

 ture by republishing work already unquestioned. 



IL Dutrochet, in the work quoted in the introduction, says 

 that two forces cooperate to raise the sap : one of repulsion, 

 osmotic in its nature, in the roots ; and one of attraction, which 

 is also osmosis — Dutrochet calls it endosmosis — in the leaves. 



The driving force in the ascent of sap is root pressure. Seve- 



