io8 TIMBER AND SOME OF ITS DISEASES, [chap. 



Capillarity and the friction of the air bubbles must 

 be important, as supporting the columns of water. 



Imbibition (in Sachs' sense) will be useful when the 

 lumina of the elements are nearlydcprivedof water in 

 summer. 



Osmosis is active in giving a push from behind 

 especially at certain seasons. 



The differences of air-pressure as expressed by the 

 contractions and expansions of the bubbles, must have 

 effect in moving small quantities of water from lumen 

 to lumen. 



Elfving then proceeds to show that the "gas- 

 pressure theory" ofHartig is but a development of the 

 air-pressure theory of Boehm, though the only point 

 which remains the same in both is the assumption that 

 the water moves in the lumina. 



Water is always present in the tracheidcs accord- 

 ing to Hartig, Boehm, Russow and Elfving ; though 

 of course this does not itself prove that this water 

 is any other than a reserve supply, as Dufour alleges 

 it to be. 



The pressure of the atmosphere could have nothing 

 to do with the matter, because the membranes are 

 impervious to air: if it could, the hypothesis as it 

 stands would be absurd ; for a continuous column 

 could only be ten meters high. There is no com- 



