THE MECHANISM OF WATER-TRANSPORT 223 



pass upwards by a forward movement of the entire chain of air and water, 

 or the chain remaining stationary, individual particles of water may wander 

 round the air-bubbles from water-column to water-column 1 . A trifling 

 difference of pressure does indeed suffice to cause a movement of the entire 

 chain ~, but nevertheless the main current of water cannot be produced in 

 this manner, for since air diosmoses only slowly through the moist tracheal 

 walls (Sect. 32), the bubbles would collect at the upper ends of the tracheae and 

 tracheides and thus interpose a marked resistance to the passage of water, 

 while whenever the tracheae and tracheides are only of moderate length, as 

 is often the case, this resistance will be interposed at very many points. 



As a matter of fact various observations show that the air-bubbles, 

 although they may vary in size, remain at rest even when a current of 

 water may be flowing past them 2 . It is, however, impossible as yet to say 

 whether the water passes from column to column through the tracheal walls 

 or through the film of water (adhesion-water) which surrounds each air- 

 bubble. We do not even know what thickness this film of water attains, 

 or what resistance it interposes to any movement in mass. A full discus- 

 sion of these and related questions is impossible here 3 , but it may be 

 mentioned that, under the given conditions, the thickness of the film may 

 be such that the greater portion is beyond the influence of the energetic 

 attraction exerted by the surface molecules of the organized membrane 

 which it touches. Moreover, one of the surprising results due to surface 

 tension is that in a very thin film of fluid a rapid movement may be 

 produced by a corresponding difference of potential. 



As the distance between the water-columns increases, the transit of 

 water will apparently become more and more difficult, and hence withering 

 commences when the amount of free water present in the wood (Sect. 33) 

 sinks below a certain limit. It is by no means immaterial whether the 

 water passes the same length of air as short bubbles with the water- 

 columns close together, or as long bubbles with them far apart, but further 

 research is still needed in this direction. In the wood of transpiring trees 

 fully supplied with water, Schwendgner 4 found that the water-columns were 

 from o-i to 05 mm. long, whereas the interposed bubbles of rarified air may 

 be as much as i mm. in length. 



For purposes of simplicity we have confined ourselves to the con- 



1 That the transference across the air-bubbles could take place rapidly enough by evaporation 

 and condensation is impossible. Cf. Dixon and Joly, Ann. of Bot., 1895, Vol. IX, p. 419. 



2 Observations and literature by Strasburger, Bau u. Verricht. d. Leitungsbahnen, 1891, and 

 Uber das Saftsteigen, 1893, p. So; Schwendener, Sitzungsb. d. Beil. Akad., Bel. xxxiv, p. 591, 

 and 1892, Bd. XLIV, p. 920; Askenasy, Uber das Saftsteigen, 1895, p. 17 'Sep.-abdr. a. d. Ver- 

 handlungen d. Naturf.-Vereins in Heidelberg). 



3 [See Kamerling, Oberflachenspannung u. Cohesion, Bot. Centralbl., Bd. LXXIII, 1898, p. 470.] 

 * Schwendener, Sitzungsb. d. Berl. Akad., 1886, Bd. xxxiv, p. 567; 1893, Bd. XL, p. 842. 



