1902] RISE OF THE TRANSPIRATION STREAM 183 



that in tubes of such size as these lumina the lifting power of 

 capillarity is not exceedingly great. In a glass tube io/a in 

 diameter this should be about 3"". Strasburger (1891:875^7.) 

 found it less in tracheae than in glass, sometimes less than half 

 as much. While the lifting power is determined by the form 

 of the meniscus, it is possible enough that this should be influ- 

 enced by the nature of the wall. In the absence of direct con- 

 trary evidence on the same subject, which is wanting in 



Schwendener's criticism (1892:912), Strasburger's statement 

 must stand. 



The efficacy of capillarity is limited again by the dilution of 

 the air in the tracheae. Naegeli's classical work on capillarity 

 (1866, 1, 11) showed experimentally that with diminishing atmos- 

 pheric pressure the lifting power of a meniscus in a glass tube 

 decreases; slowly at first, but rapidly when the pressure 

 approaches zero. The theoretical ground for this is evident 

 from the intimate connection between surface tension and vapor 

 tension. Naegeli found the details to be remarkably dependent 

 upon the temperature. Since the air in the tracheae is always 

 rarefied when transpiration is at all active, and with excessive 

 transpiration the loss of pressure is great enough sometimes to 

 lead to the formation of bubbles of water vapor, the service 

 which capillarity can perform in the elevation of water is most 

 insignificant when the demand for water is most urgent. A 

 repeated, but so far as I can see entirely unwarrantable assump- 

 tion (Bohm, Ann. Sci. Nat. 1878:232; Elfving, 1885:536; R. 

 Hartig, Bot. Zeit. 1883; Sanio) has been that each wall across 

 the path of the transpiration stream furnished a fresh starting 

 point for capillarity to work from. There is no reason w^hy a 

 wall should offer any greater resistance to movement of water 

 downward in response to gravity than upward in response to 

 capillarity. 



The Jamin's chain has never been shown to have any advan- 

 tage over a single meniscus in causing a movement of water. 

 For every added meniscus with an upward pull there is one of 

 the same shape whose concavity is dow^nward. The resistance 



