No. 474] 



SAP FLOW IN MAPLE 



433 



walls, and the slowness of the return of water to its original location 

 in the walls. 



The objections to this theory are very serious. Firstly, the 

 forces of imbibition are sufficiently strong ordinarily to insure a 

 comparatively rapid return of water to the cell walls. The return 

 of water to the walls in thawing winter buds I have observed to 

 take place almost immediately. Flow and pressure throughout 

 the entire day could not be accounted for in this way. Secondly, 

 ice does not form in wood at the temperatures obtaining during 

 the season of flow. A tree temperature of -3° C. was the lowest 

 obtained by the ^^ermont Station during this period, while on many 

 good sap days the tree temperature at night was only -0.5° or 

 -1° C. The overcooling point, not the true freezing point, is of 

 importance in determining whether ice will be formed, and this 

 is always several degrees lower than the freezing point in all 

 plant tissue. Both are lower than the freezing point of pure water. 

 Miiller-Thurgau * found these temperatures for various sorts of 

 wood as follows: — 



)vercool. pt. 



One year shoot of pear -2. IS 



Old wood of grape -G . 05 



The overcooling point seems to be greater in the more dense 

 woods than in the grape probably because of the large vessels and 

 watery' sap offering little resistance to the inception of ice for- 

 mation. ]Maple wood would be of the closer grained type. Ice 

 formation would probably not commence before a temperature of 

 -4° to -7° C. was reached. Dixon and Joly ^ found that ice began 



• Miiller-Thurgau. "Ueber das Gefriereu und Erfrieren der Pflanzen." 

 Landw. Jahrb., vol. 15, p. 492, 1886. 



^ Dixon and Joly. "The Path of the Transpiration Current." Ann. Bot., 

 vol. 9, p. 416, 1895. 



