1885.] On the transpiration-stream in cut branches. 337 



of which could be varied by raising one edge of the jar. When 

 the results calculated from the potometer and psychrometer read- 

 ings, under these circumstances, were represented graphically, it 

 was found that the two curves were strikingly similar. 



Precautions. 



Sachs* has pointed out that the amount of water absorbed by 

 cut branches suffers a gradual diminution so that long-continued 

 observations are not trustworthy. But since our results chiefly 

 depend on sudden changes in the rates of absorption, changes 

 which are made evident in a few minutes, we are not inclined 

 to believe that this diminution introduces any serious error into 

 our results. For similar reasons we have not thought it necessary 

 to give the readings of the psychrometer which were noted in 

 nearly all cases. 



Another phenomenon which occurs when a cut branch is placed 

 in the potometer must be noticed, since, unless regard is had to it 

 in experimenting, serious errors must arise. When a branch (e. g. 

 of Portugal laurel) is cut, and fitted into the potometer, the 

 readings are at first very high but rapidly sink, until after an hour 

 or so they reach approximate constancy. The following is an 

 example : 



Sept. 16, 1884. Portugal laurel {Primus lusitanica) cut under 

 water and at once fixed in the potometer : 



p.m. rate 



337.. ....263 



43... ...208 



50. .....167 



54... ...159 



413. .....118 



5-3 ... ! ... 87 



37 76 



41... ... 80 



During this experiment the air was becoming somewhat 

 damper, which would slightly increase the fall in the rate of 

 absorption. 



§ 3. On Du four's experiments. 



The rival theories of water transport, i. e. the imbibition and 

 the " intracavital " theory, have been tested by Dufour in a paper 

 published in Sach's Arbeiten, 1884. If Dufour is right, the intra- 



* Flora, 1856. 



