1S85.] On the transpiration-stream in cat branches. 367 



Avater of imbibition. Thus Dufour says : "Als Grundsatz gilt : 

 das in Zellwandern imbibirte Wasser ist durch gewohnliche Druck- 

 krafte iiberhaupt nicht versckiebbar." 



It seems to us that the above results of experiments on bending 

 and squeezing are only explicable ou the theory that the chief 

 stream of transpiration is in the cavities and not in the cell-walls. 



We cannot understand the great difficulty found by Dufour in 

 forcing water under pressure through bent twigs. In the only 

 experiments which we made on this point the difficulty was not 

 found to be so great. 



A bramble 213 cm. in length, was cut under water, and was 

 fitted to the potometer by its basal end, while the apical end was 

 attached to the water air-pump. It was subjected to a suck of 

 60 to 65 cm. of mercury and the readings of the potometer taken 

 at various pressures. Then the branch was bent in two places, the 

 bends being tied as closely as possible. The result was that with a 

 pressure of 60 cm. mercury the potometer rate for the unbent 

 branch was 45 - 4 ; for the bent branch 19"0 ; or as 



Straight 100 

 Bent 41-9 



In another similar experiment the result was greater. The 

 branch was 76 cm. in length : the pressure varied between 29*5 

 and 30'5 cm. of mercury. 



Rate 

 Before bending 60*9 or ]00 

 Doubly bent 13-6 22*3 

 Straightened 61-3 100-7 



These experiments show the similarity that exists between the 

 transpiration and filtration currents, and they can only be explained 

 on the supposition that the water of transpiration travels in the 

 cavities of the wood-elements. 



