i\ J VARIOUS THEORIES, &c 135 



an air-bubble as indicator, the length of time which 

 it took to traverse a given length of tube measured 

 the rate of flow of the column of water attached to 

 the cut branch. These results are quite independent 

 of Kohl's, and nevertheless bear out the same con- 

 clusions, but much more exactly and in detail. The 

 authors show that Dufour (i) was wrong in his 

 estimate of the great obstruction to transpiration pro- 

 duced by the double-sawing, and (2) exaggerated 

 the difficulty of forcing water through doubly-sawn 

 branches. 



A single cut produces far less diminution of the rate 

 of flow, than the double cuts. Moreover, the obstruc- 

 tion is greater at first than later on — a recovery of the 

 rate of flow occurs to some extent as the absorbing 

 power in the leaves makes itself felt more and more, 



I must refer the reader to the original for further 

 details, merely pointing out that the results are 

 distinctly in favour of the theory that the water passes 

 through the cavities of the vessels and tracheides, and 

 they are the more valuable because low pressures and 

 actual transpiration were employed. 



An interesting test of the validity of Godlewski's 

 theory was devised by Janse,^ who set himself to ask — 



^ *' En experimenteel bewys voor de theone van Godlewski omtreut 

 de bewegung van het water m de planten," Maand, hlad voor Natum- 



