20 



root towards the head. I found by 

 experiment, in 1832, that the roots of 

 birch will bleed either upwards or down- 

 wards ; and if a root of a birch is cut 

 across near the stem, so that it cannot 

 absorb moisture itself, it will bleed freely 

 from the stem. 



These facts are at least sufficient to 

 prove that the woody parts of trees im- 

 bibe without the aid of the small fibres or 

 ends of roots. From the following experi- 

 ments, also, I am satisfied that roots do not 

 imbibe at all from the small mouths (capil- 

 lary stomata), or from the small sponges 

 (spongioles), which are supposed to exist at 

 their ends. 



In March 1836 I made some horse- 

 chesnut seeds grow in water, in my barrack 

 room, and found that when the root was cut 

 off when it was several inches long the plant 

 would still grow, and would continue to 

 throw out fresh roots as fast as they were 

 cut off ; but if the plant was divided from 



