CH. IV] FILTRATION EXPERIMENTS. 89 



spirit is allowed to evaporate slowly, the wood (according to 

 Strasburger^) is in a more suitable condition for experi- 

 ment than if it is allowed to dry without being treated 

 with alcohol. We have not compared the two, but the 

 silver-fir wood dried after prolonged soaking in alcohol 

 certainly answers well. 



The three following observations must be made : 



(i) Take a turned cylinder of dry wood, or a piece of 

 Pinus sylvestris treated as above described (which answers 

 perfectly and is more easily prepared), and attach it by a 

 strong rubber tube to the short arm of a U tube. By 

 pouring mercury into the long arm it ^\i\\ be found easy 

 to force air through under a pressure of about 10 cm.; to 

 make this obvious paint the upper surface of the wood 

 with olive oil in which the escaping bubbles are visible. 

 The surface ought to foam all round the circumference of 

 the section. A chain of bubbles coming from a single 

 spot suggests a faulty bit of w^ood. 



(ii) A similar cylinder which has been thoroughly 

 soaked in water is fitted into the U tube: the wet cell- 

 membranes will be found to be extremely, though not 

 absolutely, impermeable to air. 



(iii) If the U tube is now filled with w^ater it will be 

 found that a very slight pressure of water forces water 

 through. 



(102) Oozing of water from the lower end of wood. 

 The experience gained in experiment 101 enables us 



^ LeitungsbaJnu'ii, p. 734. 



