[lynde-dupké; capillary LIFT OF SOILS 121 



poured into the funnels and the soil was settled in the centrifuge. 

 More soil and water was then added to the funnels and the soil settled 

 in the centrifuge. This was continued until the soil was well above 

 the edges of the cotton cloth filters. 



The capillary tubes, with the rubber tubes attached, were then 

 filled with water previously boiled to expel air and the funnels were 

 inserted into the rubber tubes. In doing this we took care to prevent 

 air from entering the funnels or tubes, also we carefullly avoided dis- 

 turbing the soil. The lower ends of the capillary tubes were then 

 placed in cups of mercury and the water seals were filled with bolied 

 distilled water. 



How the experiment works. — If, at the beginning, there is a layer 

 of water above the soil, it gradually percolates through the soil and 

 water comes out on the surface of the mercury in the mercury cup. 

 That is, percolation takes place. 



When the water disappears from the surface, however, the meni- 

 scuses in the soil pores begin to act. We picture this action as fol- 

 lows. Water molecules leave the surface of the meniscuses by evap- 

 oration and are replaced by water molecules from beneath the meni- 

 scuses. This brings about a decrease in pressure beneath the meni- 

 scuses and mercury is forced up in the capillary tube by atmospheric 

 pressure. 



This apparatus is similar to that used by Askenasy,* the chief 

 difference being that we use a layer of soil instead of a plate of 

 gypsum. 



The maximum capillary lift produced by the soil is found as 

 follows. Measure the length of the mercury column in cms., 

 multiply this by 13- G and add the length of the water column 

 in cms. from the top of the mercury column to the middle of the 

 soil layer. 



To test this method we used a- clay subsoil, the mechanical an- 

 alysis of which gave the following results: — 



Table 1, giving results of the mechanical analysis of the soil used. 



Clay 74-3% 



Silt 20-0% 



Sand, less than 2 ni.ni 1 «4% 



Sand, greater than 2 m.m -6% 



Organic matter 3-7% 



*Verhand. a.d. Naturalist. Med. Vercins Heidelberg, N.F., Hd. V. 1895; and 

 N.F., Bd. V, 1896. 



Sec. Ill, 191.3—8 



