262 TRANSFER OF WATER THROUGH THE PLANT. 



absorbed to cause a swelling of the tissues and a closing of th« 

 crack. In this condition it is safe to assume tliat the cell-walls 

 tliemselves are saturated, but that there is no liquid water in the 

 cavity of the cells. Tlie difference between the weight of the dr3- 

 and that of the saturated disc gives the weight of the water 

 taken up and held ; tliis, converted into volume, is found to be 

 approximately one half tiiat of tlie space occupied by the cell- 

 wall itself. 



702. The water which is taken up in relatively small amount 

 and held in the micellar interstices of lignified cell-wall is in the 

 state of equilibrium previously described. When, however, this 

 equilibrium is disturbed by evaporation at aiij- point, there is an 

 immediate transfer of the imbibed water to that point, and the 

 loss from this transfer must be made good at once by the recep- 

 tion of more water. This interstitial transfer may take place 

 through any length of woody tissue, provided there is a con- 

 sumption of the water at one extremity and an adequate supply 

 at the other. When the consumption of water is only that which 

 is due to the opening of growing buds, or to some chemical pro- 

 cess, a slow transfer of water to the point of consumption ' must 

 take place. When, however, it is duo to evaporation from the 

 leaves, the transfer is exceedingly rapid. 



703. Boehm''' considers the ascent of water in ligneous tissue ' 

 to be "a phenomenon of filtration caused by differences in pres- 



1 A similar transfer can be ilernonstrated to take place in porous inorganic 

 matter, for instance powdered hydrated gypsum. If a long tube be filled with 

 this material and well saturated with water, one end being placed in water 

 and the other exposed to a dry atmosphere, the continual loss by evaporation 

 above will be made good by water brought up from below. 



Jamin's apparatus for demonstrating the pressure exerted by the imbibition 

 of water by a porous substance consists of a cj'linder, in the mouth of which 

 can be placed a tightly fitting plug of wood, through which passes a ma- 

 nometer tube. The pulverulent substance, for instance zinc oxide, is closely 

 packed in the interior of the cylinder, around the open end of the manometer, 

 and the whole apparatus is then placed in water. With zinc oxide the ma- 

 nometer shows a jiressure of five atmospheres ; with powdered starch, more 

 than six atmospheres. If a manometer is similarly placed in a block of dry 

 chalk, and the chalk is then submerged, a pressure of three to four atmos- 

 pheres is indicated (Le9ons ])rofessees devant la Society chimique, S&nce du 

 8 mars, 1861, quoted by Deheiain: Cours de Chimie Agricole, 1873, p. 165). 



2 Ann. des So. nat., ser. 6, tome vi., 1878, p. 236. 



3 As might be expected, woody tissues never conduct water so readily in a 

 transverse as in a longitudinal direction. Experiments with regard to this have 

 been conducted by Wiesner (Sitzungsb. d. Wien Akad., Bd. Ixxii. 1 Abth., 

 1875) upon cubes of wood. Four sides of these were protected by varnish 



