MOVEMENT OF MATERIALS IN THE PLANT 1 49 



ment, thus alternately compressing and expanding the steam within the porce- 

 lain cylinder to an extent indicated by a quick rise and fall of the mercury in the 

 tube. At the end of this treatment the connection between mercury reservoir 

 and suction pump is closed, and the apparatus is allowed to cool with the 

 cylinder still submerged in water. 



In view of the impossibility of demonstrating tension in a water system con- 

 taining air bubbles of any considerable size, the apparatus is designed to facilitate 

 removal of undissolved air during the preliminary treatment just outlined. 

 (Figure 84a is referred to in the following statements.) The vertical J-shaped 

 tube (abed), in which the mercury column rises, is connected through the short 

 arm at its upper end to the porous cylinder (C), which is thus supported verti- 

 cally with its closed end downward. This arrangement makes it possible to 

 surround the cylinder with hot or cold water held in a tall beaker (B), to leave 

 the prepared apparatus in a resting state for an indefinite number of hours or 

 days after the suction treatment, and finally, by merely lowering the beaker 

 and directing a current of warm air upon the cylinder or by simply exposing it to 

 room air, to complete a demonstration at any time, without disturbing either 

 the cylinder or the vertically clamped tube. For weeks after the suction treat- 

 ment and before a demonstration the cylinder may be left submerged in water 

 which is freely exposed to the atmosphere, thus allowing dissolved air to permeate 

 the entire system. The method is consequently well suited to show tension in 

 solutions of air in water, which must in some instances be as concentrated as any 

 that may occur in the vessels and tissues of plants. 



When a demonstration is to occur the suction outlet (e) of the mercury 

 reservoir (R) is opened to the air, and the beaker (B) is lowered. Evaporation 

 from the surface of the cylinder then begins and a column of mercury extends up 

 the tube, progressively accommodating the decreasing water volume. At any 

 level in the tube the liquid is under a net pressure equal to the current barometer 

 reading decreased by (1) the hydrostatic pressure (expressed in centimeters of 

 mercury) of the liquid column extending downward from the level in question to 

 the surface of the mercury in the reservoir, and by (2) a small correction to 

 account for the capillary depression of the water-mercury meniscus in the tube. 

 As mercury replaces the water in the tube, the net pressure at the level of the 

 ascending mercury-water juncture decreases until it reaches zero, when the column 

 of mercury in the tube just balances the external pressure effective at the base 

 of the tube. Further rise of the mercury is evidence of tension, which appears 

 in the calculation as a negative value for the net pressure. The additional 

 mercury is not supported at all from below; it hangs on the water above it and 

 adheres to the water film lining the tube wall. The water above the mercury 

 hangs in turn on the water-impregnated walls of the porous cylinder, adhering 

 to mercury, glass, rubber and porcelain. The water system breaks sooner or 

 later, often within the cylinder but sometimes in the tube above the mercury, 

 and the mercury column then falls quickly to a height corresponding to the 

 current reading of the barometer, where it may remain for a considerable 

 time. It eventually falls entirely, as air enters the drying cylinder. At any 



