Dixon — Note on the Role of Osmosis in Transpiration. 773 



to the collapse of the leaf. The result of this will be that the loss of 

 water by transpiration will be lessened. That this mechanism is very 

 perfect, may be seen from the experiments quoted above, by which it 

 is shown that leaves killed in steam lose about twice as much water by 

 evaporation as leaves which have drooped owing to a partial loss of 

 turgescence, although the external surface exposed for evaporation by 

 the steamed leaves must be very much less, owing to the way in which 

 they fall together, than the leaves which have simply drooped. 



It may be urged that if the osmotic attraction suffices to draw 

 water from the conduits into the osmotic cells, the same attraction will 

 prevent its escape from the surface of an osmotic cell which is exposed 

 to the atmosphere, and there is no reason to suppose that the membrane 

 formed by the protoplasm in these cells is semipermeable on the side 

 towards the conduits and not so towards the atmosphere. Although 

 it seemed probable that when the pressure rose in the cells that it 

 would force pure water through the membrane while the salt in 

 solution would be retained in the cell, experiments were made to test 

 this. A semipermeable membrane was formed in the walls of a porous 

 pot by rinsing it out, first, with copper sulphate solution and after 

 washing it with water, with a second solution formed of Ferrocyanide 

 of potassium in water. The porous vessel, so furnished with a semi- 

 permeable membrane, was next filled with a saturated sugar solution 

 and a capillary tube one metre in length sealed hermetically into it 

 so that the end in the vessel was almost in contact with the bottom. 

 To the upper free end of the tube a small thistle-funnel was attached. 

 The pressure in the vessel could now be raised by pouring mercury 

 into the funnel. The downward motion of a small bubble of air 

 in the capillary tube showed that liquid must be leaving the vessel, 

 and since the outer surface remained dry to the touch it is evident 

 that the loss was due to evaporation taking place in the pores of the 

 vessel. The experiment was repeated with a solution of sodium 

 nitrate replacing the sugar solution.^ In these cases a pressure of 

 about an atmosphere and a-half was sufficient to bring about evapo- 

 ration of the outer surface of the semipermeable membrane ; so 

 that we may safely assume that evaporation can take place from the 

 osmotic cells of the leaf. 



We may diagrammatically represent the action of an osmotic cell 

 in the leaf by the following arrangement. A small porous pot A, in the 

 walls of which a semipermeable membrane has been formed, is sealed 



' Op. Oswald, " Outlines of General Chemistry," p. 127. 



