178 THE BIOLOGY OF FLOWERING PLANTS 



channel seen in Nipafruticans, where the effect on diffusion 

 must be important. Experimental confirmation of Renner's 

 results on the actual leaf is impracticable, but Renner has 

 substantiated his calculation by evaporation experiments 

 with physical models in which the observed depression is 

 in good agreement with that calculated. We have thus 

 every reason to believe that the protected stoma gives up 



Fig. 19. — Structure of Stomata : sections through, i , Agave ameri- 

 cana, X 450; 2, Hakea suaveolens, x 500; Nipa fniticam ; 3, Verbena 

 ciliata, x 600. (i and 2 after Renner, 3 after Bobisut, 4 after Lloyd.) 



much less water than one freely exposed. Recently Grad- 

 man (1923) has shown that the pit is specially important 

 in protecting the stoma from the drying effects of wind. 

 Diffusion of carbon dioxide is not depressed to the same 

 extent as is transpiration. 



Depression of transpiration also occurs if the stoma does 

 not open directly into the mesophyll, but is separated from 

 it by internal cuticularised cells or by sclerenchyma. 

 Occasionally prolonged drought, or the death of the guard 

 cells, leads to a complete stoppage of the passage by thick- 



