PROTECTED STOMATA 177 



greater thickness, and in the neighbourhood of the guard 

 cells may take on the form of special overarching outgrowths, 

 forming more or less deep pits often of complicated form. 

 In Dasylirion filifoUum it forms a double chamber over the 

 stoma ; in Euphorbia Tirucalli, where it attains no special 

 thickness in general, it builds up a deep well over the 

 stoma. 



The general effect of this mode of protection, whether 

 due to individual pits, to grooves, or to an inrolling of the 

 whole leaf surface, is to interpose between the diffusing stoma 

 and the dry, outer air a space in which a high degree of 

 humidity prevails, so that the rate of diffusion is slowed 

 down. For, instead of the whole diffusion gradient from 

 saturation to the minimum humidity being accomplished 

 in the short distance from the mesophyll cell to the external 

 orifice of the stoma (in moving air), it is extended for a 

 considerable distance, and consequently the rate of diffusion 

 falls. The conditions of diffusion for such stomata have 

 been worked out by Renner with such exactitude as is 

 possible for openings so irregular. For Agave americana, 

 a leaf in which the pit is of a fairly regular shape, and is 

 formed by cuticular extension, we may represent its effect 

 as that of a wide tube in the bottom of which diffusion shells 

 are formed in the usual way, and through which diffusion 

 proceeds to the outer air, where again diffusion shells are 

 formed. This can be treated by an extension of the formulae 

 we have already studied, but into the details of which we 

 need not go. Renner calculates that the superposition of 

 the pit leads to a depression of transpiration by 31 per cent. 

 The more complex, almost globular pit, of Hakea suaveolens 

 gives a depression of 37 per cent. In Nerium, with its 

 stomata grouped in common hollows, the depression is as 

 much as 77 per cent. Further complications arise when the 

 stomatal pore is not straight. We have noted that, to a 

 slight extent, this is the common case, the actual pore being 

 divided into fore- and after-courts by slight cuticular 

 protuberances. This irregularity may be much exag- 

 gerated, giving rise to such an extreme case as the sinuous 



N 



