X] ENDODERMIS AS PHYSIOLOGICAL BARRIER i8i 



ultimately be reached where the facility for interchange through the endo- 

 dermis will cease to suffice for the needs of the tissues within. This facility 

 for interchange will then become a " limiting factor." Either some means of 

 increasing the surface-area of the stele, and so of increasing the means of 

 transit, must be supplied, or the conical enlargement of the stele must be 

 checked, and the later-formed regions of the stele will then be cylindrical. The 

 increase cannot be continued indefinitely in the form of a cone. But on the 

 other hand, any deviation from the simple conical form, by involution or 



Fig. 173. Helininthostachys zeylanica : part of transverse section of root (Gwynne-Vaughan Collection, 

 slide 589; X 66). The endodermis, recognised by the characteristic structure of its radial walls, 

 marks a boundary between the outside cortex, with large starch-grains (here above), and the inner 

 conjunctive parenchyma (here below), with "small grains. (Drawn by Dr J. M. Thompson.) 



by excrescence, will give an increase of the proportion of surface to bulk, 

 and thus tend to overcome the difficulty. We may proceed now to see how 

 these demands following on increase in size have been met in the stems of 

 Ferns, putting upon the facts already stated in Chapters VII — IX an inter- 

 pretation in terms of the proportion of surface to bulk. 



It is generally admitted that the protostele is the most primitive stelar 

 type. It is present in the juvenile stage of all Ferns, and it is permanently 

 retained in the adult stems of some of them (Fig. 174). The stele receives 



