i8o SIZE A FACTOR IN STELAR MORPHOLOGY [ch. 



upon size, would be progressively simplified, and the evidence of this appears 

 in their structure. The same result is seen in the roots of Palms. (See 

 Cormack, 125.) 



In the young stems of vascular plants, and in those of the Ferns in 

 particular, the conducting tracts are strictly delimited from the surrounding 

 tissues by endodermis. The same holds also for their roots. This endo- 

 dermis forms not only a morphological, but also a physiological, boundary 

 that when normally developed is without any gap or imperfection. Its 

 physiological importance consists in the fact that the existence of even the 

 simplest type of endodermis places the contents of the conducting tract 

 under strictly protoplasmic control. All the lateral walls of its cells are 

 so specialised in substance that instead of being permeable like ordinary 

 cellulose walls, they are impermeable to fluids. Thus all possible leakage is 

 stopped, and the only channel for transit of substances into or out of the 

 stele is under the control of the protoplasts of the endodermal cells. This 

 control applies not only to salts, sugars, and other soluble substances, but 

 also to gases. Since in young and primitive plants the mantle is unbroken 

 by intercellular spaces, even the respiration of the living cells within the 

 barrier can only be conducted by interchange of gases passed in solution 

 through the cells of the endodermis. These structural facts, which can 

 be verified by sections of the stem or root of any young Fern, form the 

 foundation of a theory which may account for the extraordinary vascular 

 developments seen in these plants. 



Evidence of the effectiveness of the endodermis as a physiological barrier 

 is afforded by comparison of the cell-contents outside and inside of it. 

 Marked cases may be found of difference in size of the starch-grains on 

 either side of the barrier. This is well seen in the storage-rhizomes of 

 Pteridium, or oi Helininthostachys (Fig. 173), and a still more striking case is 

 seen vsxAcrostichnin aureiim. Such facts indicate that the endodermis controls 

 the passage of soluble sugar. That it is an effective barrier to the passage 

 of such substances as are incapable of penetrating the protoplasm, but whose 

 passage through the walls can be followed by their colour or by staining 

 reactions, has been shown by de Lavison i^Rcv. Gen. de Bot. 1910, p. 225), 

 and by Priestley (^Ne%v PJiyt. 1920, p. 192). Having such evidence before 

 us pointing to the endodermis as a selective screen, or even an effective 

 barrier to transit between the outer tissues and the conducting tract, the 

 constant diminution of the proportion of surface to bulk as the stele increases 

 in size becomes a matter of the utmost importance. The conical form of the 

 stele, noted above for the stem of all young Ferns, starts from the minute 

 stele of the sporeling. The stele expands upwards as a support to the 

 successively larger leaves of the established plant. Often the increase is 

 rapid, especially in Ferns with short internodes. In each case a limit must 



