VIII] 



ACCESSORY STRANDS 



157 



Similar blind strands are seen also in the pith of Schizaea dichotoma, and 

 of Platyzoma. Such examples show how futile it is to attempt to derive all 

 vascular tissue, or all tracts of endodermis, from some pre-existent source: 

 for P . podophylla is as complex in vascular structure as any known Pterid. 



Accessory strands are also recorded in the pith of large plants of Cerato- 

 pteris thalictroides, but neither their origin nor their connections are clear. 



It is thus seen that in the Leptosporangiate Ferns there is a wide range 

 of stelar structure in the stem. The leading character of it is the disintegra- 



Fig. 151. Transverse sections of stems, drawn to the same scale, showing that stelar complication 

 does not depend directly upon size alone. ( x 2.) A = CibotiHm BaroDtetz; B^Heniitelia setosa. 



tion of the stele, so that as the plant enlarges it is broken up into numerous 

 smaller tracts, instead of the single stele itself growing proportionately as 

 a solid cylinder, in the manner seen in so many other plants. Naturally the 

 more complex structure appears in the larger stems. But there is no exact 

 proportion between size and complexity. This is shown by comparing the 

 wide and simple solenostele of Cibotiinn Baronietz with the complex soleno- 

 stele and accessory medullary strands found in the rhizome of Hemitelia 

 setosa (Fig. 151,^, B). But the section of the latter here shown is of no greater 

 size than that of the simpler Cibotium. The whole series of the Leptospo- 



