12 Lang, Vascular Anatomy of the OpJiioglossacece. 



appear to justify the conclusion that the endodermal 

 characters of the internal endodermis are due to indepen- 

 dent specialisation of cells of a stelar pith near to the 

 inner margin of the xylem ; in this case the endodermis 

 could not be taken as proof of an intrusion of cortical 

 tissues to form the pith of the stele. This view of the 

 internal endodermis in HebiiiiitJiostadiys receives further 

 support from the occasional development of a similar 

 boundary to the pith of the root in this plant, and from 

 the fact that the internal endodermis in the rhizome often 

 abuts directly on elements of the xylem. 



In the case of some species of Botrychiuni and Ophio- 

 glossiun an internal endodermis continuous with the 

 external endodermis at the leaf-gaps is present in the 

 lower region of the young plants but is wanting in the 

 older plant. At first sight the appearance of this internal 

 endodermis in the young plants seems strong evidence of 

 an intrusion of cortical tissue at the leaf-gaps. Bower, 

 who has discussed this question at length,'^ provisionally 

 accepts the endodermis as marking the limit between 

 stele and cortex and interprets the pith in young plants of 

 BotrycJiiuvi lunaria as originally intra-stelar and later 

 partly intra-stelar and partly intrusive. I venture to think 

 that it is impossible to rest at this compromise in inter- 

 preting the stelar structure of the OphioglossacejE. The 

 uniformity of plan of construction of the stele in the 

 group makes it probable that one explanation of the origin 

 of the pith must hold throughout. 



A full discussion of the difficult morphological question 

 of the intrusive origin of pith is impossible here, but some 

 considerations may be mentioned which lead me to place 

 more weight on the interpretation suggested by Hchnin- 

 thostachys, viz., that the pith in the Ophioglossaceae is not 



« Bower, Annals of Botany, Vol. XXV. (191 1). 



