i 9 o8] THIESSEN—DIOON EDULE 373 



way vertically through the parenchymatous cells from one bundle to 

 another. They are in all senses true reticulated vessels, without 

 cross-walls, establishing a connection between the bundles of the vascu- 

 lar system. 



Discussion 



In the embryo of Dioon edule the vascular cylinder is a protostele, 

 which in some specimens contains a solid xylem mass. From this 

 solid cylinder all gradations are found to the siphonostele. The cells 

 of the pith are actively meristematic, as shown by the mitotic figures, 

 and often in older specimens xylem elements are found in the central 

 part at the level of the vascular plate. The xylem cylinder also con- 

 tinually increases in size, new elements being added to it constantly 

 between bundles already existing, as well as by the cambium. 



It should be emphasized that the vascular cylinder in the embryo 

 and seedling does not consist of the short xylem cylinder only, but is 

 continuous in a tissue, very different from the cortex outside and the 

 pith within, which gives rise to strands of procambial tissue running 

 into the leaf primordia. The pith is also a well-defined tissue from 

 the vascular plate to the stem tip, and nowhere suggests that it arises 

 from an intrusion of the cortex through the leaf gaps. An inspection 

 of fig. 8 shows that it has its beginning at the very tip of the growing 

 point, where it is seen to consist of a single row of cells or a tier of a 

 few cells, gradually expanding as the stem grows into the large pith 

 found in the older part. 



The transition of the protoxylem from the endarch to the exarch 

 position was first described by Mettenius (4), who also suggested the 

 descriptive terms centripetal and centrifugal xylem. The situation 

 was interpreted by Bertrand and Renault (5), who also established 

 that the centripetal wood is in the same relative position throughout 

 its whole length; that it increases in bulk toward its upper end; and 

 that the centrifugal elements are reduced more and more. They say 

 that the centripetal wood is intercalated between the pole and the 

 outer face. It is regarded as the primary wood (developed from the 

 procambium), while the centrifugal wood is regarded as secondary 

 (developed from the cambium). 



The present investigation shows that this statement needs modi- 

 fication. While all centripetal wood is primary, all primary wood is 



