350 BOTANICAL GAZETTE [may 



between the two (figs. 4, B, and 7). At first these extend the entire 

 distance between the two and give the appearance of expansion of 

 the stele in the plane of the leaf trace; but later they separate from 

 the stem stele, as shown by fig. 4, C. These wings may be considered 

 to be adaxial extensions of the centrifugal xylem of the leaf trace; 

 there is at no point any evidence of centripetal xylem. Soon after 

 the separation from the stem stele, the leaf trace disappears, having 

 been cut off by the absciss layer and carried out by periderm forma- 

 tion. This leaves two distinct wings of xylem and each of these 



Fig. 4. — Transverse sections of a branching rhizome of Bolryckium ramosum; 



by extension of its margins rounds up into a definite branch stele, 

 as shown in figs. 4, D, E, F, and 8. One of these branch steles 

 soon disappears, having apparently been unable to meet the com- 

 petition; its apical region has been entirely obliterated by peri- 

 derm formation. There can be little doubt that both these branch 

 steles developed from a single axillary bud. 



In the other branching specimen of this species, a wing of 

 xylem extends between the leaf trace and the stem stele on one 

 side only. It separates from the stem stele and at a slightly higher 

 point the leaf trace disappears. By extension of its margins, the 



