720 MR W. T. GORDON ON 



first (fig. 9) a root (rt.) is passing towards the upper part of the figure from the left-hand 

 lower corner : in the second (fig. 10) another is observed about to depart from the centre 

 of the right-hand side (immediately below the petiole in the drawing, but really above 

 the departing petiole-trace); while in the third (fig. 11), the root shown in the last 

 figure is seen detached though still in the cortex of the stem, and another root is seen 

 going towards the lower left-hand corner. This irregular distribution of the roots will 

 be referred to again when we discuss the question whether the stem was a rhizome or 

 an erect aerial shoot. 



Branching of the Stem. 

 Several stems were examined, and in some cases they continued for a considerable 

 distance. In one case, continuous serial sections were cut for nearly seven inches with- 

 out reaching either end of the specimen. In the case of other stems occurring in the 

 same block a considerable length could be followed, but the preservation was very bad. 

 In these last examples, gaps occurred in the xylem-tissue, and in places it had all decayed 

 away, though the same stem could be again recognised a few sections further on. In 

 the parts of the stem which were continuous, branchings were sometimes met with, and 

 in all cases the branching was dichotomous. Even at these bifurcations the outer 

 xylem-ring remains unbroken, so that every departure from the stem, whether of root, 

 petiole, or branch, is protostelic. In Plate I. figs. 1 to 4, the various stages of the 

 branching are figured. The first figure shows an almost circular cross-section of the 

 stem-xylem, with no sign of an approaching bifurcation. In the next section the xylem 

 becomes elliptical in shape and much larger. The inner zone increases in size along 

 with the outer, and forms an ellipse concentric with that formed by the outer zone. In 

 the following section (fig. 3), the inner xylem has become contracted in the centre and 

 thus assumed a dumb-bell shape, while the outer xylem remains elliptical. At the 

 next stage the outer xylem also becomes dumb-bell shaped, and the inner xylem divides 

 through the middle into two groups (fig. 4). The outer xylem of the stem is here 

 8 -shaped with the inner xylem in the loops of the 8. With the outer xylem dividing 

 through the middle, two equal branch strands are produced. It is interesting to note 

 that the inner xylem never, on any occasion, abutted directly on the periphery of the 

 stele. It will be seen that this branching is distinctly protostelic, and also that it is a 

 true dichotomy. In one case examined, the inner xylem showed a tendency to diminish 

 in size just before the bifurcation, and in one of the branches of this particular dichotomy 

 there is an almost entire absence of inner xylem for a short distance above the point of 

 bifurcation. In no case, however, is there any branch gap. 



Histology of the Petiole. 

 A typical transverse section of the xylem of a petiole is figured in PL IV. fig. 42. 

 This may be taken as the normal Diplolabis petiole-trace as described by Renault.* 



* " Bass, houill. et perm. d'Autun et d'Epinac," pp. 17 and 18, Giles mine'raux, 1896. 



