1920] CRIBBS—TODEA BARBARA 289 



they persist throughout the axis. Frequently, however, the lower 

 portion of the stem cortex is free from sclerenchyma, in which 

 instance it is found to appear in the stem at the place of attachment 

 of the second root, from the base of which it extends about the stele. 



Wh 



of the leaves 



below the point of introduction most generally lack the character- 

 istic cortical wall thickening of the cells surrounding the central 

 bundle. After its first appearance in the stem it is sparingly 

 developed or absent from the base of the next one or two petioles, 

 although quite abundant in the stem axis and the leaves at higher 

 levels, which would indicate that the rachis has been the last 

 region of the plant to develop the sclerenchyma. An endophytic 

 fungus was found to occur frequently in the cortical tissues of 

 the root, external to the endodermis and internal to the scleren- 

 chymatous cells of the peripheral region. It was found to gain 

 entrance by way of root hairs, and also by dissolving its way 

 through the epidermal cell wall at the edge of the root cap (fig. 42). 



The first leaf originates at a point about 0.1 mm. above the 

 foot attachment. The trace is isolated by an embayment, and 

 carries off from the main axis about four cells (varies from three to 

 six), which usually form a narrow band (fig. 5) which, once in the 

 petiole, usually becomes endarch in its arrangement (fig. 33). 

 Subsequent traces depart in a similar manner, carrying out an 

 increasingly greater number of xylem elements until the appearance 

 of a definite medulla tissue. The first petioles are practically 

 wingless, but with an increase in leaf size and number they over- 

 lap and become conspicuously winged. The metaxylem elements 

 meanwhile become more numerous, spreading out so as to form a 

 semicircle with the opening toward the stele (fig. 34). The poorly 

 defined phloem of the first trace becomes more definite and abundant 

 in the second and third, where it may form a complete circle about 

 the xylem. 



In the petioles of the lower leaves the xylem sometimes assumes 

 a distinctly mesarch arrangement, but is not completely invested by 

 phloem (figs. 4, 25). The occurrence of primitive structures in 

 the basal petioles has long been recognized, and the appearance of 

 mesarch strands here is further evidence for the origin of our 



