432 BOTANICAL GAZETTE [may 



trace in its relation to the central cylinder and its passage through 

 the cortex (fig 14). Not until it reaches the outer part of the cortex 

 or the decurrent base of the bulbil does it exhibit any essential devi- 

 ation from a leaf trace. Here it widens into a narrow tangential 

 band of cells, and presently gives off laterally two small traces to 

 the first leaves of the bulbil (fig. 12). The middle part continues 

 and presently widens into a radial band whose outer portions be- 

 come the traces of the second pair of leaves (figs. 13, 15), and so 

 on until four pairs of traces have separated from it. These facts 

 make it clear that the " supporting ' ' leaf is undoubtedly a lateral 

 organ. In the bulbil proper in connection with the appearance of 



Figs. 16, 17. — Fig. 16, cross-section of vascular bundle of bulbil where first root 

 is given off; X320; fig. 17, cross-section of bundles of bulbil and root below origin of 

 root: b, bulbil bundle; r, root bundle; X320. 



the first root a much greater change occurs. In a young bulbil, 

 even before the vascular tissues are' differentiated, the tip of this 

 root is outlined as a mass of deeply staining meristematic cells. 

 A second and smaller root develops nearer the tip of the bulbil 

 (fig. 5). Both of these roots connect with the vascular strand of 

 the bulbil on its adaxial side and, turning downward and laterally, 



imbedded 



slime 



Fig. 16 shows the junction of the root cylinder and the axis 



the 

 can 



In both of these 



trace is lost. 



junction of root and axis all resemblance 



somewhat 



