366 BOTANICAL GAZETTE [November 



which ensheaths the younger primordium with its more slender edge, 

 ascends in the petioles on the inner or dorsal portion, and branches 

 repeatedly. Trace L 1 has its origin near and to the right of the 

 protoxylem group px 2 , describes a girdle of approximately the same 

 magnitude as girdle L\, and runs into the base of the leaf on the right 

 hand. Traces L\ and L\ have their origin respectively on the right 

 and left of the protoxylem group px 1 (quite close to it), and after 

 ascending for a short distance run directly into the outer or ventral 

 portion of the leaf, where they ascend without further branching. 



These facts seem to indicate that there is some relationship between 

 the protoxylem groups of the vascular plate and the origin of the trace 

 of the first leaves. The traces of the third leaf (L 3 ) are followed 

 with some difficulty, but the same condition noted for the first and 

 second leaves is clear. When a fourth leaf is far enough advanced, 

 exactly the same conditions are also presented. The girdling habit 

 of the very young strands is already marked out; but their origin 

 appears no longer to be restricted to the neighborhood of the protoxy- 

 lem groups, but may occur anywhere in the plate, from which they 

 now ascend vertically for a longer distance before turning out into the 

 cortex. It is very difficult, however, to follow the younger strands 

 while they are still in the potential vascular tissue, since no xylem 

 elements have as yet been formed; but in their passage through the 

 cortex into the primordium their path may be made out clearly, and it 

 shows that the girdle is already established at this stage. Even in 

 the absence of xylem elements, the bundle is clearly marked off by 

 the character and arrangement of its cells (figs. 26, 27). 



Although the specific case described represents the general state 

 of affairs, many variations are found in the place of origin of the traces. 

 Of the many specimens examined probably no two traces were found 

 to be exactly alike in this respect; also anastomoses between adjoin- 

 ing traces were found here and there. The strands thus ascending 

 vertically for a short distance into the procambial tissue above the 

 vascular plate are the first to assume specific characters. Although 

 the young cylinder is made up of separate traces, it must be observed 

 that these are in the procambial tissue, which is very different from 

 that which surrounds the procambium. 



The transition from endarch to exarch xylem. — When a cotyle- 



