372 BOTANICAL GAZETTE [November 



sections of a maturer leaf of a seedling clearly shows the transition 

 from endarch, through mesarch, to exarch, and shows that it is quite 

 independent of secondary wood. The transition begins at the point 

 where the strand leaves the central cylinder, and in the seedling pro- 

 gresses uniformly and is completed in the petiole; but in the old plant 

 the larger proportion of the transition is completed in the lower part 

 of the petiole, where only a few strands of secondary xylem remain 

 and continue uniform until the bundles enter the pinna, when the 

 transition is completed (fig. 25). 



The cause of girdling. — No particular cause has been assigned for 

 this phenomenon. If such a leaf primordium as is represented in 

 figs. 3, L 2 , and 8, L, be selected, it is possible by careful staining to 

 detect four strands which are ultimately developed into vascular 

 bundles. The outer (abaxial) ones pursue a more or less direct course, 

 but in following them from the stem up, their course at first, after 

 leaving the procambium, is vertical; but in the base of the primordium 

 they turn inward to a considerable extent (fig. 3, L 2 ). The two inner 

 or adaxial strands after leaving the procambial cone pursue quite a 

 vertical course, but on reaching a region at the level of the base of the 

 primordium, they begin to turn toward it, one on each side, in some 

 cases ascending rather obliquely, but generally horizontal from the 

 start. Figs. 34, 35, Ug 1 shows the girdling of quite a young primor- 

 dium, the youngest one in that specimen. It is plain, therefore, that 

 the girdle is established very early in the development of the leaf to 

 which it belongs. 



Tracheae.- — In examining older plants, the oldest one being three 

 or four years old, a singular phenomenon is noticed. Between the 

 various bundles of the stem, which at this time do not yet make a com- 

 plete ring, vertical connections are found, consisting of irregular, 

 reticulated elements, branching and anastomosing, but forming con- 

 tinuous vessels. In the upper part of the plant, where they are devel- 

 oping, long and narrow cells are found winding and crowding through 

 between the parenchymatous cells. A little farther down in older 

 parts of the stem, these cells are found to be multinucleate, with only 

 here and there an ill-defined cross-wall. A little farther down still the 

 nuclei disappear, and soon lignification appears, developing finally 

 into well-defined reticulated tracheae. These wind in a verv tortuous 



