

1889. I BOTANICAL GAZETTE. IO^ 



a series of heavily thickened cells, apparently collenchymat- 

 ous, by which the bundle seems to be directly connected with 

 the group of hypodermal sclerenchymatous cells heretofore 

 described. 



The xylem also diminishes laterally in the same manner, 

 and is connected on either hand with an irregularly curved 

 line of tracheides which project into the surrounding par- 

 enchyma. The line of the M fringe of tracheides," if,indeed,it 

 can be called a fringe, can not be said to follow the contour 

 of the leaf. In many cases it terminates abruptly, although 

 in one or two instances in the sections made it was a prom- 

 inent feature. Its presence, however, was not so constant, 

 nor so marked, as might have been expected from the studies 

 of Frank 15 , and later those of Mohl 16 . In other respects 

 the xylem seems in no wise different from that usually found 

 in collateral bundles among gymnosperms. 



The phloem, which shows regular rows of similar ele- 

 ments, in all the sections examined, seemed to have thicker 

 walls than were to be expected in this section, but it is prob- 

 able that the apparent thickness was due to swelling. In 

 spite, however, of this somewhat unusual thickness of the 

 membranes of the phloem, the boundaries between it and the 

 xylem are sharp and distinct. 



The phloem, as the xvlem, diminishes somewhat lateral- 

 ly, and aids in giving the pointed appearance to the bundle 

 in cross section. 



As the xylem is connected by a series of thickened cells 

 to the sclerenchymatous strengthening cells above it, so the 

 phloem is connected with the resin duct by a number of rows 

 of thick-walled cells, and thence by way of the collenchyma 

 surrounding the duct, to that lying between the duct and the 

 epidermis. In the vertical line from the phloem to the duct 

 and thence to the lower leaf surface, the thickness of the cell 

 walls seem unchanged, but in all other directions they dimin- 

 ish somewhat rapidly until they pass into the surrounding 

 parenchyma. It is, as stated at the first, extremely difficult 

 to say where the bundle ends, and the non-equivalent ele- 

 ments of the same form begin (figs. 4 and 2). 



In longitudinal vertical section, the elements of the bun- 

 dle present no marked features unless it be in the absence of 

 spiral vessels, which were not detected in any of the sections 

 made. I am not as yet prepared to say that they do not oc- 



"Ibid, 1870, p. 10. 



