914 Ml! A. C. SEWARD AND MR A. W. HILL ON THE 



The occurrence of the uarrovver tracheids is in all probability merely the expression 

 of the result of harmful conditions, and not an indication that the secondary thickening 

 had reached its maximum extent. It might be suggested that the Lepidodendroid stems 

 possessed a limited power of secondary growth, and that the stem before us represents 

 the final stage of wood development ; we find, however, that such continuous bands of 

 smaller tracheids occasionally occur in the secondary wood of Lepidodendron * and 

 Stigmaria,t and are succeeded by larger tracheids of the normal size. A section of one 

 of the Arran stems in the Binney Collection shows an almost continuous zone of small 

 tracheids succeeded externally by larger elements near the periphery of the secondary 

 wood. The elements in the region t" of fig. 19, PL III. correspond more closely in size 

 with the large xylem tracheids than do the small elements in contact with the large 

 tracheid a. In forming the small tracheids the segments cut off from the meristem 

 cells must have divided radially ; if the radial walls were not formed, the meristem 

 segments would approximate more closely in breadth to the tracheids of normal size. 

 But the mechanics of secondary growth in Lepidodendroid stems has still to be satis- 

 factorily explained. 



Passing beyond the small tracheids (if, figs. 16 and 19), we find a few rows of 

 thinner- walled elongated elements, which are succeeded further out by several rows of 

 delicate parenchymatous cells (cm, PI. II. fig. 10, PI. III. figs. 16 and 19). The cells 

 composing this band of thin-walled tissue are rectangular, slightly elongated in a tan- 

 gential direction, and somewhat flattened, as seen in a radial longitudinal section (PI. II. 

 fig. 10, cm). The thinnest and most recently formed cell- walls in this tissue occupy a 

 radial position. In a few places the band cm may be traced continuously through the 

 more elongated elements t" to the narrow tracheids, but owing to the delicate nature of 

 the parenchymatous cells the continuity of the tissue is frequently interrupted by 

 tearing. 



In PI. II. fig. 10 we have an oblique radial section of the narrow tracheids and more 

 external tissue ; beyond the thicker tracheids t', corresponding to t' in the transverse 

 sections (figs. 16, 17 and 19), we have shorter and thinner elements t" (also in the 

 transverse sections, figs. 16, 17, 19, and the radial longitudinal section, fig. 5), which 

 we regard as partially developed tracheids, with transverse or slightly oblique walls, 

 and beyond these the band cm, composed of flattened cells in a state of active division. 

 At mr, in fig. 10, a strand of flatter cells is seen passing across the tracheids : this 

 represents the outward extension of a medullary ray (vide PI. I. fig. 5). This broad 

 band of rectangular cells we regard as the mcristematic zone, which is responsible for 

 the growth in thickness of the xylem cylinder. From the examination of the Dalmeny 

 stem and other species of Lepidodendroid plants, we are led to the conclusion that, as a 

 general rule, in Lepidodendron and Lepidophloios there was no well-defined cambium in 



* Williamson (72), pi. xliii. fig. 21. 



t Ibid. (81), pi. liii. fig. 21. Section No. 1922 H. (Williamson Coll.) shows three complete rings of smaller 

 tracheidfl in tin.' outer part of the secondary wood. Vide also Williamson (87), pi. iv. fig. 19. 



