334 NEW YORK STATE MUSEUM 



rows; and the orthostichy is therefore approximately 90 . Lines 

 connecting the centers of any group of four proximal scars will 

 form a rectangle in which the hypothenuses are vertical and hori- 

 zontal. The passage from the favularian to the lepidodendroid 

 type of leaf cushion arrangement is primarily due to the com- 

 bined effect of the rapidity and direction of growth on the 

 one hand and the phyllotaxy on the other. 



In the lower portions of the trunk, in which the ribbed type 

 of cortex is most strikingly developed, plate 6, the cushions in 

 the same vertical row are very close as the result of relatively 

 slow linear growth. At the same time the lateral expansion 

 or thickening of the trunk has widely removed the rows, espe- 

 cially near the base, as seen in plate 5. Thus, in the plate just 

 cited the leaf scars may be but 3 mm, or less, distant from 

 center to center in the same row, while at 90 cm from the 

 base they are 7 mm distant. Xear the top they become about 

 11 mm distant. At the same time, in passing upward we find 

 the ribs growing narrower, and the vertical rows coming more 

 closely together, not merely as the result of the introduction 

 of new rows, but by reason also of greater linear growth as- 

 compared with the transverse increase in the size of the tree. 

 As the cushions in the vertical rows become more distant and 

 the lateral cushions crowd closer we find the rhomboidal type 

 appearing near the periphery of the impression, and later in 

 the median zone. 



Referring again to the Knorria stage illustrated in plate 10,. 

 it will be seen that the narrow ribs correspond to the A'ertical 

 rows of nearly erect and imbricated nerve-trace sheaths whose 

 more resistant structure makes their presence known even when 

 the outer cortex and cushion tissue is but partially shrunken, as 

 in figure 2 of the same plate. When the hollow cortical cylinder 

 was flattened and the median portion of the fossil foreshortened 

 by lateral pressure, the rows of more rigid sheaths undoubtedly 

 controlled the clear alinement and strong topography of the 

 median ribs, whose present narrowness is due to the lateral 

 pressure. There is little room for doubt that the presence of 

 ribs in the median portion of the upper part of the trunk is 

 chiefly due to the resistance of the nerve sheaths. No doubt 

 they also contributed to the prominence and rigidity, of the 

 costae in the lower part of the trunk. Where the pressure was 

 oblique to the surface, as near the lateral borders in the areas 



