88 



Prof. W. C. Williamson on the Fossil [Dec. 19, 



through the stem, and which, as they alternately approach and recede 

 from, one another, divide this part of the bark into a series of lenticular 

 or rhomboidal areas, occupied by various forms of parenchyma. No 

 vascular bundles enter these areolae ; hence they are something altogether 

 different from the leaf-scars of the Lepicloclendra. Externally to this pro- 

 senchymatous layer some specimens exhibit detached traces of a very thin 

 external layer of parenchyma, apparently derived from the cells of the 

 rhomboidal areolae, which have extended beyond the fibrous laminae and 

 spread themselves over the surface of the bark as a continuous layer ; 

 but this condition appears to be confined to very young stems. 



Vascular bundles of large size ascend vertically through the two 

 inner parenchymatous layers of the bark. In some instances each of 

 these bundles exhibits, in the transverse section, an oval outline, with faint 

 traces of a vertical division into two parts. But ordinarily the two halves 

 of the bundle have separated, forming two distinct bundles, which are 

 some distance apart. They exhibit little or no tendency to diverge from 

 the ligneous cylinder as they ascend, and in some instances actually 

 become incorporated with it. It is remarkable that the position of each 

 of these double bundles, at the exterior of the ligneous zone, corresponds 

 with the spaces intervening between the detached masses of the medullary 

 cylinder tuithin it, as if the former were designed to act as buttresses 

 strengthening these weaker points in the vascular axis. It not unfre- 

 quently happens that exogenous additions are made to such of these 

 bundles as are encompassed by the innermost layer of the bark, in the 

 shape of a few radiating laminae of vessels developed on their outer or 

 peripheral surfaces. 



One specimen of the vascular axis is, as already mentioned, so large as 

 to demonstrate that the plant became arborescent. 



Though Dictyoxylon was not a dichotomizing plant, like Lepidodendron, 

 it gave off lateral bundles of vessels. Some of these are simple bundles, 

 consisting of numerous vessels intermingled with some cellular tissue. 

 Others have this central bundle invested by a thin ring of radiating laminae 

 with intervening medullary rays ; this exogenous ring sometimes be- 

 comes developed into a relatively large and distinct woody zone, like that 

 of the parent axis. The vessels of these lateral growths appear to be 

 wholly derived from the radiating woody zone. 



A second form of lateral appendage appears to spring from the me- 

 dullary rays, and consists of a cylindrical mass of reticulated cells, which 

 are chiefly prosenchymatous, but of an elongated type. It is suggested 

 that this structure may have been prolonged into an adventitious root. 



The structure of the central or medullary vascular axis of the former 

 of these two kinds of lateral appendages seems to indicate that the history 

 of the development of the medullary vascular cylinder in these plants 

 corresponds with what the author described in his preceding memoir 

 (Part III.) as taking place in the similar parts of the Lepidodenclra, viz. that 



