﻿328 ME. E. A. NEWELL ARBER ON" THE [May 1905, 



while in others again no such neck can be recognized (PI. XXX, 

 fig. 1 & PL XXXI, fig. 4). 



In many of the specimens, especially those in which the preserva- 

 tion is particularly good, the cell- walls of the outer limiting-layer 

 of the sac are conspicuous (PL XXX, figs, 1-2 & PL XXXI, 

 figs. 3-4) ; and the sac has the appearance of being strongly 

 striated in the direction of its greater axis. These walls are 

 united by very short cross-walls, placed more or less obliquely, 

 and thus the cells are rhomboidal in shape, and of much greater 

 length than breadth (PL XXXI, fig. 4). At the apex of the sac, 

 which is probably the extremity opposite to the neck, the cells 

 appear to be somewhat smaller and narrower. 



Such evidence as there is to be found in these specimens, points 

 to the fact that these sporangium-like bodies were probably attached 

 by the neck- like prolongation. In the few, and perhaps not wholly 

 trustworthy instances, in which fragments of these organs seem 

 to be still in continuity with the scale-leaves (PL XXX, fig. 4), 

 this appears to be the case ; and supplementary evidence may be 

 found in the way in which the detached bodies are often grouped 

 together, which distinctly suggests in many cases an aggregation 

 into sori, like the sporangia of many fossil and recent ferns. 

 In PL XXX, fig. 3, portions of four of these bodies are seen, 

 in one of which (a?) the neck is evident, and is turned towards 

 what was probably the common point of attachment of ail the 

 organs of this group. Another such group, similarly arranged, is 

 figured in PL XXXI, fig. 2. 



The evidence for the attribution of these sac-like structures to 

 Glossopteris Broivniana will be dealt with in the following section 

 of this paper. It will be shown that there are some grounds, 

 although not as conclusive as one could wish, for the view that 

 these sporangium-like organs were borne on the concave surface 

 (possibly the lower surface) of the scale-leaves. 



These organs are not solid bodies. They were undoubtedly sac- 

 like structures, and, in the living state, must have contained 

 something. Their appearance is essentially that of a distended 

 sac (PL XXX, fig. 2). Further, they seem to have opened or 

 dehisced longitudinally, that is, in the direction of the major axis 

 (PL XXXI, fig. 1). The exact mode of dehiscence cannot be- 

 ascertained. The splitting may have begun first at the apex 

 (PL XXX, figs. 1 & 2), but this is not quite certain. The 

 evidence is, however, sufficient to show that such dehiscence was, 

 in all probability, a natural feature at a certain stage in their 

 development. The wall of the sac, so far as one can judge from 

 specimens in which the anatomical structure is not preserved, was 

 probably more than one layer of cells in thickness. Xot only does 

 it appear to be comparatively thick (PL XXX, fig. 1 a), but the cell- 

 walls of the inner limiting -layer are much less conspicuous than 

 those of the outer, and in many cases can hardly be distinguished 

 (PL XXXI, fig. 1). The three sporangium-like bodies figured in 

 PL XXXI, fig. 1 have probably all dehisced, and are viewed from 



