40 PENHALLOW AND DAWSON 



indicates a distinct want of unity or of compactness in the structure as a whole. The 

 want of parallelism thus shown is often seen in cross section in the strictly transverse direc- 

 tion taken by some of the cells. In longitudinal section this becomes much more obvious 

 (Plate I, fig. 3), and in both sections is most pronounced in the immediate vicinity of 

 the radial openings, into which the large, tubular cells penetrate at all angles ' (Plate II, 

 iig. 6). 



A close examination of the transverse sections, also shows that the intercellular areas 

 are more or less closely occupied by a system of much smaller, rather thin-walled fila- 

 mentous cells, which are cut at various angles, sometimes exposing considerable length, 

 at other times showing a transverse section. This structure becomes much more evident 

 in longitudinal section, in which they are seen to cross the large cells in all directions. 

 This featvire, though not exhibited very clearly, is shown in one of the figures given by 

 Mr. Carrnthcrs", and is also conspicuous in Plate II, fig. 5, just to the right of the centre, 

 on the upper margin of the figure, where the small cells traverse a large cell diagonally 

 downward to the left. 



Longitudinal sections show that the principal part of the structure is composed of 

 tubular cells of indeterminate length. In none of the various sections examined haA'e we 

 yet been able to detect either a transverse septum or a tapering extremity ; but that the 

 cells are not organically united is very obvious, not only from the variable intervals which 

 separate them, but also from their very sinuous course, in consequence of which they wind 

 in and out among one another in a great variety of ways (Plate I, figs. 1 and 3). It 

 is also found that the myceloid filaments referred to, as seen in cross section, are very 

 numerous, and that they traverse the structure in all directions, forming a most conspicu- 

 ous feature, the proper nature of which it would be difiicult to mistake. These filaments 

 have a diameter of 5.33 /', and are the same as those found in the specimen to be described 

 later. As these filaments cross the cells in all directions, they give to the wall of the latter 

 a somewhat striated appearance, though a careful examination with good light cannot fail 

 to discover their true nature. Nevertheless, we have observed that under certain condi- 

 tions, they might be mistaken for striation of the cell wall. These filaments are the struc- 

 tures which gave rise to the original description of " a double series of spiral fibres," 

 being figured as such by Sir Wm. Dawson.' They are more correctly shown in the 

 figure by Mr. Carruthers,^ which may be compared with photomicrograph in Plate II, 

 fig. 0. 



So far, all our examinations have failed to show any structure in the cell wall. It 

 is also to be remarked that our specimens have so far presented no evidence whatever, to 

 show that a proper intercellular substance or primary cell wall was present in the original 

 striicture, to unite the cells now visible. On the contrary, the entire aspect of the cells 

 most forcibly suggests that they were either entirely free from the beginning, or that their 

 iiniou with one another coiild not have been established by anything more durable than a 

 mucilaginous modification of the cellulose substance. 



The ray-like openings in the tissue traverse the structure at right angles to the layers 

 already noted. Sometimes they are continuous through two or more layers (Plate I, 

 fig. 3), but more frequently their continuity is interrupted at short intervals, sometimes 



' M. Mic. Jn'l viii, plate xxxi, b. " M. Mic. Jn'l viii ; PI. xxxi, c. * 



' Geol. Soc. Can., Fossil Plants, plate ii, 26. * M. Mic. Jn'l viii, 166, plate xxxii, 6. c. 



