352 
PROFESSOR W. WILLIAMSON ON THE ORGANIZATION 
Cordaites. 
Two fragments alone, both from the Halifax deposits, whence they were sent to me 
by Mr. Binns, seem to resemble objects figured by M. Geand-’Eury which belong 
to Cordaites. One of these closely resembles the section of a leaf represented in 
Plate 18, fig. 1 of the work cited. The other is a fragment of epidermis (fig. 89) 
with numerous large and closely grouped stomata. The epidermal cells have dis- 
appeared, but the stomata are clear and distinct. From the large size of the area 
enclosed between each pair of guard-cells I presume that there has been a second inner 
pair to each stoma, but I cannot trace any division line in that area separating 
such cells. It is possible that this specimen may have belonged to some other plant 
than Cordaites, but it corresponds most closely with M. Geand-’Eury’s fig. 1", in his 
Plate 18. Fig. 89 is enlarged 360 diameters; the vertical length of each stoma is 
about ’0009, and its diameter about '0006. 
I am indebted to Professor Young and Mr. J. Young, of the Glasgow University, 
for a fragment of very curious wood from the Volcanic Ash of Arran ; fig. 90 represents 
a portion of a transverse section of this specimen enlarged 16 diameters. Fig. 91 is part 
of the same section further enlarged to 50 diameters, and fig. 92 is a tangential section, 
enlarged like fig. 90, 16 diameters. The structure of this specimen will be best under- 
stood by first examining the tangential section. It exhibits a series of vascular laminae, 
a, composed of rather large reticulated vessels, and which form a very irregular network, 
a, a, enclosing a multitude of areas, b , b, which are equally irregular in size and form and 
which are filled with intersected cells. On turning to the transverse section, fig. 90, 
we find that these vessels are arranged in rather regularly disposed laminae which 
have radiated outwards from the central axis of the stem. This is seen clearly 
wherever these laminae have been intersected at right angles to the direction followed by 
the vessels, as at a, a ; but in other places, as at a, a, the vessels are cut through nearly 
in the plane of their course as they bend round to rejoin other neighbouring laminae, and 
thus enclose the cellular areas, b, &.+ The same section shows that the cellular tracts, b, 
are composed of a modified form of prosencliymatous cells, grouped in masses which have 
radiated from the centre of the stem to the bark, the long axis of each of the component 
cells following the radial direction of these masses as a whole. These cell masses are, 
in fact, huge medullary rays of a most extraordinary form. Fig. 91, b, illustrates the 
character of these oblong cells, as well as the minutely reticulated structure of the 
walls of the vessels, a. Some of the cells have square ends ; more frequently they 
have oblique overlapping ones. They have the appearance of possessing thickened 
walls, which are translucent, whilst them cavities are occupied by carbonaceous 
matter, but whether this effect is due to original lignification of the cell-walls, or 
* Flore Carbonifere du Departement de la Loire, &c. 
f The upper and lower margins of this figure severally represent the medullary and cortical borders of 
tbe section, 
