OF THE FOSSIL PLANTS OF THE COAL-MEASURES. 
sections, since the vessels thus transversely divided closely resemble the parenchymatous 
cells, both in size and form; but on turning to such longitudinal sections of similar 
specimens as are represented in figs. 2-5 & 6, we see that they are abundantly present ; 
hut the largest vessels in such specimens do not exceed ’0025 in diameter. The number 
of vessels in each such bundle is also smaller than in the larger petioles. It is thus 
evident that the vessels in a bundle increase in number ; and as they do so, either some 
of them increase in size, or those added at the later period expand into larger dimensions 
than was the case with those first developed. Fig. 4* is a transverse section of one of 
the smallest of these petioles that I have yet met with. On comparing its general form 
with that of the terminal rachides of Angiopteris erecta and other ferns, the projecting 
angles seen at the lower right and left margins of the fossil section, separated from the 
rounded central ridge by two deep flanking grooves, are seen to correspond exactly with 
those parts of the recent rachides which actually bear the leaflets. Hence I think I can 
scarcely be wrong in concluding that fig. 4* at all events is a section belonging to the 
terminal leaf-bearing portion of the petiole to which it belonged. It exhibits two very 
distinct gum-canals at c, c \ the several larger openings nearer the centre of the section 
appear to have been similar canals which have become enlarged by some contraction of 
the surrounding parenchyma, as was probably the case with figs. 3 and 4. 
On turning to vertical sections of these petioles, we discover that nearly all the vessels 
are either of the barred or of the spiral type. The large ones are chiefly of the former 
class, the spirals being mainly found amongst the smaller ones. Fig. 8 represents a 
longitudinal section of a bundle from the same specimen as the transverse section 
(fig. 7). The largest vessels are here seen at d , yet smaller ones at d', and others still 
smaller at d". Fig. 12 represents a single spiral vessel from the same section as fig. 7. 
The spiral is here formed not by one continuous thread, as in most Exogens, but by at 
least four parallel ones, as is so common amongst Palms. Carefully prepared sections 
of vessels like fig. 12 demonstrate that these are not modifications of reticulated tissue, 
but true spirals, since the threads of lignine seen crossing each other are shown by such 
sections to belong to the opposite walls of the vessels. In Captain Att ke n’s plant I 
found a very few unmistakably reticulated vessels. I have already mentioned the fact 
that these vascular bundles are aggregated with varying degrees of compactness. In 
fig. 7 a considerable quantity of cellular tissue is interposed between the larger vessels ; 
and on turning to fig. 8, a' we see two of these sets of intervascular cells, each of which 
consists of three or four vertical rows of rectangular cells having a diameter of from 
•002 to -0006. Occasionally these cells are almost cubical, but they are usually elon- 
gated in the direction parallel with the vessels between which they are interposed. At 
figs. 8, e & 10, e we find a similar series of cells constituting an imperfect sheath to the 
vascular bundle. In the transverse section this sheath is often either absent or at least 
imperfect. In fig. 13 we have portions of two of the large vessels (d) of a vascular 
bundle from Captain Aitken’s specimen with part of its investing sheath at e, e. 
The cells of the latter graduate rapidly into the ordinary interfascicular parenchyma. 
