THE CARBONIFEROUS FLORA OF BERWICKSHIRE. 265 
nected. In spite of these occasional fusions, the three lobes of the primary xylem very 
clearly retain their individuality throughout all the sections, as may be seen by comparing 
the above-mentioned figures,—so much so indeed that they may be regarded as practically 
independent of one another. 
These three lobes are obviously in relation to the leaf-traces, one of which, as will be 
shown later, departs from the extremity of each lobe in turn. It is clear, therefore, that 
the divergence must have been 3. The sequence in which fusion occurs between any 
two of the lobes is possibly also related to the order of the leaf-trace departure, but our 
series of sections was not sufficiently extensive to settle this point. 
The elements of the primary xylem are fairly large, attaining a diameter of 160z. 
In longitudinal section they are seen to be typical elongated trachez with finely pointed 
ends, and beautiful porose pitting on all their vertical walls (fig. 12). Towards the ex- 
tremities of the lobes the tracheze become somewhat smaller, and their pitting becomes 
typical scalariform (fig. 13). There appear to be no protoxylem strands proper to the 
stem itself, but in the neighbourhood of the departing leaf-traces a pair of definite exarch 
protoxylem strands are usually to be seen (fig. 11, prw.). These are evidently decurrent 
from the leaf-trace, and lower down in the stem they fuse to a single strand (fig. 10, prz.). 
The secondary xylem evidently appears first opposite the side bays of the triangle of 
the primary xylem, and when the bays are filled up it crosses over in front of the angles, 
so that the outline of the whole xylem mass, which was at first triangular, eventually 
becomes circular. The earlier elements of the secondary xylem are narrower than those 
formed later (fig. 11 and figs. 6-10), and its tracheze are usually separated from those of 
the primary xylem by a few cells of parenchyma. Often, however, primary and secondary 
tracheze are found to be in direct contact. The trachez of the secondary xylem are 
arranged in very regular radial rows, which are separated by the medullary rays into 
wedges of from 2—5 elements in width (fig. 11). The medullary rays widen out con- 
siderably towards the periphery, and in this region secondary medullary rays also make 
their appearance. The rays are all very long vertically, and are from 1-6 cells broad 
(fig. 14). A few specially wide medullary rays occur on the inside of the departing leaf- 
trace (fig. 10, m.R.). In radial section the cells of the rays have the ordinary brick-like 
appearance, with the long axis radial (fig. 15). All the trachez of the secondary wood 
exhibit porose pitting, but the pits are confined to the radial walls; the tangential walls 
are quite smooth, being entirely without pits (fig. 14). 
No trace was found of any tissue resembling phloem, neither around the stele of the 
stem nor in relation to the leaf-traces in the cortex. However, patches of a very curious 
tissue have been preserved in the immediate neighbourhood of the stele, which are clearly 
parts of a zone that in life completely and closely surrounded it. The inner portion of 
this tissue consists of small cells with very dark contents. They are rectangular in form, 
and are arranged in very regular radial series. In longitudinal section they have 
exactly the same appearance as in transverse, only their long axes are vertical instead 
of being tangential (fig. 16, 7.z.). There can be no doubt that the cells arose from the 
