Anatomy of the Cone and Fertile Stem of Equisetum. 673 
fourth and fifth ones of the fourth whorl of Cone A would have closed 
the meshes between these traces, and we should in both cases have had 
two meshes alternating, though somewhat irregularly, with one another. 
In many cases the amount of xylem present at the node, though not so 
great as in the two cases mentioned, yet produces a marked constriction of 
the mesh in the nodal region ; where there is no perceptible increase of the 
xylem in the nodal region the mesh remains of the same width at this level. 
Another phenomenon, only observed once in E. palustre , occurs at the apex 
of Cone A of this species. Here two meshes originating above the tenth 
whorl fuse in the upper part of their course by the dying out of the strand 
between them. The conditions are here complicated by the fact that the 
strand, by the dying out of which the two meshes come into contact, is 
unusually wide, and gives off two traces ; it is clearly one of those stretches 
of xylem that extend uninterruptedly through a whole internode. Allowing 
for the convergence of the lines of superposition of the diagram (Text- 
fig. 3) owing to decrease in diameter of the stele, we can see that the two 
convergent meshes originate above the third and sixth traces of the tenth 
whorl. 
Another tendency when the xylem is less developed is for a parenchyma- 
tous mesh to be slightly decurrent on one side and below the trace above 
which it may, phylogenetically, be considered to have originated ; in such a 
case a trace-bearing strand appears to branch just below the departure 
of a trace. Such a trace would have been median and have subtended 
a mesh had no premature branching of the strand bearing it occurred ; 
it now appears to be lateral and adnate to one of the branches of the strand. 
This expression of less lignification is common in E. limosum , and an 
example of it may be noted at the eighth node of Cone A of E. palustre , 
where a mesh, apparently associated with the fourth trace, is decurrent some 
way below the latter and the third trace. (This mesh might at first sight 
appear to be between the third and fourth traces, but this appearance is 
due to the fact that the third trace is very nearly lateral ; and as the middle 
line of the mesh is very nearly directly over the fourth trace the mesh 
presumably belongs to the latter.) 
Only one clear case of the decurrence of a parenchymatous mesh was 
observed in E. arvense , and it was of a different type. The second whorl 
of Cone B of this species had nine traces, of which two were about half the 
size of the others and entered one sporangiophore. The smallness of these 
two traces and the fact that the whorl below and the three succeeding whorls 
each have eight sporangiophores indicate that we are here dealing, not with 
two fused sporangiophores, but with one of those cases, to be considered 
later, in which the first division of the vascular supply of the trace is so 
premature that the trace is double at its point of departure from the axial 
stele. Such cases of traces originating as two bundles are not rare, but this 
Y y 
