452 Brcnune. — A Fourth Contribution to our Knowledge of 
neighbourhood of the node , the direct and immediate effect would be for 
some of the meshes to persist into the internode above. I think that this is the 
explanation of the greater reduction of the vascular system in the cones of 
the species of Equisetum that have relatively wide steles.’ I then pointed 
out that of the four species so far studied the two with the widest steles, 
E. maximum and E. limosum , had the most reduced vascular system, while 
within the limits of the single species, where the examples studied had 
steles showing a marked difference in diameter, the more irregular and re- 
duced vascular network was always found in the specimens with wider steles. 
I added : c Of course the width of the stele is not the only, nor the principal 
factor causing reduction of the vascular system ; it is, for instance, charac- 
teristic of E. arvense to have cones containing more axial xylem, both 
actually and relatively to their size, than those of E. palustre , though the 
steles of the former species are usually much wider than those of -the latter 5 
(Browne ( 2 ), pp. 259-60). 
The generalization contained in the former quotation should, I think, 
be modified by the intercalation after the words ‘ at the neighbourhood of 
the node 5 of the following passage : ‘ or with the reduction in number 
of the members of the whorls.’ This qualification is necessary to cover 
cases, like that of E. debile , in which the whorls of the cone consist of few 
members, so that though the diameter of the stele is actually small the 
narrow internodal strands are separated by relatively wide parenchymatous 
meshes. We naturally find that in the species in which the meshes are wide 
compared with the internodal strands (E. maximum , E. limosum , E. varie- 
gatum , and E. debile ) they are less regularly closed than in the species with 
narrow meshes relatively to the strands ( E . arvense and E. palustre ), since 
the junction of a strand with its neighbour or neighbours can here be 
effected by the differentiation of comparatively little xylem additional 
to that present in the internode. E. sylvaticum and E. giganteum occupy 
a middle position between the types with relatively wide and relatively 
narrow meshes. E. hyemale , with its wide stele and narrow internodal 
strands separated by wide tracts of parenchyma, is an exception to this 
generalization ; for it shows a large number of meshes, many of them of the 
first or low orders. In the cone of this species the difficulty of anastomosis 
between widely separated narrow strands has been, to a large extent, met by 
the oblique course of the tracheides on each side of the base of the meshes. 
Speaking phylogenetically, we should say that in spite of the reduction of 
the xylem the connexion of the nodes of the vascular strands has been in 
many cases maintained by the oblique course of the halves of the forking 
strands. The divergence of these branches of strands frequently results in 
their fusion with the neighbouring fork of a strand (as in the vegetative axis 
of Eqziisetum generally), or, if the neighbouring strand be unforked, with the 
latter itself. In either case this causes the closure of a mesh slightly above 
the level of departure of the trace. 
