668 Browne . — Contributions to our Knowledge of the 
traces of a whorl we should have an internodal ring of wood ; occasionally, 
when all the strands become united at the node — and this of course 
involves the closure of all the parenchymatous meshes of the internode 
below — we get a nodal ring of xylem and bast (PI. LXIV, Fig. i). 1 Such a 
local ring is present twice in Cone A, twice in Cone B, and once in Cone E ; 
it was not met with in Cones C, D, F, G, and H, but may have occurred 
in them, since the series of sections did not include the whole of any of 
these cones. To judge from isolated hand -sections a ring of xylem occurs 
locally at the nodes of many cones of E. arvense. Milde in 1865 stated 
that at the level of the fertile whorl the xylem-strands become united 
into a ring. This is too sweeping a generalization, and curiously enough 
Milde figures this condition diagram matically in E . limosum , a species in 
which I have never seen even a near approach to a ring (Milde, i, p. 162, 
PI. XVI, Fig. 18). 
The cases in which a parenchymatous mesh arises some distance up 
the internode are clearly intermediate between the forms in which the mesh 
arises at the usual small distance above the node and those rare cases in 
which no parenchymatous mesh occurs above a trace. The parenchymatous 
meshes of E. arvense do not, as a rule, attain any considerable width ; they 
are frequently narrower, even in their widest part, which is usually in the 
middle of the internode, than the strands of xylem between which they 
pursue their course. As regards their vertical extent, by far the greater 
number are closed at or before the next node, and these meshes I shall, for 
convenience’ sake, term meshes of the first order. Other parenchymatous 
meshes that extend into two, three, or more internodes will be termed 
meshes of the second, third, and higher orders. In Cone A there were 
90 meshes originating above fertile whorls and closed before we reach the 
apex of the cone; of these 81 were of the first, 8 of the second, and 1 of 
the third order. Within the limits of Cone B 37 meshes originated and 
were closed : 33 of the first, 3 of the second, and 1 of the third order. As 
only seven whorls of Cone C were cut serially many of the meshes were 
still unclosed when the series ended ; if the figures of all the meshes closed 
within the series of sections were given, the proportion of meshes of the 
first order would appear to be larger than it is in nature ; for, of the meshes 
arising above the upper whorls, more of those of the first order would have 
space, so to speak, to be closed before the end of the series. But all meshes 
of all orders that arise above the lowest four whorls are closed before the 
end of the series of sections ; by considering these only we may see the 
relative proportion of meshes of different orders in this Cone C. On this 
basis we find that out of 22 meshes 10 are of the first, 9 of the second, 
1 Owing to the fact that the sporangiophores are inserted at slightly different levels, all the 
meshes may be closed at a node and yet no single transverse section will show a complete ring of 
xylem. This tendency is increased by the fact that occasionally meshes are not closed before, but 
at, or even just after, the departure of the trace. 
