24 
Isabel Browne. 
can only be very briefly touched upon here. After Qu^va’s 
researches there seems every reason to regard it as primary (15, 
p. 21). In that case it may represent a fresh development of wood, 
or, as seems to me more likely, it may be a modification of the 
ordinary primary centrifugal xylem of the axis. But whether it be 
regarded as a modification or a fresh development, it is difficult not 
to connect its retention during reduction of vascular tissues, or its 
development, with the presence of numerous appendages at the 
node. In such British species as I have examined, and in the 
small cone-hearing branches of E. giganteum , I have never seen 
any indication of the separate existence of the lateral strands inside 
and over the ring of reticulate nodal tracheides, which Mr. Gwynne- 
Vaughan claims to have observed in the nodes of E. giganteum. 
But such persistence of the lateral bands would not be incompatible 
with the phylogenetic unity of the Equisetaceous bundle. The fact 
that such a large number of appendages requiring vascular supply 
are initiated, even though they are not always fully developed, at 
most of the nodes of the vegetative axis of Equisetum , in spite of 
the relatively poor development of xylem in the internodes of these 
axes, has hardly received the attention that it deserves. 1 As 
regards the usual position of the lateral metaxylem of the internode 
it may be pointed out that the vascular supply of the branches is 
inserted on the ring of xylem vertically above the point at which 
two adjacent groups of metaxylem, each belonging to a different 
bundle, have fused. It seems not unlikely that the position of the 
persistent metaxylem at the sides of the bundle is correlated with 
the need for emission of a vascular supply to the branches, each of 
which lies midway between two leaf-traces. In this connection it is 
significant that the tendency of the lateral metaxylem to reduction— 
a tendency sometimes leading to complete abortion—is relatively 
pronounced in the rhizome, where comparatively few branches 
develop. Again, in E. limosum and sometimes in E. giganteum , 
the internode below the annulus seems to contain little or no 
metaxylem. It is possible that this observation should be brought 
into relation with the fact that at the reduced node of which the 
annulus marks the position (Browne, 2 and 3), no lateral branches 
are normally developed or even initiated. In other words, if, as 
seems to be established, the annulus is a reduced leaf-sheath, the 
reduced node at which it is inserted probably lost the capacity to 
1 The frequent presence, in unbranched or poorly branched axes, of 
numerous dormant branches, provided with a fully developed vascular supply, 
seems to show that reduction, or loss, of branching is a recent character 
in the phylogeny of the genus. 
