— ———— 
COMPARATIVE ANATOMY OF THE CYCADACEA. 115 
very far indeed in the modern Cycadean direction, by adopting a collateral structure in 
its extrafascicular strands *. My fig. 5 (Pl. 16), a diagrammatic representation of the 
structure in Cycas Seemanni, Al. Br., shows portions of three rings of vascular tissue 
outside the central stele. One of the segments of the innermost ring is almost concentric 
in structure, the inner inversely-orientated portion being in intimate connection with the 
outer. Attached to the second ring, on its inner or ventral side, are seen about six (and 
there are in reality smaller ones present, which are not drawn in the figure) tiny 
inverted strands; these represent the last sporadic remnants of the inverted portion of 
the well-developed ring immediately outside them. As the cambial activity has become 
(for obvious economical purposes) at length almost exclusively confined to this latter 
part of the ring, the inner inverted portion is naturally tending to entirely disappear 
(as it already has done throughout the greater part of the axis), leaving these mere 
rudiments to testify to its former normal presence in the ancestors of the plant. 
In the third ring a well-developed inverted strand occurs attached to an isolated 
normal segment of the ring. 
Fig. 6 (Pl. 15) represents a segment of the vascular tissue of the stem in which the first 
extrafascicular ring (7?) consists of widely-separated concentric strands, exhibiting the 
same structure as, and perfectly homologous with, the concentric strands forming 
the outermost ring of the stem in most species of Cycas. 
In the present case the ring of concentric strands is intercalated between the central 
cylinder and a normal collaterally-constructed extrafascicular ring, the first instance of 
the kind I have ever seen. 
The structure described by Gregg t in the root of this same plant is merely a variant, 
though a somewhat freakish one, on that above detailed by myself. Neither his 
structure nor mine is to be regarded as, in essentials, an abnormality, much less a freak 
or sport: both are but vestiges of the former norma! structure of the ancestors of the 
plant. The strand in Gregg's fig. 1, formed by the cambium och), is simply the 
inverted ventral portion of the part, opposed to this latter, of the first extrafascicular 
ring formed by the cambium a.cb?; these two are homologous with the inverted strand 
is! of my fig. 5 together with its opposed outer strand. The variation appearing in 
Grege’s structure consists in the curious backward folding of the cambial tissue nc on 
either side of the central stele so as to become continuous with that of the cambium a.c! 
of the inverted strand of each side, instead of, as is the normal case of my figure, curving 
round and enclosing the protoxylem at each end of the stele. This mode of development: 
-gives rise to the false appearance of two distinct concentric strands possessing a central 
phloem and an external xylem (see Gregg's fig. 3). But no importance whatever is to be 
attached to this phenomenon. If the cambium och) had become continuous with the 
cambium «.cb? of the first normal extrafascicular ring, instead of with the cambium of 
the central stele, the structure would have been attained such as Gregg found in another 
root and represented in his fig. 4, that, viz., of two concentric steles of normal structure 
outside the central stele. This involves the breaking up of the first collateral 
* Weber & Sterzel, loc. cit. pl. viii. fig. 1. 
+ “Anomalous Thickening in the Roots of Cycus Seemanni," Ann. Bot. vol. i, 1887. 
