WorsdelL — The Structure and Origin of the Cycadaceae . 135 
possessing a thick arc-shaped mass of secondary xylem, passes off ; it soon 
divides into a pair of strands within the pericycle : these lose their 
secondary tissues, and, on passing into the leaf, again fuse into a single strand. 
Certain abnormal structures in the cylinder were also observed. Occasion- 
ally a cambium (which, in some cases, may form a continuous arc stretching 
from the normal cambium of one side of the bundle, round the inner side 
of the centripetal xylem, and joining the normal cambium of the opposite 
side) arises on the inner (ventral) side of the primary strand, there forming 
xylem and phloem with inverted orientation. Sometimes this abnormal 
cambium forms parenchymatous tissue only. 
The foliage borne by Lyginoden dr on-stems was that termed, before its 
real relationships were established, Sphenopteris Hoeninghausii ; it was very 
Fern-like, with finely subdivided segments. The petioles bearing these 
laminae bore the name of Rachiopteris aspera . A transverse section of this 
latter revealed a single large bundle of mesarch structure, with three or more 
protoxylem-groups situated near the periphery of the xylem ; the whole 
was completely surrounded by phloem. The outer part of the cortex 
exhibited a Dictyoxylon-structure. 
In Heterangium there is a single large solid stele in the stem, the 
structure of which is precisely that of one of the separate steles of the 
stem of Medullosa anglica , &c., but the whole is on a larger scale. 
The characters of the cortex, leaf-traces, and petiolar bundles are similar to 
those in Lygviodendron. The foliage belonging to the stem is that known 
as Sphenopteris elegans . 
The Origin of Cycadean Structures. 
1. Origin of Axial Structures . 
Most botanists are agreed that modern Cycads had their origin 
somewhere in the plexus of the Carboniferous or Permian Pteridosperms 
(Cycadofilices). If this is the case some indication of that origin should 
be discovered in the anatomical structure of the caulome or shoot. But 
opinion is divided as to which of the Pteridospermic types of stem-structure 
that of modern Cycads has been derived from. There is, on the one hand, 
the view which regards the central cylinder of the Cycad as derived from 
that of Lyginodendron or Heterangium , while the ‘ extrafascicular ’ cylinders 
of certain Cycads are held to be peculiar to them and of no ulterior 
phylogenetic importance or value. On the other hand, there is the 
view which regards the Cycadean cylinder as derived from that of such 
a form as Medullosa porosa . This latter is the view I shall adopt in this 
paper. 
Now, in the first place, I will start out with the primary proposition that 
if ancestral characters are present at all in the vascular tissue they must be 
