Worsdell — The Structure and Origin of the Cycadaceae. 139 
in the interior of those strands being reduced to a minimum. I regard 
these primary centripetal tracheides (whose rudimentary nature is further 
betrayed by the half-hearted way in which they assume the stain) as 
of probably dual origin : I hold that for the most part they represent 
the central primary tracheides of the stele of M . anglica ; in this case 
we must imagine the protoxylem, centrifugal xylem and phloem of that 
part of the stele nearest the pith of the peduncle to have completely 
died away. Occasionally, however, these three tissues are present on 
the inner side of the bundle, while, in the wide area between them and 
the normal bundle, numerous ‘ centripetal ’ tracheides are seen. In such 
a case these inner tissues are well developed and not rudimentary in 
appearance ; they merely represent an incurved portion of a horse-shoe- 
shaped stele which has become detached and in- 
dependent (Fig. 5). Such a strand sometimes 
anastomoses with the ventral portion of the bundle 
to which it is opposite. In other cases, however, 
a group of centripetal tracheides is united, on its 
pith-side, to an extremely rudimentary phloem- 
strand, or this latter may be replaced by a rudi- 
mentary cambium. It seems possible, therefore } 
that in some cases the ‘ centripetal xylem ’ may 
represent the vestiges of the centrifugal xylem 
(primary or secondary) of the ventral part of the 
original and primitive stele, the protoxylem be- 
longing to which has become extinct. 
As regards the mode of ontogenetic origin 
of all or any of these structures, it has probably 
not the remotest connexion with their phylogenetic history. The chief point 
about this remarkable structure has hitherto been, apparently, missed. 
Scott, however, lays great stress on his belief that the ‘‘centripetal 5 
xylem, as found in the peduncle of Stangeria , Bowenia , and Ceratozamia , is 
the homologue and derivative of the centripetal primary xylem occurring 
in the bundles of the vegetative stem of Lyginodendron , Heterangium , 
Poroxylon , Calamopitys , &c. In this we entirely concur. But we differ as 
to the mode of origin of the vascular tissue in these fossil plants. Scott 
regards the solid monostele of the Heterangium- type as the primitive form 
in this group ; the Lyginodendron-type he holds to be derivative therefrom. 
According to his view the centripetal xylem of Lyginodendron represents 
the last peripheral remnant of the great mass of primary xylem which, 
along with parenchyma, occupies the entire internal area of the Heter- 
angium-stele. My view, on the contrary, embodies the belief that each 
primary xylem-group in Lyginodendron , along with its mass of secondary 
centrifugal xylem and phloem, is the one-sided remnant of an entire stele 
Fig. 5. Stangeria para- 
doxa : transverse section of 
a single bundle from vascular 
ring of peduncle, showing 
small strand on inner or pith- 
side of large strand, with 
groups of centripetal tra- 
cheides lying between the two 
(diagrammatic). 
