224 
A. G. Tansley. 
the internal strands of the stem entirely devoted to the petiolar 
vascular system, partly supplying the lateral elaborations of the 
trace, as in Pteris, and partly supplying the incurved free edges of 
trace as in Matonia. Traced downwards, the internal strands end 
blindly in the ground-tissue of the pith. Whether they represent a 
comparatively new accessory system, or whether they are to be 
regarded as a reduction from a more highly developed internal 
system in some ancestor of the Cyatheaceae, in which the external 
vascular ring was narrower than at present, we have scarcely the 
data to decide. In other species of Cyathea the relations 
are more complex. In the young plants of Alsopliila excelsa 
(Gwynne-Vaughan, ’03) the internal strands first appear at about 
the level of the tenth leaf in the form of internal thickenings 
of the xylem of the stele just below the point of origin of the two 
adaxial leaf-trace strands (representing the free edges of the trace- 
system). In the case of later nodes these thickenings become 
free during their downward course, and end blindly in the central 
ground-tissue. 
In addition to the medullary vascular strands certain accessory 
cortical strands exist in some species of Cyathea (Fig. 68) and 
Alsopliila, but the nature and origin of these is not understood. 
Ceratopteris thalictroides is an annual aquatic fern of isolated 
systematic position with a polycyclic dictyostelic adult stem evidently 
much reduced, and bearing crowded leaves. The vascular system 
of the young plant is a simple dictyostele, but the older plant 
appears to possess a perforated dictyostele, with a number of weak 
internal accessory strands, some of which end blindly below, and 
which appear to pass into the base of the petiole and there to fuse 
with the outer strands derived from the primary cylinder (Ford,’02). 
Whether the internal cauline strands have any direct connexion 
with the internal petiolar strands, of which there may apparently 
be a good many, is not clear from the available accounts, nor is it 
obvious whether there is or is not a connexion between the internal 
accessory strands and the outer cylinder of the stem. 
Marattiales. This very old group of ferns stands apart, in 
many respects, from all the rest, and although its members conform 
to Bower’s type of the Simplices, it may well be maintained on 
anatomical as well as on other grounds, that they are better kept quite 
apart from the Leptosporangiate families that we have hitherto 
been considering. From the purely anatomical point of view they 
exhibit the most complicated structure of any known types, the 
