688 CONCLUSION 
the modern Lycopods, but it is quite different in origin from medullation. 
Intrusive bands of phloem invade somewhat irregularly the central xylem, 
giving it sometimes the form of a fluted column, or of a series of 
plates connected at intervals, or of a continuous xylem-sponge (Fig. 171, 
p- 329). Such conditions, which are characteristic of modern Lycopods, 
are probably secondary derivatives of the simple protostele, since they are 
absent in the early fossils, as well as in the early condition of the plants 
that show them when adult. 
A somewhat similar intrusion of tissues from without leads, in many 
Ferns, to the condition which is described as the solenostelic. But here 
it is regularly at the point just above the exit of the foliar strands from the 
stele that the intrusive tissues enter; it thus comes about that phloem 
and endodermis and ground parenchyma come to occupy continuously 
the centre of the stele, which accordingly takes the form of a hollow 
tube, with openings opposite each leaf-base (Figs. 95, 97, 100). This 
formation of a solenostele has probably occurred along more than one 
phyletic line, and it lies at the base of those complex types of dictyostelic 
structure of the stem seen in Leptosporangiate Ferns. These follow upon 
the overlapping of the foliar gaps, which results in dictyostely formerly 
described as a polystelic state (p. 190). A similar condition in some 
species of Selaginella, though phyletically quite distinct, shows interesting 
analogies; but its origin appears to be in relation to the departure of 
supplies to axes, not to leaves; these are, however, referable also by origin 
to a primitive monostelic structure. 
Still further complications occur in certain Ferns which are associated 
with the formation of accessory vascular tracts ; these arise in relation to the 
foliar gaps as described on pp. 568, 600, and lead to a doubling or even 
trebling of the solenostele (Figs. 319, 342), or accessory strands may arise 
in pith or cortex (Fig. 339). The condition of the modern Marattiaceae 
and of the fossil Psaronius may also be mentioned as extreme cases of 
complexity of vascular structure based probably on a scheme allied to 
those above noted (p. 525). Into these details it is not necessary to enter 
further here, they concern us chiefly as illustrating some of the extreme 
methods of amplification of the vascular system seen in the axes of 
Pteridophytes. 
In some degree parallel with this progressive dilatation and disintegration 
of the stele goes also the disintegration of the foliar trace. In all the 
smaller-leaved, and in many of the larger-leaved forms, the leaf-trace consists 
of a single strand; in the Lycopodiales this is uniformly so, with exception 
of certain Sgz/arias described by Kidston! It is a single strand also in 
Lsoetes,? and in the L£guisetales. In the Sphenophyllales and Ophio- 
glossales (except § Ophioderma, and perhaps § Chetroglossa), the leaf-trace 
comes off always as a single strand, but branches frequently while still 
1 Proc. Roy. Soc., Edin., vol. xxvii., part iii., p. 203. 
2 Studies, ii., Fig. 105. 
