THE PTERIDOPHYTA : FILICALES, THE FERNS 561 



segment of the stele, it breaks right through the ring, forming a gap above 

 the point where the trace is given off. Through these gaps pith and cortex 

 become continuous, and the gaps may become Hned not only with phloem 

 but also by the pericycle and endodermis. As an example of this we may 

 cite the case o{ Loxsoma cunninghamii. 



In relation to the leaf gaps we find that in some examples the phloem and 

 pericycle spread round the xylem, so that the original xylem ring is bounded 

 both inside and outside by successive layers of phloem, pericycle and endo- 

 dermis. Such a condition is termed an amphiphloic solenostele. We see 

 examples of this in Dennstaedtia and Dkksonia (Figs. 569 and 570) among the 

 present-day genera, and also in Ahophila, though in this genus the solenostele 

 soon becomes split up into a more complex form of dictyostele such as is 

 referred to below. 



In such a type of stelar anatomy the central tissue is isolated from the 

 cortex by a continuous layer of endodermal cells, except in the leaf gaps at 

 the nodes. Where the stem is short and the leaves are set closely together 

 the leaf gaps overlap, so that several appear in a section taken at any level. 

 Thus the tubular stele is reduced to a network, as in Dryopteris, and this is 

 called a dictyostele (Fig. 564, j). A further step is the formation of open- 

 ings, or perforations, which produce a dictyostelic condition even where 

 the leaves are separated by long internodes, as in Pteridiwn. The stele of 

 such Ferns consists of a circular column of vascular tissue resembling a tube 

 of wire netting, the holes in the network being formed partly by the per- 

 forations in the vascular tissue of the stele and partly by the leaf gaps. This 

 form of the stele is obviously much more favourable to physiological inter- 

 changes between the tissues than is the continuous solenostelic tube. It 

 may be regarded as a means of increasing the total surface of the vascular 

 tissue, analogous to the subdivision of the leaf into filaments in submerged 

 water plants (Fig. 564, k). 



Like the stele, the leaf trace has a history of progressive elaboration. 

 In the most ancient types it is a simple mesarch strand of xylem and phloem, 

 with an endodermis. This very early became a flattened band, which adopted 

 a curved outline, like a U, the opening being towards the axis, that is, to 

 the upper side of the petiole. Mechanically such a form is favourable to 

 rigidity, and it became the fundamental Fern type of leaf trace. In soleno- 

 stelic Ferns it usually remains a single band, but in dictyostelic Ferns it is 

 divided into meristeles, sometimes arranged in a simple U-form, sometimes, 

 as in Pteridium and the larger Ferns generally, arranged in a complex pattern 

 of many meristeles, which can only with difficulty be related to the fundamental 

 horseshoe type (Fig. 571). 



It has been stated by one authority (Sinnott) that the leaf trace in the 

 Pteridophyta appears to have followed a course of evolution which was 

 independent of the evolution of the stem stele and has had little influence 

 upon it. 



The final elaboration of the stelar anatomy in the Filicales consists in the 

 development of a number of separate steles one inside the other, a condition 



19 



