226 DIPTEROID FERNS [CH. 



beyond them (Fig. 732, 2). Sections show that here the whole vascular 

 development lies in one plane : the condition is essentially the same as has 

 been seen in Neocheiropteris, though the details are slightly different (com- 

 pare Fig. 730). A still more interesting structure for comparison with the 

 diplodesmic state of Cheiropleuria or Christopteris is seen in PL scJiraderi: 

 for here the vascular development is not restricted to a single plane. The 

 primary venation of the leaf extends in a plane above, and parallel to the 

 secondary network, which supplies the receptacle (Fig. 732, 3). It is in 

 fact diplodesmic, after the manner which has been seen initiated in Cheiro- 

 pleuria^ and amplified to an elaborate double system in Platyceriwn and 

 • Christopteris. 



These observations on species of Pleopeliis indicate states of increased 

 elaboration of sori, probably in the first instance punctiform. Von Goebel 

 raises the question of soral fusion being sometimes involved, and this cannot 

 be excluded. But a lateral spread has clearly been a factor, leading even 

 towards an Acrostichoid state, wherever the insertion of sporangia extends 

 beyond the area of vascular supply. Thus in these Ferns, with sori unrestricted 

 by indusial protections, there has been extension of the soral area, together 

 with increased elaboration of the receptacular conducting system. This has 

 progressed in Pleopeltis along lines generally comparable with those seen in 

 Dipteroid Ferns, but with some independence of detail. When this is put in 

 relation to the similarity that exists in other respects, such as form, venation, 

 and habitat, the probability appears strong that Pleopeltis is also a derivative 

 from the Matonioid-Dipteroid phylum. How much further such comparisons 

 may extend must be left for the present open. Yoss\h\y Phlebodium may also 

 prove to be of Dipteroid affinity, and the suggestion may also be extended 

 to Niphobolus. Certain observations of Giesenhagen point in that direction 

 (Die Gattung Niphobolus, Jena, 1901, p. 61, Fig. ii,B). Though such questions 

 will require much critical and detailed research before they can be advanced 

 beyond the region of suggestion, it seems not improbable that a considerable 

 part of the collective genus " Polypodiuni" will ultimately have to be allocated 

 to a Matonioid-Dipteroid origin. 



It has been seen how in Dipteris and Cheiropleuria there is a relatively thick- 

 stalked sporangium : it has four cell-rows, while in Platyceritim and Christo- 

 pteris there are only three rows. The sporangia of "■ Polypodium^' produced in 

 large numbers in the crowded sori, illustrate a further step of simplification. 

 This is seen in Phlebodium aureum (Fig. 733). Compare also Vol. I, Fig. 240. 

 Here the sporangium is initiated as a narrow unicellular papilla, in which the 

 first oblique segmentation may often extend below the level of the epidermal 

 wall, or it may impinge on the lateral wall distinctly above that level. In the 

 former case the stalk at its base may be found to consist of more than one 

 cell : but in the latter the lower part of the stalk would be a simple cell-row, 



