124 DRYOPTEROID FERNS [ch. 



origin of the later sporangia. The whole receptacle is at an early stage 

 completely covered in by the arched indusium, which is massively thickened 

 at its base. But even the median section does not give a full conception of 

 the form of the receptacle as a whole. It is horse-shoe shaped and, as Davie 

 has pointed out, the origin of this is prefigured by the sori oi Peraneina and 

 Heinitelia. The lop-sidedness may first be traced in the semi-lunar indusium 

 of the latter genus, where it is attached on the side away from the leaf-margin. 

 This encourages a tilt of the previously radial receptacle over towards the 

 marginal side, which is free. The tilt is accentuated in Peraneina (Figs. 653, 

 654). Development of the sporangia will consequently be restricted to the 

 free side by the tilt of the receptacle, and this tends to diminish the fertile 

 area: but the loss is compensated by an extension of the receptacle right 

 and left into pouches protected by the indusium, as is seen in Davie's figure 

 {I.e. 1916, PI. Ill, Fig. 6). Such conditions already indicated in Peraneina diVe 

 more fully realised in Diyopteris, where the receptacle has accordingly 

 assumed a definitely horse-shoe shape, with the heel directed away from 

 the leaf-margin, and almost encircling the indusial stalk. By such steps 

 illustrated in closely related Ferns, the origin of the lop-sided sorus of 

 Dryopteris, with its kidney-shaped indusium and semi-lunar receptacle, can 

 be traced from a radial type such as is seen in the superficial series, starting 

 from the Gleicheniaceae. Heinitelia, Diacalpe and Peraneina, but especially 

 the last of these, are the synthetic types that link the sori of the superficial 

 series together in their increasing lop-sidedness. As Dr Davie said, between 

 them Peraneina and Diacalpe appear to unite all the intermediate soral steps 

 between the Cyatheaceae and the Aspidieae. 



It has been seen that with transient traces of a gradate sequence the sorus of Dfyopteris 

 is of mixed type, which is as a rule associated with a vertical annulus interrupted at the 

 insertion of the stalk. Examples have been seen in several distinct phyla of the gradual 

 steps of obliteration of the obliquity of the an^nulus. First the induration is seen to be in- 

 complete, though the oblique ring of cells is still there {Thyrsopteris, Vol. 11, Fig. 529; 

 Loxsoma, Fig. 521). Then the sequence of cells is almost \r\terrupied{Denns/aed/la, Fig. 539; 

 Metaxya, Fig. 553; Dipteris, Fig. 576). Where the interruption is complete, as it is in most 

 advanced Leptosporangiate Ferns, and as it is seen to be in the Male Shield Fern, it might 

 be assumed that all structural trace of the original obliquity was lost. But it is not so : and 

 the very accurate drawings of Miiller, made for Kny's IVandtafeln, show the point (Fig. 657). 

 The distinctness of the two faces of the sporangium has already been discussed in Vol. i, 

 p. 254 (Figs. 250-252). The distal face is that which lies encircled by the ring where it is 

 complete: the proximal represents the part of the sporangial wall that lies between the 

 ring and the stalk in primitive types. The two drawings, 4 a and 4 b, represent two opposed 

 sides of the biconvex head. Two rows of cells of the stalk run up to that shown in 4 a, only 

 one row runs up to that in 4 b. The former represents the distal or peripheral face of the 

 sporangium, the latter the proximal or central. Moreover, the two sides of the sporangium 

 are not of exactly the same curve or extent. Fig. 657, 3 shows this correctly, the distal side 

 to the left having fewer tabular cells and a slightly different curve from the proximal, which 



