278 SUMMARY OF RESULTS [CH. 



sori has been lost by extension of fertility over an extended leaf-area, giving 

 the Acrostichoid character (pp. 1 31-13 5). 



It thus appears that, following the Superficial Series upwards, the sorus 

 was first simple, radial, and exposed : it then became gradate, and acquired 

 a basal indusium: then followed the mixed Condition, zygomorphic in re- 

 lation to the leaf-margin, with first a reniform, then an orbicular indusium: 

 next followed the ex-indusiate or Polypodioid state: and finally the 

 Acrostichoid. The parallel between this and what has been seen in the 

 Marginal Series is obvious though not exact. It is closest between Dryoptcris 

 on the one hand, and such marginal derivatives as Nephrolepis and some 

 species of Lindsaya, though detailed comparison in other features would 

 negative a real affinity. On the other hand, there can be no doubt that the 

 Polypodioid state of Hypolepis and the Acrostichoid state of A. aureum 

 were attained by lines phyletically distinct from those ending in Phegopteris 

 or Polybotrya. These are examples of that homoplastic likeness that 

 frequently recurs in distinct phyla of Ferns. 



A collateral step in the progressive series of changes that started from 

 the non-indusial radial sorus is seen in the Asplenieae, which may be held 

 as derivative from the Nephrodioid type of sorus (Chapter XLIl). In 

 Didymochlacua the sorus appears as a horse-shoe, elongated so as to follow 

 the course of the underlying vein. The primary receptacle corresponds to 

 the point of strongest curvature, here directed towards the leaf-margin. If the 

 fertility at that region were lost, but retained on one side or the other, or on 

 both of the elongated shanks of the curve, the result would be the state seen 

 in Asplenimn or Diplazium. Intermediate conditions are frequent in A thyrinm, 

 but they are particularly well seen in Diplaziinn lancemn (Fig. 672). Finally, 

 the indusium itself may be abortive, giving the Pseudo-Gymnogrammoid 

 state of CeteracJi or of Pleurosoriis. Possibly the Acrostichoid state of 

 RJiipidopteris may have had its origin from such P'erns as these, by spread 

 (jf fertility over the reduced surface of the fertile blade. 



Starting afresh from the Onocleoid Ferns, Matteiiccia and Onoclca, which 

 are best separated as another distinct line of derivatives from the Cyatheoid 

 type, a phylum may be traced parallel to that of the Nephrodioids, but still 

 distinct (Chapter XLlIl). A recently described species, Matteuccia intermedia 

 C. Chr., serves as the starting point. It is a coarse-growing Fern with massive 

 dictyostelic stock, binary leaf-trace, very scaly surfaces, leaves dimorphic, 

 open venation, and superficial non-indusiate sori ranged in regular rows, one 

 on either side of the mid-rib. The sori are gradate, and the sporangia 

 relatively large. These characters point to a relation with Alsophila and 

 Loplwsoria rather than with Woodsia. It stands alone among the Onocleoids 

 in being non-indusiate: protection is afforded by the strongly reflexed leaf- 

 margin. It is but a step from this to linkage of the lines of sori by a vascular 



