124 MARRATTIACEAE [CH. 



marginal sorus, of few sporangia seated round a central point, as in Etapteris 

 or Corynepteris. The nearest to this of living genera is Angiopteris, while 

 stages of secondary spread of the sorus along the veins would lead to 

 the state seen in Macroglossum and Archajigioptei'is. The Rhaetic fossil 

 Danaeopsis shows that the spread from the margin all the way to the midrib 

 was arrived at early when the sporangia were still separate. A like spread 

 from the margin carried out in sori with the sporangia fused into synangia 

 may have given a parallel sequence, as seen in Marattia, Protomarattia, 

 and Danaea. The sole living representative of a third line is seen in 

 Christensenia, which is clearly specialised for growth under forest shade, its 

 broad pinnae showing advanced reticulation. The mode of origin of the 

 scattered rosette-like sori is suggested by half-fertile pinnae of Danaea. 

 Thus, however nearly they may resemble the sori of Ptychocarpus, it seems 

 probable that the individual synangia of Christensenia are really only 

 fragments of an originally intra-marginal sorus. 



The characters of the conducting system do not follow parallel with those 

 of the sorus. But in this Size has doubtless been one determining factor. 

 Angiopteris with a relatively primitive sorus grows into a large plant, and has 

 the most elaborate vascular system, while Danaea with an advanced sorus is 

 relatively small, and has a less complex stock. Nor is even the presence of a 

 suspensor a secure guide. Danaea, Macroglossum, and Angiopteris, in which 

 a suspensor has been seen, occupy divergent positions according to soral 

 characters. All the living genera except Christensenia have dichotomous and 

 open venation. But the fossil Danaeopsis has vein-fusions. Among living forms 

 the reticulation in Christensenia clearly supports the derivative character of its 

 sorus. The simple dermal appendages of Angiopteris and Marattia mark 

 them off as primitive in this feature, as against the scales of Protomarattia, 

 ArcJiangiopteris, Danaea, and Christensenia which, being held as derivative, 

 support the conclusion that the sori of these genera are derivative also. From 

 such comparative considerations a tentative phyletic grouping of the living 

 Marattiaceae may be indicated as follows: 



MARATTIACEAEi 



I. Sporangia separate. 



ii) Angiopteris {Ylo'iim-A.xwi, lyc^G) ... ... 62 species. 



(2) Macroglossum (Campbell, 19 14) ... ... i species. 



(3) A rchangiopteris {Chnst, 1899) ••• ••• ^ species. 



(4) Danaeopsis — a fossil (Halle, 1921). 



1 A protest should be entered against names that assume too much, such as Airhangiopteris and 

 Protomarattia. The assumption has been that these Ferns are primitive as compared respectively 

 with Angiopteris and Marattia. But both bear sori more extended than those of the older genera, 

 and both bear scales. If, as argued in the text above, the original sorus was rosette-like and marginal, 

 then their names are misleading. That their sori are highly derivative is the conclusion clearly 

 indicated by comparison with the early fossils, such as the Coenopteridaceae. 



