XX] COMPARISON 123 



characteristic of all the later and derivative Leptosporangiatae, where the 

 embryo bursts through the lower surface of the prothallus (Fig. 413). 



Comparative Treatment 



The relatively primitive character of the Marattiaceae follows naturally 

 from the features of form and structure above detailed, as well as from the 

 special comparisons with the Ophioglossaceae, and ultimately with the 

 Coenopteridaceae. The fossil record proves the great antiquity of the type, 

 and the wonder is that the features of the sorus have been preserved through 

 the ages with such precision. Nothing in Palaeobotany is more striking than 

 a detailed comparison of the Palaeozoic sori o{ Ptychocarpus or o{ Scolecopteris 

 with those of the living Christetisenia and Marattia. 



Nevertheless, it is difficult to place the Marattiaceae with certainty in 

 relation to other F'erns. There appears to be no other family of living Ferns 

 that could be directly referred to them as its ancestral source. They appear 

 as survivals to the present day of a Palaeozoic and Mesozoic stock that 

 was not readily adaptable except in minor features. But as regards the 

 structure of the sorus certain analogies do exist with the Gleicheniaceae and 

 Matoniaceae: these receive no support, however, from the vegetative system 

 of those Ferns. On the other hand, in Corynepteris, which Kidston regarded 

 as belonging undoubtedy to the Zygopterideae, we already see a sorus of 

 a simple Marattiaceous type: and it has been shown how the various forms 

 of sorus of the living Marattiaceae might readily have been derived from 

 such a type, first by a slide of the sorus from the margin of the narrow 

 pinnule to its broadening surface; and then as the surface became further 

 expanded (probably with progressive webbing as the venation clearly suggests) 

 the sorus may be held to have extended along the veins. Further, in certain 

 types the sporangia, originally distinct from one another, became fused to 

 form synangia, with as a natural consequence the loss of the annulus, which 

 could no longer be mechanically effective. All this appears to be biologically 

 probable, and morphologically reasonable. It may be the actual truth, but 

 there is no clear evidence that it actually is so; though the presence of 

 scales in place of hairs in Danaea and Christensenia, Protomarattia and 

 ArcJiangiopteris, is a significant parallel, suggestive of advance. 



The problem of the phyletic position and arrangement of the living 

 members of the Family among themselves presents hardly less formidable 

 difficulties than appear in placing the Family as a whole. They arise partly 

 from the absence of exact parallelism of the various characters, partly from 

 differences of opinion as to their relative value. Priority should be given to 

 the characters of the sorus, its position, the number of its sporangia, and the 

 degree of its extension along the veins: also to the relation of the sporangia 

 to one another. A prototype may probably be held to have possessed a 



