XXI] COMPARISON 147 



suggestion of a primitive character of the Family. Nevertheless, the embryo 

 has the prone position, as against the erect habit of the Marattiaceous 

 embryo. 



Comparison 



Ever since the distinction between Eusporangiate and Leptosporangiate 

 Ferns was drawn by Goebel it has been growing more and more evident 

 that the Osmundaceae hold an intermediate position between those broadly 

 distinct types. It was shown relatively early that the general constitution 

 of the sporophyte, as revealed by the apical segmentation, placed the Family 

 in a middle position (Bower, 1884, 1885, 1891), and that the spore-output of 

 the relatively large sporangium was high- (Russow, 1872). The vascular 

 structure of Osmunda was held to be relatively primitive by Zenetti (1895). 

 But point was added to all such comparisons when Campbell definitely 

 stated that the Eusporangiate type was relatively primitive for Ferns, and 

 the Leptosporangiate derivative (1890), and still more when it was shown 

 by Palaeontological evidence that the former antedated the latter in 

 geological time (Bower, 1891). How important then becomes the position 

 of the Osmundaceae in the whole system of Ferns when viewed phyletically, 

 and how significant the fact that the Family dates back to the Palaeozoic 

 Period (Kidston and Gwynne-Vaughan, 1907-1914). The Permian fossils 

 referred to this Family have been described above. It is true that the 

 fructifications have not been found in relation to such early stems as 

 Zalesskya or Tlianinopteris, and so the final proof of their correct reference 

 to the Osmundaceae is wanting. But, as noted in the introductory chapter 

 (p. 6), numerous sporangia showing in section a structure closely resembling 

 those of living Osmundaceae existed in the Coal period, though possibly 

 they may not have corresponded in detail to the fully elaborated Osmunda- 

 ceous sporangia of the present day. The remarkable similarity existing 

 between the Permian fossils and living Osmundaceae as regards the whole 

 structure of the shoot (though the stelar state of the Palaeozoic fossils 

 is simpler than in the later types) leaves little doubt of the general correct- 

 ness of the reference. The Osmundaceous type may therefore be held to 

 date from Palaeozoic times. It is incumbent on those who withhold their 

 assent to produce evidence of the existence of Ferns having this construction 

 of the shoot, but bearing fructifications that indicate some other afifinity. 

 Till this is done their objection does not appear decisive, and the conclusion 

 may be accepted meanwhile as the probable truth. 



It is open to us to support this probability by the wealth of comparative 

 fact and argument which we now possess. It will be found that the evidence 

 from such sources is quite consistent. The importance of this question to 

 the general presentment of the phylesis of Ferns is so great that a full 



