ii6 CELLULAR CONSTRUCTION [CH. 



Hymenophyllaceae into an order, which he styled the Bryo-pterideae, and 

 placed them between the Mosses and Ferns. Later writers at one time 

 shared this view of a Bryophyte affinity with the Hymenophyllaceae, basing 

 it upon the filamentous gametophyte. It was held that we are able even 

 now to follow, at least in part, the phylogenetic development of the sexual 

 generation from the Bryophyta to the Pteridophyta. But it is a palpable 

 departure from the usual methods of Natural Classification to give decisive 

 weight to vegetative characters while the propagative organs, and especially 

 the spore-producing parts, are so divergent as between the Mosses and 

 Liverworts and any Hymenophyllaceous Fern. It is now recognised that 

 similarities in respect of the filmy character of the leaf, the filamentous 

 prothallus, projecting sexual organs, and definite single initials, may all of 

 them find a ready explanation as results of parallel hygrophytic adaptation 

 in races phyletically quite distinct. 



Those who advanced these comparisons did not take the Palaeontological 

 evidence sufficiently into account. Already Campbell had argued compara- 

 tively in favour of the Eusporangiate Ferns being relatively primitive (96). 

 But over and above the difficulty of comparison there loomed large the 

 impossibility of harmonising a belief in the Leptosporangiate Ferns as 

 primitive with the growing knowledge of the fossils. The dearth of evidence 

 even of the existence of true Leptosporangiates comparable to those of the 

 present day in Palaeozoic times was pointed out. At the same time the 

 existence of numerous fossils then believed to be referable to a Marat- 

 tiaceous affinity was held to indicate a priority of the Eusporangiate type 

 (102). The comparative study of the development of the vegetative organs 

 and of the sporangium had meanwhile been actively pursued. On the basis 

 of such facts it came to be held as probable that the more delicate structure 

 seen in the Leptosporangiate Ferns was not primitive, but that it resulted 

 from progressive specialisation. The ground was thus open for recognising 

 the Eusporangiate type, whether of Ferns or of other Pteridophytes, as 

 of prior existence. Certain Pecopterids with Marattiaceous sori, the Botryo- 

 pterideae, and certain forms allied to the most primitive of living Lepto- 

 sporangiate Ferns, were recognised as the representative homosporous 

 Filicales of the Palaeozoic Period. The early existence of homosporous 

 Ferns, which evolutionary theory would suggest or even demand, is shown 

 to be beyond reasonable doubt. But the ferns which existed in the Primary 

 Rocks appear to have been mostly if not exclusively Eusporangiate. 



If this be so then we shall contemplate for the Filicales a progression 

 from a more complex to a simpler but more exact construction, such as is 

 illustrated by comparison of living Eusporangiate with Leptosporangiate 

 Ferns. That such a progression does not stand alone for Ferns is shown by 

 the Lycopodiales, though it is less completely carried out in them. In the 



