XXI] COMPARISON 151 



phyllaceae and Gleicheniaceae, though the Mertensia-typQS, of the latter 

 may show an even larger output. The figures speak for themselves when 

 compared with those of the whole range of the Filicales. 



There remains, however, the outstanding question of the sporangia of those 

 early fossils which have been ranked with the Osmundaceae on anatomical 

 grounds. Their leaves and sporangia are mostly unknown as yet, so far as 

 actual connection with the stocks is concerned. But it is a significant fact 

 that a number of fertile leaves are known, from the Coal Period onwards, 

 bearing sporangia the structure of which is such as to accord generally with 

 that of the modern Osmundaceae. Examples are seen in the Carboniferous 

 fossil Kidstonia (Vol. I, Fig. 249, E), described by Zeiller {Bull. Soc. bot. de 

 France, xliv, p. 195, 1897), and in Bozveria from the Barnsley Coal, fully 

 investigated and figured by Kidston (see Fig. 310, Introduction, p. 7)\ 

 The former has a massive annulus composed of several rows of cells, and is 

 clearly a relatively primitive type of sporangium. In Boweria the sporangia 

 are isolated and marginal, being seated on the ends of the veinlets of the 

 pinnule-segments. The annulus consists of two slightly irregular rows of 

 cells, which pass as a band across the apex of the sporangium, and 

 extend a very short distance down the sides, Kidston compares these 

 sporangia with those styled Pteridotheca by Scott {Progressus, Vol. i, 

 p. 183, 1906: also Studies, 3rd. Edn. Vol. i, p. 265): and he remarks, "The 

 plants included in the genus may possibly be referable to Ferns." The 

 alternative would be that they are Pteridosperms. If the two genera last 

 named were actually Ferns, their sporangia and sporophylls would supply 

 a prototype from which by condensation of the pinnule-segments, and 

 simplification of the annulus, the type seen in the modern Osmundaceae 

 would readily follow. Indeed Zeiller has concluded that "the sporangia of 

 Kidstonia heracleensis show the nearest afifinity to those of the Osmundaceae" : 

 but they are not without analogies with those of Senftcjibergia and Lygodiuni. 

 Thus Kidstonia would be grouped with the Osmundaceae, but at the outer 

 fringe of the family, forming a link between them and the Schizaeaceae. 

 A careful re-examination of the sporangial structure in the living Osmun- 

 daceae would help in the elucidation of this question, which at the moment 

 must be left as an open one. 



It remains to consider the probable relations of the living genera. The 

 first question relates to the validity of Leptopteris as a substantive genus. Its 

 filmy habit appears distinctive, though in other features the species fall 

 clearly within the genus Todea. A filmy habit is not held elsewhere as 

 a sufficient basis for generic distinction. The filmy Danaeas and Aspleniums 

 are not separated from those substantive genera. Following the practice 



1 " Veget. houil. Hainaut Beige," Mdin. Miis. Koy. d'Hist. iiat. de Belgiqiie, iv, p. 51 ; and Fossil 

 Plants of the Carboniferous Rocks of Great Britain, iv, p. 291, Plates LXXI, LXXII, 1923. 



