HYMENOPHYLLACEAE 587 



and radial construction in Zygoptcris and the creeping dorsiventral rhizome 

 in these Hymenophyllaceae, the structural resemblance is very close : and 

 with this go the strikingly similar facts of structure and insertion of the 

 leaf-trace, and of the mode of supply to the axillary buds. 



Next, as to the leaf-texture, H. dilatalum and T. reniforme are both 

 species with the lamina composed of several layers, and occasionally showing 

 in their development the alternate segmentation seen in the leaves of 

 ordinary Leptosporangiate Ferns. According to the argument advanced 

 above, the filmy texture is an adaptive character shown in various families 

 of Ferns : the peculiarity of the Hymenophyllaceae is that they show it in 

 the highest degree. But the species named appear less specialised in the 

 hygrophilous direction than others of the family, and thus they serve to 

 connect it with the ordinary types. 



The marginal position of the sorus is shared with the Schizaeaceae, while 

 it is to be remembered that the Botryopterideae and Osmundaceae may also 

 bear sporangia in their leaf-margins. But the Hymenophyllaceae differ from 

 any of these in the basipetal sequence of the sporangia, the elongated 

 receptacle, and the cup-like indusium. The basipetal sequence may be 

 held to be a secondary condition, bringing with it the advantage of spreading 

 the drain of spore-production over a longer period than if all were developed 

 simultaneously : the elongation of the receptacle, a consequence of intercalary 

 growth, is almost a necessary condition of its adoption. The basal cup-like 

 indusium, imperfectly represented in the Schizaeaceae, has probably been 

 a new formation : its efficacy in protecting the youngest sporangia at the 

 base of the sorus amply justifies its existence. It is thus possible to conceive 

 of the origin of the Hymenophyllaceous sorus from some Fern-type with 

 marginal sporangia, by initiation of a basipetal sequence, and establishment 

 of a protective indusium. The type from which they might have originated 

 would probably be found among some protostelic types with large sporangia 

 marginally produced, of which the Botryopterideae, Osmundaceae, and 

 Schizaeaceae are the known representatives. 



A comparison of the sporangia themselves confirms this reference to 

 the Ferns with large sporangia, rather than to simpler forms such as the 

 Polypodiaceae. For there is an oblique annulus corresponding in position 

 on the one hand to that of the other Gradatae, but also to that of certain 

 of the Simplices. A comparison of Fig. 325 of Hymenophyllum with Fig. 310 

 of Gleichenia shows plainly the close similarity of the sporangia : and it 

 has been shown that if the peripheral face of the Gleicheniaceous sporangium 

 be reduced the Schizaeaceous type is the result, both being variants of the 

 same form. If finally the point of dehiscence were shifted from the median 

 plane to the side a practical necessity where there is a basipetal sequence 

 the Hymenophyllaceous sporangium would be the result. Further, in the 

 spore-output certain of the Hymenophyllaceae approach the Simplices : it has 

 been shown that in H. TitnbriJ^cnse the output per sporangium is 256-512 : 

 in T. reniforme and sericeum it is typically 256, while other species of 



