;44 



HYMENOPHYLI.ACEAE 



[CH. 



elongation of the receptacle, so that a larger number of sporangia may be 

 formed in succession. On such facts the Hymenophyllaceae might be laid 

 out in series, and it would be roughly a series of specialisation carrying with 

 it a decrease in size of the sporangia, and in the number of spores produced 

 from each. 



The number of sporangia in the sorus tends to increase as their indi- 

 vidual size diminishes : thus the actual output of spores per sorus may be 

 maintained at an almost uniform figure, as the following table will show: 



From this it appears that, notwithstanding great variations of output per 

 sporangium, the result per sorus may remain approximately uniform for 

 certain species within this very natural family. The facts suggest a true 

 biological progression: for the production of spores is a drain upon the 

 resources of the plant: that drain is relieved by production of a succession 

 of smaller sporangia within the sorus, the demand being thus spread over 

 an extended period. This will be peculiarly important in Ferns growing as 

 these do in deep forest shade, and with their leaves reduced to a filmy 

 texture. But the principle of the Gradate Sorus is not dependent on such 

 extreme conditions. It is found to be illustrated in several distinct families 

 of Ferns of normal habitat. It may be held as a biological amendment upon 

 the primitive state of the Simplices and as marking, where it occurs, an 

 evolutionary advance upon those Ferns which Palaeophytology and com- 

 parison both indicate as the earliest. 



The Gametophyte 

 The simple structure of the sporophyte of the Hymenophyllaceae finds its 

 correlative in the simplicity of the gametophyte. But here it may be a question 

 whether this is due to reduction or to a primitive state. In Trichomanes the 

 prothallus consists of coarse branched filaments, in habit like a green Alga 

 (Fig. 516). Sometimes, as in T. alatwn, the filament may widen out into 

 flattened expanses, one layer of cells in thickness: but these are of very 

 irregular outline. They may have a definite apical segmentation at first, but 

 they do not take the cordate form usual in Leptosporangiate Ferns. They 

 may be regarded as amplifications of the simple filament (Fig. S17,A,B, C). 

 In Hymenophylliwi this form appears to be dominant, and the prothallus 

 consists of strap-shaped expanses one layer of cells in thickness: but there 



