3o6 MOSSES AND FERNS chap. 



tropical forest, and the enormous size to which many of them 

 grow, is very quickly disabused of any such notion. 



The fossil record is also extremely instructive as bearing on 

 this point. According to Solms-Laubach (2) there is but one 

 certainlv authentic case from the Carboniferous rock which can 

 be regarded certainly as a leptosporangiate form, all of the 

 other sporangia discovered being of the eusporangiate type. In 

 the later formations the Leptosporangiates increase in number, 

 but according to Luerssen ((7) II, p. 574) undoubted Poly- 

 podiacese are not found before the Tertiary, where a number of 

 living genera are represented. 



Potonie (3) cites several examples of Palaeozoic Ferns 

 probably allied to the lower leptosporangiate families, but the 

 number is very small compared to the eusporangiate types. 



Except in the few heterosporous forms there is, on the 

 whole, great uniformity in the gametophyte. The most 

 marked exception to this is the filamentous protonema-like pro- 

 thallium of some species of Trichomanes and Schizcea. Except 

 in these, however, the germinating spore, either directly or after 

 forming a short filament, produces normally a flat, heart- 

 shaped prothallium, growing at first by a two-sided apical cell, 

 the prothallium being at first one cell thick, but later producing 

 a similar cushion to that found in Marattia but less prominent, 

 and the wings always remain one cell thick. Upon the lower 

 side of the cushion are produced the archegonia, wdiich have 

 always a projecting neck, sometimes straight, but more com- 

 monly bent backward. The antheridia are produced upon the 

 same prothallium as the archegonia in most forms, but a few 

 species of Ferns are dioecious, and usually there are small male 

 prothallia in addition to the large hermaphrodite ones. The 

 antheridia, like the archegonia, always project above the surface 

 of the prothallium. 



The first divisions in the embryo always divide it into 

 regular quadrants, and the young members always grow from 

 a definite apical cell, which, with the possible exception of some 

 of the Osmundace^, is also found at the apex of the later roots 

 and always in the stem. In size the sporophyte varies ex- 

 tremely. In some of the smaller Hymenophyllacese the creep- 

 ing stem is not thicker than a common thread, and the fully- 

 developed leaves scarcely a centimetre in length. The other 

 extreme is offered by the giant tree-ferns belonging to the Cya- 



