XVI FOSSIL ARCHEGONIATES 581 



terminal instead of arising from the face of the leaf. In OpJiio- 

 giossmn, however, a study of the earlier stages of the fertile 

 leaf makes it not improbable that the si)ike may be interpreted 

 as a truly terminal organ, and the sterile segment as a lateral 

 appendage of it, comparable to the condition in Archccoptcris. 

 Dimorphic leaves are of common occurrence also in the later 

 Palaeozoic Ferns. 



From the numerous studies that have recently been made 

 upon the stem-structure of the fossil Ferns, it appears (Scott 

 (i), p. 303) that the monostelic stem is relatively commoner 

 among the Palaeozoic Ferns than it is at present. Among the 

 existing Ferns, monostelic stems are especially characteristic 

 of the Gleicheniaceae, Hymenophyllaceae, and most Scliizx^aceae. 

 There were, however, many Palaeozoic Ferns in which the stem- 

 structure closely resembled that prevailing among living Ferns. 

 Some stems closely resembling those of modern tree-Ferns have 

 been described under the name Psaronius. A study of the 

 leaves and sporangia of these shows that their affinities were 

 with the Marattiaceae rather than w^ith the Cyatheaceae, to which 

 family belong nearly all the living tree-Ferns. 



The characteristic sporangia of Ferns are the most certain 

 means of determining their affinities, and unless these are 

 known, the identification of the fossils must be more or less 

 doubtful. While fossil sporangia are of comparatively rare 

 occurrence, still enough has been made out concerning the na- 

 ture of the sporangia of the fossil Ferns to make perfectly clear 

 the affinities of many of these with the living forms. 



As might be expected from a comparative study of the ex- 

 isting Filicineae, it is found that the Eusporangiatae, while 

 showing every indication of being more primitive than the 

 Leptosporangiatae, are really much older geologically. While 

 at the present time these constitute probably less than two per 

 cent, of the living Ferns, among the Palaeozoic fossils they far 

 outnumber all others, if they do not actually include all Palae- 

 zoic Ferns. 



Of the two living families, Ophioglossaceae and Alaratti- 

 aceae, it is the latter which is especially abundant in a fossil 

 condition. Whether the scarcity of the Ophioglossaceae as 

 fossils is due to their lack of firm tissues in the leaf, or w^hether 

 the living forms have become more modified than the Maratti- 

 aceae, it is not possible to decide. The former view seems to 



