582 MOSSES AND FERNS chap. 



the writer the more probable, as there are very strong reasons 

 for considering the type of sporangium found in OpJiioglos- 

 sum as the most primitive occurring in the Fihcineae. 



Very few fossils have been found that can be referred with- 

 out hesitation to the Ophioglossace?e. The early Palaeozoic 

 genera Rhacoptcris and Archccoptcris were apparently very 

 much like BotrycJiiitui, but it is by no means agreed by all 

 Pal?eobotanists that they really were related to the Ophioglos- 

 saceae. There are also other Paheozoic genera, which perhaps 

 are quite as much like Botrychhim as they are like the Marat- 

 tiacece, with which they are usually associated, but all of these 

 forms are very doubtful. OpJiioglossifcs oiifiqua from the 

 Permian is said to resemble closely the spike of Ohiloglossiim, 

 and Chiroptcris digit at a from the upi)er Triassic has been com- 

 pared to O. paUnatuin. In a later formation (Eocene) there 

 has been found a species of Ophioglossinn, O. ococenum 

 (Potonie (3), p. 91). 



If the existence of the Ophioglossaceae during the earlier 

 geological epochs is somewhat doubtful, this cannot be said of 

 the second family of the Eusporangiatse, the Marattiacese. 

 These evidently comprised the greater part of the Palaeozoic 

 Ferns, and many of them were very much like their living de- 

 scendants. The few existing Marattiaceae are mostly tropical 

 Ferns, some of great size, such as most species of Marattia and 

 Angioptcris. 



The ]\Iarattiaceae have much firmer leaves than the Ophio- 

 glossaceae, with distinct and conspicuous venation, admirably 

 fitted to leave a clear impress in the rocks, and indeed the casts 

 of these, in many cases, might almost have been made from 

 leaves of the living species. The close relationship of many of 

 these fossil IMarattiaceae with the living ones is perfectly evi- 

 dent. Of these undoubted Alarattiaceae may be mentioned the 

 following genera : Ptychocarpus, Astcrothcca (Scott (t) Figs. 

 91, 92), Scolecoptcris and Danccitcs (Potonie, (3), Figs. y6, 

 79). The two former genera resemble in the form of the sori 

 (synangia) the living genus Katilfussia. .Danccitcs resembles 

 so closelv the jjenus Daucca that it mav verv well be considered 

 identical. All of the genera mentioned occur in the Carbonif- 

 erous rocks, but also are found in the early Mesozoic. The re- 

 cent genus Marattia has been found in the latter formations; 

 and of about the same age are Dancca-Yike forms which have 



