1502 C0RM0PHYTA. 



secondary shafts are developed; tripinnate, as in Sfihenopteris (fig. 1369), 

 when tertiary shafts occur ; or multipinnate. The coverings of the sori 

 are termed indusia ; and when the stem is creeping it is known as a 



7'hizome. 



Order i. Stipulate. — This order now contains only the two 

 families Ophioglossacece. and Marattiacece, characterised by the pres- 

 ence of stipules at the base of the petioles of the leaves. Of the 

 former Ophioglossum (Adder's-tongue), in which the sporangia are 

 embedded in the tissue of the partly sterile and partly fertile leaves, 

 occurs in the Middle Eocene of Monte Bolca ; while it is probable 

 that Chiropteris, of the Keuper of Wiirtemberg, is not really separ- 

 able. In the second family, where the sporangia are external, as 

 in the Ferns, the existing genus Marattia occurs in the Rhaetic and 

 Lias of the Continent. Danceites and Danceopsis, on the other 

 hand, are extinct types, the former occurring in the Cretaceous of 

 Europe, and the latter in the Keuper, and probably also in the 

 Permian, of the Tyrol, as well as in the Rajmahal stage of the 

 Indian Gondwanas. Finally, the living genus Dancea occurs in the 

 Lias of Verona. Several of the fossil forms have been confused 

 with the Tceniopteridecz, while Schimper has referred to this family 

 some of the forms noticed under the Pecopteridece. According to 

 Mr Kidston, the Archceopteridece. should probably be placed in this 

 order. 



Order 2. Filices. — In existing Ferns stipules are wanting ; and 

 there is only one kind of spores ; so that the sexual stage is not 

 reached till the prothallium. 



Existing forms are classified by the nature of their fructification ; but 

 since this is generally unknown in the earlier fossil types, we are com- 

 pelled to classify them by the nature of the pinnation and venation of the 

 leaves, and as these are subject to great variation among allied existing 

 genera, it is evident that this classification is more or less of an artificial 

 nature. It will be convenient to follow Schimper's plan of treating first 

 of those existing families which are known to be represented in a fossil 

 state, and then to notice the forms of which the position is more or less 

 uncertain, but which are provisionally arranged in families. 



Existing Families. — In the existing series the family Osmun- 

 dacecz has the paniculate fructification borne on specialised leaves, 

 which may or may not be like the sterile ones. The type genus 

 Osmunda, which is mainly characteristic of warmer regions, al- 

 though represented in England by the Royal-Fern, occurs fossil in 

 the Cretaceous of Westphalia and Greenland, in the Laramie beds 

 of North America, and in many of the European Tertiaries. An 

 allied form, from the freshwater deposits of Chemnitz, has been 

 named Asterodcena. In the Schizeacece^ which usually have the 

 fructification borne in spikes or panicles on the laanice of the 



