I20 



VASCULAR CRYPTOGAMS 



or rhizome, being extremely rare. The great majority of species appear 

 to have been herbaceous ; or at all events the stems of tree-ferns are 



not of very common occur- 

 rence, even in the coal 

 measures. And although the 

 fructification has frequentl}- 

 been met with, the vast ma- 

 jority of the leaves of which 

 the remains or the impres- 

 sions have been preserved 

 are barren. The only avail- 

 able system of classification of 

 the greater number of fossil 

 ferns is based on the mode of 

 venation, on which character 

 a number of families have 

 been founded by Brongniart ; 

 but it is doubtful whether 

 this has any great value as a 

 natural system of classifica- 

 tion. A form of heterophylly 

 different from anything which 

 occurs among existing ferns 

 is found in a few species from 

 the Carboniferous formation, 

 where, in 'addition to the 

 normal pinnae of the frond, 

 themselves again pinnate, im- 

 perfect pinnae of much smaller 

 size and simpler structure are 

 intercalated between them. 

 These imperfect pinnae, known as aphkbice, were described as distinct 

 species before their true character was known. Thus Rhacophyllum 

 adnascens (L. and H.) is the aphlebia of Sphenopteris crenata (L. and 

 H.) ; while various so-called species of Cyclopteris are abnormal pinnae 

 springing from the rachis below the normal pinnae of Neuropteris. 

 On the whole, the leaves of ferns belonging to the Carboniferous period 

 bear a striking resemblance to those of our own day ; in many cases 

 they might belong to living genera. 



The remains of the fructification of fossil ferns that have come down 

 to us lead us to believe that the existing orders of Filices may have been 

 represented in the earlier geological periods ; and none have as yet been 



Fig. 92. — Aphlebia, from the Carboniferous formation. 

 I, Sphenopteris creiiata L. and H. ; 2, 3, Rka£o~ 

 •^kyllum adnascens L. and H. (After Sclmnper.) 



