144 THE EVOLUTION OF PLANTS 



most ancient type of Fern-sporangium; it oc- 

 curred not only in fossil Osmundacese but in their 

 predecessors, the Primofilices, so that evidence 

 from the reproductive as well as the vegeta- 

 tive organs connects the recent family with this 

 ancient group, of which a slight sketch will now 

 be given. 



The effect of the discovery of the Pterido- 

 sperms has necessarily been to reduce greatly 

 the number of true Palaeozoic Ferns. There 

 remain, however, two considerable groups of 

 plants of that period, which may be accepted 

 as Ferns. The simpler of the two groups we 

 have called, following Mr. Arber, the Primo- 

 filices, because they appear to go further back 

 in geological history than any others of their 

 class. They are best known from their structural 

 characters, for there are not many instances in 

 which we have good evidence for the external 

 habit of members of the group, though doubtless 

 some of the fossil impressions of Fern-fronds 

 really belonged to them. 



Most modern Ferns have a number of distinct 

 vascular cylinders in the stem; the Primofilices, 

 like the Royal Ferns and one or two other recent 

 groups, always had a single vascular cylinder; 

 there was also one bundle only in the petiole of 

 the frond. In some, the stem had a remarkably 

 simple structure, the single vascular cylinder con- 



