EVOLUTION OF SEED-PLANTS 111 



considerable variety of Fern-like plants of the 

 period, we may infer that the family Lyginoden- 

 drese was an extensive one. There is nothing in 

 the pollen-bearing organs themselves to suggest 

 that they belonged to a seed-bearing plant; hi 

 fact, until their connection with Lyginodendron 

 was discovered, they were confidently referred to 

 Ferns of the Marattiaceous group (p. 147). 



We see then that Lyginodendron (which we 

 may certainly take as the type of a great family 

 of "seed-bearing Ferns") shows a very curious 

 combination of characters. In the structure of 

 the young stem and of the frond it was like a 

 Fern; in the structure of the older stem and in 

 other details it was like a Cycad. The seeds are 

 quite highly developed and clearly rank with 

 those of Cycads, but the pollen-bearing organs 

 resemble the fertile fronds of a true Fern. The 

 seed is, of course, the decisive character; Lygino- 

 dendron must be classed among the Seed-plants 

 or Spermophyta, but it was more primitive than 

 any Seed-plant of later periods, for both seeds 

 and pollen-sacs were borne, as it appears, on 

 little-altered ramifications of the ordinary vege- 

 tative fronds. Thus the Lyginodendron type goes 

 even beyond the female plant of Cycas in the 

 want of any special reproductive leaves (sporo- 

 phylls) distinguished from the ordinary foliage. 

 For this primitive group of Fern-like Seed-plants, 



