THE FERNS 153 



sented in the genus Kaulfussia. In some of the 

 fossil spore-fruits the sporangia were only con- 

 nected at the base, or they may even have been 

 nearly as free as in the recent Angiopteris; in 

 others they were completely united, as in Kaul- 

 fussia itself. Sometimes the synangium was 

 stalked, as in a part of the recent genus Marattia. 

 In other cases there was a close agreement in 

 form with the fructifications of Dancea and 

 Angiopteris. 



On the whole of the characters, it seems to 

 be perfectly established that in Palaeozoic, 

 especially late Palaeozoic times, there was a 

 large group of plants, extending probably to 

 all parts of the world, which had a clear affinity 

 with the recent Marattiaceous Ferns. 



The question that still remains is: were 

 these Palaeozoic forms true Ferns or were they 

 Pteridosperms? The pollen-bearing fronds of 

 Lyginodendron were at one time classed among 

 Marattiaceous Ferns; another supposed member 

 of the latter family has been found laden with 

 seeds ! Even in the highly-developed Bennettiteae 

 of Mesozoic age, the stamens might well be taken 

 for the fertile fronds of some unknown member 

 of the Marattiaceae. There is, in fact, no means 

 of telling for certain whether a given fructifica- 

 tion represents the sporangia of a true Fern or 

 the pollen-sacs of a Seed-fern. On the other 



