THE FERNS 151 



pollen-sac of Bennettites. When ripe, the two 

 kidney-shaped halves of the fruit split apart, like 

 opening a book, and then each of the numerous 

 compartments in either half releases the spores by 

 means of a slit on its inner surface (fig. 17, A). 

 Dancea has much the same arrangement, except 

 that the spore-fruit does not split open and the 

 spores escape through pores instead of slits. This 

 genus shows some slight difference between sterile 

 and fertile fronds, which are not usually distinct 

 in the family. 



In Kaulfussia, which has fronds like the leaves 

 of a Horse-chestnut, the synangia scattered on 

 the under side are circular, each having the form 

 of a neat rosette, with a cup-shaped hollow in the 

 middle, into which the compartments (about a 

 dozen in number) open by slits (see fig. 17, B). 



These curious spore-fruits are no doubt an ad- 

 vanced form of fructification compared with the 

 simple spore-sacs of the ordinary Ferns. Yet the 

 family Marattiacese, so highly-organised both in 

 vegetative and reproductive structure, represents 

 a very ancient race. A certain number of Meso- 

 zoic Ferns have been referred to this family, and 

 some of these, from rocks of Triassic and Liassic 

 age, are so like the recent forms that they have 

 been placed in the genera Marattia and Dancea. 

 When we get back to Palaeozoic times we meet 

 with a surprising number of Ferns, which, both in 



