THE FERNS 143 



pith, but in another species of Zalesskya it is 

 probable that the inner, large-celled wood ex- 

 tended to the centre, while in the magnificent 

 Fern Thamnopteris, the wood certainly formed a 

 solid cylinder with no pith at all, the middle being 

 occupied by the short and wide wood-elements 

 already mentioned. We have here got back 

 to a structure of the vascular system very dif- 

 ferent from what we know in recent Osmun- 

 daceae, but the other characters, especially the 

 structure of the petiole, leave no doubt as to 

 the relationship. The interesting thing is that 

 these oldest Osmundaceous stems closely re- 

 semble some stems of the most ancient known 

 group of Ferns, the Primofilices, characteristic 

 of Palaeozoic rocks. 



Thus a modern family has been traced back 

 till it almost becomes merged in an ancient 

 group, which probably represents the common 

 source of most of our living Ferns, as well as 

 of the Osmundacese. The evidence for the 

 history of this latter family does not rest only 

 on anatomical structure, important as it is. 

 In various Mesozoic rocks, down to the Rhsetic 

 and Triassic, Ferns have been found with the 

 characteristic sporangia of Osmundacese, and it 

 is a most interesting fact that sporangia of the 

 same general type are common in petrified material 

 of the Palaeozoic period. We might call this the 



