654 GENERAL COMPARISON OF THE FILICALES 



the overlap represents those which are in course of transition, or are 

 believed to have recently passed from the one state to the other. With 

 these explanations the scheme may be taken as representing the con- 

 clusions arrived at in the preceding pages. 



Of the Simplices the Marattiaceae stand somewhat isolated from modern 

 Ferns, both anatomically and sorally. They are approached most nearly 

 by the Gleicheniaceae in point of the sorus, but in anatomy and in habit 

 they stand widely apart from them. Their stock probably never progressed 

 beyond its present state. If they had any near relation with other living 

 Ferns it must have been very far back. 



The other very ancient family, the Botryopterideae, shows obvious 

 relations with the Osmundaceae, both in type of shoot, in anatomy of 

 the earlier species, and in sporangial characters. There appears to be 

 an anatomical resemblance, on the other hand, to the Hymenophyllaceae, 

 which extends to the external characters of the shoot also ; but the sorus 

 of the Hymenophyllaceae is widely apart, having passed to the basipetal 

 type. It is perhaps in relation to their consequent increase in number 

 that reduction of size of the sporangium is here shown, and especially 

 illustrated in Trichomanes. None of these three related families appear to 

 have progressed further than their living representatives : they are held 

 to be blind branches of descent. 



The Schizaeaceae appear as an isolated family, though nearest to the 

 Gleicheniaceae : their solitary marginal sporangia show analogies of structure, 

 but not of position, with those of Gleichenia on the one hand and of 

 Osmunda on the other : it is, however, probable that their monangial 

 sorus is a result of reduction from a radiate type, such as is seen in 

 some species of Gleichenia. Anatomically also their simpler types approach 

 Gleichenia, but the upright genera, Aneimia and Mohria, appear to have 

 progressed along a line of their own to a dictyostelic structure. The 

 Schizaeaceae also appear to have been a blind branch. 



The Gleicheniaceae are somewhat isolated, from the fact that they 

 show cross characters : their sori compare most nearly in position and 

 structure with those of the Marattiaceae, but their type of shoot and their 

 anatomy correspond rather with the Schizaeaceae. Thus among the Ferns 

 which show their primitive character by their large spore-output per 

 sporangium, including the Simplices and the Hymenophyllaceae, there 

 appear to be several fairly distinct lines : it is possible to link these 

 together by hypothesis as to a common ancestry, but there is no distinct 

 evidence of their common descent from any known Fern-type. This is 

 indicated in the scheme by the convergent but disconnected lines, the 

 longer lines of the Botryopterideae and Marattiaceae indicating their 

 priority in the fossil record. 



From the Simplices, though with uncertainty as to their definite 

 reference to any exact origin, at least two other main lines in addition 

 to the Hymenophyllaceae appear to have proceeded to the basipetal type 



