282 SUMMARY OF RESULTS [cH. 



manner analogous to the like change in the Pteroids. With these genera 

 Pellaea and Doryopteris may be linked, while Trachypteris and Saffordia 

 present a type like the latter, but with an Acrostichoid spread of the sorus 

 over the surface of the blade. Superficially this is like what has been seen in 

 the Pteroid Ferns; but the two series, the one indusiate and the other non- 

 indusiate, appear to have progressed along homoplastic lines, separate but 

 parallel, or even convergent. The Gymnogrammoid P^erns as a whole 

 probably represent a plexus of non-indusiate phyletic lines, all traceable 

 back to Ferns with marginal sporangia larger than their own, such as 

 are seen in the ancient Families of the Osmundaceae and Schizaeaceae 

 (Chapter xxxix, pp. 92-97). 



As epiphytic derivatives from the central group of the Gymnogrammoids, 

 the five genera now included as the Vittarioid Ferns take their natural 

 place. They appear to be forms arrested or specialised in relation to their 

 habitat. In Anethtni a step towards an Acrostichoid state is again seen 

 (Chapter XLVii). 



It must not be assumed that by essaying some clearer phyletic grouping 

 of Ferns the writer has undertaken to place all genera in some probable 

 relation by descent. Anyone who makes the attempt will very soon find 

 problems rising before him that are insoluble for want of precise data: or 

 questions in the resolution of which a personal estimate of the importance 

 of relevant facts necessarily takes the place of proof Illustrations of this are 

 given in Chapter XLViii, which indicate the sort of difficulties that are apt 

 to arise. They also suggest the wide intervals that lie between possibility, 

 reasonable probability, and demonstration. That Chapter was intentionally 

 introduced to show that the present work is a mere Tentamen, not in any 

 sense a finished task. Some of the examples chosen illustrate how a better 

 knowledge of detail may lead to an assured conclusion {Prosaptia, Dcpajia), 

 or to a definite bias in a certain direction {Cystopteris): others leave 

 alternative views still in suspense iTamitis): others again show how slight 

 are the present grounds for any definite opinion {Azolla, Salviiiid). 



The analysis contained in Vol. i, in respect of the twelve criteria of com- 

 parison used in the phyletic treatment of Ferns, has led to the recognition 

 of states respectively primitive and relatively advanced for each of them. 

 P^or instance, the progressions from equal dichotomy to dichopodial, and 

 finally to monopodial branching: from protostely to vascular disintegration: 

 from dermal hairs to scales : from open to reticulate venation : from more 

 complex to simpler cellular construction: from a marginal to a superficial 

 position of the sporangia: from simple to gradate and mixed sori : from 

 complex and sunk to simpler and stalked sporangia: from a large spore- 

 output to a smaller: from larger to smaller antheridia: from embryos with 



