rOLYrODIACE.E. 331 



witli the edgc moi'e or less elevatecl ; then the sporaogia come iuto sight. 

 When fiilly grown, it becomes a soriis cliflereut iu aiipearance from that of 

 Cryptoso>-us, but very similar to that of Davallla. Yet, gr.irlually transitional 

 forms between this DavaUia-\\kc soriLS ancl the normal C'njptosorus-\Wi.e sorus 

 are to be found in the clevelopmeut of the fructificatiou iu Prosaptia, antl also 

 are to bo seen in the full grown sorus of Polypodium urceolare, as I have stated 

 before. Conscquently, Prosaptia is in its vegetative as well as its propagative 

 organs so closely relatcd to Pohjpodium phylogeneticall}' that it ls c^uitc propcr 

 to nnite them iuto one geniLs. 



Now lct us considcr -whether thc rescmblancc of Prosajjtia aud Davallia 

 which sccms apparcntly veiy close is really an indication of phylogenetic kiu- 

 ship or a merc accidcntal featurc. To decide this question fuudamcntally 

 I made the same study ou DavaUia, as I liad dono before on Cryptosorus aud 

 Fi'Osaptia. DavaUia is a fcrn not of indefinitc, but of definite growth, aud 

 all the sori in one frond mature simuItaneoiLsIy. I took a yoLLug shoot of D. 

 buUata ucarly 5 cm. long, just coming out from tlio rhizome, partly coiliug 

 and partly unfolded, yct bcaring bcautiful sori of a vcry youug stagc. I 

 examined the froud Liuder a biuocular microscope with objcct-glasses a^ and 

 foimd thc sori iu a stage jiLst beforc sporangium-formation. IndiLsium-foi-ma- 

 tion was jiLst bcginuing a littlc below tho apcx of thc lobcs. The indasium 

 (Fig. 137, k) was coming out like a broad quadrangidar scalc, attached at its 

 base to the surface of the frond, but leaving its margin quite frce from thc 

 lattcr. Therc was no depression whatevcr. Thc tjiic of this kiud nf iudasium 

 is represented in a fidl grown sorus of Uuinata. It is totally differcnt from 

 the type secn in Prosaptia. It is, therefore, highly probablc that DavaUia is 

 derived dircctly frnm Humata ; aud Prosaptia, from Polypodium. Gousequently 

 it is clear that thc rcsemblance of DavaUia and Prosapdia is a merc accidental 

 featurc. Thc two can ncver be unitcd iuto one, nor should thc one bc treated 

 as a SLibgeniLS of thc other. Oue might as wcll takc a pscudomorph for a true 

 ciystal, as to take Prosaptia for a DavaUia. 



After considcring all the alx)ve mentioned cascs, I have been led to cou- 

 clude that Prosaptia iu the first placc should bc taken iuto Polypodium ancl 

 for convcnicnce sakc rctained as a suligcuus of thc lattcr, as is the case with 



