590 Pteridophyten. 



two main series are distinguished: the Marginales, in which the 

 sori originate from the margin of the leaf, and the Superficiales, 

 in which they originate from the surface of the leaf. The different 

 families of Ferns have been grouped thus (but several large groups 

 have not been examined from this point of view and some of them 

 may have to be divided between the two series): Marginales: Schi- 

 saeaceae, Loxomaceae, Hymenophyllaceae, Dicksoniae, Thyrsopterideae , 

 Davalliae, Olea)idreae and Ophioglossaceae ; Superficiales: Gleichenia- 

 ceae, Matonineae, Cyatheae, Woodsiae, Onocleineae, Aspidiae, Blech- 

 ninae, Aspleniae, Ptcrideae and Marattiaceae. Both series contain 

 members of the Simplices, Gradatae and Mixtae. The value as a 

 distinguishing character of the point of origin of the sorus will 

 depend on its constancy as a phyletic criterion. In these series sori 

 originating marginall}^ may become displaced when mature as the 

 result of biological adaptation [Thyrsopteris, Cibotium, Saccolonia, 

 Odontoloma and Davallia); in Nephrolepis and in Oleandra the mature 

 sori are much more markedly intra-marginal and it is possible that 

 they originate superficiall)'-, indicating "that there had been a sort 

 of "phjietic slide" of the originally marginal sorus to the lower 

 surface of the pinna". Their position is, however, clearly a secon- 

 dary modification and ma}?- be compared to the modification in the 

 Position of the sori of certain Pterideae which though originating 

 superficial!}'' may later approach to a marginal position. "With these 

 exceptions and with the further exception of abnormal or anoma- 

 lous isolated types no family except the Osmundaceae has sori which 

 are indeterminately marginal or superficial in position and in this 

 respect the latter order may suggest the indeterminate ancestry 

 from which the marginal and superficial series of the Leptosporan- 

 giatae sprang. 



The following phylogenetic conclusions are drawn: Metaxya 

 should be placed with the simpler CycUheaceae ; the combination in 

 this genus of dorsiventrality with such primitive characters as sole- 

 nostely, an undivided leaf-trace, a sorus of the Simplices type, the 

 possession of hairs not scales as dermal appendages strongl}' Sup- 

 ports the view that in the Cyatheaceae the dendroid type is a deri- 

 vative one. Metaxya also shows similarities to the Polypodiae, with 

 which it has been classed. A further elaboration of the Metaxya-iy^e , 

 involving the breaking up of the trace and stele by the formation 

 of more numerous perforations such as occur occasionally in the 

 axial Stele of the genus, the replacing of the simple hairs by scales 

 and the assumption by the sorus of the mixed habit and the deve- 

 lopment of a complete oblique annulus would lead to the evolution 

 of a form that would be included in the genus Polypodium. In 

 Opposition to Professor Goebel who regards Loj0ma as a reduced 

 form nearest to the Cyatheaceae, Professor Bower brings forward 

 arguments for believing that the dosest afiinities of this genus lie 

 with the Schisaeaceae, Thyrsopteris and Dicksoniae. Finally it is held 

 that Nephrolepis with its "phyletically" marginal sori and the Aspi- 

 diae with their "phyletically" superficial sori afford a remarkable 

 instance of the evolutionär^' convergence of phylogenetically inde- 

 pendent series. the similarity in the position of the sorus and in 

 the form of the indusium in these two genera being very striking. 



Isabel Browne (London). 



