Bower. — Studies in the Phytogeny of the Filicales. 471 
If the course of events has really been as thus sketched, and the 
distinctness of the two phyla be as stated above, it will be apparent that in 
the phyletic treatment of Ferns the criterion of position of the sorus must 
take precedence, in point of early appearance and of constancy, over several 
others to which importance has habitually been attached. For instance, all 
the stelar types may be illustrated in either of these sequences ; also all the 
three conditions of the sorus, simple, gradate, and mixed. The size and 
spore-output of the sporangium vary through a wide range in each. 
In each the more primitive types show hairs, while scales appear in the 
more advanced. Various minor characters also fluctuate within the series 
named. In fact, excepting the most fundamental features of the protostele 
and of the sporangium itself, there is no character so deeply seated in point 
of early appearance and constancy as the position of the spore-bearing 
members relatively to the leaf-margin. Accordingly, in the systematic 
treatment of the Filicales full weight should be given to it. The Lepto- 
sporangiate Ferns (exclusive of the indeterminate Osmundaceae) will then 
be naturally divided into two sequences which may be fitly styled the 
Leptosporangiatae Marginales and the Leptosporangiatae Superficiales. 
And the grouping of the families will be tentatively thus : 
Simplices. 
Gradatae. 
Mixtae. 
Marginales. 
Superficiales. 
^ Schizaeaceae 
Gleicheniaceae 
Matoniaceae 
(Loxsomaceae 
J Hymenophyllaceae 
Cyatheae 
1 Dicksonieae 
Woodsieae 
\Thyrsopterideae 
Onocleinae 
Davallieae 
Oleandreae 
Aspidieae 
Blechninae 
Asplenieae 
Pterideae 
The above table does not attempt to indicate phyletic sequences with 
accuracy or completeness. It is constructed merely to show the main 
constituents of the two great series. Moreover, certain large groups are 
intentionally omitted, such as the Polypodiinae and Acrostichinae, the 
Vittarieae and Taenitidinae, since it is quite uncertain how these will have 
to be split up and phyletically grouped, until much further detailed work 
has been done upon them. It is seen that in both the series there are 
representatives of the three soral conditions. But the lines of limitation of 
these cut athwart the two main series ; in fact, the distinctions which they 
