652 GENERAL COMPARISON OF THE FILICALES 
but in the Leptosporangiates this became the rule, and with it is associated 
the more definite segmentation of the projecting and delicate sporangium, 
and a diminution of the output of spores. In fact, the character of the 
sporangium may be taken as an index of the meristic character of all 
the embryonic tissues. 
PHYLOGENY OF FERNS. 
The comparative study which has now been made of the Filicales 
provides a basis for views as to their phylogeny. Several different lines 
of comparison have been adopted, such as that on external form, on 
anatomy, on the spore-producing members, and on spore-output: and 
the results have been found to be substantially parallel along them all, 
though with some exceptions. Moreover, these results are found to harmonise 
with the geological record: from this it is learned that, speaking generally, 
the Ferns with simultaneous .formation of the sporangia in the sorus 
(Simplices) were those present mainly, or perhaps exclusively, in the 
Palaeozoic period: the Ferns of the Mesozoic Period included, in addition 
to these, a large proportion of those with basipetal sequence of the sporangia 
(Gradatae), while the bulk of the more recent Ferns are characterised 
by the mixed type of sorus (Mixtae). While these three types are 
found to coincide generally with three great periods of time, it must not 
be assumed that every phyletic line ran through them all: at the moment 
it is impossible to demonstrate in any clearly defined family of Ferns - 
that the sorus has passed successively through all the three phases. What 
is recognised is a general trend of development, carried out in many 
phyletic sequences, rather than any single progression. It may indeed 
be said that no connected phyletic tree of the Filicales can at present 
be constructed with any degree of certainty: it must suffice to give a 
provisional arrangement af the Ferns, divided in the first place according 
to their soral condition, which also tallies approximately with their geolo- 
gical age (Fig. 354). The three recognised grades are limited by horizontal 
lines in the graphic table. The several sequences of Ferns are indicated 
by vertical or oblique lines so placed as to indicate relationships, but 
disconnected so as not to convey necessarily a common descent. Where 
one of these lines crosses the limit between the three grades of soral 
construction it is intended to convey the idea of the derivation of the 
higher from the simpler plan. Further, the arrangement is such as to 
give some general idea of the position of the sorus: the forms with marginal 
sori are placed to the left side of the plan, and those with superficial 
sori to the right. Finally, no attempt has been made to represent separately 
the numerous sequences of the Mixtae: the unravelling of the lines will 
have to be deferred till a better knowledge is gained of their anatomy, 
and of other details. They have been represented conventionally in the 
scheme by overlapping areas, the one to the left stands for those 
with marginal sori, that to the right for those with superficial sori, while 
