248 Lady Isabel Browne. 
The former suggests that Ceratopteris has followed a different course 
of evolution from other dictyostelic ferns (15), but Mr. Chandler 
regards the splitting of the cauline strand “ merely as an anticipation 
of the characteristic double nature of the leaf-traces” (13). He 
points out that in the complex forms the leaf-trace usually originates 
from two separate strands and that a splitting of the cauline strand 
to give rise to these strands from which the trace originates is the 
most obvious preparation for the departure of the latter (13). This 
view, bringing Polypodium aureum and Ceratopteris into line with 
other dictyostelic Ferns, is more satisfactory than Miss Ford’s. 
Ceratopteris has undergone reduction, probably owing to its aquatic 
habit, and this may account for the suppression in its ontogeny of 
some of the stages hypothecated in its phylogeny. In some cases 
the solenostele seems to have become dictyostelic by the formation 
of gaps, generally in the ventral part of the stele, unconnected with 
the departure of the traces (17). For these gaps Mr. Tansley has 
suggested the name “ perforations ” (34). 
Polycycly also occurs in the orders under consideration. In 
its development it resembles the polycycly of Matonia (q.v.) and 
every stage in its evolution from the thickening of the edges of the 
leaf-gap in a solenostele to a tetracyclic condition has been recorded 
within the two orders. 
The fronds of the orders are usually pinnately divided ; but 
Mr. Seward and Miss Dale have shown that in Dipteris the frond 
is dichotomous. That this is a primitive character is borne out by 
the presence of Dipteris- like fronds from the rocks ranging from 
the Triassic to the Cretaceous (28). 
Professor Bower has shown that whereas in all Cyatheaceae, 
and some Polypodiaceae, the sporangia arise in basipetal succession 
on a receptacle, in other Polypodiaceae they develop without any 
definite order, forming a “mixed sorus ” (7). Miss Armour has 
recently shown that in some species of Dipteris the sporangia 
originate simultaneously, while in other species the sorus is of the 
mixed type (3). Professor Bower has shown that the sorus in 
which the sporangia originate simultaneously is the most primitive, 
that the sorus showing basipetal succession of the sporangia 
originated from this type and itself gave rise, in several series of 
forms, to a “ mixed sorus.” Another point brought out by him is 
that the oblique annulus, retained by most Cyatheaceae, has become 
vertical and incomplete in most Polypodiaceae (7). In a few, notably 
in Dipteris , the annulus is somewhat oblique and the geologica 
