450 Bower . — Studies in the Phytogeny of the Fi lie ales. 
parallel these : and in point of fact Metaxya was included in Polypodium by 
several of the early authors. But it is to be remembered that there the 
sorus is a mixed one, while here all the sporangia arise simultaneously. 
The further discussion of this point will be reserved for the concluding part 
of the memoir. 
From the characters detailed above there is sufficient ground for 
upholding Presl’s view that Metaxya should be regarded as a substantive 
genus. The characters which mark it off from Alsophila, the genus to which 
it had been attached by Sir William Hooker and others, are the creeping 
habit, the unbranched hairs and the absence of scales, the solenostelic 
structure of the axis without medullary strands, the undivided leaf-trace, 
and, most important of all, the characters of the sorus and sporangium ; 
these are the flattened and enlarged receptacle, the simultaneous origin 
of the sporangia, the almost vertical annulus, and its interruption at the 
insertion of the stalk. The Fern is technically one of the Simplices, 
for all its sporangia arise simultaneously, a character which it shares with 
Lophosoria . In fact, these two monotypic genera lie aloof from the Cya- 
theaceae, and show characters both anatomical and soral which are more 
primitive than theirs, while the fact that the dermal appendages in both are 
hairs, not scales, points to the same conclusion. On the other hand, the two 
genera are clearly allied to one another. This comes out not only in the 
characters named, but also in the peculiar relation of their runners to 
the leaves, and in the attachment of their vascular supply to that of the leaf. 
They differ, however, in their sori and sporangia ; for while Lophosoria has 
few large and stout sporangia with a complete oblique annulus, Metaxya 
has a very large number of smaller sporangia in each sorus, but the annulus 
is almost perfectly vertical, and is interrupted at the stalk. It has been 
shown in a previous memoir that Lophosoria possesses characters reminiscent 
of the Mertensia section of Gleicheiiia . Metaxya shares with both the 
creeping habit, solenostelic structure, and undivided leaf-trace, dermal hairs, 
and simple simultaneous sorus. But the divergence from that type is seen 
in Lophosoria in the upright habit and sometimes divided leaf-trace ; in 
Metaxya it appears in the soral characters rather than in habit or anatomy. 
Hemitelia setosa, (Klf.) Mett. 
Having thus seen two Ferns with undoubted Cyatheaceous affinity, 
both showing a solenostelic structure, and most markedly in axes with 
a prone habit, it became a question of interest whether any of the true 
Cyatheaceae show similar characters. Gwynne-Vaughan, in his second 
memoir on Solenostelic Ferns (‘ Ann. of Bot.,’ xvii, p. 710), has pointed out 
how Alsophila excelsa in its sporeling stage actually passes through such 
a condition of solenostely in its upright axis, but rapidly becomes dictyo- 
stelic by overlapping of the leaf-gaps. It would seem probable that if any 
