592 FILICALES 
of the sporangia are intermixed. The probable position of these genera 
will be considered later. There remains, then, only the old comprehensive 
genus Duzcksonta. This was divided in the Synopsis Filicum into three 
sections—Ciéotium, LEudicksonia, and FPatania (= Dennstaedtia). While 
Cibotium and Eudicksonia have obvious relations to Zhyrsopterts, Patania 
(Dennstaedtia) clearly approaches the genus Davad/ia, and especially to that 
section of the old genus which has been separated as the independent genus 
Microlepia: these relationships will now receive the support of develop- 
mental and anatomical evidence. I shall follow Prantl! in separating 
Dennstaedtia and Microlepia from the position given them in the Synopsis 
Filicum, and recognise them as constituting, perhaps with certain other genera, 
a natural sub-tribe under the name of the Dennstaedtiinae Prantl, having a 
position between Dicksonta and Davallia, and constituting with them a 
natural sequence. It will be shown that following this series from Zhyrso- 
pteris to Davallia we shall pass from a type with basipetal sequence of the 
large, short-stalked sporangia, with oblique annulus, to forms with a mixed 
sorus, smaller, long-stalked sporangia, and a vertical annulus. The receptacle, 
which is a prominent feature in the former, is reduced, or represented only 
by a flat surface in the latter types. The gradual nature of these parallel 
steps seems to indicate that the whole series is one of natural affinity, as 
indeed has always been recognised by systematic writers. 
Dicksonia (EXCL. § Patania). 
The genus Dicksonza includes some large ‘ree Ferns, and others of smaller 
stature but with prevalent radial construction. Some of the smaller species 
closely resemble small plants of Zhyrsopieris: like it they have leaves 
repeatedly pinnate, with numerous sori borne at the margins, but without 
any differentiation of sterile and fertile pinnae. The sori are protected by 
a two-lipped indusium, but the lips are unequal, and their character has 
been used as a basis of division of the genus. It will be shown that the 
receptacle itself is marginal, and that the lips of the indusium are develop- 
mentally outgrowths from the surface of the pinnule, just as in the 
Hymenophyllaceae and in Zzhyrsopterts. Thus there is essential corre- 
spondence with these Ferns, and the differences are rather of habit and 
size than of the more fundamental features of the sorus. 
SPORE-PRODUCING MEMBERS. 
The sorus in this genus has already been investigated by Gliick,? 
who points out that the receptacle arises from the original leaf-margin, 
while the two lips of the indusium spring from the upper and lower 
leaf-surfaces. The structure of the young receptacle, as seen in Dizcksonta 
(Cibotium) Schiedei, Baker, is like that of a leaf-margin, with a marginal 
1 4rb. Kénigl. Bot. Gart. su Breslau, 1892, p. 18. 2 Flora, 1895, Heft 2. 
