Bower . — Studies in the Phytogeny of the Filicales. VI. 3.7 
Gymnopteris has been sunk in Leptochilus } All that is necessary 
is to restore Beddome’s designation for this Fern, thus reviving the genus 
Gymnopteris to receive it. This appears to be an appropriate method for 
marking the distinction from Leptochilus , and the designation will then 
be Gymnopteris tricuspis (Hook.), Bedd. 
The genus Leptochilus itself remains to be considered. It has recently 
been revised by Frau Schumann . 2 She segregates it into two main 
divisions, which she regards as naturally apart. They are distinguished on 
the basis of simple as against compound leaves. The former she regards as 
Polypodioid derivatives, the latter as Dryopterid. This, on the whole, 
I should be prepared to accept, as a first recognition of the polyphyletic 
origin of the Ferns grouped as Leptochilus. But it seems not improbable 
that the sources of the species so designated may have actually been more 
than two. In particular, for the reasons given above, Leptochilus tricuspis 
should be referred more directly to a Dipterid origin, as indicated by com- 
parison with Neocheiropteris and Cheiropleuria. It is possible that investi- 
gation of L. various may lead to a similar conclusion for it also. In fact it 
seems likely that Leptochilus , like Polypodium , Acrostichum , and Gymno- 
gramme , may not be a phyletic genus at all. Such genera have been 
convenient groupings for the classification and recognition of species. 
Probably none of them are pure genera in the phyletic sense, but mere 
assemblages ot forms having some obvious characteristic in common, which 
may itself have been evolved along more than one phyletic line. 
Note on Sporangial Segmentation. 
It has been found that in Cheiropleuriap in Dipterisp and in Metaxyap 
the segmentation of the primordium of the sporangium is a two-sided one, 
and the segments themselves are arranged in two rows, while the*sporogenous 
cell itself is two-sided, like half of a biconvex lens. In the large majority of 
Ferns the segmentation of the sporangial head is three-sided, and the shape 
of the sporogenous cell is tetrahedral. Among the Ferns here discussed the 
latter type is found in Platycerium and in Gymnopteris tricuspis , and it is 
almost certainly so in Neocheiropteris. In the former type the stalk of the 
mature sporangium is four-rowed, in the latter it is three-rowed or less. 
The question may be raised whether this difference of segmentation does not 
militate against the phyletic relations which have been suggested. 
The view which may be held as probable is that the two-sided 
segmentation, though apparently simpler, is really the more primitive. 
It readily gives the four-rowed stalk, which is a more massive construction 
than the three-rowed. And thus those sporangia which have the two-sided 
segmentation are nearer in the character of their stalk to the relatively 
1 See Christensen’s Index, p. 342. 2 Flora, 1915, p. 250. 
3 Studies, V, PI. XXV, Fig. 17. , 4 1 . c., Fig. 18. 5 1 . c., Text-fig. 19, p. 521. 
