38 Bower. — Studies in the Phytogeny of the Filicales . VI. 
primitive types. This is further illustrated by the structure of the sporangial 
head in Metaxya , Dipteris , and Cheiropleuria , which all show with more or 
less plainness the oblique and continuous annulus, while Neocheiropteris , 
Gymnopteris , and Platy cerium have the annulus vertical and interrupted. 
The conclusion is that the former group, notwithstanding the apparently 
more simple segmentation, represent in truth a more primitive type, while 
the latter are derivative. A phyletic transition from the one to the other is 
no less easy to comprehend than a transition from a two-sided to a three- 
sided segmentation of the apex of axis or leaf in a Fern, or in Selaginella . 
Accordingly the difference of segmentation of the sporangium appears to 
be no serious obstacle to the comparisons instituted above, and Dipteris , 
Metaxya , and Cheiropleuria take their place as more primitive in sporangial 
construction than Gymnopteris , Neocheiropteris, or Platycerium. 
Summary of the Chief Results. 
1. Cheiropleuria , Platycerium , and perhaps Neocheiropteris share with 
the Fern hitherto styled Leptochilus tricuspis (Hook.), C. Chr., an extension 
of the sorus, with a special vascular supply, spreading in a plane below and 
parallel with the venation of the sporophyll. This is most extensive in 
Platycerium and L. tricuspis , and the condition may be described as diplo - 
desmic . 
2. These Ferns in external morphology, venation, anatomy, sorus, and 
.sporangia are regarded as Dipterid derivatives, and may be grouped phy- 
letically as Dipteroideae. To these may probably be added, later, many 
Polypodioid Ferns, especially Phlebodium , Phymatodes , Niphobolus , and 
Drynaria , and some simple-leaved species of Leptochilus. 
3. L. tricuspis stands alone in the latter genus in various features, but 
especially in the diplodesmic character. It should, therefore, be removed, 
and by reviving its old generic name, now merged in Leptochilus , it may be 
styled Gymnopteris tricuspis (Hook.), Bedd. Of that genus it will be at 
present the only species. 
4. A parallel series to the Dipterioideae, but differing in venation, and 
probably distinct phyletically, is related to Metaxya as its probable source. 
It includes Syngramme and Elaphoglossum. These may be styled the 
Metaxyoideae. 
5. Both of these progressions illustrate advance from circumscribed 
sori to an ‘ Acrostichoid 5 spread of the sporangia over the leaf-surface. 
This runs parallel with changes of leaf-form, disintegration of the vascular 
tracts, passage from dermal hairs to scales, increasing areolation'of veins, 
and changes of sporangia from the continuous oblique to the interrupted 
vertical annulus. Since there is substantial parallelism in these various 
characters of advance, the progressions are firmly established : but they are 
constantly distinguished from one another by their venation. 
