64 Bower.—Studies in the Phytogeny of the Filicales. VJI. 
specially quotes Anopteris heterophylla as illustrating this. 1 The conclusion 
stated above takes the opposite view, that the inner indusium is a down¬ 
grade structure in these Ferns, and in course of elimination for reasons 
easily understood. This seems to be the only interpretation of the facts 
drawn from a comparative study of closely-related Ferns, provided that it is 
extended to a consideration of soral characters wider in its scope than the 
genus Pteris itself. 
The sequence of sporangia has been found in Saccoloma to be gradate. 2 
The same is seen for young stages in Lindsay a (Fig. 13), but later the 
‘ mixed * state supervenes. It is clearly seen also in some cases in Pteridium , 
but this order is not maintained, and the mixed condition becomes 
a regular feature in matured specimens of Pteridium (Figs. 19, 20), as it is 
also from the first in Pteris (Figs. 31, 32). The conclusion follows that 
Saccoloma and Lindsay a most closely follow the more primitive gradate 
type ; P’teridium takes a middle position, and Pteris is the most advanced 
in respect of the succession of sporangia. 
It has been seen that in the Dicksonioideae there is a progression 
in sporangial character from the short-stalked type of Thyrsopteris with 
oblique and complete annulus through the intermediate types of Cibotium , 
Dennstaedtia , and Microlepia , to the long thin-stalked type with vertical 
interrupted annulus of Davallia. The sporangia of the Pteroideae show a 
similar though less complete progression. It has been shown that in 
Lindsay a the sporangium is rather long-stalked, with a complete annulus. 3 
The same is the case, though less perfectly, in Odontosoriat But in 
Pteridium and Pteris the sporangium is of the ordinary Polypodioid 
type—long-stalked, and with interrupted, vertical annulus. This indicates 
that Lindsaya is in this character also relatively primitive, corresponding 
most nearly with the state seen in Cibotium , while Pteridium and Pteris 
are relatively advanced. 
There is nothing calling for special remark in the spore-numbers of 
this series. They seem to have settled down to figures between 32 and 
64. The spore-form appears to be inconstant. The fact, as stated by 
Diels, 5 that they are bilateral in Histiopteris and Paesia , but tetrahedral in 
Pteridium , Anopteris , Pteris , and Monachosorum , shows that in this affinity 
spore-form is not a dependable character for comparison. The character 
of the spore-coats, however, brought into prominence by Hannig, 6 presents 
some points of interest. Obviously the presence or absence of a perispore 
is not a dependable character for comparison in all cases. It is shown by 
Hannig’s own table 7 that it cuts athwart the very natural group of the 
1 l.c., Fig. 17. 
2 l.c., p. 458. 
3 Studies, III. Ann. of Bot., vol. xxvii, PI. XXXIV, Fig. 22. 
4 1 . c., Fig. 19. 
6 Flora, 1911, p. 321. 
5 l.c.,pp. 287-97. 
7 U, p. 339 • 
