462 Bower .—Studies in the Phytogeny of the Fi lie ales. 
a continuation of the flattened pinna, consists of tissue produced by inter- 
calary growth, signs of which are already seen in Fig. 24. 
It may be a question what exact relation the two indusial flaps bear 
to the marginal segmentation of the pinna. Some sections give the appear- 
ance as though the succession of segmentation was continued directly into 
the marginal segmentation of the upper flap, a condition not improbable, 
seeing that it becomes ultimately so much the larger. The exact genetic 
relations in Fig. 23 seem uncertain on this point. But whatever these rela- 
tions may actually be, there can be no reason to doubt the substantial 
correspondence of the receptacle and the two flaps with the correlative 
parts seen in Lindsay a and Saccoloma , and ultimately in Dicksonia and 
Thyrsopteris. It may, in fact, be considered as conclusive that the sorus in 
Davallia is of the marginal type. 
Such questions in Davallia become more insistent in N ephrolepis and 
Oleandra , genera which have usually been classified with the Davallias, 
though their sori appear in the mature state to be much further intra- 
marginal than they are in Davallia . The latter of these genera has not 
been examined developmentally. But in Nephrolepis biserrata , (Sw.) Schott, 
the matter has been investigated, though, owing to certain technical diffi- 
culties, the development has not been fully followed. Fig. 25 shows a com- 
paratively early stage of the sorus, which clearly corresponds in essentials 
to that of Davallia , but the inequality of the two lips is here so strongly 
marked as to make it a still more open question whether or not the receptacle 
itself originated in a marginal position. It may be held that the lip (/), in 
Fig. 24, of Davallia is the phyletic counterpart of the body marked (/), in 
Fig* 25, of Nephrolepis , which is often described as the ‘ indusium ’. The 
hollow behind it is the receptacle. The large body marked (u) in Fig. 25, 
which appears as a continuation of the leaf-surface, is the correlative of the 
upper lip (u) in the Davallia. The difference between the two sori appears 
to lie in the much greater inequality of development of the two sides in 
Nephrolepis , and especially in the very extended development of the upper 
lip (u). In the Davallia this soon loses its marginal segmentation, and its 
greater part results from intercalary activity. But in Nephrolepis , as is shown 
by Fig. 25, an active marginal segmentation appears to account for the 
predominance of the upper lip. It must remain for the present uncertain 
whether or not this activity was developmentally continuous with that at 
the margin of the pinna before the sorus appeared. If that were shown to 
have been the case, it would indicate that there had been a sort of ‘ phyletic 
slide * of the originally marginal sorus to the lower surface of the pinna. 
That such a transition can occur will be shown in the next memoir of this 
series, which will deal with the Blechnineae, and it may be hoped that before 
long such questions will be set at rest for the Ferns now under consideration 
by a careful comparative study of the development of their sori. Meanwhile, 
