640 GENERAL COMPARISON OF THE FILICALES 



archaic types the annulus was vertical or oblique, and the dehiscence 

 was mainly, though not exclusively, 1 in a plane including the axis of the 

 sporangium : this is seen in all the surviving Simplices, excepting the 

 Matonineae, and also in Loxsoma. But in the Gradatae the annulus was 

 oblique and the dehiscence lateral, while in the Mixtae the annulus is 

 again vertical, but the dehiscence transverse. If we contemplate 

 a derivation by descent of Ferns with a lateral dehiscence from 

 those with median dehiscence, we shall have to enquire whether 

 there are any evidences of shifting of the annulus itself as well as 

 of the point of dehiscence. One material point is that the formation 

 of the annulus in Ferns at large does not stand in any constant relation 

 to the segmentation of the sporangium-mother-cell, though that segmentation 

 itself shows so singular a constancy. This fact leaves the question of 

 a shifting of the annulus more open than it would otherwise appear. The 

 more primitive type of complete annulus is that seen in the Gleicheniaceae 

 and Schizaeaceae, with oblique position and median dehiscence : Loxsoma 

 maintained the oblique position and median dehiscence, but part of its 

 annulus is incompletely indurated. In others, while the complete oblique 

 ring was maintained, the point of dehiscence was shifted laterally, the 

 result being as in the Cyatheaceae, Hymenophyllaceae, and others with 

 a basipetal sorus. With the transition from the basipetal sorus to the 

 mixed came also a further change of the annulus : maintaining the 

 lateral dehiscence, the annulus became vertical, stopping short on either 

 side of the stalk, which interrupts it. But in many cases a slight 

 obliquity was retained, as seen in Dennstaedtia apiifolia (Fig. 333 c) 

 and Diaca/pe, the two sides being so far dissimilar that it is possible 

 still to distinguish the "central" from the "peripheral" face: this is 

 also the case in Davallia, Lindsaya, Nephrodhtm (Fig. 6), and many 

 others. But there are other outstanding cases of an oblique annulus among 

 Polypodiaceae which have been the subject of discussion, and have even 

 been considered a sufficient reason for rejecting the oblique or vertical 

 positions of the annulus as characters which are not dependable : for 

 instance, the genus Lomaria, in which the Plagiogyria has a well-marked 

 continuously oblique ring of the annulus, somewhat similar to that seen 

 in the Dicksonieae. So far from looking upon such cases as these as 

 being subversive of views based on the character of the annulus, they are 

 exactly what might have been anticipated if the types with a vertical annulus 

 were derived from forms in which the annulus was oblique : it is hardly 

 to be expected that the transition would be carried out completely in all 

 cases : these exceptions may be regarded as being occasional survivals of 

 the earlier oblique type. 



It would appear thus probable that the simple annulus of the Leptosporan- 

 giates is prefigured by the vertical, many-rowed hoop of the Eusporangiates. 

 That in the course of descent, as the bulk of the sporangium was reduced, 



1 Scott quotes a porous dehiscence for Stauropteris (Progress us AV/ Bot'., i., p. 186). 



