276 



affords a just ground for subdivision, and if so to what amount. In 

 order to judge fairly of the vakxe of characters derived from these 

 sources, let us compare their constancy. 



The median or dorsal attachment of the capsules is common to the 

 vast tribe or suborder which comprises the genus Polypodium of Lin- 

 neus. The large section subtracted from the genus by Swartz under 

 the name of Aspidium, offers no objection in this respect. The chief 

 groups into which the British examples of this tribe are divisible are 

 these : — 



If we carefully examine these, we shall find no exception to the rule 

 that the attachment of the capsules to the receptacle is median or 

 dorsal. But if we turn our attention to the presence or absence of 

 the involucre, the result of the investigation will be widely different. 

 In Ilvensis the hairy fringe around the base of the capsules is consi- 

 dered an involucre : in fragilis a hoodlike involucre springs from the 

 receptacle at the point where the capsules are attached : in aculeatum 

 it spreads like an umbrella over the capsules, standing on a central 

 stipes like a mushroom : in multiflora the capsules are about half co- 

 vered, the involucre being forced on one side, and the capsules pro- 

 truding on the other : in Oreopteris the involucre may be said to have 

 reached a minimum as regards size, when present, but its presence 

 seems rather the exception than the rule : in Phegopteris it is so rare- 

 ly observed that most authors deny its existence, and in vulgare I 

 believe no author has asserted that he has found it. Those who are 

 acquainted with exotic ferns are well aware how perfect a series is to 

 be found filling the gap between aculeatum and multiflora : in some 

 instances the involucre is so exactly intermediate, that it is a most 

 difficult problem to solve whether its attachment is central or lateral. 

 Again, some exotic ferns exhibit the involucre in every intermediate 

 state between that of multiflora and that of Oreopteris ; while others 

 most completely connect Oreopteris and Phegopteris. It is remark- 

 able, that in Oreopteris the same plant will produce fronds with and 

 fronds without involucres, and even the same frond may not unfre- 

 quently be found having some of its clusters accompanied by an in- 

 volucre and some of them perfectly naked. 



