6o PTEROID FERNS [CH. 



Fig. 26). The leaf-blade is coarsely reticulate, with hexagonal areolae, after 

 the manner of Litobrochia, bearing very numerous stomata on the lower 

 surface: it is invested with scaly hairs when young. The Acrostichoid soral 

 areas may cover the whole lower surface of the fertile pinna, or only parts 

 of it: the sporangia are large, having about 20 indurated cells of the 

 annulus, while the stomium consists of about six. Abnormalities of the 

 sporangia are not uncommon, particularly about the limits of the fertile area. 

 Protection is given to the sporangia while young by branched hairs (Schumann, 

 I.e. p. 220, Figs. 10, 11). The spores are tetrahedral, and are without any 

 perispore: a fact that distinguishes them from those of other Acrostichoid 

 Ferns in which a perispore is present (Hannig, Flora, 191 1, p. 338). 



It is interesting to note that the gametophyte of A. aureiuii, after 

 developing first a spathulate form, establishes a lateral meristem, in the 

 manner described by Von Goebel for Pteris longifolia (Vol. I, p. 276, 

 Fig. 267): but later it may become many-lobed. 



The facts thus summarised for Acrostichiun aiirewn L., and A. praestantis- 

 simwn Bory, fully substantiate the claim that they are Pteroid Ferns. The 

 steps in soral origin of such a tj^pe may be traced from the bi-indusiate 

 sorus as it appears in the Dennstaedtioid Ferns (Vol. II, Fig. 540). These 

 already show a bias of the sorus towards the lower surface of the blade, 

 which is accentuated in Pteridiiivi (Fig. 609). A definite transit to the lower 

 surface together with a loss of the lower indusium resulted in the state seen 

 in Pteris (Fig. 619): the receptacle is seen flattened and widened out in 

 Litobroehia (Fig. 620), and still more fully spread over the lower surface in 

 Aerostichwi praestantissiuuini (Fig, 621); finally it covers the whole of the 

 lower surface of the blade as in A. aureuin. The two species last named, 

 together with A. fascieulatum C. Chr., are now the sole representatives of 

 that great mass of Ferns that were congregated, by systematists who 

 followed a single leading character, under the comprehensive name of 

 AcrostieJium. It is now clear that the Acrostichoid sorus is not the preroga- 

 tive of one natural group alone, but of many. It cannot be held as a generic 

 character from the phyletic point of view: it connotes rather a state or 

 condition of the sorus that has been achieved along many distinct evolu- 

 tionary lines. It is well that the steps of its origin can be so conclusively 

 demonstrated in the Fern to which Linnaeus gave the name in 1753: so 

 that in it we return, after the lapse of nearly two centuries, and great 

 systematic confusion, to a generic designation which can be upheld in its 

 original purity. 



Returning now to the list of genera included by Diels under his heading 

 of Pterideae-Pteridinae, here designated the Pteroid Ferns, those which 

 have been accurately examined from the phyletic point of view may be 

 disposed according to the results obtained, as follows: — 



