616 FILICALES 
of the sori and the habit, as well as by certain anatomical features. The 
genus appears to hold an intermediate position between the Dennstaedtiinae 
and some Ferns referred to Polvpodium: the relationship to’ the latter 
has already been the subject of remark; for instance, in the Synopsis 
Filicum, p. 130, Dr. Griesebach is quoted as writing of A. Purdieana, Hk. : 
“Not to be distinguished from P. rugudoswm but by the specially transformed 
involucral appendages, and probably passing into that widely-ranging species.” 
Again,! under Polypodium (Pheg.) punctatum, Thunb. (which Hooker 
regarded as including P. rugudosum, Labill), he remarks : “Very closely 
related to ZLuvhypolepis.” All this seems to indicate a probable sequence 
which would consist of (a2) some Dennstaedtiinous Ferns with basipetal 
sori, (4) some type with mixed sorus, and receptacle within the margin 
which is curved over as an indusium, as in A/ypolepis: (c) such a type as 
Polypodium punctatum, Thunb., with 
its definitely superficial, unprotected 
sorus, having sporangia with ages inter- 
mixed, and no regularity of orientation. 
It may next be enquired how far 
the anatomical data will support the 
results of examination of the sori in 
this series. It cannot be assumed that 
characters so distinct as those of the 
sorus and of the vascular system 
Fic. 342. must necessarily run parallel; but if 
 Pleriselata,v. Narsteniana. Diagram showing they do, it 1s a strong support of the 
the arrangement of the vascular tissue at the ae q ‘ 
insertion of a leaf. A piece is supposed to be cut Correctness of recognition of a phyletic 
out of the side of the solenostele, so as to show the : 
internal vascular system. Note that a small strand, line. Gwynne-Vaughan has found that 
lying within the second vascular ring, is also present. 
(After Gwynne-Vaughan.) in every species of the  sub-tribe 
Dennstaedtiinae in which the anatomy 
is known the same type of primitive vascular system, the solenostelic, is 
seen: this indicates the primitive nature of Prantl’s sub-tribe as a whole. 
The statement applies for all species of Deznstaedtia examined, but in 
D. rubigtnosa the solenostele is not quite typical, additional vascular 
strands being also present: this is, however, the very species in which 
an approach to a mixed sorus is found: thus, the two characters indicate 
that plant as an advance upon the rest. The approach is towards a 
condition seen in species of Fver’s, where with a mixed sorus there is a 
still more elaborate accessory vascular system within the original soleno- 
stele (Fig. 342). 
Turing to the Aypolepis-Polypodium line above noted, the anatomy 
again supports the relationship. A7yfolefrs is solenostelic : so is P. punctatum : 
in fact this species stands structurally isolated among the Polypodieae, and 
is evidently related closely to Ayfoleprs, which in turn is related to the 
other Dennstaedtiinae.* 
1 Syn. Filte., p. 312 “Gwynne-Vaughan. Zc., p. 735. 
