XXXVI] HYPOLEPIS ii 



comparing such types on the spot in New Zealand, says ( Trails. N. Z. Inst. 

 Vol. XLVII, p. 85), "In some forms oi Hypolepis the spurious involucre (= inner 

 or lower indusium) is hardly or not at all developed, and sometimes it appears 

 slightly in Polypodium, while the sori of the latter are frequently and dis- 

 tinctly marginal." The fact appears to be that these Ferns are variable in 

 the features of the sorus as well as in its exact position. From a phyletic 

 point of view there is no obligation to draw any sharp line between Hypolepis 

 and the ferns styled Polypodiiim rugulosum or pimciatum. That they are 

 very closely related, notwithstanding soral differences, is shown by the habit- 

 similarity, by the absence of scales and the presence of simple hairs, and by 

 the practically identical vascular structure, as noted by Gwynne-Vaughan 

 {Ann. of Bot. XVII, p. 694). We may take it that in this nearly related series 

 those forms which are more conservative of the Dennstaedtioid characters have 

 been designated Hypolepis, while the more advanced forms have been referred 

 to PJiegopteris ox Polypodinvi; but still the series consists of Ferns naturally 

 akin. 



The sporangia of Dicksonia and Thyrsopteris are relatively thick-stalked, 

 with complete oblique annulus consisting of numerous cells: in fact they 

 have archaic features. Those of Dennstaedtia approach the usual Lepto- 

 sporangiate type in having a long thinner stalk composed of three rows of 

 cells, while the annulus is almost vertical, being partially or even completely 

 interrupted by the insertion of the stalk, and it consists of relatively few 

 cells (see Vol. Il, Fig. 539, C, D). In Hypolepis also the sporangia have thin 

 stalks and a vertical annulus. There is thus evidence of a progression in the 

 detail of the sporangium itself from an archaic to a more modern type. 

 These Ferns constitute in fact a series, which may be traced in a number 

 of features, from such a source as Cibotiuin: but the vegetative characters 

 remain more constant than the soral, and these features clearly indicate the 

 unity of the whole series, as well as its primitive origin. 



The systematic result of these comparisons, if classification is to be 

 phyletic, can only be the removal of the Ferns designated Polypodiiim 

 {Pheg.) punctatum Thunbg., and Dryopteris (^PJiegopteris^ punctata (Thunbg.) 

 C. Chr. from either Polypodiuni or Dryopteris, and their inclusion under 

 Hypolepis. This follows from the argument given above: and a particularly 

 distinctive fact is that here the dermal appendages are hairs and not scales, 

 such as are a marked feature of the genera designated Dryopteris and Poly- 

 podium. 



We may conclude then that the series Cibotium — Dennstaedtia — Hypolepis 

 provides an example of phyletic advance within a circle of affinity which 

 has been regarded by many systematists as a close one. The chief interest 

 lies in the changes of soral character. In particular, (i) this series carries 

 out in a very convincing way the transition from a typically gradate to a 



