Gymnogramma rutœfolia, Br. G. alpina, Potts, is certainly referable to this species. Specimens collected on the Upper Rangitata and kindly given to me by Mr. Potts, do not present any characters differing from those of G. rutœfolia sent from other parts of the colony. I have it from several localities in Canterbury and Otago, but have seen no North Island specimens. The New Zealand plant is, however, quite distinct from G. pozoi with which it was united in the “Syn. Filicum.”
Lastrea invisa. The Nephrodium thelypteris var. squamulosum of the “Handbook” is undoubtedly the Aspidium invisum of Forster, and the specific
name must therefore be changed as I have no doubt whatever as to its distinctness from the British plant. It has not, so far as I am aware, been found in the Southern Provinces.
Cystoptoris novœ-zealandiœ. This plant is certainly distinct from C. fragilis, Bernhardi, to which it has been referred by most authors. The plant is always much smaller and more fragile with an erect underground miniature branching caudex, and is always less divided than the European plant. The indusium is often entirely absent from Canterbury specimens, and nearly always becomes obsolete with age. I have examined large numbers of British specimens together with living plants of the true C. fragilis, Bernhardi, and have found none approaching this. Some states of it, however, are much like C. dentata in aspect, but differ much in the rhizome and sori. The description in the “Handbook,” copied from the “Species Filicum,” was drawn up from European specimens, and does not apply to the New Zealand plant. It is found from near the sea-level (close to Christ-church) up to 4,800 feet or higher, but presents scarcely any variations, thus offering another point of difference from the British plant, which is exceedingly variable if I may judge from the numerous specimens in my own herbarium.
Dicksonia fibrosa. I am glad to find that Mr. Baker is disposed to admit this old species of Colenso's. We have cultivated the Tasmanian D. antarctica, and this for many years, and I have no hesitation in pronouncing them distinct if examined in the living state, as all ferns ought to be before any opinion is given as to their distinctness.
D. antarctica is the fastest growing tree-fern we have as yet experimented with. Plants only seven and a half years old from spores have made trunks varying from 3 to 10 inches high and 3 to 4 inches through. It is therefore likely that the rapidity of growth of tree-fern stems has been very much under-estimated by most writers on ferns.
Cheilanthes kirkii. An examination of cultivated plants of the true C. tenuifolia, Swzs., has convinced me of the distinctness of the New Zealand plant referred to that species by Mr. Kirk.* In my Sketch of the Flora of Canterbury† I considered this to be a variety of C. sieberi; but on further consideration I am disposed to acknowledge it as a distinct species and to attach to it the name of Mr. T. Kirk, who has done so much to increase our knowledge of New Zealand plants. Very good descriptions of this and C. sieberi will be found in the volume of the “Transactions” referred to above. C. kirkii is rather a common plant in Canterbury, but appears to be somewhat rare in the North.
[Footnote] * “Trans. N.Z. Inst.,” Vol. VI., p. 247.
[Footnote] † “Trans. N.Z. Inst.,” Vol. XII., pp. 325—353.
Asplenium, sp. Probably another species of this genus, found in the lower Canterbury Alps, will yet be added to our list of New Zealand ferns, and admitted to specific distinction. It is intermediate in appearance between A. flaccidum and A. hookerianum, and frequently produces dense masses of sori covering the whole under-surface of the frond. It has been erroneously referred to A. richardi by collectors of living ferns.
Hymenophyllum. The great increase in the number of species of this genus is noteworthy. In the “Handbook” only fifteen species are given, whilst I here acknowledge twenty-six, and I am by no means sure but that I ought properly to have admitted several more. This is undoubtedly the most difficult New Zealand genus of ferns, and requires to be carefully studied in the living state. I think that most of the species here given will be generally admitted in the future, seeing that most of the New Zealand species suppressed in the first edition of the “Synopsis” have been re-admitted in the second. Probably also several new ones remain to be discovered. It will be seen that I have restored Mr. Colenso's old species H. revolutum, which I consider quite as worthy of specific standing as H. unilaterale, H. montanum, and others. H. cheesemanii is closely allied to H. minimum, but differs in the ciliated margins and costa. It is the least satisfactory of all the species adopted here.
H. armstrongii is also allied to H. minimum, but differs essentially in the remarkable border, and in the columnar receptacle being frequently exerted, which led Mr. Baker to place it in the genus Trichomanes (see Synopsis Filicum, first edition in appendix). It is, however, more naturally allied in habit to Hymenophyllum, and was first referred to this genus by Sir. J. D. Hooker on its discovery in 1867.
Of Mr. Colenso's H. erecto-elatum I have not yet seen specimens, but he is such a careful observer, and is so well up in New Zealand plants, that I have little reason to doubt its distinctness. The same author's H. pusillum is certainly very near to H. tunbridgense, but I accept it for the present. I have placed H. lyallii in this genus, although it was referred to Trichomanes in the “Synopsis,” because it appears to me to have more of the habit of Hymenophyllum than of Trichomanes; but it cannot be denied that the New Zealand species of the two genera present a graduated series of forms, and the limits of these two genera are anything but clearly defined. Still, it is convenient to keep up both generic names, and I think it will be found best to refer those species with long, distinctly-exerted, columnar receptacles and densert issue to Trichomanes, and the remainder to Hymenophyllum. H. malingii is a difficult plant to place, but apparently it comes nearest H. œruginosum.
Trichomanes venustulum. This new species of Mr. Colenso's is, I believe, not uncommon in the south, but has hitherto been confounded with T. venosum.
Osmunda. I have here departed somewhat from Smith's arrangement, and united Todea with Osmunda of Linnæus. This genus has given me more trouble than any other, and consequently I do not expect New Zealand pteridologists to agree with my views at once, but I have little doubt that many of them will do so when they have studied the question as much as I have done. There is really no natural character to separate the typical species of Todea from Osmunda. The character usually given is that the fronds of Osmunda are contracted in fruit, but the fronds of Todea are often partially contracted, though never so much as those of Osmunda regalis. Consequently this character is of no value, as we cannot distinguish genera by the amount of contraction of their fertile fronds. In the variation, texture, and in the sori and venation, we find no difference beyond what is sufficient to distinguish them as species of one genus. In the case of the two species of the sub-genus Leptopteris, however, we find characters of more importance, such as the membranous texture of the fronds and the scattered sori; but these characters are not now considered by authors to be of sufficient importance for defining genera, or even for separating Leptopteris from typical Todea, and therefore I feel justified in placing all three plants in Osmunda. I find that Sir J. D. Hooker has, in the “Handbook,” given an opinion that Todea should merge into Osmunda, thus anticipating my action.
In conclusion, it is necessary to remark that I have made a very free use of Smith's generic descriptions;* but I do not suppose that the author would object, as his descriptions are so admirably drawn that it would be quite impossible for me to improve upon them, and I shall conclude with a hope that this paper will be the means of bringing before the New Zealand student the least difficult and most natural system of fern classification ever presented to the public.
[Footnote] * All descriptions omitted, the subject of the paper being “arrangement.”—Ed.