1. Aciphylla Scott-Thomsonii Ckn. et Allan in Trans. N.Z. Inst. 57 (1926) 48. The flowers were described as “albi.” Fresh material shows that this should be amended to “luteoli.”
2. Celmisia Lyallii Hook. f. in Handbk. N.Z. Fl. (1864) 133. The localities cited by Hooker are: “Dusky Bay, Lyall; Hurrumui [Hurunui] valleys, 1200–1600ft., Travers. Abundant in the Rangitata, Two-thumb, and Malvern ranges, alt. 3–5000ft., and alps of Hopkins River, alt. 2500ft., Sinclair and Haast; Otago, lake districts and Lindis Pass, displacing the tussock-grass at 4000ft., Hector and Buchanan.”
All the specimens that Hooker had under examination, except that of Lyall, belong to C. Lyallii as described by Cheeseman in Man. N.Z. Fl. (1925) 947, and as understood by all local botanists. Lyall's specimen, however, belongs to the distinct species C. Petriei Cheesem. in Man. N.Z. Fl. (1906) 311 and (1925) 947. Hooker's description is so drawn up as to include the leaf-measurements of Lyall's specimen. As the bulk of his material comes under C. Lyallii as at present understood, it appears best to select the type from this, thus allowing the retention of the name C. Petriei Cheesem. for the species to which Lyall's specimen belongs. I take as the lectotype, therefore, the specimens in Herb. Kew collected on Mount Alba by Hector.
3. Clematis foetida Raoul var. B? depauperata Hook. f. in Fl. Nov.-Zel. i (1853) 7. This variety was based on specimens collected by Colenso at Lake Rotoatara. In the Handbook (p. 2) there is added a specimen collected by Travers in Canterbury. Both gatherings belong to C. marata J. B. Armstrong.
4. Forstera Mackayi Allan, sp. nov. Affinis F. sedifoliae, sed foliis linearibus, subacutis differt.
Herba perennis; caules subrepentes, radicantes, c. 4–15 cm. longi, c. 4 mm. crassi, foliis sessilibus, confertis vestiti. Folia imbricata, laminis primo patentibus demum reflexis, glabris, c. 10 mm. longis et c. 2.5 mm. latis, acutis vel subacutis, linearibus, coriaceis, marginibus incrassatis supra nitidulis, subtus basi versus costa prominente c. ·75 mm. lata. Pedunculi axillares, sub apicibus caulorum dispositi, tenues, c. 6 cm. longi, 1–3 flori, bracteis 1–3, linearibus, sub floribus suffultis, c. 4.5 mm. longis et 1.5 mm. latis. Calyx subturbinatus,
tubo c. 3 mm. longo, c. 25 mm. lato, lobis c. 6 mm. longis et 1.5 mm. latis, obtusis vel subacutis, apicibus ciliatis, basi versus sparse capillis mollibus vestitis. Corolla tubo c. 2 mm. longo, lobis (5) 6 (7), c. 8 mm. longis, 4 mm. latis, obovato-cuneatis, albis, basi versus rubro-maculatis, glandibus 2–3, elevatis linearibus. Columna tener, c. 4 mm. longa, antheris albis, c. 1.5 mm. longis, subreniformibus. Glandes epigenae subulatae.
South Island: North-western Bot. Dist.—Summit of Sewell Peak, Paparoa Mountains, W. Mackay! Type in herb. Plant Research Station, No. 8961.
In the Cheeseman herbarium at the Auckland Memorial Museum this species is represented by the following collections (placed under F. sedifolia by Cheeseman): Mount Frederic, near Westport, P. G. Morgan!; Mount Rochfort, near Westport, W. Townson 291! It thus appears to be endemic to the North-western South Island Botanical District. The leaf-bases are appressed to the stems, and at leaf-fall remain investing them. Mr Mackay informs me that the mid-ribs are obscure when the leaves are fresh, and draws attention to the prominent gland on the back of each leaf near the apex. This gland is probably in the nature of a hydathode. I take pleasure in associating this new and handsome species with the name of one who has done much to advance our knowledge of the botany of the north-west of South Island.
5. Gaimardia minima (T. Kirk) Cheesem. in Man. N.Z. Fl. (1925) 288; Centrolepsis minima T. Kirk in Trans. N.Z. Inst. 23 (1891) 441.
If the species be maintained in Gaimardia the specific name is invalidated by Gaimardia minima Colenso in Trans. N.Z. Inst. 22 (1890) 491, although, as the Director of the Royal Botanic Gardens, Kew, has informed me, Colenso's specimens in the Kew herbarium show his plants to be the single-flowered forms of the grass at present included in New Zealand floras under Zoysia pungens Willd. But Cheeseman's species belongs to the genus Alepyrum Hieronymus (non Alepyrum R. Br.). Until a more satisfactory revision of this difficult group of the Centrolepidaceae is forthcoming, it does not seem wise to give a new name to Cheeseman's species.
6. Hebe acutiflora (Benth.) Ckn. in Trans. N.Z. Inst. 60 (1929) 468; Veronica acutiflora Benth. in D. C. Prodr. x (1846) 460. Cockayne and Allan (Trans. N.Z. Inst., 57, 1926, p. 23), basing their views on a series of specimens collected by Petrie at Kerikeri Falls (Bay of Islands), coupled with the “variability” of the specimens in Cheeseman's herbarium, rejected this species, considering it based on hybrid material. An examination of the type, however, shows that the species is valid, having the distinctive features pointed out by Bentham, as compared with H. angustifolia and H. ligustrifolia, “pedicelli longiores et imprimis calyces multo majores, 2 lin. longi corollae laciniae acutae.” Specimens sent from Kerikeri Falls by Mr G. Reid agree with the type specimens, but his gatherings show also that a polymorphic series of forms occurs there in which probably all the species H. acutifolia, H. ligustrifolia, and H. salicifolia are involved.
7. Hebe pimeleoides (Hook. f.) Ckn. et Allan in Trans. N.Z. Inst. 57 (1926) 38. Cockayne and Allan, in transferring Veronica pimeleoides Hook. f. to Hebe, treated the type as Hebe pimeleoides var. minor (Hook. f.) Ckn. et Allan. Cockayne was familiar with the jordanon growing about Lake Heron, whence came the specimens on which Hooker's var. minor was based (Handbk. N.Z. Fl., 1867, p. 738), and considered it different in no way from the common more or less prostrate jordanon on which the original V. pimeleoides was based. An examination of the material that Hooker had at his command shows that the jordanon Cockayne and Allan had in mind agrees with the type specimens from “Port Cooper, Lyall” (Fl. Nov.-Zel., i, 1853, p. 195). But the specimens from Lake Heron collected by Haast, though closely allied, are not identical with the type. They differ in the points Hooker stresses, and the habit of the plant appears to be nearer that of H. Raoulii than of H. pimeleoides sens, strict. Specimens collected by Mr R. M. Laing on Mount Potts (herb. Plant Research Station, N.Z. No. 10, 750) rather closely approach the type specimen of var. minor Hook. f. The status of var. minor remains uncertain, and further examination should be made of the locality whence it came.
8. Ranunculus Buchanani Hook. f. This was described by Hooker (Handbk. N.Z. Fl., 1864, p. 5) from specimens collected by Hector and Buchanan in “Otago, Lake district, in large patches, alt. 5–6000ft.” He doubtfully adds specimens collected without flowers by Haast at “Macaulay River and Waimakeriri country, alt. 2–5000ft.” There is therefore no doubt as to what should be treated as the type.
Investigations by Messrs Geo. Simpson and J. Scott Thomson show that two well-marked jordanons occur, which may be differentiated as follows:—
(a) var. latilobus Allan var. nov. Folia basalia reniformia latilobata, lobis 2–5 segmentis; segmenta ± 1 cm. lata, abrupte acuta.
The upper leaves are also comparatively broadly lobed or cut. The specimens of Hector and Buchanan belong to this jordanon, as do specimens in the Kew herbarium labelled “Cockayne 7006” from the Minarets. Messrs Thomson and Simpson have observed it about the headwaters of the Estuary Burn (Lake Wanaka). Thus the known localities lie within the south-western corner of the Eastern South Island Botanical District.
(b) var. multifidus Allan var. nov. Folia basalia angustilobata, lobis 2–7 segmentatis; segmenta linearia, ± 5 mm. lata, sensim attenuata. Type in herb. Plant Research Station, no. 5338, coll. J. Scott Thomson and G. Simpson, Bold Peak (Fiord Bot. Dist.). The whole foliage is much more finely-cut than in var. latilobus, giving the jordanon a very distinct appearance. It occurs also on the mountains to the west of Lake Wakatipu, and thus extends into the South Otago Botanical District. Hybrids between this jordanon and R. Lyallii and R. Scott-Thomsonii are illustrated by Allan (Genetica VIII, 1926, p. 535).
9. Raoulia bryoides Hook. f.
The species was described by Hooker (Fl. Nov.-Zel., ii, 1855, p. 332) from specimens collected by Monro on the summit of Gordon's Knob, Nelson. The type specimen in the Kew herbarium well fits Hooker's description (Pl. 26, Fig. 1). He makes no mention of leaf-nervation. Kirk (Student's Fl., 1899, p. 307) says, “1–3-nerved.” Beauverd (Bull. Soc. bot. Genève, 2me série, 4, 1912, p. 51) states that the leaf “présente généralement de chaque côté de la très forte nervure médiane une ± faible nervure latérale s'arc-boutant sous l'anastomose apicale,” and places the species in his section “trinerves.” Cheeseman (Man. N.Z. Fl., 1925, p. 968) accepts this position, but in his description makes no reference to the nervation. After an examination of very many specimens I have been unable to find any example of true R. bryoides bearing three nerves, and
Text-Figure 1.—Venation in Raoulia leaves. Mag. 6 diam. 1. R. bryoides; 2. R. mammillaris; 3. R. Brownii; 4. R. eximia (type); 5. R. eximia (narrow-leaved form); 6. R. eximia var. lata.
consider that the species belongs to the uninerves, where it forms a very natural group with R. eximia and R. mammillaris. Owing to pressure in the congested branches the leaves have a tendency to crease on both sides of the middle line. The creases often simulate real lateral nerves, if not examined by transmitted light.
10. Raoulia eximia Hook. f.
Hooker (Handbk. N.Z. Fl., 1864, p. 149) based the species on specimens from “Riband-wood range, Mount Arrowsmith and Dobson, alt. 5500–6000ft., Sinclair, Haast.” The Ribbon-wood Range specimen of Sinclair, now in fragments, fits Hooker's description and is here taken as the type (Pl. 26, Fig. 2). Beauverd (Bull. Soc. bot. Genève, 2me Série, 2, 1910, p. 231) described and figures a specimen
collected by Berrgren on Mount Torlesse. This is longer—and much narrower—leaved than the type, and has the inner involucral bracts more distinct. This narrow-leaved form is represented in my collection by several examples (Pl. 26, Fig. 3). Kirk (Students' Fl., 1899, p. 304) described a var. lata, remarking: “This is only known to me from a specimen collected by Enys and forwarded by the Director of the Royal Botanic Gardens, Kew, without mention of locality, as R. eximia, with the following note by Mr N. E. Brown: ‘This may be a variety of R. eximia, but I doubt it; the heads seem smaller, the hairs shorter, and the phyllotaxis different.’ In the absence of flower-heads I am unable to separate it from R. eximia.” Cheeseman (loc. cit., p. 970) had not seen the specimen. I was unable to find any collection by Enys at Kew, but a collection of Raoulias discovered in the herbarium of the Department of Agriculture when I took over its charge included the specimen to which Kirk refers, in its original envelope (Pl. 26, Fig. 4). The statement on the envelope is “Dr. Haast, from specimens in case.” Kirk also (loc. cit., p. 307) very briefly described his R. Brownii and remarks: “Sent to me as R. mammillaris by the Director of Kew, with the following note by Mr N. E. Broun: ‘This is certainly not R. mammillaris Hook. f.; it appears to be a new species.’ I have not met with it. In all probability it was collected by Enys on Mount Torlesse or on one of the lofty peaks above the Broken River.” The specimen is still in its original Kew envelope as sent, and is labelled “R. Brownii T. Kirk,” in Kirk's writing (Pl. 27, Fig. 5). I can find no features sufficient to separate it from R. eximia. While many of the leaves may be described as subacute, others are more broadly rounded. A like difference may be met with in many specimens of R. eximia, even in the specimen labelled var. lata. In the investiture of hairs both var. lata and R. Brownii agree with eximia.
Apart from these two aberrant forms there are clear-cut differences between the type and the form described by Beauverd. Until a more exhaustive study is made of the polymorphy of R. eximia it does not seem wise to give varietal names.
11. Raoulia mammillaris Hook. f. in Handbk. N.Z. Fl. (1864) 150.
The type is a specimen collected by Haast on Mount Torlesse. It agrees well in structural details with Hooker's description (Pl. 27, Fig. 6). Beauverd (loc. cit., 1910, p. 233) based his description and illustration on a specimen collected by Berrgren. A specimen of Berrgren's (Pl. 27, Fig. 7) from “Mount Torlesse, Feb. 1874” in the British Museum of Natural History very well agrees with the type. Beauverd states (loc. cit., 1910, p. 233): “Comme caractère différentiel important, la présence de poils glanduleux (cf. fig. xiv: 11) noirs mélangés aux poils laineux roussàtres (cf. fig. xiv: 10, a, b) du P. mamillare mérite d'autant plus d'ětre relevée qu'elle était, sauf erreur, restée inédite jusqu' à ce jour.” These hairs with blackened tips are very noticeable in the Berrgren specimen in the British Museum, but they are not a constant feature of R. mammillaris, nor are they confined to that species. They may be met with to a greater or less extent also in R. eximia and R. rubra at least.
They are in fact diseased hairs, attacked by one of the fungi imperfecti. As Beauverd had only seen the narrow-leaved form of R. eximia he naturally laid stress on the differences in leaf-size and shape between it and R. mammillaris. The figures here given show the marked differences in the hairy covering. In comparing the leaves of the different species of Raoulia it is important to choose leaves in a similar stage of maturity. I have selected the uppermost fully developed leaves. Over-mature leaves often have the hairy covering in part removed. There are in my collection a number of specimens not typical of any of the three species here remarked on, and I consider it very probable that further study will show that the species hybridize.
Wall (Rec. Cant. Mus., 2, 1923, pp. 105–109) has paid considerable attention to R. mammillaris, and after detailed field study has proved that R. mammillaris is not, as Hooker thought, a large, but a comparatively small plant, “not more than 2 feet or 18 inches long, and about 6 inches high.” My own observations fully support this. In addition to the stations noted by Wall R. mammillaris is frequent on Mount Hutt, in company with R. eximia.
12. Raoulia Goyeni T. Kirk.
Kirk (Trans. N.Z. Inst., 16, 1884, p. 373) describes the leaves as with “close set whitish hairs on the apical half of the upper surface.” In his later description (loc. cit. 1899, p. 306) he says: “The apical portion of the upper surface clothed with a dense brush of short white uneven hairs slightly exceeding the margins, glabrous at the base.” Beauverd (loc. cit. 1912, p. 47) says, “glabres extérieurement, sauf quelques long cils épars vers la base, très glabres interieurement, sauf une louppe dense de très long cils au sommet.” Cheeseman (loc. cit. 1925, p. 970) says, “glabrous beneath.” I find the hairs on the lower half of the dorsal surface to be constantly present, and in fully developed but not over-mature leaves the brush may be fairly dense (Pl. 27, Fig. 8). This relates the species more closely to R. Buchanani than has so far been recognized.
13. Nomenclature of the New Zealand Species of Rubus.
In his Prodromus of 1876 G. Forster described his Rubus australis from New Zealand specimens. In his Precursor of 1839 A. Cunningham described two further species R. schmidelioides and R. cissoides. In the Flora of New Zealand of 1853 Hooker united all the New Zealand forms known to him under the name R. australis Forst., assigning varietal rank to Cunningham's species, and giving the varietal name glaber to R. australis Forst. f. sens. strict. This treatment he maintained in his Handbook of 1864. Although he distinguished the “varieties” essentially as the “species” had been, he remarks [Fl. N.Z., i (1853), p. 53]: “I am quite unable to distinguish the above varieties specifically, and, indeed, as varieties they present inconstant characters.” Field studies by local botanists did not support this lumping of all forms under one species-name. Kirk in his Students' Flora of 1899 accordingly restored R. cissoides and R. schmidelioides to specific rank, but quite misunderstood the application of the names. The mistake appears to have arisen from his assuming that the plate in the Flora Novae-Zelandiae represented
Hooker's variety glaber, though neither Hooker's description nor his citation of the plate suggests this. This incorrect application of the names was followed by Cheeseman in 1906 and in 1925, as well as by all local botanists. An examination of the type material revealed this confusion, and unfortunately makes it necessary to reapply the names as originally intended. The following is intended to clear up the situation concerning these three species, as a preliminary to a revision of the genus now in hand.
a. Rubus australis Forst. f. in Prodr. (1786) 40; A. Cunningham, in Ann. Nat. Hist. iii (1839) 245; R. australis Forst. var. α glaber Hook. f. in Fl. Nov.-Zel. i (1853) 53, and in Handbk. N.Z. Fl. (1864) 54; R. schmidelioides Kirk, in Students' Fl. (1899) 126, non R. schmidelioides A. Cunn.; R. schmidelioides Cheesem. in Man. N.Z. Fl. (1906) 125 and (1925) 501 (excl. var. coloratus T. Kirk), non R. schmidelioides A. Cunn.
Forster's diagnosis is utterly inadequate in view of the multiplicity of forms of Rubus found later on in New Zealand, but his specimens leave no doubt as to the group to which the name should be applied. The specimen in the British Museum of Natural History labelled “J. R. Forster. Herb. Pallas” has two pieces (Text-fig. 2a). In the herbarium of the Royal Botanic Gardens, Kew, there are (1) a sheet labelled “Forster Herbarium. Presented by the Corporation of Liverpool. R. australis”; (2) a sheet labelled “R. australis Forst. f. No. 567 A. Cunn.,” and further labelled by Hooker, “var. glaber Hook. fil.” These gatherings quite definitely belong to the same species as that of the British Museum specimen, and fix the type of R. australis Forst. f. and var. glaber Hook. f. Cunningham collected his Rubus specimens in the neighbourhood of Whangaroa and the Bay of Islands. He lists R. australis Forst. f. (No. 567) with a fuller description, which along with the specimen above cited, shows that he correctly understood Forster's species. He gives as the locality of his specimens, “margins of forests, Wangaroa.—1826, A. Cunningham.”
R. australis Forst. f., then, is the species treated in the floras of Kirk and Cheeseman, and in recent botanical literature generally, as R. schmidelioides.
b. Rubus schmidelioides A. Cunn. in Ann. Nat. Hist. iii (1839) 245; R. autralis Forst. f. var. β schmidelioides Hook. f. in Fl. Nov.-Zel. i (1853) 53, and in Handbk. N.Z. Fl. (1864) 54; R. cissoides Kirk var. coloratus Kirk in Students' Fl. (1899) 126, non R. cissoides A. Cunn.; R. schmidelioides Cheeseman var. coloratus (Kirk) in Man. N.Z. Fl. (1906) 126 and (1925) 501.
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
Cunningham gives for his species R. schmidelioides (No. 568) “Forests at Wangaroa and Bay of Islands.—1826, A. Cunningham.” To his description he adds the remarks: “Habitus praecedentis, sed differt foliolis omnino ternatis rugosis venosis subtus (discoloribus) valde ferrugineo-tomentosis, racemoque multo breviore.” There is a specimen in the herbarium labelled “R. schmidelioides A. Cunn. 38/1826. Allan Cunningham's New Zealand Herbarium. Presented by Robert Heward, Esq. 1862. No. 568.” (Text-fig. 2c.) There is another specimen of the same
species labelled “R. Cunningham, Whangaroa.” In the British Museum herbarium is a specimen labelled “R. schmidelioides, Whangaroa. 1826, A. Cunningham. Presented by the Linnean Society.”
The description and specimens prove clearly that Cunningham's species belongs to the group understood by Kirk and Cheeseman as “var. coloratus.” Kirk (loc. cit. 1899, p. 126) says of his variety, “most plentiful in the South Island.” The three named specimens
Text-Figure 2.—All leaves natural size. a. Two leaves of Rubus australis Forst. f. Type. b. Leaf of R. cissoides A. Cunn. Type. c. Two leaves of R. schmidelioides A. Cunn. Type.
in his herbarium, however, were gathered in the neighbourhood of Wellington. Kirk included both the broader-leaved (answering to Cunningham's type of R. schmidelioides) and the narrow-leaved form in his conception. This narrow-leaved form was later described by Cockayne as R. subpauperatus.
c. Rubus cissoides A. Cunn. in Ann. Nat. Hist. iii (1839) 245; R. australis Forst. f. var. γ cissoides Hook. f. in Fl. Nov.-Zel. i (1853) 53, and in Handbk. N.Z. Fl. (1864) 54; R. australis Kirk in Students' Fl. (1899) 125, non R. australis Forst. f., non R. australis Forst. f. var. glaber Hook. f.; R. australis Cheeseman in Man. N.Z. Fl. (1906) 125 and (1925) 500, non R. australis Forst. f., non R. australis Forst. f. var. glaber Hook. f.
Cunningham gives for his R. cissoides (No. 569), “Dense forests of Wangaroa.—1826, A. Cunningham,” and adds to his description: “Habitus omnino Cissi. Folia praesertim quinata longe petiolata, valde nitida et pulcherrime venosa, serraturis acuminatis.” In the Kew herbarium is a sheet labelled “Wangaroa New Zealand. 1826, A. Cunningham. Presented by the Linnean Society 1915.” (Text-fig. 2b.) There are other sheets of the same species in the Kew and the British Museum herbaria labelled as collected by A. Cunningham, and by R. Cunningham. The narrow-leaved male plant figured by Hooker (Fl. Nov.-Zel. i, t. xiv) well represents R. cissoides A. Cunn., while the female spray figured represents a closely allied broader-leaved form. These, and similar forms, make up the compound species understood by Kirk and Cheeseman to be R. australis Forst. f. The cissoides group, however, is perfectly distinct from the Forsterian species.
d. Rubus squarrosus Fritsch, in Oesterr. Bot. Zeitschr. xxxvi (1886) 259; R. cissoides Kirk in Students' Fl. (1899) 126, non R. cissoides A. Cunn.; R. cissoides Cheeseman in Man. N.Z. Fl. (1906) 125 and (1925) 500, non R. cissoides A. Cunn.
Fritsch's species is based on individuals formerly cultivated in the Vienna Botanic Garden. It is further illustrated in Kerner's Pflanzenleben and in Weisner's Organographic und Systematik. It is wrongly given in Index Kewensis Supp. 1 (1906) 372 as a synonym of R. australis Kerni (a South European species).
From the description and figures there is no doubt that R. squarrosus Fritsch is the species wrongly taken by Kirk and Cheeseman, and by New Zealand botanists generally to be R. cissoides A. Cunn., and as Fritsch's name appears to be the earliest, it is here taken up.
e. Rubus parvus Buch. in Trans. N.Z. Inst. vi (1874) 243, and R. subpauperatus Cockayne in Report Bot. Stewart Id. (1909) 42. There has not been any confusion in the application of these names, and they require no comment here.
The following table summarises the necessary changes:—
[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]
|Names as in Cheeseman's Manuals of 1906 and 1925.||Names according to Type Specimens.|
|1. R. australis||R. cissoides A. Cunn.|
|2. R. cissoides||R. squarrosus Fritsch.|
|3. R. schmidelioides||R. australis Forst. f.|
|4. R. schmidelioides var. coloratus||R. schmidelioides A. Cunn.|
14. Senecio Solanderi Allan, nom. nov.; S. latifolius Banks et Sol. ex Hook. f. Fl. Nov.-Zel. i (1853) 145, non S. latifolius D.C. in Prodr. vi (1837) 387. The name S. latifolius for the New Zealand species is invalidated by the earlier homonym of De Candolle.