Flora of New Zealand Series

Veronica tumida Kirk in T.N.Z.I. 28, 1896, 521.

Original localities: "Nelson-Mount Rintoul and Ben Nevis, F. G. Gibbs! Mount Starvation, W. H. Bryant! Otago-J. Buchanan! 3,000 ft.-4,500 ft." Type: ?

"Stems prostrate, forming compact depressed patches 6 in.-24 in. diameter; branchlets excessively numerous, 1 in.-3 in. long, strict, erect." Lfy twigs 1·5-2 mm. diam., tetragonous but serrate rather than smooth in outline because of projecting lf-tips. Lvs erect, 1·5 × 1·5-2 mm., shortly and bluntly deltoid above the rather narrow base, the whole tumid and rounded with slightly cucullate tip; marginal cilia inconspicuous. Infls to 12 mm. long; peduncle c. 2 mm. long. Fls in up to 4 pairs, rather laxly arranged, pedicels to 1 mm. long. Bracts 2 mm. long, broadly oblong-ovate. Calyx-lobes 2.5-3 mm. long, narrow-oblong. Corolla-tube = calyx, rather narrow, lobes c. 3 × 2 mm. ♂ and ♀ sts less distinct than in the last three spp. Capsule 3·5 × 2 mm., narrowed towards tip.

DIST.: S. Wairau, St Arnaud and adjacent mountains.

HYBRIDISM

Hybridism plays a great part in the polymorphy exhibited by Hebe in N.Z. Hooker, as long ago as 1854 (Fl. N.Z. 1, 190), stated "the species hybridize with great facility", and hybrids arise freely in gardens. Little has been published about artificial hybrids and their progeny but limited results from interspecific crosses between plants of known chromosone number are recorded by Frankel and Hair (N.Z. J. Sci. Tech. 18, 1937, 669-687). Cockayne and Allan (Ann. Bot. 48, 1934, 39-43) list 43 wild hybrid groups and state "if we err, we feel it is on the side of caution". Cooper (Rec. Auck. Inst. Mus. 4, 1954, 259) presents measurements of specimens from two populations of Hebe, illustrating variation presumably due to hybridism.

Below are assembled formally published names considered to be based on wild hybrid material. A hybrid basis is suspected for several of the names listed above under INCERTAE SEDIS, and references to probable hybrids appear also in notes under species descriptions.

H. adamsii (Cheesem.) Ckn. et Allan in T.N.Z.I. 57, 1926, 15.  Veronica adamsii Cheesem. Man. N.Z. Fl. 1925, 786. Type locality: "North Cape district, open hillsides by the road [sic] between Spirits Bay and Tom Bowline's Bay". Known only from a few flowering twigs collected by Cheeseman and Adams in Jan. 1896, A 7665, 7666, 1545 to Kew. Branches rather stout, lvs c. 4-5 × 1-1·5 cm., subcoriac. with cartilaginous glab. margin; fls rather crowded, large, "purple", the broad corolla-tube little > narrowly acute calyx-lobes, corolla-lobes acute in some, rounded in other specimens. Suggested parents are H. ligustrifolia and H. macrocarpa var. brevifolia which both occur in this general area, and between which the specimens are about intermediate.

H. × affinis (Cheesem.) Ckn. et Allan in T.N.Z.I. 57, 1926, 20.  Veronica macrocarpa var. affinis Cheesem. Man. N.Z. Fl. 1906, 505. Original localities, as given in Man. N.Z. Fl. 1925, 792 : "Headlands in the Waitemata and Manukau Harbours, and the adjacent coasts". The only Cheeseman specimen from the vicinity of Auckland is A 7730, cliffs at Northcote, Auckland Harbour, 1875, T. F. C." Cheeseman states that of the many intermediates between macrocarpa and salicifolia "the one which I have called var. affinis might be referred to either species". Comparing with macrocarpa he says "leaves smaller, 2-4 in, long. Racemes shorter, 3-5 in.; flowers smaller, white. Capsules 1/5-1/4 in., twice as long as the calyx." Calyx-lobes of A 7730 are ± pubescent.

H. × amabilis (Cheesem.) Ckn. et Allan in T.N.Z.I. 57, 1926, 21.  Veronica salicifolia var. gracilis Kirk For. Fl. 1899, 244, t. 120.  V. amabilis Cheesem. Man. N.Z. Fl. 1906, 506. Specimens of Kirk's var. gracilis are in both A and W; lectotype: W, 5321, T. Kirk, The Bluff. "A form of singular beauty forming a small tree 14 ft high with trunk 8" diameter." Lvs to 8 × 2 cm.; corolla-tube short, broad, lobes to 6 × 4 mm. V. amabilis var. blanda Cheesem. Man. N.Z. Fl. 1906, 506 (H. blanda (Cheesem.) Pennell in Rhodora 23, 1921, 39) was distinguished by smaller, more close-set and more distinctly petiolate lvs; lectotype: A, 7804, Port Chalmers, Otago, D. Petrie. Rather similar specimens appear in herbaria from many parts of the coast of Otago and Fiordland, and from Stewart Id, and all could be H. elliptica × salicifolia. Wild hybrids of this group from Blanket Bay, Dunedin, are discussed and figured by Allan, Simpson and Thomson (Genetica 8, 1926, 375-388). H. blanda has been recorded from Southern Patagonia by Pennell (loc. cit.) and from Puerto Balladares, Chile, by Gunckel (Revista Universitaria, Univ. Catolica de Chile 23, 1948, 162); from each of these places H. elliptica is also recorded.

Veronica × bishopiana Petrie in T.N.Z.I. 56, 1926, 15 from "Rocky knobs between Huia Hill and Little Huia, near Manukau North Heads: J. J. Bishop! H. Carse! E. Jenkins! April, 1924"; lectotype: W, 5329. Low straggling rupestral shrub, stems slender, glab., dark purple, widely branched; lvs narrow lanceolate-elliptic, 5-7·5 × ± 1·5 cm.; racemes to 10 cm. long, slender, peduncle long, fls lavender-colour. Recorded by Petrie as V. obtusata × salicifolia. Affinities with H. obtusata are obvious but calyx-lobes are often less pubescent than would be expected if the other parent were H. stricta var. stricta, the commonest representative of "H. salicifolia" sens. lat. in that district.

H. carsei (Petrie) Ckn. in T.N.Z.I. 60, 1929, 469.  Veronica carsei Petrie in T.N.Z.I. 55, 1924, 96. Based on specimens from "Margins of forest and woods, Waimarino Plain" where "the plant, though not uncommon, is rarely found in any great quantity". Cockayne and Allan (T.N.Z.I. 57, 1926, 27) considered this to be (using the names of the present treatment) H. stricta × venustula, vegetatively closer to H. venustula, but with infl. more like that of H. stricta. The group is rather amply represented in A and in W by ± uniform specimens from Waimarino Plains and Tongariro National Park. Lvs. c. 2-3 cm. × 5-8 mm., elliptic or elliptic-oblong, ± polished above, drying red-brown; lf-bud with narrow elongated sinus; racemes 4-8 cm. long, fls ± 8 mm. long, corolla-tube > ciliolate calyx-lobes and corolla-lobes. Only two specimens show tripartite branching of infls as in H. venustula; in few is there any sign of pubescence on surface of calyx-lobes that would be expected if one parent were H. stricta var. stricta which is probably the commonest long-lvd Hebe in that area.

H. cassinioides (Petrie) Ckn. in T.N.Z.I. 60, 1929, 469.  Veronica cassinioides H. J. Matthews ex Petrie in T.N.Z.I. 47, 1915, 52. Described from a plant (W 5324) cultivated in Dunedin by H. J. Matthews and said by him to have originated in the Takitimu Mts. Petrie delayed publication until he could cite not only cultivated plants but also wild specimens collected on the Garvie Mts by Poppelwell and Steadman. Specimens, also in W, from the Garvie Mts plants cultivated at Waikaia School do not in fact well match those of Matthews. Cockayne and Allan (T.N.Z.I. 57, 1926, 42) and Cockayne (loc. cit.) consider that the two represent similar but not identical hybrids resulting from the crossing of a whipcord Hebe with a small-lvd sp. such as H. buxifolia, and they restrict the epithet cassinioides to the commonly cultivated form originally grown by Matthews. Later (Ann. Bot., Lond. 48, 1934, 40) they refer the cultigen to the hybrid group H. buxifolia × lycopodioides, noting again that the wild Garvie Mts plants probably belong to a different group.

Various rather similar hybrids result from crosses between whipcord hebes and small-lvd spp. such as H. odora, H. pauciramosa, and H. venustula. Characteristically these hybrids are low much-branched shrubs with sessile, keeled, ± imbricate lvs 3-4 mm. long, and compact terminal, sts branched, infls with large bracts and sessile fls. Cheeseman (Ill. N.Z. Fl. 2, 1914, t. 153) figures a plant of this kind which he refers to as a persistent juvenile form of H. tetragona; on other occasions non-flowering hybrids have been mistaken for juvenile whipcords. Probable parents can us. be suggested only when the spp. growing in company with the hybrid are known. Cockayne and Allan (loc. cit. 1934) list the following in addition to H. buxifolia × lycopodioides : astonii × buxifolia, astonii × laevis (× H. laevastonii Ckn. et Allan), buxifolia × hectori, buxifolia × tetragona, laevis × tetragona. Simpson and Thomson (T.R.S.N.Z. 72, 1942, 27) describe a plant collected on the Garvie Mts to which they ascribe the origin H. buxifolia var. pauciramosa × propinqua (BD 82999).

H. dartonii (Petrie) Ckn. in T.N.Z.I. 60, 1929, 470.  Veronica dartoni Petrie in T.N.Z.I. 55, 1924, 98 (also 436, a slightly different description). Type: W. 5325, "Firewood Creek near Cromwell, in bottom of valley on rocky slope, mid-Jan. 1911, D. P." a fruiting twig, the capsules glab. Also represented in Herb. Petrie by wild plants from rocky banks of Clutha River, at Roxburgh Bridge. H. L. Darton, Nov.-Dec. 1922 and March 1923, ovary and capsules slightly pubescent. BD 78104 is from a more recent collection from the type locality. Shrub 1-2 m. tall; lvs c. 25 × 8-10 mm., lanceolate, rather abruptly narrowed to acute tip; infls lateral, simple, 4-6-(15) cm. long; calyx-lobes acute, strongly ciliate; capsule acute. Considered by Cockayne to be "probably H. pimeleoides × salicifolia"; the former sp. is suggested particularly by the pilose infl. of Darton's flowering specimens. At one time well-known in gardens.

H. divergens (Cheesem.) Ckn. in T.N.Z.I. 60, 1929, 470.  Veronica divergens Cheesem. Man. N.Z. Fl. 1906, 502.  Ill. N.Z. Fl. 2, 1914, t. 148. Type: A, 7691, "Brighton, south of Westport, W. Townson". Lvs are mostly c. 2.5 × 1 cm., raceme much > lvs, corolla-tube short, lobes c. 3 × 2 mm. Cockayne says "possibly H. elliptica × salicifolia though this may not be so, since it differs greatly from any form of H. ellipsala". H. elliptica × gracillima would be expected to produce something like this.

H. × kirkii (J. B. Armst.) Ckn. et Allan in T.N.Z.I. 57, 1926, 19.  Veronica kirkii J. B. Armst. in N.Z. Ctry J. 3, 1879, 58.  V. salicifolia var. kirkii (J. B. Armst.) Cheesem. Man. N.Z. Fl. 1906, 504. Based on plants from Upper Rangitata. Specimens not now to be found in Herb. Armstrong but Armstrong specimens in various herbaria (K, W, CM) agree with the original description. Lvs of rather heavy texture, c. 25-30 × 8 mm., without evident stomata; lf-bud without sinus; racemes 5-11 cm. long; calyx-lobes ± acute, corolla-tube short, ovary ± pubescent. Specimens collected by H. H. Allan in the same vicinity and with similar lvs differ in minor points, especially in quite glab. capsules. Probably, as suggested by Cockayne and Allan, a hybrid between H. salicifolia and one of the small-lvd spp., most likely H. rakaiensis.

Veronica lewisii J. B. Armst. in T.N.Z.I. 13, 1881, 357 from "Downs near the sea in the south of Canterbury". Two sheets in Herb. Armstrong (CM) from "Timaru Downs, J. F. A." and "Timaru" and one in W 5322 "South Canterbury, near the sea, J. B. A." have lvs 3-6 × 1-2 cm., corolla-tube short, broad, lobes to 5 × 4 mm. These could well be of the same origin, H. elliptica × salicifolia, as H. × amabilis.

Veronica loganioides J. B. Armst. in N.Z. Ctry J. 3, 1879, 59. Based on specimens collected in the Upper Rangitata by J. F. Armstrong in 1869, type in Herb. Armstrong (CM). Plants were vegetatively propagated by J. B. Armstrong in Christchurch and widely distributed to other gardens. Stems woody, decumbent and rooting with bifariously pubescent ascending branches to c. 20 cm. tall; lvs spreading or ± imbricate, sessile, to c. 5 mm. long, ± ovate-acuminate, keeled and ± sheathing stem at base, entire or with 1-2 pairs of small teeth; infls us. lateral and pedunculate, occ. terminal, with large bracts and shortly pedicelled or sessile fls; corolla white, 7-8 mm. diam., tube very short; only a few ± abortive capsules seen and these laterally, compressed and emarginate. Cockayne and Allan (Ann. Bot., Lond. 48, 1934, 42) consider the plant to be "a remarkable hybrid between a whipcord Hebe and Veronica Lyallii . . . It has the habit of a whipcord Hebe cross with the capsule and flower of a Veronica." They do not attempt to name the sp. of Hebe concerned. No further information is available about the possible parents, but some features suggest H. pimeleoides. A collection from the Garvie Mts (BD 63345) referred by G. Simpson, along with a quite different plant, to H. propinqua × buxifolia var. pauciramosa appears to be a hybrid of related origin, but the lvs are broader and more distinctly toothed.

H. × simmonsii (Ckn.) Ckn. et Allan in T.N.Z.I. 56, 1926, 20.  × Veronica simmonsii Ckn. in T.N.Z.I. 48, 1916, 202. Based on plants from French Pass and Pelorus Valley and found "only when in company with V. angustifolia and V. salicifolia var. Atkinsonii". Specimens in CM and W show intermediate characters. Type: W, 5323, French Pass, shore scrub, L. Cockayne No. 8117; isotype in CM.

FL. 1.

From Herb. Kirk come two specimens without collectors' names: W 5348 without infl. from Ben Nevis and W 5349 with fl. from Mt. Starveall (? = Starvation). Buchanan's specimen, labelled Otago, is in Herb. Kirk, but the sp. is not otherwise known outside Nelson.

HORTICULTURAL FORMS

The genus Hebe contains many ornamental shrubs of easy cultivation and many spp. and hybrids are common in gardens both in N.Z. and abroad. Wall (T.N.Z.I. 60, 1929, 383) lists some 75 spp. cultivated in Britain. Crossing occurs frequently, often between spp. that would not meet in nature, and the nomenclature and origins of many of the cultigens are confused. Plants are readily propagated by cuttings and many hybrid forms have been maintained in this way for long periods. The following are some of the published names based on garden plants and not already mentioned, but no attempt can be made to list the many horticultural nomina nuda in general use, or names of cultivars. Some commonly cultivated forms originating as wild hybrids have already been discussed.

H. andersonii (Lindl. et Paxt.) Ckn. sp. hort.  in T.N.Z.I. 60, 1929, 468.  Veronica andersonii Lindl. et Paxt. in Flower Gdn 3, 1863. Stated by Cockayne to be an artificial hybrid between H. salicifolia and H. speciosa and extremely common in cultivation.

H. balfouriana (Hook. f.) Ckn. sp. hort.  in T.N.Z.I. 60, 1929, 468.  Veronica balfouriana Hook. f. in Bot. Mag. 1897, t. 7556. Raised at Edinburgh Botanic Garden from seeds from some unrecorded locality in N.Z. Type: K, "from Sir J. D. Hooker's garden June 28, 1894". Specimens forwarded from Edinburgh by Prof. J. B. Balfour (A 7967, W 2330) suggest, as recorded by Cockayne and Allan (T.N.Z.I. 57, 1926, 28) a hybrid with H. vernicosa as one of the parents. H. pimeleoides could be the other; K specimens are annotated "Leaves opaque green, the younger with a dark purple margin. Stem violaceous-brown. Flowers purple"; fls are opp., bracts = or > pedicels and like the acute calyx-lobes slightly fleshy.

Veronica carnea Hort. ex J. B. Armst. in T.N.Z.I. 13, 1881, 357 non Vitm. 1789-92 nec DC. 1822, nec Opiz ex Schultes 1822. Armstrong's description was of a plant commonly cultivated in N.Z. gardens as a native, but he was unable to obtain authentic information as to its origin or to locate it in the wild. Cockayne and Allan (T.N.Z.I. 57, 1926, 14) regard it as "evidently an artificial hybrid cultigen with H. speciosa as one of its parents".

Veronica × edinensis Lindsay in Gdnrs' Chron. 48, 1910, 103. The hybrid, a cross between H. hectori and H. pimeleoides, is briefly described in comparison with its parents and is said to be very similar to wild specimens of H. epacridea.

H. erecta (Kirk) Ckn. sp. hort.  in T.N.Z.I. 60, 1929, 470.  Veronica erecta Kirk in T.N.Z.I. 28, 1896, 517. Described from specimens grown in the nursery garden of W. Martin. "A very distinct species, remarkable for the strict erect habit and elongated interrupted racemes with the flowers often fasciculate. It should be placed near to V. raoulii, Hook. f." The plant was believed to have been collected on Mt. Bonpland, but a wild habitat was never confirmed. Cheeseman (Man. N.Z. Fl. 1925, 824) accepts it tentatively as a wild sp., placing it near H. hulkeana, but Cockayne and Allan (T.N.Z.I. 57, 1926, 43) dismiss it as a cultigen. Cockayne (T.N.Z.I. 60, 1929, 470) remarks: "I have always understood that it was either an artificial or spontaneous garden hybrid with H. Lavaudiana, or perhaps H. Hulkeana, as one of its parents." The type (W 5332) has many features of H. pimeloides, and could well be a hybrid between this sp. and H. salicifolia; it is more stiffly erect than specimens of H. dartonii for which a similar origin has been postulated.

Veronica franciscana Eastwood in Leafl. West. Bot. 3, 1943, 220-222. According to Souster (J. roy. hort. Soc. 81, 1956, 497) this is the earliest validly published name for a commonly cultivated hybrid resulting from a cross made by I. A. Henry of Edinburgh; the parents were V. decussata from Falkland Is (= V. elliptica) and V. speciosa. V. lobelioides, the name originally proposed by the raiser, has been widely used but apparently never validly published. Souster adopts H. × franciscana cv. 'Blue Gem' for the common garden plant.

H. godefroyana (Carr.) Ckn. sp. hort.  in T.N.Z.I. 60, 1929, 471.  Veronica Godefroyana Carr. in Rev. Hort., Paris 1888, 455. Described from a French garden plant of unknown origin. Souster (J. roy. hort. Soc. 82, 1957, 519) after a full consideration treats it as a cultivar of H. pinguifolia. He also (loc. cit. 83, 1958, 303) places under H. pinguifolia cv. 'Pagei'.

Veronica Hendersonii Hort. ex Vilmorin's Blumeng. ed. 3, Sieb. et Voss 1, 1895, 784. Listed in Index Kewensis Suppl. 9, 1938, 295 with note "in syn.: V. speciosa".

H. imperialis (Bouch.) Ckn. sp. hort.  in T.N.Z.I. 60, 1929, 471.  Veronica imperialis Hort. ex Vilmorin's Blumeng. ed. 3, Sieb. et Voss 1, 1895, 784 in syn. V. speciosa. V. speciosa var. imperialis Bouch. in Flore des Serres 22, t. 2317. According to Cockayne "a form of H. speciosa with, purplish-crimson, very showy flowers".

Veronica jasminoides Hort. ex Bean Trees & Shrubs Brit. Isles 3, 1933, 493. Listed in Index Kewensis Suppl. 9, 1938, 295 with note "in syn." V. diosmifolia".

Veronica kermesina Loud. Encycl. Pl. t. 1546. Cited in Index Kewensis 1895 and by later authors as a synonym of V. speciosa.

Veronica lawtonii Lawton in J. roy. hort. Soc. 51, 1926, 313, fig. 73. Briefly described in comparison with H. hulkeana from which it is said to differ in being more rigid with erect branches, lvs dull without crimson margins, panicles more compact, white-fld, appearing six weeks later. Photographs of cultivated specimens of both are included. No reference is made to a wild origin.

Veronica lindleyana Paxt. Mag. Bot. 12, 1846, 247 non Wall. Cited in Index Kewensis 1895 and by later authors as a synonym of H. salicifolia, though specimens so labelled in W are quite unlike that sp. Armstrong (T.N.Z.I. 13, 1881, 356) makes the new combination V. stricta var. lindleyana, citing V. lindleyana Hort. as a synonym, and applies the name to wild plants, giving as distribution "Both islands, common in lowland forests".

Veronica × myrtifolia Lindsay in Gdnrs' Chron. 48, 1910, 103 non V. myrtifolia Banks et Sol. ex Benth. in DC. Prodr. 10, 1846, 460. The result of a cross between V. balfouriana and V. salicifolia, briefly described.

Veronica Reidii Hort. ex Kavka in Mollers Deutsch. Gart.-Zeit. 48, 1933, 100. Listed for N.Z. in Index Kewensis Suppl. 9, 1938, but no further information available. Garden specimens so named in CM resemble H. subalpina.

H. rotundata (Kirk) Ckn. sp. hort.  in T.N.Z.I. 60, 1929, 472.  Veronica rotundata Kirk in T.N.Z.I. 28, 1896, 530. Original localities: Near Wellington; near Southbridge. Based on a series of specimens collected at Newtown Park, Wellington in 1893-94 and distinguished from H. stricta var. atkinsonii (well represented by specimens from the same locality and date) by larger lvs and by capsule "suborbicular, compressed, broadly rounded at the apex". Cockayne says "almost certainly an artificial garden hybrid with H. salicifolia [sens. lat.] as one parent; it is common in cultivation".

INCERTAE SEDIS

1. Veronica biggarii Ckn. in T.N.Z.I. 48, 1916, 199 (Hebe biggarii Ckn. in T.N.Z.I. 60, 1929, 469) was "based on a plant cultivated in D. L. Poppelwell's garden, raised from a cutting collected by Poppelwell on the Eyre Mountains" on subalpine rocks at 1200 m. altitude. Cockayne labelled specimens in A (his No. 8115) and in CM "Type specimen", presumably both from the same cultivated plant; W 5310 is labelled "Eyre Mts, Otago, D.L.P." All three match a Poppelwell photograph held in BD. As distinctive characters Cockayne lists: "decumbent or prostrate habit; spreading, rather stiff, densely leafy branches; small glaucous, glabrous, almost sessile, oblong, ovate or oblong-ovate leaves with from rounded to subacute apices; numerous short, rather dense racemes; small white flowers with short calyx-segments and corolla-tube; and small slightly pilose capsules". Branches are described as 20-24 cm. long, lvs 12-18 × 6-8 mm.; infls lateral, simple, c. 3 cm. long, peduncles 2 cm.; calyx-lobes 1 mm. long, obtuse, ciliolate; corolla-tube 1·5 mm. long, lobes c. 3 × 2 mm.; capsule 4 mm. long, acute. Bracts, like calyx-lobes, are fleshly, obtuse, short; pedicels mostly 1·5-2.5 mm. long, but lowest occ. > fls.

Cockayne and Allan (T.N.Z.I. 57, 1926, 36) doubted the validity of the sp. and suggested affinities with H. pimeleoides var. rupestris, but later Cockayne (1929) recorded "sufficient evidence has been supplied by G. Simpson and J. S. Thomson to prove that this is a jordanon". Specimens labelled H. biggarii from Mt. Dick (Cockayne No. 8134 in K, = A 627, = W 5311) have obtuse, rather tumid glab. capsules and in several respects approach H. buchananii. Possibly this last-named sp. crosses with more than one of the slightly larger-lvd spp. in W. Otago to produce the rather diverse glaucous specimens attributed to H. biggarii in herbaria and gardens; few match the original plant.

2. H. brockiei Simpson et Thomson in T.R.S.N.Z. 72, 1942, 28.

Much-branched spreading shrub 20-30 cm. high. Branchlets rather stout, bifariously pubescent; length of internodes 2-3-(4) × diam., lf-scars remaining rough. Lvs ± closely imbricate, deeply concave, 10-15 × 7-10 mm., broad-ovate, coriac., bright green; lf-bud without sinus, lamina narrowed gradually to base; lamina broadly obtuse, "abruptly subapiculate", glab., entire, margin "yellowish" and very smooth with cartilaginous translucent border; midrib forming obscure keel. Infls lateral, simple, 2-4 cm. long; peduncle 5-15 mm. long. Fls crowded, pedicels very short. Bracts 1-2 mm. long, acute, ciliolate. Calyx-lobes 1·5-2 mm. long, ovate-lanceolate, acute ciliolate. Corolla white, tube not > calyx, lobes > tube, broad. Capsule erect, c. 4 × 3 mm., glab., "broadly ovate acute".

Known only from the original locality, "grassland on hills between Amuri Pass and Lake Man, at the head of the Doubtful River, Canterbury, at 1200-1500 m. altitude". Type: BD 56679 in fl.; BD 72728, garden-grown from the type plant has old capsules. "Plentiful in the type habitat."

In margin and colour of lf, and in infl. and fl. this closely approaches H . subalpina, while low habit and obovate concave lf-shape resemble rather H. pinguifolia.

3. H. carnosula (Hook. f.) Ckn. in T.N.Z.I. 60, 1929, 469.  Veronica laevis var. carnosula Hook. f. Fl. N.Z. 1, 1854, 194.  V. carnosula Hook. f. Handbk N.Z Fl. 1864, 210 non Lam. Illustr. 1, 1791, 47 (non Hook. f. in Bot. Mag. 1881, t. 6587 = V. pinguifolia). Hooker's type (K, Nelson, Morse's Mt., 5000 ft., Bidwill) shows the following characters: Branches stout, woody, pubescent towards tips, internodes short. Lvs suberect, ± concave, c. 9-12 × 6-7 mm., broad-obovate, thick and glaucous; lf-bud with distinct sinus, the petiole broad and thick; lamina broadly obtuse, entire, glab., midrib obscure. Infls lateral, simple, to 2.5 cm. long; peduncle to 1 cm. long, pubescent. Fls crowded, pedicels < 1 mm. long, bracts 2-2.5 mm. long, narrow, ciliolate. Calyx-lobes 2-2.5 mm. long, rather broad, obtuse, ciliolate. Corolla-tube slightly > calyx, lobes longer. Capsule erect, 5 × 2.5 mm., dark and acutely pointed, glab.

The only other collection mentioned in the Handbook is "upper Wairau, Munro" and this, as represented by A 8103, has stems almost glab., lvs more imbricating, infls very crowded, bracts broader and overlapping the very similar calyx-lobes; no lf-buds, ovaries or capsules are visible.

Hooker (Handbook 210) says: "from V. pinguifolia the ovary and capsules alone distinguish it". The type differs in other ways also, particularly in the distinct sinus in the lf-bud.

Though the name carnosula has been widely used few specimens have been seen that match the type. Until more is known of the glaucous hebes of Nelson it seems best, as suggested by Cockayne and Allan (T.N.Z.I. 57, 1926, 34), to regard the specimens dealt with by Hooker as "quite likely hybrids".

4. H. darwiniana (Col.) Ckn. in T.N.Z.I. 60, 1929, 470.  Veronica darwiniana Col. in T.N.Z.I. 25, 1893, 332, "on hills in the interior, Hawke's Bay; 1890-92, W.C." Colenso compared this sp. with V. colensoi from which it was said to differ in "leaves smaller, sub-concave and of another form; flowers pedicelled, bracts thin and longer than pedicels; lobes of calyx thin and ciliolate; corolla lobes very different, with puberulous throat; capsule larger, etc." A 7968 "Veronica Darwiniana Col., Hawke's Bay, W. Colenso" [labelled by Cheeseman] has (a) two fruiting twigs evidently belonging to H. venustula with dry gaping capsules like those described by Colenso; (b) one twig with several simple racemes of unopened fls and glaucous lvs "minutely and thickly dotted with whitish specks [stomata] on both surfaces". Colenso's specimens in K, two twigs, have glaucous lvs like (b) and some dozen racemes with many open fls. The glaucous-lvd specimens which agree with the main part of Colenso's description closely resemble H. glaucophylla except that the ovaries examined are glab.; they are not matched in any recent collections from Hawke's Bay or other North Id stations. Cheeseman's fuller description (Man. N.Z. Fl. 1906, 518; 1925, 803) was apparently based partly on South Id plants, and he included the type of V. glaucophylla Ckn. under V. darwiniana. The name H. darwiniana remains of uncertain status.

5. H. dilatata Simpson et Thomson in T.R.S.N.Z. 73, 1943, 164, from "debris slopes near Blue Lake, Garvie Mts, Otago, altitude 1370 m.", is known from the type BD 63426A and a few isotypes. Prostrate, much-branched, spreading to form close mats to 1 m. diam.; adventitious roots amongst oldest lvs; branchlets purplish when young, sparsely bifariously pubescent, length of internodes 3-6 × diam. Lvs patent to recurved, ± distichous, c. 10-15 × 5-6 mm., obovate-oblong, coriac. to almost fleshy, pale green; lf-bud with narrow sinus; lamina subacute, glab., margin entire and distinctly bevelled. Infls lateral, simple, few > lvs. Bracts rounded, ciliolate, c. = very short pedicels. Calyx-lobes c. 1·5 mm. long, oblong, obtuse, ciliolate. Corolla-tube c. = calyx, lobes slightly longer, broad, rounded. Capsule "obovate, acute, 4 mm. long", glab.

Infls are rather poorly developed and many abort in lf-axils in both wild and cultivated specimens; no capsules on specimens seen. Habit and size and texture of lvs suggest comparison with H. biggarii which has more glaucous lvs with less bevelled margins, and us. longer pedicels.

6. H. leiophylla (Cheesem.) Ckn. et Allan in T.N.Z.I. 57, 1926, 23.  Veronica parviflora var. phillyreaefolia Hook. f. Fl. N.Z. 1, 1854, 192.  V. leiophylla Cheesem. Man. N.Z. Fl. 1906, 509. Hooker's original record for his var. phillyreaefolia was "Nelson, abundant, Bidwill" and the type is in K. Differences from V. parviflora were stated to be: lvs narrowly linear-elliptic-oblong, racemes dense-fld, longer than lvs, sepals small, obtuse, corollas somewhat larger. Seven Bidwill specimens in K show some range in size (2-3·5 cm. × 4-7 mm.) and shape of lf, and in length of infl. (4-9 cm.), but all are reported to share the following characters: stems uniformly minutely puberulent; lf-bud with sinus; lvs uniformly brownish, thin, coriac., midribs scabridulous above in rather wide band, margins densely shortly scabridulous; ovary glab. where visible.

Rather numerous collections from the northern part of South Id can be grouped near Bidwill's specimens but they are far from uniform, those from the west showing more resemblance to H. gracillima, those from more easterly localities approaching H. traversii, and evidence of crossing with H. salicifolia appears in many specimens.

For his V. leiophylla Cheeseman gives the distribution "South Island. Apparently not uncommon throughout, from Nelson to Otago. Sea-level to 3000 ft." and the description is broad enough to cover specimens on some 25 sheets which he labelled "leiophylla". Plate 149 of Ill. N.Z. Fl. was "drawn from specimens collected in the central portions of the Nelson Provincial District" and three sheets, mentioned individually below, are marked as having been used for illustrations. According to the present treatment Cheeseman's specimens fall into four groups: (a) Ten sheets from localities in Nelson, Marlborough and Amuri showing ± resemblance to Bidwill's var. phillyreaefolia specimens but not at all uniform amongst themselves; of the two marked "Ill. N.Z. Fl." (both also marked "1559 to Kew"), A 7859, Awatere Valley, approaches H. traversii but has corollas more like those of figs 1 and 4, while A 7869, Buller Valley, has on two of its three twigs the pubescent ovary (reminiscent of H. glaucophylla) and the rather large calyx shown in fig. 3. (b) Four sheets, all from Banks Peninsula, referable to H. strictissima; an isotype of Kirk's var. strictissima is annotated "Probably a state of V. leiophylla". (c) Four sheets from Otago of H. rakaiensis or hybrids involving this sp. (d) Six sheets of H. traversii; of these A 7857 Marlborough, J. H. Macmahon, is marked "Ill. N.Z. Fl." and it has the long glab. capsule and small calyx typical of H. traversii as shown in Pl. 149.

In a note sent with specimens to Kew Cheeseman states: "For several years I referred it to V. Traversii, under which I suspect it was included in the 'Handbook', but if the plate of Traversii in the Bot. Mag. t. 6390 represents the type it cannot possibly be merged with it." The specimens accompanying this note are from Hanmer and are certainly much nearer to H. traversii sens. strict. than to either Bidwill's Nelson specimens or the plant figured in t. 6390 (= H. brachysiphon Summerhayes). There are good grounds for designating as lectotype a specimen from group (d) so that the name leiophylla would fall into the synonymy of H. traversii; the other alternative seems to be to regard the type of var. phillyreaefolia as the type of V. leiophylla but this would serve no immediately useful purpose since no reliable circumscription can be provided for that taxon. The name is therefore left in abeyance until more is known about the medium-lvd hebes of Nelson.

7. H. longiracemosa (Petrie) Ckn. in T.N.Z.I. 60, 1929, 471.  Veronica longiracemosa Petrie in T.N.Z.I. 49, 1917, 52 non Col. in T.N.Z.I. 20, 1888, 203 [= V. arvensis]. Type locality: Awatere Valley, Marlborough, H. J. Matthews. Type: W, 5312, "Plant grown by H. J. Matthews, Mch 1909". The only specimens are two flowering and three fruiting twigs: lvs 10-12 ×1-1·5 cm., lf-bud without sinus; infls to 16 cm. long; calyx-lobes subacute, ciliolate only; corolla-tube > calyx, rather broad, lobes = tube; capsule erect, c. 3·5 × 2.5 mm., glab. Placed by Cheeseman under salicifolia var. paludosa but the lf-base differs from the type of that var. and as Cockayne and Allan point out (T.N.Z.I. 57, 1926, 24) the locality is improbable. Most likely it is, as suggested by Cockayne and Allan, a hybrid between H. salicifolia and one of the smaller-lvd spp. even though Petrie says "it seems impossible that it can be a hybrid". It is quite distinct from H. salicifolia var. longiracemosa Ckn.

8. H. menziesii (Benth.) Ckn. et Allan in T.N.Z.I. 57, 1926, 25.  Veronica menziesii Benth. in DC. Prodr. 10, 1846, 461. Bentham's description is: "fruticosa, glaberrima, foliis breviter petiolatis oblongis acutis rigidis, racemis brevibus ad apices ramorum corymbo-ramosis, calycis segmentis 4 lanceolatis acutis, capsula calyce plus duplo longiore. In Nova Zeelandia (Menzies!) V. diosmaefoliae simillima sed capsulae majores et calyces diversissimi. (v. in herb. Hook.)"

This name has long troubled N.Z. botanists. In Herb. Hooker (K) a specimen bearing an infl. with dehisced fr. is marked, presumably in Hooker's hand, "V. Menziesii Benth. the original, New Zealand, Menzies"; N. E. Brown's accompanying note says: ". . . This specimen had been cut off from some other sheet & mounted on a sheet of V. elliptica, I presume by Hooker fil. . . ." The specimen matches Bentham's description as far as can be seen but the calyces are for the most part broken and the number and shape of the lobes cannot be determined. The specimen is extremely like some robust and closely lfy plants of H. diosmifolia not only in general appearance, dimensions and infl. but also in every small detail where comparisons have been possible; the pubescence on the branchlets agrees in kind and distribution, most of the lvs have a serrature of 1 to several small incisions, the texture, margin and peculiarly narrow acute lf-apex are exactly those of H. diosmifolia. A smaller specimen mounted on the same sheet and agreeing in lf-characters and kind of branchlet-indumentum has fls and young fr. and here the anterior calyx-lobes are united for about ⅞ of their length. There is every reason to believe that both these specimens belong to the earlier-described (1836) diosmifolia, a species which is known, apart from one unconfirmed Hawke's Bay record, only from North Auckland. Menzies collected in Dusky Bay, Fiordland, a district where diligent searching has failed to reveal any Hebe with diosmifolia -like infls. Such infls are common on plants in the northern part of South Id and some Nelson specimens which were neither mentioned, by Bentham nor in any way connected with Menzies fit the description of V. menziesii; among these are "Mr. Bidwill's specimens from a garden in Nelson, into which the plants were brought from the neighbourhood" (Hook. f. Fl. N.Z. 1, 1854, 193) though the portion of that gathering sent to Cheeseman (A 7905) has fls only and no capsules. Other names are available for the Nelson plants and the puzzle of the type of V. menziesii Benth. remains unsolved.

9. H. myrtifolia (Benth.) Ckn. in T.N.Z.I. 60, 1929, 472.  Veronica myrtifolia Banks et Sol. ex Benth. in DC. Prodr. 10, 1846, 460.  V. macrocarpa var. β myrtifolia Hook. f. Fl. N.Z. 1, 1854, 192. Bentham quotes specimens of Edgerley and Logan, presumably both from N. Auckland, and states that the lvs are like those of V. salicifolia, the racemes more lax and the corolla-lobes shorter and broader, the calyx-lobes ciliolate only, and the capsules 4 mm. long and rather obtuse. The plant of Banks and Solander's unpublished description and plate was distinguished by its short broad lvs (2-3 × 1-1·5 cm.), large white fls, and subrotund-oval, ± acute capsules; the localities there listed are "Tigadu, Tolaga, Opuragi, Motuara".

10. H. obovata (Kirk) Ckn. in T.N.Z.I. 60, 1929, 472.  Veronica obovata Kirk in T.N.Z.I. 9, 1877, 502. Kirk described this as "An erect glabrous shrub, 4-5 feet high, sparingly branched, branches erect, when old strongly marked with the scars of fallen lvs . . ." from Broken River, Canterbury. Specimens in K and in CM, apparently from the same plant, show the following characters: stem stout, woody, glab. except for a few short hairs in bifarious bands; length of internodes < 2 × diam.; lvs suberect, loosely imbricate, slightly concave, c. 2.5 × 1 cm., obovate, of rather heavy texture; lf-bud without sinus; lamina narrowed evenly to broad base, broadly obtuse, entire with quite glab. rather thick margin; midrib minutely hairy on upper surface, projecting slightly below; infls lateral, simple, 3-4 cm. long; peduncle to 1 cm. long; pedicels to 2 mm. long, bracts shorter, obtuse, ciliolate; calyx-lobes 2-2.5 mm. long, obtuse to subacute, ciliolate; corolla-tube not > calyx, lobes c. 2 mm., almost orbicular; ovary apparently with a few short hairs.

Resemblances are to H. pinguifolia from which it differs in taller habit, larger lvs, pedicellate fls, smaller and narrower bracts and calyx-lobes.

Specimens from Mt. Cook district attributed by Cheeseman and others to this sp. differ in rather smaller and less obovate lvs and narrower and more pointed calyx-lobes, and in these characters and in the thin cartilaginous lf-margin they approach H. subalpina; ovaries are us. but not always glab., but from further south in the Lake district specimens with similar narrow obovate lvs are more inclined to have pubescent capsules often combined with broader pale-bordered calyx-lobes and very short corolla-tubes; these last are more reminiscent of H. rakaiensis.

It is suggested that hybrids of several different origins can have obovate lvs of about the dimensions quoted for V. obovata, but very few specimens have been seen that closely resemble the type. Cockayne and Allan considered the sp. "not well founded".

11. H. subulata Simpson in T.R.S.N.Z. 79, 1952, 427. "A small yellow-green shrub to about 15-20 cm. high. Stems ± 4 mm. diam., dark brown, rough by the persistent leaves, prostrate or semi-prostrate and occasionally rooting, with the branchlets erect. Branchlets 2 mm. diam., numerous, quadrate. Leaves densely quadrifariously imbricate, the opposite pairs connate for about 1/2 their length and the connatures scaly pubescent, rounded at the tip, 2 mm. long, 2 mm. broad, thick and coriaceous, convex, orange, ribbed at the tip and the rib continued to a subulate point. Habitat. Sub-alpine grassland at 1,200 m. altitude. Type specimen from a plant in cultivation collected by Mr. Owen Fletcher from the Old Man Range, in the Herbarium, Plant Research Bureau [now BD], Wellington. The plant from which this description is drawn was collected by Owen Fletcher some twelve years ago on the summit of the Old Man Range and cultivated at Christchurch Botanic Gardens and at Dunedin by Mr. W. B. Brockie and the author respectively. No flowers have appeared and the description is incomplete."

The type specimen has not been found and the name remains of uncertain status. The only whipcord Hebe seen from the Old Man Range (Herb. Otago Univ. 002741, J. E. Holloway) has stout, terete or obscurely tetragonous branchlets 3 mm. diam. with thick convex lvs tipped with a small blunt apiculus and appears to belong to the H. hectori complex. Similar plants occur on the Garvie Mts, but also others with strongly tetragonous orange-green branchlets 3 mm. diam. and lvs tapering to a distinct long spine which may match the plant described by Simpson. These differ from H. lycopodioides in having lvs connate for c. 1/2 length with narrower, more abrupt spine and no visible ribbing.

12. H treadwellii Ckn. et Allan in T.N.Z.I. 56, 1926, 27 from "stony ground at about 1200 m. on Mount Ollivier (Sealey Range), and near the Mueller Glacier at a similar altitude: L.C. This species . . . comes into the group which contains Hebe pinguifolia and its allies, but it is distinguished by its green leaves, its fewer few-flowered racemes almost hidden by the leaves, its subulate bracteoles, its smaller flowers with acute calyx-segments, and its glabrous ovaries." The description includes: lvs closely imbricate, c. 25 × 12 mm., obovate-oblong, subacute, strongly concave, coriac., margins pale; racemes c.12 mm. long, few-fld, fls shortly pedicellate below, sessile above; corolla small, tube very short, ± = calyx.

The only authentic specimen seen is BD 6336, "Mount Cook area, Jan. 1928, H. H. Allan" Herb. Cockayne No. 8186, labelled H. treadwellii by both Cockayne and Allan. This is very close to H. subalpina though probably more glaucous. The Mt. Cook area is the source of many specimens approaching H. treadwellii and all showing affinities with H. subalpina; H. buchananii might in some cases be suspected as the other parent. Comparisons should be made also with H. brockiei and with plants like those from Poulter Hill (Cockayne No. 7973) which Cockayne annotated as "probably one of the hybrids between Hebe pinguifolia and Hebe montana or some species related to the latter".