Flora of New Zealand Series

S. viridis Kirk Stud. Fl. 1899, 350 non Phil. in Anal. Mus. nac. Chile 1891, 43.

Type locality: Kirk cites a number of localities and indicates no type.

Shrub up to 1·5 m. tall, less compactly and more slenderly branched. Lvs 3·5-7·5 × 1·5-2.5 cm., on petioles up to 2 cm. long, less coriac., oblong-obovate. Panicles larger; pedicels more slender, up to 1·5 cm. long.

DIST.: S. Montane to subalpine scrub from lat. 41° to 44°.

POLYMORPHY AND HYBRIDISM

Cheeseman (Man. N.Z. Fl. 1906, 368) remarked of the genus in N.Z.: "The herbaceous species are exceedingly variable, and some of them difficult to determine; but the shrubby varieties are remarkably distinct." The complex of forms found in the herbs with lvs in rosettes and scapose infls is still far from resolved, especially as to how far the linking forms are due to hybridism and how far to true-breeding vars. There is good evidence that hybrids occur of the origins S. lagopus × saxifragoides and S. haastii × southlandicus. It is probable that detailed study of the occurrences of glandular hairs and of trichomes would yield evidence of "introgression", as Wall's examination of the populations on Banks Peninsula (T.N.Z.I. 50, 1918, 198) suggests.

There is abundant field evidence that S. lyallii and S. scorzoneroides cross where they meet, with the resultant progeny showing differences in fl.-colour and lf-form. Du Rietz (Svensk bot. Tidskr. 24, 1930, 383) remarks: "If I may rely upon my own observations, this hybrid quite often out-numbers the pure species. In this case it appears equally probable that the 'species' are segregates of the hybrid as that the hybrid was formed by the meeting of two species originally isolated." Study of the populations throughout their range does not lead me to support this suggestion.

It will be obvious from the treatment in the text that much remains to be done before we have a satisfactory picture of the complexes grouped under S. lautus, S. colensoi, and S. banksii, especially as regards the possibility of the existence of separate geographical vars.

Occ. aberrant forms are met with that appear to be sterile resultants of the hybridization of a shrub with a rosette herb. Cockayne (T.N.Z.I. 49, 1917, 60) described his S. christensenii from a plant collected on dry rock on Leslie Hills, near Hanmer, and remarked: "This species bears a superficial resemblance to S. bellidioides Hook. f. and to small forms of S. lagopus Hook. f.; but it differs from either in its shrubby or perhaps suffruticose habit." Subsequently Cockayne and Allan (Ann. Bot., Lond. 48, 1934, 52) suggested the origin as S. bellidioides × monroi. Simpson and Thomson (T.R.S.N.Z. 71, 1941, 98) listed S. cassinioides × haastii for a plant with lvs similar to S. haastii and long stems with thin papery bark similar to those of S. cassinioides. S. pottsii J.B. Armst. in T.N.Z.I. 4, 1872, 290 is known only from the description and a scrap in the Armstrong Herbarium, labelled as from Mt. Potts, not Mount Jollie as stated in the description. It appears to be of similar origin. The suggestion that these and other forms are of hybrid origin is supported by the spontaneous occurrence in the garden of the late Mr. W.A.. Thomson, Dunedin, of hybrids between S. hectori and S. southlandicus (T.N.Z.I. 60, 1929, 265).

Allan (J. roy. N.Z. Inst. Hort. 12, 1943, 62) described a hybrid growing in the garden of Mr. S. Duncan, at Tawa Flat, Wellington. The plant grew from a cutting originating in the garden of the late Mr. Alfred Atkinson at York Bay, Wellington. Atkinson's plant died before flowering, but Duncan's grew into a bushy very floriferous shrub. The very probable origin was S. hectori × perdicioides. The horticultural name Senecio "Alfred Atkinson" was given to it, and the shrub is now well-known in cultivation.

Needing further study are the "intermediate" forms, for which the origin S. cassinioides × revolutus has been suggested, and S. bidwillii × elaeagnifolius. S. buchananii J.B. Armst. in N.Z. Ctry J.3, 1879, 56 was described from specimens from (a) Arthur Pass, South Id, and (b) Mount Egmont. Kirk (Stud. Fl. 1899, 349) followed by Cheeseman (Man. N.Z. Fl. 1925, 1026) reduced this to var. buchananii of S. elaeagnifolius. The status of the forms remains uncertain; North Id forms ascribed to var. buchananii belong to S. elaeagnifolius, those of the South Id to S. bennettii, which replaces S. elaeagnifolius in that region.