Flora of New Zealand Series

Cercodia erecta Banks ex Murr. in Comm. Soc. Gott. 3, 1780 (1781), 3, t. 1.

H. alata Jacquin Ic. Pl. rar. 1, 1781, 7, t. 69.

Cercodia alternifolia A. Cunn. in Ann. nat. Hist. 3, 1839, 30.

Perennial somewhat scabrid herb or subshrub, up to 1 m. tall; stems many, four-angled, angles often rough with irregular thickenings. Lvs opp., lanceolate, acute, lamina 2.5-5 cm. long, on petiole up to 1 cm. long, smaller and more nearly sessile in upper parts, serrate, shining and somewhat scabrid. Infl. racemose to paniculate, lfy bracts subtending clusters of up to six fls with narrow bracteoles (sts reduced to one). Fls ± drooping, on very short terete pedicels. Calyx-lobes 4, acute, persistent. Petals 4, boat-shaped, deciduous; stamens 8, falling early, filaments very short and slender, anthers longer, oblong, yellow; ovary 4-winged, styles 4, very short; stigmas 4, oblong, obtuse; fr. with wings on the 4 angles, 4-locular, with one seed in each loculus.

DIST.: K., N., S., St., Ch. Forest margins, banks, cliffs, roadside, etc.

FL. 12-2. FT. 1-4. Apparently endemic.

The name H. erecta has been applied also to Australian and Juan Fernandez plants which are now regarded as belonging to different spp. Of the two other spp. in subsect. Cercodia that have been recognized in N.Z., H. cartilaginea seems clearly distinct in its broadly oval and carti-laginous lvs, in the kind of pubescence, and in fr. shape; the status of H. colensoi is more un-certain.

The original diagnosis of Cercodia erecta from which the above description is taken rather directly mentions specifically scabrousness and winged frs. The Banksian specimen at A has scabrid lvs but the frs are scarcely winged. Skottsberg's diagnosis of H. colensoi emphasizes the perfect smoothness of the whole plant and the small barely alate frs. Plants matching these two descriptions can be found but there is a much wider range of diversity. In some plants frs are quite without wings, in others all frs are strongly winged, but many individuals development of wings varies from fr. to fr.; transverse rugosity occurs in both winged and wingless frs or may be absent in either. Lvs vary in shape, texture, and details of toothing and they range from 1 to 8 cm. long and from 5 to 25 mm. wide; upper lvs are mostly smaller than lower ones but some large plants have only small lvs. Habit ranges from erect to procumbent, and the amount of red colouring varies. No one of these characters is at all obviously correlated with any other, or with roughness or smoothness of lvs. In specimens examined it has been found, however, that smooth-lvd completely glab. plants can be clearly separated from those with lvs scabrid on the upper surface and with minute hairs on sepals, ovaries, pedicels, young stems and petioles. On the angles of the stems, particularly in lower parts, rough irregular thickenings, which are not hairs, are almost universally present (though us. better developed where lvs are scabrid) so that few plants can be called strictly smooth all over. The single specimen (a small-lvd low-growing plant) on which H. colensoi was based seems to be just one of the long and almost continuous series of forms to be found amongst smooth-lvd plants. Pending further study it seems best to adjust the content of H. erecta so that it includes all N.Z. scabrid-lvd plants, which, as far as is known, also invariably show some pubescence, and to use the name H. colensoi for all smooth-lvd plants, these apparently being always quite hairless. In these senses both spp. are found over wide geographical and ecological ranges here; the two occur together in some places (e.g. on Wellington coasts).

The two glab. spp. described by Skottsberg (Nat. Hist. Juan Fern. and Easter Id 2, 1922, 155-156) from Juan Fernandez Is both have almost wingless frs which seem to differ from those of all N.Z. specimens, being narrow and 8-costate in H. masatierrana and very broad and rounded in H. masafuerana. Scabrid-lvd N.Z. plants with almost or quite wingless finely pubescent frs need to be critically compared with H. asperrima Skottsb. from Masafuera (which has less rugose and more nearly conical frs) and with all forms of H. exalata F. Muell. from Australia. It is not certain whether smooth-lvd N.Z. plants with quite wingless frs (a combination not very commonly found) could be matched amongst Australian plants attributed to H. laevis Schindler.