Myosotis saxosa Hook.f.
Exarrhena saxosa Hook. f. Handbk N.Z. Fl. 1864, 196 in part only.
Type locality: "Crags at Titiokura, east coast". Type: K, Colenso 1711.
Rosette-lvs ∞ and tightly bunched, c. 2-3×0·5-1 cm., petiole narrow, c. = lamina-length and sheathing at base, lamina broad-ovate to obovate, tip apiculate often down-turned; hairs long, crowded, spreading, on undersurface similar but more appressed, the whole plant hoary. Lateral branches ascending to erect, c. 7 cm. long, lfless below cyme, lower internodes < lvs. Stem-lvs up to 15×5 mm., elliptic, sub- acute to acute, sessile; hairs softer and more appressed than those on rosette-lvs. Cymes mostly simple, ebracteate, up to 12-fld, up to 2 cm. long; internodes < calyx; pedicels up to 2 mm. long. Calyx c. 5 mm. long, lobes > 1/2 length, subacute; hairs long and strong, an under-layer shorter and finer, ± appressed, flexuous but not hooked. Corolla white, 7-8-13 mm. diam., tube 3-5 mm. long, with vertically elongated scales at mouth, flaring above into rather wide funnel-shape with lobes c. 3 × 3 mm.; filaments fixed between scales and reaching to level of lobes, anthers c. 1·5 mm. long, versatile; styles c. 2 × calyx in fr., stigma capitate. Nutlets c. 2·2 × 1·3 mm., slightly convex outside, bluntly keeled within, slightly winged round narrow tip.
DIST.: N. Known only from one place in Hawke's Bay.
FL.- FT. 11-12.
Hooker based his original description only on Titiokura specimens but later (Handbk N.Z. Fl. 1864, 196) listed under Exarrhena saxosa "Northern Island: east coast, crags at Tetiokura, Colenso (very white and hispid)". Middle Island: Dun mountain, in open stony places, Munro, Travers (much less hispid)". Cheeseman (Man. N.Z. Fl. 1906, 469) pointed out the distinctive characters of the Dun Mt. plant and called it M. monroi.
The only known locality for this sp. has been visited periodically by N. L. Elder over many years, his latest report (5 April 1957) being that only 3 surviving clumps could be found. He has examined comparable sites for some miles along the range with negative results.
Specimens collected at Titiokura by A. P. Druce (BD 82416) differ from all others seen in having hairs regularly retrorse on undersurfaces of rosette-lvs. Lf-fragments (at A) from the type show no retrorse hairs.
