Flora of New Zealand Series

Small tree with black bark; young lvs and branchlets pubescent. Lvs alt., on petioles ± 5 mm. long; lamina submembr., elliptic-obovate to obovate, obtuse to subacute, apiculate or not, gradually narrowed to petiole; 4-7 × 2-2.5 cm. Fls both solitary and axillary and terminal in 2-3-fld umbels. Sepals ovate-oblong, obtuse, 4-5 mm. long; petals c. 1 cm. long, narrow-oblong. Capsules ± 2 cm. long, ovoid to obovoid, 3-valved; valves pubescent, woody; pedicels 1-1·5 cm. long, decurved.

DIST.: Kawau Id. Known only from a single tree, now destroyed.

Kirk (loc. cit.) described the capsules as ovate-acuminate and the lvs as acuminate. Later (Stud. Fl. 1899, 48) he described the lvs as "rounded at the apex or shortly acuminate." The specimens I have seen do not well agree with these descriptions. The specimen at K ("115. T. K. received 8/68 Kawau") has lvs obtuse to subacute, 3·5-4 × 2-2.5 cm. The capsule valves are 1·5-2 cm. long. In W are 4 sheets: 442 Oct. 1872; 443 Oct. 1872; 444 Feb. 1868; 445 Feb. 1868. The October specimens are in full fl.; colour dark, but not now exactly determinable. The February pieces have capsules. Kirk (T.N.Z.I. 4, 1872, 266) said: "In foliage it resembles large forms of P. tenuifolium, while the capsules partake of the characters of P. crassifolium and P. ellipticum . . . I have been tempted to attribute its peculiarities to hybridization." Cheeseman (Man. N.Z. Fl. 1925, 489) was similarly tempted, but rejected the explanation as P. ellipticum is not known at Kawau. It is possible that the plant originated from the crossing of P. crassifolium with P. tenuifolium.