Hypericum latisepalum (Nomenclature)
Shrubs to 1 m or more, erect, with stems erect from base, distally outcurving; branches slightly 4-lined and ancipitous above, soon terete. Leaves 1–2 mm petiolate; laminae 30 x 15-38 mm, ovate or triangular-ovate to broadly oblong-ovate, paler beneath, chartaceous; apices apiculate-obtuse or very rarely acute to rounded, margins plane, bases rounded to truncate; venation: 5-6 pairs of main laterals from lower half of midrib, with subsidiary midrib branches above, tertiary reticulation obscure; laminar glands dense, elongate-punctiform to punctiform, intramarginal glands dense, ventral glands absent. Inflorescences (1-)3-7(-c. 20)-flowered from 1(-2) nodes, corymbiform; bracts and bracteoles broadly ovate to lanceolate; pedicels 4-8 mm. Flowers(35-)40-50 mm in diam., shallowly cupped, becoming ± stellate; buds ovoid, subacute. Sepals equal, basally imbricate, erect in bud and fruit, 5-8 x 4-5 mm, enlarging in fruit to 8-12 x 5-8 mm, ovate, acute to apiculate-obtuse or rarely rounded, with margins narrowly hyaline, entire or minutely denticulate; veins c. 9; glands linear. Petals golden-yellow, 18-25 x 12-17 mm, broadly ovate to broadly oblong, with apiculus rounded, very short, margins entire, eglandular. Stamen fascicles each with c. 50 stamens, longest 12-17 mm, 0.6-0.7 × petals; anthers golden yellow. Ovary 6-8 × 4-5 mm, ± broadly ovoid; styles 5-7 mm long, c. 0.85 x ovary, appressed, apically curving outwards. Capsule 15-20 x 10-12 mm, 2.5-3.0 × sepals, ± broadly ovoid. Seeds dark chestnut brown, c. 1.5 mm long, cylindric, usually narrowly carinate with cairns sometimes prolonged distally into narrow wing; testa shallowly linear-reticulate.
Open areas, roadsides; 1800–2700 m.
China (SW. Sichuan, NE. Yunnan).
Hypericum latisepalum is closely related to H. bellum, but not in the way that I described in Robson (1985: 274), where I treated them as subspecies. Neither does my description of their variation, given when altering their rank to species (Robson 2005: 276), provide a complete explanation of the situation. After a detailed revision of this group, I realised that H. latisepalum is a westward development from H. beanii in which the leaves become triangular-ovate to ovate with the apex obtuse to rounded, the (appressed) sepals become obtuse and then rounded with a tendency towards enlargement (‘foliose’), the petals become markedly incurved, and the ovary and capsule become ovoid. It is most variable in the Cangshan area, the type (from “Tali, top of Yin Yo Mountain”) being unusual in having triangular-ovate (not ovate) leaves and comparatively large flowers. At Dashao (near Kunming), SBEC K067has sepals almost like those of 50. H. forrestii but obtuse, not rounded, the stamens are unusually long (0.6 × petals instead of 0.4-0.5) and the petals are spreading-incurved as in H. beanii. It is possible, therefore, that this collection is a hybrid, although the occurrence of apomixis in the H. forrestii group means that other interpretations are also possible.
In western Yunnan there are two independent trends from H. latisepalum towards reduction in length of stamens relative to the petals. One, in which the leaves become large and oblong-lanceolate, leads without intermediates to 49. H. addingtonii of NW. Yunnan; the other, in which the leaves first become broadly rhombic-ovate and then smaller and ovate with an undulate margin, results in H. bellum. This latter reduction trend runs from NW. Yunnan into SE. Xizang and NE. Assam. However, there is a population in the Nam Tamai valley, N. Burma (Kingdon-Ward 13212,13350, both BM) that is intermediate between H. latisepalum and H. bellum. It has large flowers and ovate-lanceolate leaves like those of H. latisepalum, but the stamens are 0.4-0.5 x as long as the petals like those of H. bellum. It now seems best to treat this Burmese population as an extension to the variation of H. latisepalum rather than as an intermediate that necessitates reducing these taxa to subspecies, as I originally did. For a summary of the differences between those two species, see the key (p. xx).