Hypericum caespitosum
Perennial or annual(?) herb 0.05-0.2 m tall, decumbent to prostrate but not rooting, much branched from the base and often forming loose cushions, lateral branches sometimes present below inflorescence, erect to ascending, not naked below. Stems green to vinous red, 4-lined and ancipitous when young, soon 2-lined, cortex persistent; internodes 1-17 mm long, upper ones often exceeding leaves. Leaves sessile, usually erect; lamina 3-18 x 0.4-1 mm, linear, margin plane to subrecurved, paler beneath, not glaucous, chartaceous; apex rounded, base parallel-sided to rounded or narrowly cuneate, sometimes subamplexicaul but not decurrent or forming V, free; basal vein 1, unbranched or with up to 3 pairs of main laterals, tertiary reticulum absent; laminar glands dense, not prominent. Inflorescence 1-7-flowered, dichasial/monochasial, without accessory branches, usually with lateral branches from up to 9 nodes below, the whole subcorymbiform to narrowly cylindric; primary pedicels 2-7 mm long; bracts and bracteoles linear. Flowers 5-9 mm in diam., stellate; buds ellipsoid, rounded. Sepals 2.5-5 x 0.6-1.3 mm, equal, imbricate, linear-lanceolate to narrowly elliptic or rarely narrowly oblanceolate, acute; veins 5, not or only midrib prominent; glands linear, distally interrupted to punctiform. Petals yellow, always (?) tinged or veined red outside, (4?-)5-7 x 1-1.5 mm, 1.5-2 x sepals, oblong-oblanceolate to narrowly oblong; apiculus obsolete; glands linear. Stamens 10-35, irregularly grouped, longest 3.5-5 mm long, c. 0.7 x petals. Ovary 1-1.5 x 0.6-1 mm, cylindric-ellipsoid; styles 3, 0.8-1.5 mm long, 0.8-1 x ovary, spreading; stigmas broadly capitate. Capsule 2.5-5.2 x 1.5-2.8 mm, cylindric-ellipsoid to cylindric, apex rounded, 1-1.3 x sepals. Seeds 0.6-0.7 mm long; testa finely ribbed-scalariform.
Dry or wet grassland; 90-1100 m.
Chile (Valparaiso and Corico to Chiloé).
H. caespitosum is related to the southern Peruvian and Bolivian narrow-leaved forms of H. silenoides. Although perhaps superficially resembling some forms of H. brevistylum, the thinner, green (not glaucous) leaves, and red-tinged petals indicate its true affinity. The lack of recent records from the northern and southern extremes (Valparaiso and Isla de Chiloé) may indicate a contraction in the range of H. caespitosum. Rodriguez Jimenez (1973) misinterpreted Ariqué (3946'S; 7304'W) as Arica (1828'S; 7018'W) when localising Lechler 414. Her speculation that H. caespitosum has been introduced into northern Chile is therefore unnecessary. I am grateful to Prof. Clodomiro Marticorena of the Universidade de Con- cepción for drawing my attention to this error.