Hypericum empetrifolium (Nomenclature)
Subshrub or dwarf shrub, 0.1–0.6 m tall, with stems rather stout to slender, erect to prostrate, branching and rooting at base, with branches ascending to strict along stem, with short axillary shoots from median and upper nodes. Stems 3-lined, eglandular; internodes 2–20 mm, shorter than to exceeding leaves. Leaves 3-verticillate, spreading, sessile to very shortly petiolate, glaucous beneath, (2–)4–12 × 0.5–2 mm, narrowly linear, apex rounded, margin revolute, base rounded, 1-veined; laminar glands pale, scattered; intramarginal glands not seen. Inflorescence 1–40-flowered, from 1–4 nodes, lax, cylindric, 10–100 mm long, without subsidiary branches below; bracts and bracteoles reduced-foliar, entire. Flowers 8–20 mm in diam.; buds globose. Sepals equal, basally united, not imbricate, 1–2.5 × 0.5–1 mm, oblong to elliptic or oblanceolate, obtuse to rounded; veins 3, not prominent; margin with regular or irregular sessile globose black glands; laminar glands pale, linear to striiform. Petals bright yellow, not tinged red, deciduous, 5–10 × 2–3 mm, 5 × sepals, narrowly oblanceolate; marginal glands absent; laminar glands pale, few, shortly striiform or absent. Stamens 25–60, longest 5–8 mm, 0.8–1 × petals, deciduous. Ovary 1.5–2 × 1–1.2 mm, ovoid; styles 3–4.5 mm, c. 2 × ovary. Capsule 4–6 × 3–4 mm, ovoid; valves with narrow dorsal vittae and oblique swollen interrupted lateral vittae. Seeds reddish brown, c. 1.5 mm long, cylindric to ellipsoid, papillose.
2n = 18 (Contandriopoulos & Lanzalavi, 1968; Reynaud, 1980, 1981), n = 9 (Contandriopoulos & Lanzalavi, 1968; Reynaud, 1981).
Macchi, Pinus woods, dry calcareous (limestone, serpentine) rocky slopes; 0–2000 m.
Northern Albania, Greece, Crete, Aegean Islands, western Turkey, Libia (Cyrenaica).
Hypericum empetrifolium varies little throughout its range except in Crete, where there are two reduced high-altitude forms. These were described as varieties (Rechinger, 1943) and raised to subspecies by Hagemann (1987). They differ from subsp. empetrifolium and from each other in habit and size of parts, and to me these differences did not appear to be clear-cut. For this reason Robson & Strid (1986) included the prostrate var. tortuosum in the pulviniform var. oliganthum. Hagemann, however, showed that these taxa have different habits that remain distinct in cultivation. I have therefore decided to adopt her ranking and to recognise the three taxa as subspecies.