Carum Linnaeus sect. Dactylaea Franchet; Dactylaea (Franchet) Farille.

Herbs, perennial, slender, glabrous (except S. filicinum). Rootstock fusiform or elongate, usually slender. Stems erect, solitary or 2–4, caespitose. Basal leaves petiolate; petiole sheath usually broad, ovate or oblong-ovate at base; blade ternate-1–3-pinnate or 1–3-pinnate. Leaves reduced upwards. Inflorescence branching, umbels compound, terminal; bracts mostly absent, occasionally 1–4, linear or similar to uppermost leaf; rays few, usually 5–15; bracteoles present (rarely absent), usually linear, entire, apex rarely lobed; umbellules usually many-flowered. Calyx teeth obsolete, or conspicuous, triangular or subulate-lanceolate. Petals white or purple, ovate, oblong-ovate or obovate, base clawed, apex acute or slightly obtuse, rarely 2–3-lobed or palmately 3–5-lobed. Stylopodium flat, rarely low-conic; styles short. Fruit oblong-ovoid, slightly laterally compressed, smooth; ribs 5, filiform; vittae 1–3 in each furrow, 2–6 on commissure. Seed face plane. Carpophore 2-fid or 2-parted.

This taxonomically complex genus is closely related to, and sometimes difficult to distinguish from, Acronema. Sinocarum is usually circumscribed by a suite of characters: rhizome elongate, petiole sheaths expanded, petals obtuse at apex, clawed at base, flowers radiant, and fruit oblong-ovoid. By contrast, Acronema is characterized by having tuber globose or oblong, petiole sheaths narrow, petals acute to filiform at apex, cuneate at base, flowers symmetric, and fruit usually ovoid or broadly so, slightly cordate at base. However, within each genus there are species that deviate in one or more of these characters, and the generic boundaries are blurred. Revision of these two genera is hampered by a lack of complete material: specimens are usually collected in flower, and mature fruits are unknown for an alarmingly high proportion of the taxa. Initial results from molecular sequence data on Himalayan species suggest that these two genera should be combined, but further work and more collections are needed to clarify the situation across the whole geographic range.

Sinocarum pseudocruciatum H. Wolff (Repert. Spec. Nov. Regni Veg. 27: 182. 1929) was described from Sichuan (“Washan,” A. Henry 7067, holotype, K). However, it is not treated in this account as it is imperfectly known.

Pimenov and Kljuykov (pers. comm.) consider the following imperfectly known taxa to be conspecific and a species of Sinocarum: Trachydium souliei H. de Boissieu (Bull. Soc. Bot. France 53: 422. 1906), described from Xizang (J. A. Soulié 1049, holotype, P), and T. dielsianum H. Wolff (Acta Horti Gothob. 2: 300. 1926), described from Sichuan (SE of “Matang,” 4800 m, K. A. H. Smith 4375, holotype, unlocalized).

About 20 species: high-altitude Sino-Himalayan region from Nepal to SW China; eight species (four endemic) in China.

(Authors: Pu Fading (溥发鼎 Pu Fa-ting); Mark F. Watson, Ingrid Holmes-Smith)