Medium tree to 25 m, closely related to J. regia. Leaves 15–50 cm, petiole 7–12.5 cm, glabrescent, rachis glabrescent, terminal leaflet present. Leaflets 9 or 11(–15), ovate- or elliptic-lanceolate, 6–18 × 3–8 cm, glabrescent, margins entire or obscurely serrulate, apex acuminate; lateral petiolules absent or to 1 mm, terminal petiolule 2–3 cm. Male spikes 13.5–18 cm long; flowers with 24–27 stamens. Fruiting spikes with 1–3 fruits; fruits glabrescent, 3.4–6 × 3–5 cm, globose to subglobose, irregularly dehiscent; nuts valvate, valve forming a prominent, longitudinal ridge around the nut, the surface otherwise irregularly pitted; the shell forms deep ridges inside, the seed contorted and convoluted through filling the spaces between. Flowers March–April, fruits September (China) (Lu et al.1999; Grimshaw & Bayton 2009).
Distribution Bhutan China Guizhou, Sichuan, SE Xizang, Yunnan. India Arunachal Pradesh, Sikkim.
Habitat Forests in valleys and on mountain slopes; 1300-3300 m.
USDA Hardiness Zone 8-9
RHS Hardiness Rating H6
Conservation status Not evaluated (NE)
The Iron Walnut is, at least to Western eyes, a distinctive tree, although its taxonomic status is rather uncertain. It is clearly very close to the familiar Persian Walnut, recognisable by the number of leaflets (9–15, vs. 5–9 in J. regia) and by the shell of the nut (extremely hard – hence the common name – and covered with numerous deep pits, while in J. regia the shell is wrinkled, not pitted). The notably acuminate tips and, sometimes, more than a hint of teeth on the leaflets are also distinctive.
This is a species cultivated in southwestern China for its nuts (Wang et al. 2015); whether or not it is truly native or an ancient cultivated strain of J. regia is a moot point. Zhang et al. (2019), in an impressively thorough phylogenomic study, demonstrated that regia (taken by them to include sigillata) is derived from an ancient (around 3.5 MYA) hybrid between a black walnut species and a member of the J. mandshurica lineage. They consider that J. sigillata is best thought of as an ecotype of J. regia maintained as a landrace in southwest China. Wang et al. (2015), while treating them as distinct species, found evidence of introgressive hybridisation blurring the boundary between the two. We choose to follow Flora of China (Lu et al. 1999) here in maintaining J. sigillata as a distinct, horticulturally recognisable species.
The Iron Walnut reached western cultivation remarkably late. Most European specimens seem to derive from a 5 kg bag of nuts imported to the United Kingdom from the China National Tree Seed Corporation by John Grimshaw and Michiel Zwaan in 2005 (Grimshaw & Bayton 2009), who noted much variation on size and pitting, as well as the extremely hard shells. These were distributed to dendrologists across Europe, with good germination resulting. Initial growth was fast, but there has been a tendency for the seedlings to make low shrubby specimens at first (J. Grimshaw, K. Rushforth, pers. comms. 2019). The term ‘herbaceous walnut’ is sometimes heard! Increasingly, individuals are breaking away from this habit to make standard trees, and grow well thereafter. Keith Rushforth (pers. comm. 2019) feels that optimal growing conditions in a single season, rather than simply length of time established, allows this.
The species was also collected around this time in Arunachal Pradesh, India, by Keith Rushforth; it is growing in both Europe and New Zealand from this introduction. A specimen measured at 12 m × 21 cm in 2019 (The Tree Register 2019) grows in Maurice Foster’s Kent arboretum. A tree – now fruiting – at Tregrehan, Cornwall (11 m × 17 cm, 2014 – The Tree Register 2019) derives from an earlier Tom Hudson collection, from Yunnan in 1996.