Embothrium coccineum J. R. & G. Forst.

TSO logo


Kindly sponsored by
a member of the International Dendrology Society


Jack Aldridge (2023)

Recommended citation
Aldridge, J. (2023), 'Embothrium coccineum' from the website Trees and Shrubs Online (treesandshrubsonline.org/articles/embothrium/embothrium-coccineum/). Accessed 2024-05-20.

Common Names

  • Chilean Firebush
  • Notro
  • Ciruelillo
  • Fosforito


  • Embothrium coccineum var. lanceolatum (Ruiz & Pavon) O. Kuntze
  • Embothrium coccineum f. andina Hoffmann
  • Embothrium coccineum Lanceolatum Group Hort.
  • Embothrium coccineum Longifolium Group Hort.
  • Embothrium gilliesii Meissn.
  • Embothrium lanceolatum Ruiz & Pavon
  • Embothrium longifolium Hort., not Poir.
  • Embothrium valdivianum Gandoger


Pollen-producing structure of flower at the tip of the filament; part of a stamen.
Immature shoot protected by scales that develops into leaves and/or flowers.
Organism arising via vegetative or asexual reproduction.
A collection of preserved plant specimens; also the building in which such specimens are housed.
(of fruit) Vernacular English term for winged samaras (as in e.g. Acer Fraxinus Ulmus)
Lance-shaped; broadest in middle tapering to point.
The visible form of an organism.
pioneer species
Early colonists of disturbed areas.
Small grains that contain the male reproductive cells. Produced in the anther.
Act of placing pollen on the stigma. Various agents may initiate pollination including animals and the wind.
Folded backwards.
(of a leaf) Unlobed or undivided.
(in a flower) The part of the carpel that receives pollen and on which it germinates. May be at the tip of a short or long style or may be reduced to a stigmatic surface at the apex of the ovary.
(var.) Taxonomic rank (varietas) grouping variants of a species with relatively minor differentiation in a few characters but occurring as recognisable populations. Often loosely used for rare minor variants more usefully ranked as forms.
(of habitat or site) Arid. (Cf. mesic.)



Jack Aldridge (2023)

Recommended citation
Aldridge, J. (2023), 'Embothrium coccineum' from the website Trees and Shrubs Online (treesandshrubsonline.org/articles/embothrium/embothrium-coccineum/). Accessed 2024-05-20.

Evergreen (sometimes deciduous) trees with a single trunk and without suckers, or shrubs suckering to form pure thickets. Leaves leathery, alternate, variable in size and shape, occasionally on the same plant; narrowly lanceolate to ovate, to 15(–30) × 2–4 cm, hairless, green to dark green above, pale beneath. Leaf petiole to 1.5 cm. Flowers tubular, in pairs; arranged in loose, terminal (sometimes axillary) racemes. Perianth to 3 cm, bilaterally symmetric, scarlet to orange-red, very rarely yellow or white; segments separating at maturity. Anthers 4, stalkless. Style curved with conical pollen presenter. Fruit a woody pod, greenish yellow becoming red-brown when mature, with persistent style, containing numerous winged seeds (Muñoz 1966; Bean 1981; Hoffmann et al. 1998; Hoffmann 2005; Nelson 2011).

Distribution  ArgentinaChile

RHS Hardiness Rating H4

In flower, Embothrium coccineum can hold claim to being one of the most spectacular horticultural sights imaginable, its branches ablaze with scarlet in late spring, inspiring the common name Chilean Firebush.

Known locally as notro or ciruelillo, it is a widespread species in its native Chile and Argentina, with a distribution that covers some 2000 km north-to-south (Hoffmann et al. 1998). It occurs continuously from Curicó Province in central Chile to the southernmost part of Tierra del Fuego, while a population around Lago Ñorquinco in Neuquén Province constitutes the northernmost extent of its range in Argentina (Donoso 2006). Within such an extensive geographical distribution, populations can be found in a wide range of habitats, where they are influenced by the associated altitude, climate and soil conditions. With populations at sea level to 1300 m, these extend from temperate rainforests on the Pacific coast that receive 4000 mm annual rainfall to xeric ecotonal areas on the rain-shadowed eastern slopes of the Andes, bordering the arid Patagonian steppe (Rovere & Chalcoff 2010). Not surprisingly, E. coccineum a highly variable species exhibiting huge degrees of variation in flower morphology and colour but especially in habit and leaf shape.

It occurs as a pioneer species, often colonising gaps in vegetation following disturbance, particularly on acidic volcanic soils where little else can grow (Rovere & Chalcoff 2010). In habit it can vary enormously according to edaphic and climatic conditions: in high rainfall areas e.g. on the edges of the wet forests of southern Patagonia, it can make a tree to 15 m, growing with Drimys winteri, Maytenus magellanica, Nothofagus antarctica and N. betuloides; by contrast, on Volcán Osorno at 1100 m, it forms thickets no taller than 1 m high, associated with Escallonia alpina, Maytenus magellanica, Nothofagus betuloides, Ovidia andina and low suckering Gaultheria poeppigii (Gardner, Hechenleitner & Hepp 2015). Another explanation for why these stunted, shrubby individuals found in colder locations at higher altitudes naturally remain compact is that they are continually pruned by frost, which kills ascendant branches that emerge from the snow cover (Comber 1932). Plants from such populations are also more likely to be deciduous, but relatively short lived.

In terms of flower and pollination morphology, before the showy, tubular, nectar-producing flowers open, the four tepals are joined, enclosing the stamens to form a small sphere resembling a match-head, inspiring the Chilean common name fosforito – a small match (Gardner, Hechenleitner & Hepp 2015); on opening the tepals become strongly reflexed. As in many other Proteaceae, pollen is automatically transferred in bud from the anther to a swollen, non-receptive part of the stigma (the ‘pollen presenter’) from which it is brushed onto visiting pollinators (Rovere & Chalcoff 2010). Being a protandrous, highly self-incompatible species, it is fully reliant on pollinators; appearances might suggest hummingbird pollination, and in North American gardens hummingbirds are common visitors, but the situation in the wild is not so simple. In the highest rainfall areas of northwest Patagonia, hummingbirds are indeed the main pollinators although various birds, bees and long-tongued flies also visit and bring about pollination, if less effectively. Moving down a rainfall gradient into the rain shadow of the Andes, hummingbirds become less important than other visitors, with bees almost the sole pollinators in the driest areas (Devoto, Montaldo & Medan 2006; Chalcoff, Aizen & Ezcurra 2012). Flower colour typically ranges from rich crimson through scarlet to orange, even varying considerably on a single plant. Both anthocyanins and carotenoids contribute to flower colour, and the balance between the two presumably determines the precise shade (Rovere & Chalcoff 2010).

The species is known to have many uses. Its beautiful timber is used in furniture making, while dyes can also be prepared from the wood, flowers or seeds. Medicinally, an infusion of the leaves and branches is a known treatment for neuralgia or toothache, and a similar concoction is used as a laxative (Hoffmann et al. 1998; Gardner, Hechenleitner & Hepp 2015). Besides these more practical applications, it is widely grown as an ornamental tree throughout its native range where it is regularly planted in parks and gardens (Hoffmann et al. 1998). It has been used to make effective clipped flowering hedges on some large estates (Harper Goodspeed 1941), while they have also proven successful grown as street trees. On a visit to Frutillar, south of Osarno in the Chilean Lake District, Roy Lancaster saw a whole street lined with scarlet Embothrium, regularly pruned to maintain them as ‘standards’ with compact, rounded crowns (R. Lancaster pers. comm. 2023).

Embothrium coccineum was first introduced to cultivation in Europe by William Lobb in 1846, on his second trip to Chile collecting for Veitch & Sons (Veitch 1906), understood to have been gathered near Valdivia. Grown under glass, it flowered for the first time at Veitch’s Exeter nursery in 1853 (Shepard 2003), while some years later, another collection was made by Richard Pearce for Veitch between 1860–62 (Bean 1981). Around this time, two distinct forms were recognised in cultivation, one with brighter-coloured flowers than the other (Veitch 1906), although is unclear whose collections were referable to which form. By this time it had already made an impression: after flowering for the first time outdoors at Kew, one contemporary writer in The Gardener’s Chronicle mused that ‘perhaps no tree cultivated in the open air in the British Isles gives so striking and brilliant a display as this does!’ (Shepard 2003). It proved to be a fast-growing evergreen, with the flowers produced in terminal inflorescences (Bean 1981), but alas proved too tender for all but the milder maritime gardens of Cornwall, Ireland or Scotland’s west coast (Arnold-Forster 1948). For these early, low-altitude collections this remains largely true to this day, the exceptions being mature plants growing in obliging locations, likely from early Veitch introductions, for example at Kilmacurragh, Co. Wicklow (O’ Brien 2010).

More widely, the species is represented by plants derived from high altitude collections from the Andes, which ultimately proved hardier and more floriferous. For the most part these were made by Harold Comber collecting for the Andes Syndicate between 1925–7. He made several collections (as Embothrium lanceolatum) in Argentina on the slopes of the mountains surrounding the valley around Lago Ñorquinco, near the head of the Pulmari river, and close to the border with Chile (Comber 1928). Plants raised from his introductions (and their subsequent progeny) were found to be more amenable to growing and flowering in colder inland gardens (Aberconway 1947). A second reputedly hardy form has been known in cultivation for many years under the invalid name E. longifolium. It is said to differ from Comber’s introductions in having more substantial, persistent evergreen foliage and scarlet inflorescences at the ends of the branches (Arnold-Foster 2000). Another early 20th century introduction by Clarence Elliot of Six Hills Nursery was also attributed to E. longifolium (O’ Brien 2010).

Given the remarkable natural variation of Embothrium coccineum, it is not surprising to find a similar level of inconsistency throughout its taxonomic history. Authors have adopted a plethora of names to define this morphological or geographical variation, at species (e.g. E. lanceolatum) or infraspecific level, as in var. lanceolatum (Ruiz. & Pav.) O. Kuntze or f. andina (Hoffmann 2005). Meanwhile, attempts to recognise this variation in cultivated taxa has created much confusion surrounding the names attached to plants in gardens. While historic variants are still commonly encountered on labels and in nursery catalogues, and are unlikely to ever disappear from horticulture entirely, these have given rise to two Cultivar Groups as a means of defining the ‘horticultural’ variation within the species. They were first used in the 6th edition of the Hillier Manual of Trees & Shrubs (Hillier Nurseries 1991). Plants attributed to Lanceolatum Group include the ‘least evergreen but hardiest forms, with linear lanceolate leaves’ – as introduced by Comber (syn. E. coccineum var. lanceolatum (Ruiz. & Pav.) O. Kuntze), while those attributed to Longifolium Group differ in having ‘longer, somewhat broader, usually persistent leaves’.

Ultimately, these distinctions based almost purely on leaf size and shape can make identification of plants in gardens challenging at infraspecific level. It has long been recognised that Embothrium coccineum and var. lanceolatum (originally E. lanceolatum) could not be separated except by the shape of the leaves. However, even before Munõz (1966) amalgamated the two taxa, it was argued that this discrimination could not be upheld on account of the broad spectrum of intermediate leaf forms observed in herbarium specimens (Comber 1936). Since these observations, subsequent introductions and the raising of plants from seed has accounted for the presence of numerous intermediate leaf forms in cultivation (Nelson 2011; pers. obs.). Variation in leaf shape among seedlings is not unknown, with broad-leaved seedlings reportedly raised from lanceolate parents (Armytage Moore 1939); this is now widespread as a result of subsequent raisings from seed (pers. obs.). More recent introductions of wild origin material have broadened this gene pool, while further confusion is caused by the fact that the mature foliage can be seen to vary even on a single branch (Bean 1981; Schilling 1984; pers. obs.).

Other distinctions attributed to the Cultivar Groups are not much further help. Observations on habit, flowering and foliage persistence can be unreliable characteristics, subject to variation according to the growing environment. For example, it is more accurate to describe the species as semi-evergreen in cultivation, depending on local climate and conditions. Add to that much greater variation seen in cultivated material as described above, it is impossible to attribute hardiness to a particular foliar form, as was once possible when fewer distinct individuals could easily be recognised. On the basis of providing clarity, it seems futile to prolong this uncertainty. For the reasons outlined above, we believe use of the aforementioned Cultivar Groups should be avoided, and as such they are not upheld in this treatment.

Notwithstanding the confusion surrounding pre-existing Cultivar Groups, there is much to be said for singling out individual clones of horticultural merit, with proven hardiness or of floral quality. In fact, it is surprising that in a species that displays such remarkable variation, only a few selections have been made to date. This has not been helped by the habit of raising plants from seed (greatly favoured from a commercial nursery perspective), which has further led to the effective loss of distinctness of plants grown as a single clone – see discussion about ‘Norquinco’, treated here as Norquinco Valley Form. New selections of outstanding garden clones are needed to replace these and should be actively encouraged. Investigation into micropropagation may also provide a solution to producing cultivars in greater numbers, with a more guaranteed commercial success rate than other asexual methods of propagation.

Since its introduction the species has performed best in mild oceanic gardens along the western seaboard of Europe; perhaps nowhere more so than in the gardens of Devon and Cornwall, the west coast of Scotland and throughout Ireland, where it grows luxuriantly, but the oft-repeated advice that is only growable in the milder western extremities of Britain is misleading. In current climatic conditions it is safe to say that Embothrium coccineum could be planted throughout Britain and Ireland with the expectation of success. Exceptional examples include an erstwhile 18 m specimen at Brodick Castle on the Isle of Arran off the west coast of Scotland, another at 17 m at Fota Arboretum, Cork, Ireland and trees c. 15 m at Mount Congreve, Waterford, Ireland, and Marwood Hill, Devon (The Tree Register 2022), but very respectable specimens grow at Kiftsgate Court, Gloucestershire, high on the cold Cotswold escarpment, and in Ray Wood at Castle Howard, North Yorkshire (J. Grimshaw pers. comm. 2023). Many other examples could be cited from all over the UK and Ireland. The UK’s only (very small) National Plant Collection of Embothrium is held at Bodnant in North Wales (Plant Heritage 2023).

In continental Europe, however, records suggest good specimens are scarce (e.g. Plantcol 2019). The species is thought to have been introduced to North America by 1908, although hadn’t become established in the nursery trade until at least the 1940s (Jacobson 1996). It remains a rare plant there; its absence from key American horticultural literature (e.g. (Dirr 2009) is somewhat telling. As in Europe, its distribution is confined to appropriate climatic zones, almost exclusively on the Pacific coast. Many good examples can be found in Washington State, including an impressive 17 m specimen in a private garden in Seattle (Wott 2016). There are also several accessions at Washington Park Arboretum, from a variety of wild and cultivated sources (University of Washington Botanic Gardens 2019).

While always considered subjects for the woodland garden, Embothrium need good light to flower well, so perform best when they receive full sun for at least part of the day. The best flowering plants are always seen in an open sunny position (pers. obs.). On balance, they are also intolerant of high soil temperatures at the roots; less of an issue in a benign climate but should be considered in areas with hotter summers. In terms of siting, a north-facing aspect and some local shade can help mitigate this, as can the use of companion planting of groundcover plants or low-growing shrubs, to act as a ‘living mulch’ and keep the roots cool: evergreen azaleas are an option (Haworth-Booth 1970). This is critical in climates with high summer heat, where the combination of elevated soil temperatures with (necessary) summer moisture can lead to fungal infection and root death. Irrigating in cooler periods and planting in low organic matter soils can also help (Hogan 2008). Plants exposed to cold winds are more likely to be deciduous.

Growing in almost sterile volcanic mineral soils in habitat, they produce masses of fine proteoid roots, allowing them access the few available nutrients otherwise inaccessible to other species, particularly phosphorous; a useful adaptation as a keystone pioneer species (Piper et al. 2013). Therefore, Embothrium prefer relatively infertile, well-drained acidic soils in cultivation, resenting applications of inorganic fertilisers, although will benefit a mulch of leafmould. Being pioneers, they can be very quick growing in optimum conditions under cultivation – one reportedly grew 5.5 m in four years in San Jose, California (Jacobson 1996) – and will often flower at a young age, but are short-lived plants as a result.

Contradicting reports of strongly self-incompatible wild populations (Rovere et al. 2006), even single isolated individuals of Embothrium coccineum in cultivation will often set viable seed (J. Sutton pers. comm. 2023); and raising from seed is probably the easiest method of propagation. Semi-ripe cuttings taken at the earliest possible opportunity, rooted under polythene with bottom heat are also often successful in small numbers. Growing them on as seedlings or rooted cuttings presents more of a challenge, given the often-fatal effects of root disturbance on Proteaceae. Keeping this to a minimum, best results are obtained by sticking cuttings individually in plug trays or small pots, kept cool and dry over winter: they should only be potted on when growth resumes in spring (P. Catt pers. comm. 2023). Not allowing young plants to become checked in pots is also critical. It has often been advised to plant out small, even as 15 cm tall seedlings. Suckering plants can present a propagation opportunity on a small-scale, in that suckers can be detached and transplanted in situ, or potted-up, ideally before they leaf out early in spring. Typically ther have rather sparse roots and Jeff Layman (pers. comm. 2024) notes that he lifts and pots suckers in autumn, placing them in cool shade so that they are rooted-out before growth recommences in spring. By the same measure, it is suggested root cuttings may also be successful (Hogan 2008).

'Inca Flame'

Synonyms / alternative names
Embothrium coccineum (Lanceolatum Group) 'Inca Flame'

Selected at Duncan & Davies Nursery, New Zealand and introduced in 1990, this floriferous clone is described as a selected form with a tree-like form and ‘profuse brilliant orange-scarlet fiery catherine wheels all along the branches’ (Duncan & Davies Ltd 1991). It has also been observed to have a more rounded habit than other selections, with ovate leaves (O’ Brien 2010).

It is available in Britain, where Liss Forest Nursery, Hampshire were responsible for bringing this cultivar into wider circulation, having obtained it originally from Hewton Nursery, Devon (P. Catt pers. comm. 2023). Reports of general hardiness vary (Thompson 2001). Plants grown under this name at RHS Garden Wisley are decidedly semi-deciduous in cold winters, but are yet to live up to the catalogue description (pers. obs.). It has also been introduced to North America, where in Seattle it is reportedly a precocious clone, flowering reliably ahead of most in late April (Wott 2016).

'Inca King'

Synonyms / alternative names
Embothrium coccineum (Lanceolatum Group) 'Inca King'

While there is little mention of this cultivar in literature, it is known to be a another selection from New Zealand, said to be less hardy but more floriferous than the type (Thompson 2001). With a similar name to ‘Inca Flame’, there is much to suggest it was also selected by Duncan & Davies, but does not appear in their sales catalogues from that period (Duncan & Davies Ltd 1991).

Besides New Zealand, it has had a small distribution in Britain, infrequently listed by a handful of nurseries in RHS Plant Finder since the mid-1990s, and seldom seen so-labelled in gardens. A plant sourced from a now-defunct specialist nursery in Hampshire has grown well in Jeff Layman’s Hampshire garden since 2015, despite being planted in poorly drained clay soil, and suckers freely (J. Layman pers. comm. 2024).

Lanceolatum Group

For the reasons outlined in the main text, this Cultivar Group and its synonyms are no longer upheld.

Longifolium Group

For the reasons outlined in the main text, this Cultivar Group and its synonyms are no longer upheld.

Norquinco Valley Form

Synonyms / alternative names
Embothrium coccineum 'Norquinco Valley'
Embothrium coccineum 'Norquinco'
Embothrium coccineum Norquinco Form
Embothrium coccineum (Lanceolatum Group) 'Norquinco'

AGM 1992

‘Looking as though it had donned a pair of scarlet “plus fours” when in flower’ was the evocative description for an outstanding plant exhibited at an RHS show in 1947 by Lord Aberconway as E. lanceolatum Norquinco Valley Form (if treated as a cultivar the name should be ‘Norquinco Valley’ to reflect this first publication, rather than the more frequently seen ‘Norquinco’). It represents the best of the progeny from Comber 387 raised at Bodnant, North Wales, where it was selected for its floriferousness (Aberconway 1947). Collected at c. 1000 – 1500 m in the Ñorquinco valley in Argentina in 1927 (Comber 1928), this collection’s relatively high altitude provenance accounts for its reported hardiness in cultivation, another attribute that makes this a superlative garden selection. It was highly regarded by all that saw it: ‘the finest form… with flowers in great fox-brushes of flaming scarlet, even more numerous and effective’ (Haworth-Booth 1970).

As such, with a glowing reputation accompanying it, plants labelled ‘Norquinco’ are still sought after, although unfortunately there is no certainty that plants bearing the various iterations of that name (see synonyms) represent a single clone, nor indeed the one exhibited from Bodnant. From there, it was propagated vegetatively from suckers, as occasionally distributed by Hillier Nurseries through the last century (A. Postill pers. comm. 2023), but it is suggested that second-generation seedlings were more often sold under one or other of the variants of the name that have been applied (often inconsistently) over the years (Lancaster 1987). On the basis of its popularity, this is not unlikely given the relative ease of raising Embothrium from seed in quantity, compared to vegetative methods. This remains true today, where seed-raised ‘Norquinco’ are still sometimes available in nurseries (C. Williams pers. comm. 2023). While many of these are themselves excellent plants (not unlike the original), there is some inevitable inconsistency and, over time, this has diluted the authenticity of this as a named clone. It further highlights the need for new selections to be made and propagated, as suggested in the main text.

We are therefore reluctant to accept ‘Norquinco’ as a cultivar, as it is clear that any connection to the original plant is now at best tenuous, and even to define characters that would enable the establishment of a Cultivar Group is challenging. Nor, in our opinion, should the RHS Award of Garden Merit be attached to such a set of entities. Further research is needed to establish whether it is possible to define a Cultivar Group from extant examples that also match what is offered in the trade.

'Windwhistle Weeping'

This truly extraordinary weeping form of E. coccineum was recently found in the unlikely location of Cricket St Thomas Golf Club (then Windwhistle Golf Club) in Somerset, UK, as reported in The Plant Review (Armitage 2018). Growing on the 18th hole, it first attracted attention when images were sent to the RHS Advisory Service in 2011. How it came to be there remains a mystery, but it is nothing short of striking, flowering profusely to the very tips of its strongly weeping branches, which express clear positive geotropism. Its habit is so distinct – especially when compared with two ‘typical’ plants growing nearby – that it warranted clonal recognition; it was later named ‘Windwhistle Weeping’. After three attempts, nurseryman Peter Catt (pers. comm. 2023) successfully rooted it from cuttings. Young plants have been distributed to gardens but are proving to be ver slow growing.

Similar individuals are not unknown in cultivation, inevitably occurring as anomalies among batches of seedlings; aberrant forms are not wildly unexpected within a species known for its morphological variation. Such plants are found in Chile, including one with a ‘crooked trunk and pendent twigs’ that grew on a street in Bariloche, Argentina, as described by German botanist Hermann Sleumer in a letter to Royal Botanic Gardens, Kew (Sleumer 1953).

Dan Hinkley (pers. comm. 2023) collected seed from another growing in a town square, just north of the Lake District in Chile. Interestingly, most seedlings he raised retained this characteristic, evident even from an early age as young plants in pots. Although none of his own original plantings remain, some were sold by Heronswood Nursery, Washington as ‘E. coccineum weeping form’ (Heronswood Nursery Ltd. 2001). There is a chance one may still exist in a North American garden somewhere.

In the United Kingdom, at least one other weeping clone is known, with others perhaps not unlikely to occur in future. While probably not a valid name, an old plant labelled E. coccineum ‘Pendulum’ grows at Glenwhan in Scotland, where it has made a tall, upright tree, flowering on exceptionally lax, trailing branches (J. Sutton pers. comm. 2022). It is thought it may have originally come from Claymoddie, Wigtownshire, where Robin Nicholson kept an interesting collection of southern-hemisphere plants (T. Knott pers. comm. 2022). While this is the only known mature example, interestingly young plants under that name have been distributed by the late nurseryman Alan Clark in recent years (P. Shotter pers. comm. 2022).

Yellow-flowered forms

Embothrium coccineum var. luteum Nom. invalid

Very few woody plants have excited the imaginations of enthusiasts like the canary yellow-flowered forms of Embothrium coccineum. They have remained highly elusive, making brief but tantalising appearances in cultivation throughout the last century. Attempts to introduce them, from both wild and cultivated sources, have been largely unsuccessful with plants lost to unfortunate circumstances. Meanwhile written accounts of these efforts (e.g. Hillier & Beckett 1977; Mitchell 1999; O’ Brien 2010) have only served to further fuel the desire to see this striking colour form represented in gardens outside Chile.

Success at long last came when, in a recent article, Martin Gardner and Paulina Hechenleitner confirmed that a single yellow-flowered plant is established in the open ground at Royal Botanic Garden Edinburgh. It has been given the clonal name ‘Yellow Flame’ and, while its true provenance remains undisclosed, it is said to represent the same clone as a group of trees that grew at Universidad Austral de Chile (Gardner & Hechenleitner 2021). Given the understanding that any propagated material wouldn’t be commercialised, it will not become a mainstay feature in obliging gardens, but likely be an exciting addition to botanic collections in the UK. It remains the only successful introduction outside Chile to this date.

Naturally occurring yellow variants are reportedly very rare, making them something of a ‘holy grail’ plant for collectors travelling in Chile and Argentina. In fact, they are so infrequently scattered throughout the species’ distribution that only two of the ten field botanists in Chile & Argentina consulted by Gardner & Hechenleitner 2021 had ever seen this form in the wild. In Argentina, they have been seen in at least two locations: in the north, in Parque Nacional Lanin, Nequen province and further south, in Parque Nacional Los Glaciares, Santa Cruz province (Donoso 2006). A low-growing yellow form seen by Sean Hogan on the lower slopes of Anconcagua, in northwest Argentina (Hogan 2008) is now thought to be of cultivated origin, rather than represent a new disjunct wild population (S. Hogan pers. comm. 2023). Over the border in Chile, yellow forms are better represented in cultivation than in habitat. In 1999, Roy Lancaster was taken to see two such plants near Frutillar, which had been transplanted by a forester from the wild into his private garden (Lancaster 2017). Some of the best specimens grew at Arboretum of the Universidad Austral de Chile in Valdivia, before they were uprooted by a storm in 2013 (Gardner & Hechenleitner 2021).

Of the few known wild examples in Chile, those growing on the island of Chiloe have been well documented; the first known successful introduction to UK cultivation was derived from these plants. Cheese, Watson & Beckett collected material from three yellow-flowered plants, seen south of the town of Ancud, on their 1971 expedition. While cuttings reportedly failed, seed was collected and distributed under Cheese & Watson 5019 (Hillier & Beckett 1977). One recipient, Alan King, germinated seedlings, raising six plants under glass in large pots at his home in Northamptonshire, UK. Of the three that survived, one plant flowered yellow (against all the odds!) for the first time in 1981 (King 1985). It was later planted out at Wakehurst Place, only to be lost along with countless other rarities in the devastating 1987 storm (J. Armitage pers. comm. 2022).

Attempts at vegetative propagation of cultivated material from Chile are a similar story of woe and misfortune. Sir Harold Hillier received scions in 1977 from a yellow-flowered tree that grew in a public park in Purranque from a Dr Wygnanski. It was likened to Comber’s ‘Norquinco’ with narrow deciduous leaves, and presumably as hardy, but with flowers a similar colour to those of Jasminum nudiflorum – a wonderful prospect! Having successfully grafted two plants, he named the clone ‘Eliot Hodgkin’ after a mutual friend (Hillier & Beckett 1977). Alas, we will never know for certain what came of the two plants – with no record they were ever planted out or flowered, it is thought they may have even been stolen! (O’ Brien 2010; J. Armitage pers. comm. 2022).

More recently, others have tried in vain; nurseryman Peter Catt successfully rooted material received from Chile from cultivated plants, but the resulting plants all succumbed to pot culture before ever flowering (P. Catt pers. comm. 2023). Meanwhile in North America, rooted cuttings from the low-growing plant in Argentina (mentioned previously) were distributed, although none of these survived (S. Hogan pers. comm. 2023).

The white flowered form mentioned by Munoz (1966) has never been seen in cultivation. It is unlikely to be as striking as the yellow.