Catalpa bignonioides
Sponsor
Kindly sponsored by
Nicholas Barber CBE
Credits
Martin Deasy, Richard Olsen & John Grimshaw (2025)
Recommended citation
Deasy, M., Olsen, R. & Grimshaw, J. (2025), 'Catalpa bignonioides ' from the website Trees and Shrubs Online (treesandshrubsonline.
Common Names
- Indian Bean Tree
- Southern Catalpa
- Eastern Catalpa
- Later Blooming Catalpa
- Petalfra
- Catawbaw Tree
Synonyms
- Bignonia catalpa L.
- Bignonia triloba T.Freeman & Custis
- Catalpa bignonioides var. aurea (Mast.) Lavallée ex Bureau
- Catalpa bignonioides var. nana (Otto & A.Dietr.) Bureau
- Catalpa bignonioides var. pulverulenta (Rob.) Bean
- Catalpa bignonioides var. variegata Bureau
- Catalpa catalpa (L.) H.Karst.
- Catalpa cordifolia Moench
- Catalpa syringifolia Sims
- Catalpa syringifolia var. koehnei Hesse
- Catalpa tibetica Forrest
- Catalpa umbraculifera Ugolini
- Catalpa vulgaris hort.
Infraspecifics
Deciduous trees attaining 20–30 m in height, bark smooth when young, characteristically platy on old trunks; young stems glabrous, sparsely glandular with elliptic lenticels. Leaves in whorls of three, distinctly unequal in size within each whorl, emitting a rank, musky odour when bruised, often purple when emerging, petioles 10–18 cm long, lamina entire (rarely weakly trilobate), ovate to broadly ovate, (10–)15–31 × (8–)12.5–22.5(−24.5) cm, base cordate, truncate, obtuse (rarely cuneate), apex acuminate, midrib with 6 or 7 arcuate secondary veins on each side and 2 or 3 at base, papery, ± glabrous above, abaxially glabrate to villous with villous veins and extrafloral nectaries consisting of dark brown areas of closely-packed glandular trichomes in the axils of 4 or 5 basal veins. Inflorescence a conical thyrse of 60–80 flowers, with 4–6 whorls of 3 branches evenly spaced along the rachis and an apical flower, rachis and peduncles ±glabrous with scattered lenticels. Flowers with pedicels 3–16 mm long, glabrous, glandular, with a whorl of 3 small narrow bracts mid-pedicel, frequently damaged or lost on older flowers; calyx calyptrate, light-bulb-shaped in bud, 8–12 × 6–9 mm, purple to greenish purple, glabrous, glandular, splitting into 2 (rarely 3) lobes; corolla white, corolla tube 1.5–2.5 cm long, widening from base to 8–12 mm wide at throat, three lower and two upper lobes spreading, frilled, lower lobes slightly longer with conspicuous purple dots along veins and inside tube, with 2 keyhole-shaped bright yellow spots at mouth of tube on inside between lower lobes, turning dark red with age; stamens and staminodes white, inserted at base of corolla tube; ovary 2–3 mm long, glabrous, green, 2-carpellate, style 12–20 mm long, white; stigma 1–2 mm long, white. Fruit a many-seeded capsule resembling a long green bean, 1–6 per infructescence, 24–56 × 0.9–1.2 cm, glabrous; seeds narrowly oblong, with connate hairs drawn out into an acute ragged tip at either end. (Olsen & Kirkbride 2017, Stephenson 1981).
Distribution United States northern Florida, southern Georgia, southern Alabama, southern Mississippi, southern Louisiana, eastern Texas, southern Arkansas
Habitat Humid areas
USDA Hardiness Zone 5-9
RHS Hardiness Rating H6
Conservation status Not evaluated (NE)
The first Catalpa species to be introduced to cultivation, and the type of the genus, Catalpa bignonioides was first described by Mark Catesby in his Natural History of Carolina, Florida and the Bahama Islands (1729) from collections made during his travels in the southeastern United States earlier that decade. As one of the first Europeans to venture further inland than the colonial settlements on the coast, Catesby encountered many unfamiliar plants on his expeditions into the interior, one of which was a tree ‘unknown to the inhabited parts of Carolina till I brought the Seeds from the remoter parts of the Country. And tho’ the Inhabitants are little curious in Gard’ning, yet the uncommon Beauty of the Tree has induc’d them to propogate it [sic]; and tis’ become an Ornament to many of the Gardens’ (Catesby 1729).
Catesby’s name for the tree, published with a Latin diagnosis (‘dirty white flowers sprinkled with purple and yellow spots, with very long, very narrow pods’) was Bignonia urucu foliis [‘Bixa-leaved Bignonia’] (not given in English by Catesby, but translated as ‘Bignonia aux feuilles de Rocou’ in his parallel French text). Urucu is one of the vernacular names for the Central and South American Bixa orellana – annato or achiote (Mabberley 2017) – whose palmate-veined ovate-acuminate leaves recall Catalpa.
Catesby’s name (which was not a binomial) was substituted in Linnaeus’s Species plantarum (1753) by Bignonia catalpa, the epithet being the word that Catesby had reported as the local Native American name for the tree (‘katalpa’ or ‘kutuhlpa’ (Olsen & Kirkbride 2017)), and not a derivation from the name of the Catawba tribe, a supposition that goes back to Nuttall (1818) and which has been repeated uncritically by many who followed.
Brought back to London by Catesby on his return in 1726, Catalpa bignonioides was germinated and sold by Stephen Bacon’s nursery at Hoxton, where Philip Miller (1735) reported its having flowered in 1732 (‘very hardy […] a handsome upright tree: the leaves are very like those of the lilac, but somewhat larger’). Miller dubbed the tree the Lilac-leaved Bignonia (Bignonia syringae caeruleae foliis), the ‘blue Trumpet Flower, or Cattalpa […] brought from the Bahama islands by Mr Catesby a few years since’ (the Caribbean association here is an obvious misunderstanding).
Miller was a curious and diligent observer, and in the revisions made to successive editions of his Gardeners Dictionary from the 1730s we are able to watch the catalpa’s progress in real time, as it were, as it established permanent residency in London’s gardens and nurseries. By the early 1750s, we find that the seeds ‘have been raised in many of the gardens near London; so that now it is propagated pretty commonly in the Nurseries near London, and sold as a flowering Tree to adorn Pleasure-gardens’; also that it is ‘now usually propagated by Cuttings’ (Miller 1752). Lancelot ‘Capability’ Brown was able to purchase plants at 1s 6d apiece in 1755, whereas two decades earlier the specimen at Carlton House, belonging to the Prince of Wales, had cost three guineas (Fraser & Sellers 2014). By the 1760s, the ‘Bignonia with Anatto Leaves’ was firmly established, though fruiting was still unknown in Britain: ‘the Flowers are in America succeeded by very long taper Pods, filled with flat winged Seeds … in England there has not as yet been any of the Pods produced; but the Seeds are annually brought over from South Carolina’ (Miller 1764, 1768). In fact the tree appears not to have produced fruit throughout the whole of the eighteenth century, since as late as 1808 John Sims stated in Curtis’s Botanical Magazine that ‘[w]e have never known it to produce seeds’ (Sims 1808).
In Paris, the tree was recorded on a garden plan of 1740 (Bailly de Merlieux 1847), though it cannot be discounted that the specimen in question – growing in a garden on the Rue Vieille-du-Temple in the Marais – could have been an early introduction of C. speciosa (see that entry for a discussion of the eighteenth-century history of the genus in France).
A vivid sense of the catalpa’s conquest of European gardens over the following century can be gained from the long catalogue of specimens and statistics in John Loudon’s Arboretum et Fruticetum Britannicum (1838, 1844), which lists scores of notable trees from London to the Crimea, ranging from a poor specimen barely hanging on at the Glasgow Botanic Garden (‘almost herbaceous, even under the shelter of a wall’) to a tree at Avranches, France reputed to have attained a scarcely credible (actually, incredible) height of 27 m after only 29 years (girth 1.33 m).
Sims (1808) reported that ‘the largest specimen we have ever seen grows in the garden belonging to the Society of Gray’s-Inn’. This was clearly a very early planting, its veteran status giving rise to the tradition (cherished by the Inn’s members) that it had been planted by the philosopher and statesman Francis Bacon (1561–1626) who had laid out the gardens in 1608 – Bacon’s essay Of Gardens (Bacon 1625), in which he explores his ideas of garden-making, is a horticultural classic). Evelyn Cecil (1907) gave this fancy short shrift, noting that the tree, whose exotic appearance would have occasioned wonder in seventeenth-century London, is mentioned neither by Parkinson (1629), Johnson (1633), nor Evelyn (1664). A further embellishment – that the plant had been given to Bacon by Walter Raleigh, as reported sceptically by Elwes and Henry (1906–1913) – is equally implausible given the species’ natural range, and the lengths to which Catesby would later have to go to find it on his travels; the 1584 voyage sponsored by Raleigh – who did not make the trip in person – did not even venture onto the American mainland, remaining ten days on the islands of the Outer Banks. Should any doubt remain, it could be added that there is no mention of anything resembling catalpa in the principal account of the voyage (Hakluyt 1589), in Thomas Hariot’s Briefe and True Report (1588), or in Gerard’s Herball (1597).
It seems obvious that the confusion must result from a conflation of Francis Bacon with the nurseryman Stephen Bacon who originally brought Catesby’s tree to market. ‘Bacon’s catalpa’ in this light, takes on a completely different light. What is unexpected – indeed quite remarkable – is the fact that the tree still exists in a living albeit vestigial form. The ancient tree was still standing in 1870, ‘its aged limbs sweep[ing] the ground’ (Gardeners’ Chronicle 1870), but by 1909 it had been reduced to a single layered branch from the original plant, rooted at one end and supported by props at the other, as shown in a contemporary photograph. Elwes and Henry (1906–1913) reported its having died, though this was evidently not quite true, and the remnant limb survives and flowers even today. Not quite the methuselah of repute, it is nevertheless plausibly datable to Catesby’s first introduction in 1726, making it comfortably the oldest catalpa in cultivation anywhere in the world. A cutting taken from the original, planted in the Walks a short distance away, had itself made a very large tree by the middle of the twentieth century, only to be crushed by a neighbouring plane tree in the Great Storm of 1987 (Gray’s Inn 2025). Like its parent, the remains of this tree – now reclining on the lawn – continue to make new growth, to flower and to fruit.
Another improbably dated tree, described by Loudon (1838) as having been ‘planted 150 years’ (i.e. around 1690), grew in the gardens of Fulham Palace, which under the patronage of the botanist bishop Henry Compton (1632–1713) had from the 1670s been the preeminent focus of new plant introductions from the North American colonies. Compton did not live to see a catalpa, dying more than a decade before Catesby’s return from America, and his episcopal successors were not known as gardeners (Loudon’s implied date almost seems a retrospective calculation based on the garden’s heyday). Nevertheless the presence of what was clearly an exceptionally ancient catalpa at Fulham is intriguing, possibly indicating an origin in the early batch of seed brought back by Catesby. Yet another very old tree, probably contemporaneous with that at Fulham, grew upriver at Syon House. Loudon (1844) reported its girth as 90 cm, but the engraving in the corresponding volume of plates gives the girth as 60 cm, height 16.5 m, and shows a fully grown, handsome tree with marked wind-pruning to its eastern side, but potentially a centurion at least. These very early-planted specimens can plausibly be associated with Catesby’s original introduction, planted while the catalpa was still little known and material was scarce. Later in the century, imported seed from American sources would become easily obtainable (Miller 1764, 1768).
For a tree that could be grown successfully out of doors, the catalpa’s subtropical aspect – large leaves, showy flowers and the capacity for startlingly rapid growth – would have astonished northern European gardeners. Indeed, the tree’s exotic appearance was just as baffling to early settlers in the New World. Its natural range fell entirely outside the boundaries of the Thirteen Colonies, within the territories that lay beyond the Proclamation Line of 1763, so it would have been unknown as a wild plant except to those few travellers of exploratory bent who ventured upcountry (George III’s edict only prohibited settlement by the colonists west of the Line, not travel). Given the catalpa’s unusual appearance, the fact that it was known only in cultivation caused some to wonder whether it might not in fact have been introduced from some far-flung location (Warder 1881). After all, Linnaeus (1753) treated Catesby’s Bignonia urucu foliis as conspecific with a tree reported from Japan by Engelbert Kaempfer several decades earlier, the as-yet-unnamed Catalpa ovata (‘habitat in Japonia et Carolina’).
The possibilities for westward expansion brought about by the political upheavals of the century that followed – notably the declaration of independence in 1776, and the Louisiana purchase of 1803 – brought travellers into increased contact with Catalpa bignonioides in its natural habitat. Thus while François André Michaux (1813) prepared his account of the tree without ever having visited ‘the thinly inhabited country in which it abounds’, his successor Thomas Nuttall, a well-travelled and perspicacious observer, travelled across this region in 1830, reporting that ‘for the first time in my life [I] beheld this tree decidedly native, forming small haggard crooked trees leaning fantastically over the rocky banks of the river [Chattahoochee]’ (Nuttall 1849).
Appreciated chiefly for its amenity value, the tree’s cultivated range crept gradually northwards as far as New England (‘propagated and dispersed […] through most of the English Settlements in North America’ (Miller 1764)). Then, as the settlement of the Midwest picked up pace in the second half of the nineteenth century, settlers brought the tree with them into Ohio, Illinois, Kentucky, Missouri and Iowa – areas in which, however, it was found not to be hardy. As discussed in the entry for Catalpa speciosa, the move westwards, and the experience of the exceptionally hard winters of 1855 and 1856, led to the realisation that there were two species of catalpa, one of which (the ‘Western Catalpa’) was significantly hardier than its eastern counterpart familiar from cultivation along the Atlantic seaboard (C. bignonioides). From this point on, the fortunes of C. bignonioides begin to decline relative to those of its hardier counterpart, though the tree’s early popularity means that outside America it is still more frequently encountered than its more recently discovered sister species.
That it should have taken so long for C. speciosa to be distinguished from C. bignonioides is perhaps not so surprising in view of the similarities between them. Seen at close quarters, however, C. bignonioides is easily told by its scaly (not ridged) bark, as well as the drawn-out tip of its seed-coma (rather than an obtuse fringe). Its inflorescences also tend to be distinctly fuller and taller than those of C. speciosa, the flowers more conspicuously spotted with purple and the overall effect more demure than its sister species with its rather theatrically frilled lobes. Seen from a distance, the trees also have a tendency to an irregular, stooping habit, though this character is less likely to be seen in well-grown ornamental specimens than in wild populations.
Bean (1976) described the tree as ‘not particularly long-lived’, suggesting that thought should be given to succession planting by the managers of collections. However, as the saga of the Gray’s Inn catalpa demonstrates, trees are able to regenerate clonally by ‘resting their elbows’ on the ground, contact with the soil inducing branches to root and produce a succession of reiterating trunks. An existing catalpa at Syon House is a layer from the tree described by Loudon (Elwes & Henry 1906–1913), and a similar phenomenon can be seen in the celebrated tree on the Insel Mainau in Lake Constance. Such specimens require both ample space in which to spread their limbs, and a degree of indulgence on their owners’ part as the trees enter senescence and lose their shapeliness. But what is lost in regularity of form is gained in individuality, and these can be remarkable trees. One in West Medford, Massachusetts, in particularly well-irrigated ground, produced ‘a small forest of young catalpas […] under the shade of the parent tree’ covering an area of 28 m2 (Brooks 1890). Brown (1995) speculated that this rooting habit might represent a distinct form of the tree, rather than a universal trait, on the basis of his observations at Kew and elsewhere.
Among the more wayward synonyms acquired by Catalpa bignonioides is C. tibetica, described by George Forrest from specimens collected (by agents on his behalf) in Tsarong, Tibet, but never seen since, even though it is included in the authoritative Flora of China (Zhang & Santisuk 1998). This species posed an intriguing conundrum for botanists until Richard Olsen and Joseph Kirkbride, having studied Forrest’s herbarium specimens in Edinburgh, Kew and Paris, concluded that they are identical to C. bignonoides and must therefore have been collected from a planted or naturalised tree somewhere in Yunnan (Kirkbride & Olsen 2011). As Grimshaw and Olsen (2011) have pointed out, given the long history of Western missionary activity in China, it is far from improbable that so attractive a flowering tree as C. bignonioides should have been planted in a mission garden or elsewhere, even in some remote outpost. The episode – one of the many man-made muddles that seem to beset the genus – serves as a timely reminder that traffic in plants went in both directions, Western plants having been introduced to China for more than a thousand years (Sheng 1979).
Aurea Group
Synonyms
Catalpa bignonioides 'Aurea'
Catalpa bignonioides 'Aurea Variegata'
Catalpa bignonioides var. aurea (Mast.) Lavallée ex Bureau
Catalpa bignonioides f. aurea (Mast.) Schelle
Catalpa syringifolia var. aurea Mast.
Representatives of this group are spreading, medium-sized trees to approximately 15 m in height, with rich yellow or golden-green leaves, smaller in all characters than type C. bignonioides, but of similar vigour. The first gold-leaved clones were seedlings that occurred contemporaneously at the nurseries of Gaujard (Wetteren, Belgium) and Van der Bom (Oudenbosch, Holland) in the late 1860s (Burvenich 1874). Cripps and Sons introduced the variety to Britain in 1870 (Gardeners’ Chronicle 1870), evidently from one of these two sources, though it is not known which. Lavallée was growing it at Segrez by the middle of the decade (Lavallee 1877), and plants under the name were being grown in the United States by 1880 (Parsons 1880). In America, the tree was sold initially under the name ‘Aurea Variegata’, though it is clear from the description that this is the golden-leaved form (Ellwanger & Barry 1886).
As the variety’s multiple origins indicate, plants going by the name ‘Aurea’ are better treated as a cultivar group, a position adopted here. Commercially produced plants are usually vegetatively propagated, but some (not all) trees produce fertile seed. Hortus Loci Nursery (Hampshire, UK) has raised numerous yellow-leaved seedlings from an isolated, apparently self-fertile specimen, while seed from a tree at Hillier Nursery proved not germinable (D. West, pers. comm. 2025). According to Jacobson (1996), the tree ‘comes partly true’ from seed.
The variability in leaf colouring, while partly attributable to variables such as exposure, latitude, light level, as well as to seasonal changes in leaf chemistry, does not seem to be wholly due to cultural factors, and may well reflect genetic differences. In some trees the foliage attains a rich unalloyed yellow, particularly in full sun, while in other instances the characteristic ‘catalpa green’ of the type C. bignonioides remains detectable, giving a more greenish tint that appears chartreuse when lit up by the sun: it is telling that a century and a half of catalogue descriptions cannot agree on whether the foliage is golden-yellow or yellowish green, ‘vert-blond’ or ‘jaune pâle’ (Audiguier, Besaucèle & Robinet 1879, Chargueraud 1889). The reddish tint of the young growth tends to pull the overall effect in the direction of rich yellow. Cripps (1871) described the plants they sold as ‘not a pale sickly straw colour, but a deep golden yellow, suffused with green near the principal veins […] Although raised from seed [i.e. by the original nursery] three years ago, it has remained very true’. In some plants in contemporary cultivation the veining on the upper leaf is reddish.
The foliage of some plants – though not others (cf. Bean 1976) – shows a tendency to revert to green towards the end of the season, a phenomenon more commonly, but not exclusively, reported in relation to American trees (Otto 1881, Burvenich 1877, Ellwanger & Barry 1886, Jacobson 1996). Curiously, the first description of the new variety reported the opposite effect – yellow leaf colour at either end of the season, but greening in mid-summer (Burvenich 1874).
The tree is commonly topworked on stocks of either of the American species and/or pollarded, leafing into a globe-shaped crown, at which point a young pollard might be mistaken for a golden form of C. bignonioides ‘Nana’. However, the extension growth quickly reveals a tree of significant vigour, though the proportions remain much more compact than the type species, and the trees branch lower, forming a lower crown. Jacobson (1996) reported a Polish tree that had attained a height of 12.8 m, 90 cm in girth, while Clarke (1988) gives the dimensions of a tree in Dulwich Park, London as 11 m × 175 cm girth. The Belgian national register records several specimens of 15 m in height; a tree in Bruges is reported to have been measured at 30 m in height (3.14 m girth) in 2003 (Beltrees 2025) – but this does not seem credible, since even type C. bignonioides does not attain that height in Europe.
Bean (1976) is rather sniffy about the golden catalpa (‘Those who admire yellow-leaved trees …’), but unlike many plants in this category, in which the yellow carries associations of chlorosis, C. bignonioides ‘Aurea’ gives not even the suspicion of sickliness. The leaves’ softness of texture complements the unusual hue, and the overall effect is quite distinguished. Its ‘wonderful luminous quality’ caught the eye (unsurprisingly) of Christopher Lloyd, who though it best used it as a shrub in a mixed border, pollarded or stooled annually in winter (Lloyd 1985). The leaves do not burn in full sun, a common problem in other yellow-leaved plants.
'Aurea Nana'
A plant under this almost certainly illegitimate name has had limited circulation in the UK and continental Europe. In England it is best known from a tree at John Ravenscroft’s Cherry Tree Arboretum in Shropshire, but its origins are obscure. It has, however been propagated by cuttings and some plants have been distributed: John Grimshaw was given one for the Yorkshire Arboretum in 2022, for example. The only other source located by Chris Sanders is the Yuflor nursery in Serbia, whose website describes it as ‘a newer dwarf Catalpa with golden-yellow leaves.’
Chris Sanders (pers. comm. 2025) says of the Cherry Tree Arboretum specimen ‘it does appear to be a slower-growing, bushier form of ‘Aurea’ with smaller yellow leaves, but is much less compact than ‘Nana’.’ It is possible to speculate that it is a selected slightly more compact seedling from ‘Aurea’: it is clearly not connected to C. bignonioides ‘Nana’.
'Koehnei'
Synonyms / alternative names
Catalpa syringifolia var. koehnei Hesse
Golden-yellow leaves with a large dark-green blaze from the lower petiole to the middle of the leaf, and dark-green colouration to some of the veins, the colour said to intensify in strong light (Hesse 1905). Introduced in 1902 by the Hesse Tree Nursery (Weener, Hannover), where the the variety originated as a chance seedling in the late 1890s (Hesse 1903). It was an early addition to the collection of the catalpa collector and breeder Émile Gallé (see C. × galleana), who was growing ‘le nouveau Catalpa Koehnei’ in 1904 (Lambert 1904). The honoree of the varietal epithet is unclear, the most plausible candidate being the German dendrologist Bernhard Adalbert Emil Koehne (1848–1918), though as yet no evidence has come to light that would associate him with Catalpa in general, or this cultivar in particular.
'Nana'
Common Names
Umbrella Catalpa
Dwarf Catalpa
Synonyms / alternative names
Catalpa bungei var. nana hort.
Catalpa kaempferi hort., not (DC.) Siebold & Zucc.
Catalpa bungei hort., not C.A.Mey
Catalpa bungei var. pumila hort.
Catalpa kaempferi var. nana hort.
Catalpa umbraculifera Ugolini
Catalpa kaempferi caerulea nana hort.
A dwarf clone to 5.5 m high, originating in Paris in the mid nineteenth century, and known under many names. Topworked onto a short stem, it has proved a versatile tree for use in small residential gardens as well as formal landscapes on both sides of the Atlantic; in such contexts it is normally treated as a pollard, producing a round ball of foliage that has provoked reactions ranging from the quizzical (‘quite a strange effect’ (Bureau 1894) to the scathing (‘an inelegant and unattractive blob on a pole’ (Grimshaw & Olsen 2011)). Once the stem has increased in girth however, much of the incongruity diminishes, and the result can be effective and decorative. If the branch scaffold is then allowed to extend, the crown fills out into a well-proportioned ‘umbrella’ form. Unpruned specimens, meanwhile, develop into remarkably characterful trees, the canopy gracefully descending to ground level in a wall of foliage, the leaves ‘laid with shingle-like precision’ (Greening Nusery Co. 1917). The effect is unique.
The plant is propagated vegetatively, and appears to be a single clone (the speculation by Bean (1976) that the weeping effect seen in some American specimens is of different origin appears to be unfounded). Young leaves are often flushed red. There is one report of flowering in a European tree – a few depauperate inflorescences with incompletely developed, ‘small, ugly’ flowers (Jan De Langhe, pers. comm. 2025), but sparse flowering and fruiting does occur in unpruned North American grafted trees grown under this name (University of Utah 2025).
There is mystery aplenty to be had in relation to the cultivar ‘Nana’, but the date of its introduction into cultivation can actually be identified with reasonable precision, since a new dwarf catalpa (‘catalpa nain’) was reported by the seedsman Mathieu Bossin in 1850 to have been grown for several years by a certain Masson (Bossin 1850) – almost certainly at the trial garden of the Société centrale d’horticulture, located in the nursery of the Palais de Luxembourg, where Étienne Masson was gardener (later head gardener). The Muséum d’histoire naturelle, based at the nearby Jardin des Plantes, acquired the same tree in 1847 (Pépin 1856); by 1856 it was 2.2 m tall with a 2.3 m spread and a girth of 30 cm at 0.3 m (the Museum gardeners had also experimented with topworking this form onto regular C. bignonioides stocks). Since Pépin’s tree must have been a cutting from Bossin’s – Pépin used the phrase ‘le premier pied introduit au Muséum en 1847’ – this would push back the origins of Masson’s dwarf into the early 1840s.
No type was described until a neotype was designated by Olsen & Kirkbride (2017), but Bossin’s detailed description of the vegetative characters of Masson’s plant – an as-yet-unflowered shrub of 5–10 years old – serves well in lieu of a protologue: shoots and leaves brownish-purple when young, and young shoots of a similar colour (flecked with slightly raised white spots), subsequently taking on a green tint, becoming pale grey bark on 2-year stems; young stems soft, glabrous, slightly viscous, quadrangular in form, slightly channeled on the four sides, internodes closely spaced; leaves 18–22 cm wide, of similar length, 3-lobed or sometimes entire opposite, petiole 15–25 cm, dark violet above, green below. The veins on the adaxial side are brown above and green abaxially. The tip of each lobe is dark green above, soft green below (Bossin 1850).
The origin of Masson’s tree is unknown, but it was clearly a dwarf form of C. bignonioides – Bossin (1850) described it as C. bignonioides ‘en petit’ – perhaps originating as a mutant seedling. The dense branching pattern, giving a bushy effect, is consistent with the abnormal proliferation of axillary shoots known as a ‘witches broom’, caused by hormonal (cytokinin/auxin) imblances. Though the phenomenon can represent a physiological response to fungal or mycoplasmal infection, stable instances – capable, as in this case, of vegetative propagation – are attributable to a genetic mutation. Witches brooms of Catalpa are not unknown, but do appear to be extremely rare – an instance was reported on a German specimen of Catalpa speciosa, not attributable to fungal infection (Neger 1921); another example, also on C. speciosa is in cultivation at the US National Arboretum (Olsen 2008). The rarity of the phenomenon is consistent with the presumption that the ‘Nana’ cultivar constitutes a single clone. Left unpruned, topgrafted trees occasionally produce fertile flowers and fruits, the seedlings resulting in a 50:50 normal : dwarf phenotype ratio, consistent with dwarfness caused by a somatic mutation (Richard Olsen pers. obs.).
Masson’s dwarf was incorrectly referred to the species C. kaempferi (synonym of C. ovata G.Don), for reasons that are easy to account for. For mid-nineteenth-century Parisian gardeners the accepted name for a diminutive catalpa was C. kaempferi, the only ‘small catalpa’ hitherto described – by Kaempfer (1712), who had given an account of a catalpa ‘the size of a pomegranate bush’, a description restated almost verbatim in De Candolle’s then recent Prodromus (1845). Masson’s dwarf tree therefore became known by this name (‘C. kaempferi ou nain’ (Pépin 1856)). The obvious resemblance of this ‘C. kaempferi’ to a small C. bignonioides posed no intellectual difficulty, since De Candolle (who had never seen living C. kaempferi) treated Kaempfer’s ‘arbuscula’ precisely as a dwarf variety of C. bignonioides. (In spite of their disjunct distribution the two taxa had been regarded as conspecific for more than a century, albeit with some puzzlement; see the entry for C. bignonioides). Since true ‘C. kaempferi’ (C. ovata) was not yet in cultivation in Europe, and since Masson’s plant did not produce flowers, there was no reason for Parisian horticulturists to challenge this assumption.
More mysterious is how the ‘C. kaempferi’ name was supplanted by another misnomer, ‘C. bungei’, a name that continues to haunt it in the horticultural trade, particularly in North America. The origins of the confusion are – as Charles Sargent concluded long ago – ‘a mystery that will probably never be cleared up’ (Sargent 1911), though considering the amount of variously provenanced, juvenile catalpa material circulating in Parisian trial gardens and nurseries in this period, it is easy to imagine how confusion might arise. Certainly, the ‘bungei’ label had become attached to Masson’s dwarf form by the late 1860s – Durand’s general catalogue lists the ‘catalpa nain de Bunge’ (Durand [1869]). Bureau (1894) noted that the dwarf clone was sold under both aliases (C. kaempferi and C. bungei), but agreed with Lavallée (1885) that it was a diminutive form of C. bignonioides. By the early twentieth century the name C. bungei was in general use for the ornamental form with rounded head, usually topworked (e.g. Ellwanger & Barry 1900).
'Pulverulenta'
Common Names
Speckled-Leaf Catalpa
Synonyms / alternative names
Catalpa speciosa 'Variegata'
Catalpa speciosa var. pulverulenta (Rob.) W.Paul & G.Paul
Catalpa speciosa f. pulverulenta (Rob.) Rehder
Catalpa bignonioides var. pulverulenta (Rob.) Bean
Catalpa bignonioides f. pulverulenta (Rob.) Paclt
This cultivar with fresh green leaves up to 30 cm long, finely sprinkled with white or cream variegation’, was introduced to cultivation in the United Kingdom by B. Paul and Son in January 1910, as a sport from Catalpa speciosa (Gardeners’ Magazine 1910). However, Bean (1914) referred it to C. bignonioides, and the broad-crowned, ground-brushing habit of a tree at the Parc floral de la source, Orléans, France – 6 m tall with a spread of 11 m (Demoly & Picard 2005) – is characteristic of that species. The unusual variegation is not unpleasing, though perhaps too closely resembling an infestation of spider mite for some tastes.
'Semiplena'
A partially double form found by Plouvier (1947) in the park at Buttes-Chaumont, Paris. Although assigned by him to C. bignonioides, a hybrid origin is not unthinkable (see C. × erubescens).
'Variegata'
This variegated cultivar was exhibited in Paris in September 1852 at the exhibition of the Société nationale d’horticulture de la Seine (Société nationale d’horticulture 1852). The variegated effect is most conspicuous in the new leaves, a broad yellow margin surrounding an irregularly-shaped central patch of normal green. Offered by Bluebell Arboretum & Nursery (Bluebell Nursery 2025).
Houtman (2004) includes a variegated form of C. × erubescens that closely resembles plants offered under this name. The 1852 French clone described above predates any known instances of C. × erubescens (C. ovata × bignonioides) by more than fifteen years, so the historical existence of a variegated form of C. bignonioides is not in question. (C. ovata itself was unknown in France until the early 1850s, and even allowing for the species’ well-known precocity, the time-scale is still too tight for a hybrid to have emerged by 1852.)
Nevertheless, it cannot be excluded that some modern material circulating under C. bignonioides ‘Variegata’ is in fact referrable to C. × erubescens (and possibly also vice versa) and further investigation and comparison is needed.
