Bignoniaceae

Primary tabs

Bignoniaceae

Description

Trees, shrubs, lianas, very rarely herbaceous (extra-Mal.); The majority of the Malesian genera are small or large trees, the only climbing genera being Nyctocalos, Hieris, Tecomanthe, Pandorea, and Neosepicaea; in subalpine heathland Tecomanthe may be forced to creep on other vegetation. Most trees are of medium size, but species of Stereospermum, Fernandoa, Pajanelia, and also Radermachera gigantea may attain quite good dimensions. Oroxylum is a short-lived nomad tree.
Very poorly known for the Malesian representatives. Diversity of stomatal type and indumentum (non-glandular and glandular hairs in a variety of forms) certainly deserves detailed comprehensive studies, which will probably yield important additional taxonomic characters.
— P. BAAS.
Leaves simple or mostly compound (digitate or impari-l-4-pinnate), (in Mal.) decussate, rarely in whorls of 3-4, often provided with glands underneath, in the New World often provided with terminal tendrils, rarely scattered or in pseudo-whorls (extra-Mal.); Stipules absent. Inflorescences bracteate, cymose, but not rarely thyrses contracted to racemiform or racemose inflorescences, or even reduced to solitary flowers (extra-Mal.), terminal, axillary or from the old wood. Flowers usually very showy, rather large, bisexual, articulate with the pedicel or not. Stamens 5 almost equal, or mostly 4 didynamous, the 5th sterile, rudimentary, adnate to the corolla tube, mostly inserted at the rim of the basal tube and not rarely (glandular) hairy at the insertion, more rarely inserted higher up. Ovary superior, 2-celled, rarely 1- or 4-celled (extra-Mal.); Capsule 2-valved, either loculicid with the septum perpendicular to the valves — sometimes provided with an additional transverse false septum — or septicid with the septum parallel with the valves, or (extra-Mal.) an indehiscent, 1-celled, soft or hard-shelled, pulpy berry. Ovules (in Mal.) in each cell on the septum in two or more rows of 3-, mostly on 2 placentas. Seeds in each cell attached to the dissepiment in one or more rows, inserted transverse to axis of fruit, anatropous, mostly on both sides with hyaline wings;

Distribution

Africa present present present present, America present present, Asia-Tropical: Maluku (Maluku present); Sumatera (Sumatera present), Australasia, Chile present, Continental SE. Asia present, E. Asia present present, E. Australia present, Indo-Australia present, Madagascar present, Melanesia present, N. America present, New World present, Old World present, Penang present, Southern America, tropical Asia present, tropics and subtropics, roughly between 40° N and 30-35° S present, warm-temperate zone present, worldwide present
About 120 genera and some 650 spp., mainly in the tropics and subtropics, roughly between 40° N and 30-35° S, very few in the warm-temperate zone; in Malesia: 14 native genera of which 2 are endemic, viz Hieris in Penang and Lamiodendron in Papuasia Among the remaining 12 one occurs through the Old World (Dolichandrone), 7 are shared with continental SE. Asia (two of which extend also to Africa and Madagascar: Femandoa, Stereo-spermum) and 4 with Australia and Melanesia; the latter occur in Malesia only in the east except Deplanchea which ranges westward to Sumatra.
In the family tropical Asia and Africa share a few genera (Markhamia, Femandoa, Stereospermum, and Dolichandrone), but Africa and America share only one, viz Tecoma. This latter affinity goes further, though very disjunct via Campsidium (Chile) and Campsis (N. America and E. Asia) to Tecomanthe-Pandorea-Neosepicaea (Moluccas to Three Kings Is. and E. Australia). Otherwise there appear to be only two other transoceanic ties, viz tribe Crescentieae which is shared by Africa and the Americas, and the genus Catalpa which occurs in E. Asia and the Caribbean area.
As GENTRY () has shown, the average number of species per genus is only 5, which is very small in comparison with many other families, but can only partly be explained by a possibly small generic concept. There are quite a number of monotypic genera (in Malesia 5), but they are well defined in many characters and stand very apart.
This, and the worldwide distribution of the family, and the disjunctions in ranges, definitely point to relict survival and ancient origin, onwards of which period the three tribes have undergone a separate, independent development on the continents, mainly leading to differentiation in Indo-Australia and in the New World, with the greatest abundance in the latter. Unfortunately, the fossil evidence (only Tertiary) is meagre and untrustworthy (), both to macrofossils and to pollen.

Ecology


Pollination

Flower-shape, -colour, -position, and -scent are very different in the mostly showy flowers of Bignoniaceae, and the syndromes attract different visitors.

Bats are frequently visiting species of certain genera, another phenomenon restricted to tropical plants. According to FAEGRI & VAN DER PIJL () the attraction syndrome is: nocturnal anthesis, whitish, creamish or drab greenish or dark purple colour, stale or sour, unpleasant smell reminiscent of fermentation at night, large quantity of nectar and pollen in large anthers, large-mouthed and coarse flowers on strong stalks sticking out of the foliage or cauliflorous to flageHiflorous flowers, thus coming into easy reach for landing. This is found in Malesia in several cultivated genera (Kigelia, Crescentia, Parmentiera, Mark-hamia, etc.) but occurs also in the native Fernandoa adenophylla, Pajanelia, and Oroxylum. . Notwithstanding the many papers and records of observation — corollas show claw marks after these visits — it is not proved to my satisfaction that visits of bats are compulsory for pollination cq. fertilisation, experimenting being in this field deplorably meagre. My doubt is strengthened by observations by HARRIS & BAKER in Ghana where Kigelia is native () and can set fruit in absence of bats; they observed also frequent visits by sphingids but they doubt effective pollination by these.

Birds, humming birds and sun-birds, frequently visit certain species, the attraction syndrome being: tubed, vividly coloured (orange, scarlet), diurnal, mostly odorless, nectar-producing tubular flowers. Here also many observations are made, e.g. in Tecoma (Tecomaria). To this class belong in Malesia some species of Radermachera (R. ramiflora), Neosepicaea, Tecomanthe, and it can be expected for Deplanchea. Also the cultivated Spathodea campanulata is frequented by birds (cf. ), notably kutilans and ?djalaks, at Bogor; they severely damage the corolla. Here again the question whether bird-visits are compulsory for pollination cq. fertilisation is inadequately supported by experiments. Caution is necessary to conclude to the necessity of cross-pollination, as e.g. HUNTER () recorded that in Tecomanthe speciosa, of which cuttings of a single plant led to its cultivation, fertilisation — that is selfing — could be effected by hand-pollination, but later also naturally by bees, although far from its native habitat.

Dispersal

Bignoniaceae occur throughout the tropics and several are still found in the sub-tropics of the whole world. One might ascribe this to their having winged seed (except Crescentieae and a few other exceptions), but against expectations they are almost absent from oceanic islands. Bignoniaceae occur all along the coasts of the West Pacific, notably in New Guinea and in Australia species of Tecomanthe and Pandorea are not rare, but the only occurrence in the West Pacific islands is a common Australian Pandorea in New Caledonia and Lord Howe I., and a peculiar Tecomanthe in a single locality of the Three Kings Is., the northernmost territory of New Zealand.

Obviously wind dispersal has not been as effective as one would expect.

Dispersal by seawater is common in Dolichandrone spathacea, a back-mangrove species, ranging from the western Deccan Peninsula to North Luzon, south to Timor and southeastwards to New Caledonia; the range is almost continuous without gaps. Fig. 14. It is most peculiar, however, that so far it has never been found in the mangroves of northern Australia. Its seeds have thickish corky wings instead of flimsy wings as usual in most members of the family (except the fleshy indehiscent fruits of the Crescentieae and a few other exceptions as e.g. Pauldopia) and are most excellently adapted to be dispersed by seawater.

Morphology

Juvenile plants of Pandorea pandorana show leaves very different from the mature foliage, in having many jugae and being coarsely dentate. Tecoma filicifolia NICHOLS, was based on such material. This led also to a serious misinterpretation of Tecoma leptophylla Bl., from New Guinea, of which the juvenile leaves () are Pandorea pandorana but the flowers belong to Neosepicaea.

Taxonomy

Since the basic work on the systematy by E. BUREAU (), the treatment of the family in Flora Brasiliensis by BUREAU & SCHUMANN (1896-97), and the treatment of SCHUMANN in the Pflanzenfamilien (1895) the traditional subdivision of the family in 5 tribes has proved satisfactory. Crescentieae with 1-celled berries occur in Africa and the Americas, two other monogeneric tribes are South American, while the bulk of the family belongs to Big-nonieae and Tecomeae, of which the latter are about balanced as to number of genera in the Old and New World, but Bignonieae are predominantly American. These two tribes are largely distinguished on the dehiscence of the fruit, loculicid in Bignonieae and septicid in Tecomeae.

In passing it may be remarked that GENTRY () recently advocated that Crescentieae of the neo-tropics and of Africa-Madagascar are of separate descent and would represent two parallel evolutionary lineages; this suggestion is more based on geographic argument and evolutionary hypotheses than on morphological arguments.

The delimitation against other families of Sympetalae is well-defined, but there are a few genera, notably Wightia and Paulownia, which are sometimes referred to Bignoniaceae, though FENZL (), BUREAU, SCHUMANN, VON WETT-STEIN, and other specialists referred them to Scrophulariaceae. A survey of opinions I gave in my paper on Wightia (), in which I excluded it from Bignoniaceae. Even recently Paulownia is sometimes casually treated as Bignoniaceous (e.g. ), although the embryo is embedded in endosperm; furthermore the stigma is different from that in Bignoniaceae, the anthers have no prolonged connective, there is no rudimentary stamen and the seeds are provided with several wings and seem to be laterally attached, not transverse as in Bignoniaceae. For Wightia I tabulated (l.c.) the relation to both families. Its seeds have no endosperm, but the absence of a staminode, the structure of the stigma, the central placenta and the absence of a produced connective on the anthers point distinctly to Scrophulariaceae. The seed is quite differently attached as compared with Bignoniaceae, viz laterally and the wing surrounds the entire seed. Its wood has two kinds of medullary rays, narrow and broad ones, a character which, at least in Malesian Bignoniaceae, is absent.

Though the capsule in Wightia is septicid and in Paulownia loculicid, both genera have the same kind of axile placentation, in which the thickened placenta becomes detached from the valves as a subquadrangular seed-cake, showing their close affinity, completely differing from the situation in Bignoniaceae.

According to SURYAKANTA () the pollen of both genera differs from that in Bignoniaceae and resembles that of Scrophulariaceae.

NAKAI () accommodated Paulownia in Paulowniaceae, probably induced mostly by its arboreous habit and fruit; they certainly merit to be placed in a separate tribe or subtribe of Scrophulariaceae. We regard nowadays the arboreous habit as primitive in herbaceous families and we might conclude that they are ancient relicts from a period when Bignoniaceae and Scrophulariaceae had a common matrix.

Also in South America there are two woody genera of the Scrophulariaceae which were at times referred to Bignoniaceae, viz Schlegelia (syn. Dermatocalyx) and Gibsoniothamnus, according to GENTRY (); see also LEINFELLNER (). They are (hemi-?) epiphytic shrubs or lianas, a similar habit as in Wightia.

Cytology

Chromosomes. DARLINGTON & WYLIE () and MOORE (ed.) () gave for 26 genera x = 20 (2n = 40) and they belong to Tecomeae, Bignonieae and Crescentieae, both from the palaeo- or neotropics. There is one higher number x = 22 (Amphilophium, South America, Niedzwedzkia = Incarvillea) and several lower ones: Pandorea, and some doubtful countings in Tecoma x = 19, Tecomanthe dendrophila 2n = 36 (Christine BRIGHTON in litt.), Jacaranda x = 18, Tecoma capensis x = 17, Oroxylum, Milling-tonia, Argylia (from South America) x = 15, Spathodea x = 13, and Incarvillea x = 11. In supplement indices Campsis is also given as 16 and Oroxylum as 14.

I have scanned the numbers of Scrophulariaceae, Gesneriaceae and Verbenaceae, but can find no reliable ties, Bignoniaceae being obviously more homogeneous than those.

The number given for Paulownia, 2n = 40, x = 10, might as well fit Bignoniaceae as Scrophulariaceae.

Hybridisation. Not many species hybrids are known to me, but those known are interesting, as there are at least two between species of East Asia and SE. North America which are now very disjunct after the Pleistocene Ice Age; it is not impossible that they formed part of more continuous populations in the warmer Pliocene via Beringia. This idea is supported by the fact that in both cases the hybrids are fertile.

E. C. SMITH () reported on Catalpa ovata DON × C. bignonioides WALT. (= × C. syringifolia SIMS). Haploid all have 20 chromosomes ().

Then there is × Campsis tagliabuana (VIVIANI) REHDER, a hybrid between the Chinese C. grandiflora (THUNB.) K. SCH. (C. chinensis (LAMK) VOSS.) and C. radicans (L.) SEEM, which produces fertile progeny (cf. ).

The third one is also bi-continental, Tecoma smithii W. WATSON (; cf. also ). This is a reputed hybrid, which E. SMITH made at Adelaide, in 1882, between T. velutina (a hairy variety of T. stans) and T. capensis. It was propagated by cuttings, but it produced seed and its offspring of seedlings diverged in size and flower colour. Curiously SPRAGUE, in a succinct note () reduced it to T. alata DC, without referring to its hybrid nature.

Uses

There are no outstanding qualities marking Bignoniaceae as useful plants, otherwise than ornamentals and these concern mostly the introduced species for which I refer to the special key and account at the end. There are magnificent native species notably of Tecomanthe but they have as yet not become in general use.

Good roadside trees are Millingtonia hortensis and Spathodea campanulata. A highly esteemed vegetable (lalab) with the Sundanese is Oroxylum indicum (flowers, buds, and very young pods).

For re-afforestation and holding terraces on slopes the pioneer qualities of species of Rader-machera and Deplanchea might be useful.

The timber is in general not valuable and in nature not available in sufficient quantity. The soft wood of Millingtonia hortensis was advertized as useful for tea-boxes. The only species yielding sizeable timber of good quality are: Fernandoa macroloba, Pajanelia longifolia, Radermachera gigantea, and the three species of Stereospermum, which all may be valuable for silviculture.

Notes

Since my thesis (), here always cited as 'Thesis (1927)', and subsequent revision in , I have remained always much interested in this family and have published some revisions and many notes precursory to the present treatment. I have to thank the late Mr. N. Y. SANDWITH (Kew) for namings of cultivated species, and Dr. A. L. GENTRY (St. Louis) for recent information on them, Dr. H. HEINE (Paris) for assistance in various matters, Mr. Michael GALORE (Lae) and Prof. E. J. H. CORNER (Cambridge) for photographs, Miss Christine BRIGHTON (Jodrell Lab., Kew) for the first chromosome count in Tecomanthe, while I gratefully acknowledge precursory work performed by Mr. J. C. DEN HARTOG on Tecomanthe and Pandorea in 1969/70 at the Rijksherbarium where he worked as a graduate student.

Phytochemo

Bignoniaceae share a number of biochemical tendencies with Verbenaceae, Labiatae, Scrophulariaceae and with several other families of WETTSTEIN'S Tubi-florae. Most of their outstanding chemical characters were already mentioned and discussed in my , to which the reader is referred. Much phytochemical information, however, became available only in more recent time. Recent results confirm the trends already apparent in 1963; they are summarized in the following pages. Chemical characters of Bignoniaceae may ultimately prove to be very useful in tracing inter- and intrafamiliar relationships.
  1. Most members seem to produce and accumulate iridoid glucosides (formerly often called pseudoindicans). Since a long time Bignoniaceae are known to contain labile glycosidic bitter principles. Such a compound was isolated in 1888 from the bark and fruits of Catalpa bignonioides WALTER and called catalpin (name changed later to catalposide). The structure of catalposide was definitely established in 1962; it is an aucubin-type (C9-aglucone) ester glucoside and one of the first pseudoindicans for which clearcut structural and biogenetic relationships with iridodial and nepentalactone were demonstrated (hence the name iridoid glucosides for a presently very large group of constituents of dicotyledonous plants). Catalposide (tastes bitter) is an ester of p-hydroxybenzoic acid with catalpol (= 7,8-epoxy-aucubin). Catalpol and catalposide occur in all species of Catalpa (leaves, stems, fruits) and catalpol (= catalpinoside) was also isolated from barks of Paulownia tomentosa STEUD. and P. fargesii FRANCH. where it occurs together with syringin (). Probably catalpol and catalposide occur in many more members of the family. In most recent times some related glucosides were isolated from Bignoniaceae. Vanilloyl-catalpol (= amphicoside) is a constituent of Amphicome emodi LINDL. and veratroyl-catalpol occurs in Tecomella undulata SEEM. 5-Hydroxycatalpol (= macfadyenoside) was isolated from Macfadyena cynanchoides MORONG. All iridoid glucosides mentioned hitherto have structures based on the aucubin-derivative catalpol. The first non-aucubin-type glucoside described from Bignoniaceae is tecomoside with a C10-aglucone; it was isolated from Tecoma capensis LINDL. (). It is to be expected that much more iridoid glucosides will be detected in the family in future.
  2. Some Bignoniaceae produce alkaloids. So far only pyridine-type and piperidine-type alkaloids with an iridoid C10 or rarely C9 were identified definitely in species belonging to this family. This fact strengthens the belief that the tendency to produce iridoid compounds is a very important character of Bignoniaceae. Thusfar simple iridoid alkaloids were described for species of Campsis (boschniakine), Incarvillea (plantagonine, indicain), Tecoma (tecomanine, tecostidine, tecostanine, boschniakine, 4-noractinidine and several derivatives of skytanthine). The basic constituents of Amphicome (now reduced to Incarvillea), Newbouldia and other genera may belong to the same group of alkaloids.
    A recent review of the chemistry, distribution and systematic meaning of all presently known main groups of iridoid plant constitutents was published by S. ROSENDAL JENSEN et al. ().
  3. Many Bignoniaceae synthesize naphthaquinones and corresponding anthraquinones by prenylation of o-succinylbenzoic acid. This pathway to quinonoid naphthalene- and anthracene-type secondary metabolites is presently known from taxa belonging to Rubiaceae, Verbenaceae, Scrophulariaceae, Bignoniaceae and possibly Acanthaceae and Gesneriaceae. In roots, woods and barks of Bignoniaceae lapachol, lapachonone, α- and β-lapachone and dehydro-a-lapachone occur frequently. These monomeric naphthaquinonoid compounds are often accompanied and sometimes replaced by more complex dimeric constituents like tectol, guayacanine and guayine and by corresponding anthraquinones such as tectoquinone and 2-methy 1-3-hydroxyanthra-quinone. Woods which contain appreciable amounts of these quinonoid compounds are more or less resistant to marine borers, white ants and Fungi. At the same time such woods may be the causes of skin irritations and of allergic skin diseases in man. Lapachol- and tectoquine-type substances are presently known from species of the genera Catalpa, Heterophragma, Kigelia, Paratecoma, Phyllarthron, Stereospermum, Tabebuia, Tecoma, Tecomella, and Zeyhera. R. H. THOMSON has reviewed the chemistry and distribution of quinones and related compounds in his book 'Naturally occurring quinones' (2nd ed. 1971). The phthalide catalpalactone from the wood of Catalpa bignonioides WALTER and C. ovata G. DON arises from the same pathway as lapachol and its congeners (). On the other hand it should be stressed that the red-coloured naphthaquinones of Boraginaceae (e.g. alkannin) which are structurally very similar to lapachol are produced along a totally different biosynthetic pathway (cf. ).
  4. The "tannins" mentioned for many Bignoniaceae in the older phytochemical literature (e.g. DEKKER, 1913) seem to be glycosides and esters of o-diphenolic compounds. Orobanchin (= verbascoside)-type ester glycosides were definitely demonstrated to occur in species of Campsis, Catalpa, Eccremocarpus and Pandorea. A review of this type of polyphenolic plant constituents which simulate true tannins in some respects is to be found in my . Orobanchin yields a molecule of caffeic acid, 3, 4-dihydroxy-phenylethanol, glucose and rhamnose each. Just as in most other families of Sympetalae true tannins are replaced in Bignoniaceae by more or less complex esters and glycosides of odiphenolic cinnamic acid derivatives. Moreover, simple esters of caffeic acid and biosynthetically related derivatives of cinnamic and benzoic acids are present in large amounts in many Bignoniaceae. The recent investigations of V. B. PANDEY and B. DASGUPTA with the bark of Tecomella undulata SEEM, (veratroylglucose = tecomin: ) and of M. SUGUMARAN et al. with leaves of Tecoma stans H.B.K. (16 aromatic acids: ) exemplify this trend. P-Hydroxybenzoic acid is present as an ester in all species producing catalposide; probably this phenolic acid is rather ubiquitous in the family. The presence of appreciable amounts of hydroquinone (in living cells as the glucoside arbutin?) in leaves of Jacaranda mimosaefolia D.DON () might be connected with a strong tendency to produce and accumulate p-hydroxybenzoic acid; if this is actually the case hydroquinone (and arbutin?) may be detected in much more Bignoniaceae in future. Jacaranone, a quinonoid compound which exhibits antitumor and cytotoxic activity was recently isolated from leaves and twigs of Jacaranda caucana PITTIER (); it seems to be derived from tyrosine and is chemically very similar to the Cornus quinol glucoside (= cornoside) which is also present in leaves of Digitalis purpurea ().
  5. According to J. B. HARBORNE () leaf flavonoid patterns of Bignoniaceae are close to those of Acanthaceae, Gesneriaceae, Labiatae and Scrophulariaceae. Features which support such a statement are the replacement of flavonols by flavones in many species, the relatively frequent occurrence of flavones with an unsubstituted B-ring (e.g. chryson, baicalein), of 6-hydroxylation of chrysin (baicalein), apigenin (scutellarein) and luteolin (6-hydroxyluteolin) and of O-methylation of flavones. The latter trend is illustrated by Zeyhera tuberculosa BUR. ex VERLOT which contains 5,6,7-trimethoxyflavone and 5,6,7,8-tetramethoxy-flavone in leaves (). The bitter principle of the fruits of Sparattosperma vernicosum BUR. & K. SCH. was shown by J. P. KUTNEY et al. () to be the 7-neohesperidoside of pinocembrin (= 2,3-dihydrochryson).
  6. Free triterpenic acids occur in appreciable amounts in leaf waxes of many families of Tubiflorae (especially Verbenaceae, Labiatae and Plantaginaceae) and related orders. It is of interest in this respect that ursolic acid was isolated in recent time from leaves of Bignonia diversifolia H.B.K., Campsis radicans SEEM., Catalpa bignonioides WALTER, Heterophragma quadriloculare K. SCH., Jacaranda mimosaefolia D.DON (not definitely identified) and Paulownia tomentosa STEUD. The bark of Jacaranda mimosaefolia yielded lupenon.
  7. Many members of Verbenaceae, Labiatae, Scrophulariaceae and Plantaginaceae replaced starch by stachyose-type oligosaccharides as storage carbohydrates. The same trend seems to exist in Bignoniaceae. Large amounts of stachyose occur in species of Catalpa (roots, wood, bark), Newbouldia laevis SEEM, (roots) and Paulownia tomentosa (stem).
  8. Most representatives of Tubiflorae produce starch-free seeds which are rich in proteins and oils. The seed oils are often characterized by a high degree of unsaturation. In this respect Bignoniaceae conform to the rule. Their seeds generally contain 20-35% of oil. In some taxa oleic and (or) linolic and (or) linolenic acid are the only major fatty acids of the seed oils (e.g. species of Crescentia, Niedzwedzkia = Incarvillea, Paulownia and Stereospermum). In other taxa the 'normal' fatty acids are accompanied or replaced by large amounts of unusual fatty acids such as conjugated trienoic acids (species of Catalpa, Chilopsis, Jacaranda), C26-keto-acids (Cuspidaria pterocarpa DC), octadeca-trans-3,cis-9,cis-12,cis-15-tetraenoic acid (Tecoma stans H.B.K.) or hexadec-9-enoic and octadec-11-enoic acid (Doxantha unguis-cati MIERS). M. J. CHISHOLM and C. Y. HOPKINS discussed the chemistry of seed oils of 11 species representing 4 tribes ().


The preceding phytochemical picture places Bignoniaceae phytochemically very close to a number of families of Tubiflorae, especially Verbenaceae, Labiatae and Scrophulariaceae. Still other constituents are known from Bignoniaceae. Lack of acquaintance with their structures and (or) with their distribution, however, does not yet allow a systematic evaluation. Saponins, which seem to be rather widespread in the family but were never investigated in detail, belong to these chemical characters. The same holds for a number of phenolic compounds isolated in recent time, such as the lignans sesamin and paulownin from Paulownia tomentosa STEUD. and Phyl-larthron comorense DC, the dilignol (a lignan-type compound) zeyherol from Zeyhera digitalis HOEHNE and the dihydroisocoumarins 6-methoxymellein, kigelin and 6-demethylkigelin from Kigelia pinnata DC.

Concluding it may be stated that the intimate relationships between Bignoniaceae and Scrophulariaceae which are indicated by genera like Catalpa and Paulownia (often placed in Scrophulariaceae) are confirmed by phytochemistry. At the same time phytochemistry stresses a very close coherence of a core of families of Tubiflorae', this core comprises Scrophulariales sensu CRONQUIST (1968) and Lamiales sensu TAKHTAJAN (1969). — R. HEGNAUER.