Butterflies in Soy Sauce

Table of Contents and IntroductionAfrican Violets: the Doyen of House PlantsAgapanthusAlstroemeriasAmerican SagebrushesAmerican WildflowersAndean Nymph - a pink mimulusAquilegias (Columbines)AuriculasAustralian AnnualsBellis Perennis (Daisies)Black-Eyed SusanBlue African LiliesBreeding DelphiniumsBreeding Fuchsias: Part 1: From SeedBreeding Fuchsias: Part 2: HybridisationBromeliads from SeedButterflies in Soy SauceCacti RaisingCampanulasChilean Bell FlowerChilly Treatment for Peruvian LiliesCollecting SeedConserving Annuals and BiennialsCosmosCucumbersCycads: Living FossilsCyclamen for IndoorsDahlias from SeedDamping OffErythroniumsF1 Hybrid - What Is it?F2 and Open Pollinated VarietiesFrom Obscurity to the Height of FashionGentiansGeraniumsGermination to Pricking OutGolden Impatiens - The StoryGrowing Alpines From Seed: Part 1Growing Alpines From Seed: Part 2Growing Bananas From SeedGrowing Cyclamen From SeedGrowing Eucalyptus From SeedGrowing Ferns from SporesGrowing Foxgloves From SeedGrowing Fritillarias From SeedGrowing From Seed: From Packet to SoilGrowing Hardy Ferns from SporesGrowing Hardy Geraniums From SeedGrowing Lewisias From SeedGrowing Magnolias From SeedGrowing Meconopsis From SeedGrowing Proteas and Related Genera from SeedHandling Small SeedsHardy Annuals: Part 1Hardy Annuals: Part 2Hardy CyclamenHardy SpurgesHelleboresIncreasing your Violas and ViolettasKeeping Records of Seed SowingLettuce - the Vital Part of any SaladLord Anson's PeaMagnolias - the Root of the ProblemMimulus from Seed to Seed in Sixty DaysNew Guinea ImpatiensOld-fashioned Sweet PeasOnions with EverythingPansies are not Difficult to Grow from SeedParsleyPerennials for American GardensPetuniasPhygelius - their cultivation and propagationPinks: Evocation of the PastPlants That Flower OncePrickly but NicePropagatorsRadishesRaising PalmsRaising PrimulasSalad Leaves - a lot to be desiredShow Stoppers Twice OverSowing SeedsSowing in Open GroundSowing to Feed our FriendsSturt's Desert PeaSweet Peas - Autumn SownSweet Peas - Sowing and GerminatingTall and HandsomeTechniques with Tiny SeedTexas BluebellThe Effect of Light on Germination and SeedlingsThe Nervous PeaThe Scarlet Welsh PoppyThe Search for the Yellow Sweet PeaToad LiliesTomatoesTreasure Hunt for TasteUnderstanding Latin NamesUnique Poppy (Mother of Pearl)Viability of VegetablesWeaning SeedlingsWill the Real Mimosa Please Stand Up

John Watson looks in detail at the genus Schizanthus and relates his experience of finding it in the wild.

Taking a cue from nature. What a highflown ideal to start off a gardening project. Yet how many painstaking, hopeful endeavours to realise and recreate too literally the full glorious impact made by some spectacle of wild flowering have been doomed to abject failure and ended up on the scrap heap of disillusion.

Take Schizanthus

To wander through a violety-pink billowing sea of thigh-high S. litoralis along Fray Jorge valley near Coquimbo in the Chilean mist belt spring is to sample earthy paradise knowing it must be lost. Reel off a portfolio of artfully composed photographs. Stuff handfuls of seed into bulging brown paper bags at a later visit for mass sowing at home. Scribble down frantic attempts to record the living sensation on paper whilst it still has a bloom of freshness in the memory. None of these will pin down more than a faint shadow of the transient magic of that moment. So what hope do mortals have of getting the most out of schizanthus in cultivation? What do we actually do? Pretty though a solitary specimen and its individual flowers undeniably are, their talent finds its fullest expression when they perform as a choir rather than as soloists. Sensing this, we have their measure, tending to grow them by the greenhouseful, or packed in on terraced ranks of nursery and local authority displays. What we miss in boundless perspective and the haunting beauty of Chile's unpopulated and undefiled countryside, we make up for in skillfully shaped bushiness and a claustrophobic riot of carefully bred and selected colours. You might search all day for just one such variation in the wild and be lucky to bag it. Which really goes to show that if we cannot beat them at their own game, we may equal, if not beat them at another of our own making. At one level, that is what plantsmanship is all about. Why, we even learn to love many of our plants in no sort of context whatever, simply for their own sweet sakes in pots and pans. And I hope to leave you feeling you could certainly do with schizanthus as well, if you so wished.

The genus Schizanthus belongs to the Solanaceae, we learn, but is a very far cry from the regular, usually rounded flowers we associate with that family: Solarium, Fabiana, Datura, Petunia and so on. In fact Schizanthus is very close botanically to Scrophulariaceae and has regularly in the past been included in that family, as you may notice in older reference books.

One of two popular names, 'poor man's orchid', obviously alludes to the cheapness and ease of growing such colourful and spectacular flowers. It can surely be no more than sheer coincidence that all schizanthus and many orchids share the trait of having their flowers twisted through 180 degrees so that they are in effect upside down (resupinate). Ironically enough, orchids have used this means to bring down a conspicuous upper segment and develop it into their almost universal trademark, that showy lower lip. Schizanthus, on the other hand, have done exactly the reverse to good effect. In fact their lower lip, formed by the two technically 'upper' petals is a fairly or very inconspicuous affair of two to four narrow concolorous lobes. The remainder of the floral outline is taken up by the familiar lateral petals, just two, but giving the impression of more as each is divided into a minimum of two, and on occasions more than eight divisions. From this their characteristic frilly, flighty appeal is drawn.

'Butterfly flower', the other popular name for them, has also been thought up by Chileans for some species (in their case maripositas, 'little butterflies'; it is a national pastime to create diminutives.') Whether the English name is a straight copy, I do not know.

Our good friend and colleague, the late Professor Carlos Muiioz, used to quote a figure of 27 species, all Chilean. This would have to include more than three-quarters of the names ever published. More conservative assessments stood at 15 or so, and the latest monograph, which I follow here, cites 12: (synonyms in parentheses), S. alpestris (3 syns.), S. candidus (S. aibi-florus), S. grahamii (5 syns.), S. hookeri (S. calycosus), S. integrifolius, S. lacteus (S. sanromanii), S. laetus (S. fallax), S, litoralis (S. splendens), S. parvulus, S. pinnatus (9 syns.), S. porrigens (S. florihundus) and S. tricolor (S. pinnatus). Interestingly, this dozen is mirrored exactly by the bulbous genus of Leucocoryne, also of 12 species, all of them Chilean endemics. Unfortunately such neat, self-contained symmetry is slightly dented on closer examination as S. grahamii has leaked over into Argentina, occurring at several points along the central province of Mendoza: and gorgeous it is there too. That Chile, such a ribbon of a country over 2800 miles (4500 km) long, yet seldom exceeding more than 200 miles (320km) in width, (only Norway, at under half the length, has a comparable geographical layout) should manage to develop almost exclusively within itself these wide ranging genera is still quite remarkable. They are products of a unique isolation, sealed in on all sides by sea, mountain, ice and desert. Schizanthus itself occupies some 1200 miles (2000km) of that total, from Tocopilla in the north almost to Valdivia in the south. The centre of distribution lies along an axis between Valparaiso and Santiago, where over half the genus may be encountered. Most are annual coastal species with a tendency to venture shortly inland. S. alpestris prefers hot foothills. S. hookeri and S. grahamii have made it to the mountains proper where they usually survive a bit longer as biennials or even short-lived perennials. Four basic groups have been recognised, each of two or three closely related species. There are three left with no clear ties.

I bring all this detail to your attention mainly because the genus would seem to be a promising subject for collectors and also because I suspect its commercial potential is far from fully exploited. 12 species, each with its own charm and personality, all of them attractive in their way, and in reasonable and accessible geographical proximity, make it a handy and attainable goal for any interested plantsman. Nor need it stop there with sufficient variation in several species and positive indications for hybridisation to keep a specialist going for many a year, if not a lifetime. It seems that all known schizanthus in cultivation will perpetuate from seed. Whether the already brilliant range of colouration and markings in commerce might be extended still further is open to question, but there are undoubtedly other aspects that could be developed, which should become apparent as I evaluate them species by species. Heights in centimetres given after species are normal maxima.

The pinnatus group

S. pinnatus (60cm) and S. tricolor (60cm). Coastal and lowland annuals from Valparaiso southwards. It seems appropriate to kick off with this group, S. pinnatus rightly or wrongly being the best-known name in cultivation. In so doing I must also kick off with a confession of my own. Of the four commonest species of schizanthus, we have never met up with S. pinnatus in the field. Nor, despite my having sauntered past displays grown under cultivar names such as 'Pansy Flowered' or 'Dwarf Bouquet', did anything more register than my fleeting admiration. My credentials are based on having observed five of the twelve species in the wild, as few others have. There I dare to speak with some authority. Elsewhere I may be forced to speculate, so please bear that in mind. Nearly all writers, however illustrious, are prone to dreadful howlers when they come to describe and draw conclusions about plants they have never seen. Why should I be any exception? Despite that I am prepared to stick my neck out and suggest that our plants in cultivation may owe part of their blood at least, through hybridisation to S. litoralis, if not the lot, rather than wholly to S. pinnatus. On what basis? Simply because they do not happen to look anything like wild S. pinnatus. Now someone is bound to turn round and point out gently that, say, florists' giant strains of Cyclamen persicum hardly match up with the wild thing either. Fair enough, but nor do they look in the least like any other cyclamen on God's earth, whereas our schizanthus do happen to bear more than a passing resemblance to S. litoralis, as will be seen when we get to it. Meanwhile, in support of the negative argument in my case, I can claim to have seen good illustrations of S. pinnatus and photographs of the cultivars. I can also draw on one witness far more expert than myself in these matters, my good friend and fellow collector in Chile, Ken Beckett, who certainly is familiar with cultivated schizanthus. He has voiced the same doubts and conclusions. This gives me more confidence to face many readers who will undoubtedly have the advantage over me here. There is every excuse for muddle. You have already seen how widely botanists' views can differ. Also, until recently, S. pinnatus was assumed to extend northwards, overlapping much of the range occupied by S. litoralis. To top it all they are probably the two most numerous and variable of all schizanthus. Of course I could be mistaken. These opinions are a gut reaction and need the back-up of a proper investigation. My reasons for bringing the issue into the open now are two-fold. First of all I believe it would make sense to drop S. pinnatus as a name for the cultivars we grow, at least until their status has been clarified. Many commercial growers do this already, and horticultural authors would do well to bear it in mind. Incidentally, names such as 'S. grandiflorus' and (S. hyhridus grandiflorus' have no validity and are not even botanical synonyms. Secondly, it could be useful in beginning to sort out the present situation for the benefit of any future breeding programmes.

So what is wild S. pinnatus like? (See below for a comparison made with the litoralis group). As far as I can gather, the following superficial features hold good. Colouration, although somewhat variable, tends to be the warmer side of light pink. The whole flower has a slenderer, more elongated aspect, much of this deriving from the side petals which are generally deeper cleft, narrower and more widely spread, so as to bring in mind a Jerusalem cross on occasions. Where obvious, the top lip marking of deep purple-violet is never as solid and distinctive on average as that of S. litoralis. Less well marked, squinnier examples of the two species are harder to separate, requiring careful observation of botanical differences. But that need not concern us here.

S. tricolor looks quite similar in outline and character except it is said to have a basically white upper lip with no yellow marking and a ground colouration of rose pink with intense pink spot markings.

The litoralis group

S. litoralis (80cm) and S. porrigens (80cm), basically coastal annuals from Valparaiso northwards. These, too, are variable as to shape, colour and markings, but in general their flowers give the impression of greater solidity and look more horizontally disposed than the previous group. In particular the uppermost lobes of the lateral petals are notably substantial by comparison.

S. porrigens has the side petals deeply divided, but once only, and they are blunt-ended, so that it has little of the 'frilly' nature about it. It is soft pinkish-violet with deeper purple markings and some white, though the markings are not especially extensive, being more by way of a stippled line. Distribution is quite limited, confined to the Valparaiso and Santiago areas.

S. litoralis, by contrast, may in its finest forms exhibit the most dramatic markings of any. The ground colour is usually a cool pastel violet or violet-pink, though it may be somewhat more rosy. The upper lip bears a central yellow marking streaked and dotted darkly to some extent with the branched veining common to several species, and may be surrounded by white. Beyond this again is a greater or lesser outer marking of striking smoky deep colouration, sometimes a near-blackish violet. Where continuous, the dark patch extends across most of the large top lobes of the side petals, joining up with a broad strip covering a substantial mid-portion of the upper lip to complete a three or even four-banded colour effect. Those familiar with the cultivars we grow may recognise how closely this description begins to fit some of the deeper colour forms, and may now begin to appreciate our reason for preferring S. litoralis to S. pinnatus as the sole or major influence in their creation. In less conspicuous forms the dark marking may be reduced to three or four heavy blobs or dots, if not almost absent, all more akin to S. pinnatus. Very confusing.

In the field, another aspect of S. litoralis used to intrigue me. We were impressed by certain broad similarities between some of its forms and one or two of the species and varieties in the Alstroemeria pulchra grouping. Nothing that could stand up to close scrutiny: more resemblance than mimicry. Enough, though, momentarily to fool us at a distance on occasions. For these unrelated plants flower at the same time and often appear juxtaposed in the same habitats. The comparisons exist in the basic colour of the flower and its structural balance, for both are strongly developed in the upper half only, where they also exhibit remarkably similar patterning, colouration and positioning of their guide markings. Now 'top heavy' flowers are comparatively rare anywhere in the world, let alone two side by side and sharing so many characteristics. In all other respects they remain as unidentical as chalk and cheese. Reading recently about this phenomenon of floral guilds brought all this back to mind.

There are various reasons why different organisms come to look wholly or partly alike. Sometimes it is because their functions are comparable, as in swallows and swifts. Vulnerable and harmless creatures which take on the fearsome appearance or bright warning signals of others more frightening or deadly than themselves greatly increase their chances of survival. Floral guilds are a third example. To illustrate it simplistically, suppose the pollinators in a given area show an overwhelming tendency to favour yellow. This is likely to affect not only yellow flowered species, but also any yellow forms and varieties, random yellow mutants, even the degree of yellow in variegated species. Over a period of time yellowness in the ecosystem ought to increase noticeably at the expense of other colours. Such simplification would confer clear benefits. The message: Go yellow! Be a winner! This can all too easily result in one species sweeping the board, which does not concern us here. Besides, in most systems there are usually complex factors favouring diversity, so this sort of alliance does not happen all that often. Colour is not the only attractant, of course. Shape, scent and guide markings also play significant parts. A very dominant and effective pollinator showing sophisticated preferences for a combination of these stimuli could, in theory, lead the flower of one species to evolve and come to resemble a more favoured flower of another in several ways just to stay in the frame. Alternatively, two equally unrelated species might converge towards a common yet quite complex general appearance. Could this perhaps be the story of alstroemerias and schizanthus, the latter originating from an ancestral type similar to Salpiglossis today? Where mutually beneficial set-ups of this kind are recognised, they have been named floral guilds. It is well known how 'clever' an orchid can be in assuming whatever of the shape and odour of an insect is needed to dupe a real insect of that species into trying to mate with it. Beside this feat, for plants to impersonate one another in the cause of sharing or competing for pollinators seems a walkover. Exact copying is not required. Just enough to fool a single-minded and probably dim-witted insect. The means need be no more extravagant than required to do the job, yet no less flamboyant. That is the economic message of evolution that shapes all life. After all, even the lip of Ophrys, however miraculous it may seem, is not going to pass itself off as a living female wasp to anything but a sex-crazed male wasp. More in the nature of your inflatable naked lady really. (Plant hunters only carry foot pumps to blow up punctured Land Rover tyres, by the way!)

During the majority of our travels through S. litoralis country in spring time, we and it were constantly wrapped in a clammy shroud of Pacific mist. At its densest, as always with cloud, mist and fog, this had a disorientating effect, leaving you cast adrift from all but your immediately visible surroundings. You would move around as it were in your personal fuzzy capsule. Time became as untrustworthy as space, so that it was easy to lose touch with reality and to feel you had entered some plant hunter's disembodied dream world. Merely to wander off looking at nearby flowers rather than hanging about waiting for that early morning kettle to boil risked mislaying the campsite. By the time you stumbled upon it again the chances were your cup of tea had gone stone cold along with most of your fingertips. Photography came off worst of all: dodgy light values which shifted by the second, not enough depth of field and too little sparkle to bring dimension to flower colours on film. Hardly, it might be supposed ideal conditions for appreciating one of the world's floral spectaculars.

Yet there were surprising compensations. These were not gloomy northern hemisphere vapours, where skeletal dripping trees, hoary lawns and bleak shadowy townscapes swallow up and regurgitate pale faced and darkly clothed humanity. They had more the exuberant air of a ghostly carnival parade, especially at the best sites, Los Vilos, Talinay and Fray Jorge. If seeing everything under sunshine in one sweeping canvas of thickly laid on colour was something of a rarity until we passed north of Coquimbo, at least the drowning of distance did elevate each plant discovery to the quality of an intimate portrait. Mystery replaced the obvious. Exploring without horizons up measureless slopes and along valleys of lost beginnings and unknown endings left the exquisite uncertainty of never knowing whether there might be other new goodies a footstep or so ahead, nor quite when to give up and turn back.

It was the floral equivalent of a bran tub dip. What might come swimming out of the mist at you next? It could be scarlet flights of dark-tipped Tropaeolum tricolorum, languidly threading up through a cracked, peeling, papery-grey skeleton of a puya. Or some multi-limbed giant sentinel of Echinopsis cactus barring the way, one arm clad in a bizarre muff of tightly packed blood-red Tristerix aphyllus mistletoe flowers. Colours, which all too often turned out flattened and dull in photographs, were wonderfully muted and subdued in life by the opaque, pearly light. An almost endless succession of delights awaited, not be gobbled up by the instant eyeful, but held back in the mist to be savoured at leisure, focussed on one by one: leucocorynes, rhodophialas, calceolarias, alstroemerias, godetias, oxalis, concantheras, sisyrinchiums, loasas, oenotheras, ground orchids, calandrinias and the rest. These were the company kept by Schizanthus litoralis, not in competition for a pointless title of outright beauty, but rather all showing off brilliantly together. There were far too many winners to pick just one.

The role played by well-marked forms of S. litoralis in these pageants was that of a Regency dazzler from a masked ball, turning heads with the vividly striking contrast of velvety dark face patches and cool pure colours. For all these special qualities, it was not singled out and identified as a separate species until 1895, by Philippi, a whole century after the genus itself and S. pinnatus had already been named and described. Today it even offers an alternative local popular name, pajarito, a general word covering small and colourful little birds.

Such conclusions as can be reached about the last two groups apply equally to them both. If schizanthus commonly in cultivation are indeed a product of S. pinnatus and S. litoralis, or else already embody their best features, perhaps those two species have little more to offer us. But who knows? With so much variation in their wild stocks, coupled with S. tricolor and S. porrigens as distant memories in horticulture, surely here is a situation crying out for some further investigation.

The candidus group (white flowered annuals).

S. candidus (60cm), S. integrifolius (80cm), and S. lacteus (70cm). These closely related, similar looking species are three of the four most northerly, inhabiting the coastal oases of the Atacama desert with S. integrifolius also being found somewhat inland. None occur south of Coquimbo (308S.). They are the only schizanthus with predominantly all-white flowers, although S. lacteus has occasionally been reported with soft violet colouration overall. (Elsewhere in the genus rare albinos occur in S. grahamii, S. hookeri, S. litoralis and S. pinnatus.) One has a notable perfume. As this is a frequent device amongst white flowers to attract night-flying pollinators, moths in particular, it is very likely possessed by all three. Another shared and unique characteristic of the group is the upper lip. In other members of the genus it is usually tongue-shaped and more or less tapering at the tip to a point or small notch. In white flowered species, however, the top is invariably broad and squared off (truncated) and divided into four or more generally unequal short, rounded lobes, similar to those of the side petals, but less prominent. S. lacteus displays the most deeply cut side lobes, S. candidus the largest flowers, but with an extremely diminutive lower lip. As indicated by the name, the definitive feature of S. integrifolius is entire, undivided leaves.

Factors likely to be worth investigating for hybridisation in this group include scent, the interesting upper lip profile and the generally frilly appearance of the flowers. Insofar as hardiness in these annuals is relevant for us, we might expect them to be the least so of the genus, coming from subtropical areas with the lowest rainfalls and highest mean temperatures.

Our sole encounter with a white came very close to sunset, as we were driving back from the exciting coastal dunes and rocky outcrops of Huasco. This natural desert garden of drought resistant xerophytes is an incredibly rich few acres surrounded by sterility; a mixture of cacti, bulbs, colourful annuals and shrublets, a number of them endemic to the spot. They owe their existence to the cool, damp invasion of localised Pacific mists responding to slight variations of land level. Such rich variety had left us mentally bloated and physically empty after a tiring day trudging around in the fierce sun recording and photographing one plant new to us after another.

On the way in one of us had noticed an abalone restaurant. We were on the lookout for it when suddenly and unexpectedly there they were, brilliantly backlit, shining bushes, quite unmissable despite being almost the only vegetation on a bleak, stony roadside rise we had actually driven straight past in the morning glare, intent only on reaching the sea. What a difference lighting can make, both to perception and to appreciation. This pure S. candidus rekindled our aesthetic responses and three neglected stomachs were forced to postpone their hunger for another half an hour or so whilst their owners tried to figure out ways of making the best of conditions far from ideal for photography. Incidentally, the much vaunted, greatly anticipated abalone dinner hardly lived up to the rest of the Huasco experience. Then how often do all the pieces fit to perfection anyway? Schizanthus seed was taken at a later date and distributed, but whether any raised, let alone flowered it is not known to me.

The hookeri group.

S. hookeri (80cm) and S. grahamii (75cm). (Annuals), biennials, (or maybe short-lived perennials), exclusively from subalpine or low alpine habitats between 1800m and 2500m. Their distribution runs from 318S to 388S. Although less widespread, S. grahamii also uniquely occurs on the eastern (Argentinian) side of the continental watershed.

In the past we have suffered confusion over the naming of this group, and I have good reason to suspect we were not alone. In a series of articles written 13 to 14 years ago, I misidentified S. hookeri as S. grahamii, gave S. grahamii var. coccineus as S. hookeri and called S. grahamii by one of its five synonyms, S. gilliesii. Anyone who reads the following account carefully should have less excuse for making such a monumental cock-up in the future!

Examine outline drawings of their flowers and the two species do seem deceptively similar. Both have fingered lateral petals and a prominent upper lip shaped like a candle flame. To separate them, look sideways on, where S. hookeri has a well-developed corolla tube, the petals of S. grahamii, on the other hand, springing straight from the calyx. In life there is no real need for such botanical niceties. S. hookeri will never threaten your blood pressure. S. grahamii stands out as something special in all its forms and varieties. No contest.

S. hookeri keeps company with Tropaeolum tricolorum, T. polyphyllum and Berberis empetrifola, amongst others, in bringing a welcome softening touch of background life to ugly jumbles of assorted boulders and coarser rock tailings at the very base of huge lower screes in central Chile. It may also be found as an element of a much more diverse high meadow flora. This almost grassless steppe mixture fills dusty soil fields or typical stony rises, reaching up from the bushy mountain shrub belt as far as the refined dwarf zone of the higher Andean ridges. Unassuming mauvey-pink, the basic flower colour of S. hookeri, always blends in rather than pulling your eyes towards it. Nor, for all its uniform bright yellow, does the main lip really make a bold enough impact to advertise its presence strongly. So in our terms the species exists primarily as a comfortably familiar and expected part of the mountain furnishings at numerous points between Illapel and Victoria under its local name of mariposa cordillerana.

Plant hunting experiences stick in the mind for all sorts of reasons. There are absurd, single-minded hunts for solitary rare objectives which often seem to smack of a Lewis Carroll epic. Any discovery of a species new to science is unforgettable, provided the plant is unforgettable as well! So is the occasional nerve-jangling or life-threatening incident. All of these serve to etch into the memory particular plants associated with the events. Not so schizanthus. They are recalled purely by their singular beauty and the superb settings they inhabit. Finding and collecting any of them could hardly be described as anything more than a routine and uneventful doddle. To make up anything thrilling about S. grahamii beyond its colours and mountain homeland, how about the Argentinian popular name for it, bocas de tigre, 'tiger's jaws'? Quite how they arrived at it is unclear. The fringed florid petals make up the toothy gape, fair enough. But is the yellow the top of the tiger's head or its tongue? What a terrible night out on the town the poor beast must have had to get a tongue that colour! In humans the physical anatomies of a dreadful bore and a great performing star may be indistinguishable. The gulf exists between their personalities. Plant personalities depend entirely on physiognomy, and measurable differences between charismatic S. grahamii and rather diffident S. hookeri are surprisingly slight. The oomph of the former comes from marginally larger flowers set closer together on stems, but above all from the roseate pink, degrees higher in temperature, and its marvellously showy egg yolk of a tongue. A free-range tongue, please note, not from some pallid battery farm excuse for an egg either! Together they form an incomparable bicoloured effect, always outstanding, and once seen, immediately recognisable on any mountainside. They get even better. Some colonies near the ski centre of Las Lenas, Mendoza, are such a smoldering deep pink as to be no more than a smidgeon away from red.

Most flowers would have to be considered static or passive, excepting those with trigger mechanisms and so on for pollination, of course. A few others always seem active; well, 'caught in the act', as a still photograph freezes action, might be a better way of expressing it. Cyclamen spinning around. Haemanthus exploding. Some aroids poised to strike. Those who dish out popular names certainly subscribe to this theory: shooting stars, angel's tears, snap dragon, flying ducks, shrimp plant and, of course, butterfly flower. S. grahamii var. coccineus looks even more active than the others. Its side petals are somewhat reduced in proportion, so you might take it either as long-tailed birds, pajaritos, or a vivid swarm of diving insects, full-bodied and bee-like. The striking yellow of the lip has spread to infuse all the flower except for an intense carmine stain along the bottom edge of the side petals and lower lip. Although appearing less permanent than the basic species itself, and therefore probably more definitely annual, this variety is most distinctive and exciting. All schizanthus are somewhat hairy, tending towards a glandular stickiness, especially on the stems, but this feature is particularly marked in the two mountain species and even more so in var. coccineus. There is no denying the attractiveness of the additional incandescent glisten to stems and foliage, over and above the flower colour, when caught or photographed against the light. However at closer quarters and in a practical sense it can be a confounded nuisance, acting like a piece of exposed Sellotape by collecting every scrap of dust, hair, straw, dead insect, etc., going. Both the sticky gunge and the detritus eagerly transfer to fingers which in turn begin to collect more dust, hair, straw. . It never seems to dry out. So, ignoring any possible benefits to the wild populations, there would be every point in eliminating - breeding out - the goo factor from schizanthus in cultivation, if possible. In S. grahamii var. coccineus this secretion is first noticed as a flaccid tackiness which turns out on closer inspection by the nose to match exactly the aroma of yes, we have arrived there at last. soy sauce! This we were able to verify by direct comparison with the sachets from our Vesta Chow Mein food packs. If anyone should tell you its popular local name is maripositas en salsa de soy, they really are having you on, or one of my Chilean friends has pulled off a monstrous crib. It is actually known there as mariposita roja, 'little red butterfly'.

Seriously, it may well face a far greater threat than from our favourite Chinese restaurant in South America. We first encountered the variety growing very locally, but in some numbers, on gravelly terraces of a black-rocked little stream near the roadside just below the embryonic ski village of La Parva, above Santiago. On further exploration it turned up again, scattered very sparsely here and there across the lower alpine reaches. In those days La Parva was little more than a ski lift and fairly harmless tight knot of pretty-pretty alpine chalets, their wooden walls stained or concealed under gaudy primary colours. A predominance of brand new corrugated roofs winked sunlight signals across the mountainside while aggressively green postage stamp lawns whirred and sparkled under a regiment of mowers and sprinklers. It was a sophisticated winter and summer resort, strict preserve of the rich and privileged. Since then alarming developments have taken place. The basic village has multiplied out of recognition. Earth-moving machinery churns parking sites and foundations and cuts winding roads deep into the mountainscape back towards the main peaks. Industrial helmeted construction workers swarm over the scaffolding of a modern hotel and ski complex to be. Even in the early days we worried about the schizanthus as a vulnerable target to be thoughtlessly uprooted and dismembered for dining room posies as well as a potential victim of trampling and disturbance.

Remember, it must perpetuate itself from seed every year, unlike virtually every other plant there. What chance does it have at La Parva now, I wonder? In her field guide, the botanist Adriana Hoffman describes it as a very rare variety, occurring only in Santiago province. I do not even know whether it may be found anywhere besides La Parva. Certainly it could be under threat of local extinction there. Even worse, if it has no other home, that could be total extinction. Let us hope not. If ever any plant cried out in need of being set up and preserved in cultivation for every good reason, surely this is it? Its renowned compatriot Tecophilea cyanocrocus, the Chilean blue crocus, already waits for it there.

On a more cheerful note, eight years after our original introduction, Kew still had a specimen of S. grahamii flowering in the alpine house. It could hardly have been one plant going all that time, though whether it had been reproduced by vegetative means or seed I cannot say. I suspect seed. It would be instructive to know whether these longer-lived schizanthus do ever produce material suitable for cuttings. One subscriber to the seed collected on the trip has reported regular setting of seed by S. grahamii which enabled him to keep a succession going for a good few years, but it was always touch and go. Eventually the crop failed one season. Without doubt the large seed reservoirs and banks of commercial sources and research stations offer the securest option to sustain these plants artificially for a more or less indefinite period along with the true annuals.

As for hardiness, although the genus ranges across latitudes the equivalent of Medina to Lisbon, or Central Mexico to San Francisco in the northern hemisphere, such exotic comparisons have little bearing. Cooler seas and smaller land masses south of the equator result in generally more temperate climates for similar latitudes. It would seem sensible to treat all ten annual species in the same way as those we already grow, which presents few problems. The standard method is to sow under glass in regular seed compost at a temperature of 16C (61F). For pot culture combined with earliest flowering, sow in autumn and after germination maintain a temperature of around 8C (46F) over winter, not less. Or sow in March and follow the same procedure. For all indirect sowings, prick off the seedlings into pots, or into boxes if they are to be planted out after the risk of frost recedes in May. Where sowing directly outdoors, April is ideal, and seedlings should be thinned out in due course.

There is no obvious reason to treat the two species of the hookeri group any differently as regards basic germination processes. However, it would pay to investigate them thoroughly for hardiness. Might they overwinter as seedlings outdoors without protection to flower the following year? Or survive from one year to the next in the open? We need to know the answers one way or the other.

S. grahamii is a hard act to follow, but here goes …

Three individuals

After the four groups of allied species come three last annuals with no obvious affinities. S. alpestris (60cm) as we found it was a relatively unspectacular schizanthus with smallish deeply slashed flowers of lilac-pink, paler towards the centre, and with a pale yellow upper lip marking. That was as a very few individuals sprinkled along the feet of the bare hot screes and soaring rocky walls of the Turbio valley above Rivadavia, a somewhat airless and not especially welcoming site for plants in general save perhaps cacti. Such, though, is its standard habitat: the blistering hot parched lower precordilliera (Andean foothill) zone between Coquimbo and Santiago, apart from one recorded break into the coastal hills. Our plants were exceptionally compact, tough and wiry, certainly the best schizanthus we saw in terms of overall shape, balance and self-support, useful assets in any plant breeding scheme. On the hardiness scale it is unlikely to stand comparison with those more committed mountaineers S. hookeri and S. grahamii.

S. laetus (50cm) from the environs of Chanaral and Tocopilla is somewhat isolated, being very much its own species as well as the most northerly, and the only one with a partially tropical distribution. The solid compact flower comes with a broad, rounded upper lip, stubby side petals and less divided than any of the others, and a small fingered lower lip somewhat reminiscent of an insect's mandibles. It is said to be rich violet with some yellow marking and therefore sounds different and desirable.

Equally so, and even more distinctive, is our final species. S. parvulus (25-75cm), probably the rarest of the lot, as it is only known from 31.758S. If I have the diagram and description right, we can expect a flower of rather elegant appearance with a narrow, graceful pointed upper lip and side petals deeply divided into two segments only, not further subdivided, and those segments somewhat tapered and swept upwards. The lower lip again is barely significant. A large marking of violet or light purple fills the centre whilst all the extremities are white, the whole effect of shape and colour suggesting some cool little flame. As with all restricted species in the wild, its status should be very carefully considered and respected when and if it becomes a target for introduction, as surely it merits.

Hybrids

For the sake of completeness, and also to provide a useful pointer, I am again forced to flounder into territory that I am not equipped by experience to guide you through: the subject of incontrovertible hybrids. Historically, a large basket of schizanthus came into cultivation between 1822 and 1843, according to Paxton, including at least half a dozen of those known today. Amongst them was an early hybrid, S. x evansianus. There are no clues to its origins except perhaps that its colour is recorded as identical to that of S. pinnatus. I assume it was not a natural hybrid and am not aware of any either. In some respects it would hardly come as any great surprise were they to occur at points where species overlap, given a common pollinator. My reason for suggesting this is the existence of S. x wisetonensis, which the reference books quote as a garden hybrid between S. pinnatus and S. grahamii. Now what 'S. pinnatus' might mean in this context I am unsure, but presumably something from the usual range in cultivation that I have already called into question. S. x wisetonensis is described as having large flowers in a wide variety of rich colours, so must certainly be a strain. I have even seen it offered in an amateur seed list. One extremely hopeful conclusion may be drawn. If it is possible to achieve a successful pairing between two schizanthus from such unrelated groups and produce fertile offspring, then anything might be possible.