Mirabilis Travels was a series of Haworthia shucked Facebook posts in 2022.

There is not much that I can now usefully do so I am going to switch attention to H. mirabilis. If any one species exemplifies the complexity of the genus it must be this. When I first started organising field information I ran headlong into host of names with very little to relate them to in the field. Now I recognize that it is that mirabilis forms a dark coloured summer flowering half of a single system that extends from southwest of Bredasdorp all the way to east of Mossel Bay, while also extending into the Little Karoo. In no particular order, here is the “var. notabilis” from east of Worcester. My great regret is that I have no digital record of a great stretch of exploration prior to about 2002.

Wheeeee!! It really stretches the imagination. Just randomly diving in was a bad idea. So I add fuel to the fire by showing these in the no-man’s-land between “systems”. This mirabilis is west of Swellendam and both this and the previous “notabilis” cast light (or shadow) on the relation of the retusa (turgida)/mirabilis system and that of herbacea/reticulata/maculata/pubescens of the Worcester/Robertson area. The near presence of a floribunda variant is added value. One could almost argue for a restatement of the situation where mirabilis and turgida are established as generic names with many species (population variants) attached. Notice how I interchange turgida and retusa – I explain that the taxonomic system trips itself with its own rules. Priority means that retusa holds sway over turgida, when in fact it is more probable that turgida is the genetically and historically dominant.

I have over 200 folders of mirabilis localities just since 2000 and no picture records for the 30 preceding years. Much posted for this species in HaworthiaUpdates.org, but nothing can help dismiss the unbelief that these could all be one species. Here is just a blind selection of 7 pictures to show how improbable it all seems … imagine another few hundred variants!

These are mirabilis “meiringii” from east of Bonnievale. There must be something significant in the similarity to H. herbacea? Bear in mind we are dealing with many populations strung out over the landscape in what Essie and I recognise as continuities. But, mirabilis is not continuous with herbacea. Flower and flowering time as well as geographical location and continuity with many other populations show “mirabiloid”.

A little further east = still “meiringii”

… some more from a bit further east before the more traditional fewer leaved forms occur …

This is what I wanted to post – more meiringii from almost at Bonnievale. Curiously this is also the only place in the Robertson Karoo where Aloe glauca occurs. It is salutory to compare the taxonomy of Aloe vs. Haworthia because the scales are quite different. However, A. glauca also has a range of variants that never seem to have been examined. I can’t imagine what the classification of either will look like after another 50 years of this limited one-man-band style of taxonomy. Meiringii does seem to have some degree of independence but just how much and guessed at by whom? It diffuses to a spineless spotted form westward, which in turn diffuses to three different versions following compass direction.

How does one get to grips with H. mirabilis? If you take all the available names for this lot of dark coloured February – March flowerers and match them up against the 400 or so known populations, you are going to have to explain why most of the mirabiloid populations are in a nameless limbo? This is a population from North Ashton is quite out of character for that area.

These are at the southern entrance to Cogmanskloof and fairly similar to plants northwards as well as to the general ensemble of mirabilis in the Worcester/Robertson Karoo, but I am having fun putting together a general perspective of H. mirabilis. It was not even considered known in the field until I started messing with the classification. That is another story I do not want to see unfold! So far there are 44 names that cover the story of the mirabiloid to help explain why so few have any idea of what the name mirabilis conveys.

I have had a bit of fun in taking all the names relevant to H. mirabilis to show what a taxonomic nightmare it has been and will remain. Those 50 extant names only represent about 75 populations at the most. In the majority of cases they do not cover even the variants in the designated populations. So not only do we have another possible 200 or more populations to consider, but also the variation within them.

This is about the most westerly record for mirabilis in an almost ecological desert. There is clear pattern in how the populations relate to one another, with just enough exceptions to overturn expectation.

There are about 10 known populations of this type of mirabilis between Caledon and Napier that gave rise to the varieties of triebneriana viz. napierenis, turgida and multituberculata. Grasping what the situation is with this system of species would go a long way to resolving the differences of opinion about the entire genus.

As we move eastward to north of Napier the plants change a bit as they get smaller. It is worth looking for that floribunda-like leaf tip!

At Napier itself the plants are smaller. Just keep in mind we are considering just what constitutes this species mirabilis. It is a questioning of trying to reconcile about 300 or more populations across a very wide area AND accommodating them each and everyone in a rational system. Remember history gives us about 50 names many of which are alternate species names. How do we make sense of this? So I am taking what pictures I have and local area by local area presenting the problem as it is – not what I imagine it to be.

Badia is a particularly striking variant of mirabilis and occurs as several populations along a 3km stretch of a recent geological surface west of Napier. Forming a part of the southwest border of distribution of the species.

These are badia from the western population – rougher surfaces.

Bredasdorp is only 15km east of Napier and has its own odd set of mirabilis. Mundula is from southwest of the town. In 1969 when I first saw the plants, there were large clumps of 30 to 40 rosettes, highly coloured. Over the years the clumps collapsed, the surrounding vegetation aged and the plants were scattered about and either single or in much smaller clumps.

In sandstone just south of Bredasdorp there is quite a strange population of mirabilis we know as var. sublineata. With narrow upright leaves it can have some resemblance to H. rossouwii and even to H. emelyae var. multifolia. I once grew it from seed and had over 70 plants that were all different. Some variation in Haworthia mirabilis var. sublineata

While generally presenting the things that constitute mirabilis, I have mentioned rossouwii. It seems to be generally assumed that the species are real discrete things that can be fitted into some sort of classification structure. This is definitely not true. At one stage I thought there might be a species heidelbergensis, and consequently named a population north of Bredasdorp as its var minor. So let us look at the issue. Before I get to that particular population, look at these plants to the west …

When my hard drive crashed I was just going to start the mirabilis heidelbergensis conundrum. It is not an area I like at all because there is no clarity as already indicated. So here is a population close to the origin of Smith’s “heidelbergensis” – but these are big plants and I file them as retusa/turgida when they could better be turgida/mirabilis or retusa/mirabilis. It is so easy to generate opinion and comment when there is not much information available. So I made a selection of photographs of 20 or so populations to indicate what has to be incorporated in a classification – somehow or other! Often pictures are poor and then I only have since about 2004. Another problem is seasonal difference. One is not always lucky to get to a population when the plants are filled out and actively growing, while they are not the same year by year either. Much is made of growing and observing these plants in cultivation. Bananas! All very well for ten, twenty even 100 populations. But impractical for hundreds.

About 800m west of the type locality for “heidelbergensis” and 400m N of the previous post of large plants, these are small and roughly equate to ‘heidelbergensis’. Here the plants are turgid and greenish coloured as opposed to the much darker and shriveled in the dry season or spell. Being dark coloured they are invariably well sheltered under other vegetation.

H. mirabilis is generally a dark coloured plant. Thus mirabiloids are dark coloured vs. the retusoids(turgidoids) that are lighter. Of course in shade or cool conditions the plants green up. Here is a “light coloured” mirabilis from Greyton to just demonstrate that making any statement to distinguish mirabiloid from retusoid is difficult. My conclusion is indeed, because they are the same species! “To avoid confusion” I must also add that when I say that plants that look different are the same species, some critics get highly apoplectic? This is the problem my posts are addressing and there is not much evidence available to me that the message gets received or is welcome. Perhaps the latter sentiment is most true because I find it uncomfortable myself.

Just east of Heidelberg is another population of these small plants that fall in the ambit of what actually is heidelbergensis. Growing in think moss and lichen it is a slightly different habitat. These pictures are of plants in cultivation.

These are a few km S Heidelberg. Now I am enthralled by what my own pictures are telling me. I write to find out and establish what my observations mean! I just deeply wish I had prepared sufficiently for the end stage when the pieces fall into place. What is happening here in floribunda is included in the compromise. Nearby we have a retusa var. ‘nigra‘ for which the story has been told. ‘Heidelbergensis’ is for me a deep insight into the connection between the mirabiloids and the retusoids (turgidaoids). Colour is a significant character, so is the bend back of the leaves as well as the rounded leaf-tip. A thing to note is that “retuse” as originally used was to mean “bent back like a thumb” and NOT the modern usage for a flat leaf in which the margins grow ahead of the leaf point. So erect vs retuse leaves is also significant, but in Haworthia it is not a species determinant and more than colour or rounded leaf-tip OR even flowering time. These pictures help to demonstrate this fact. If just one population did this we could perhaps dismiss the issue as hybridization. And this is my greatest regret. I just do not have enough or good enough pictures taken at all seasons and of more variants, or for all the populations I have been so privileged to enjoy. Because i am going to be succeeded by “taxonomists” intent on imposing their own opinions for the sake of effect rather than reality?

Eastwards the ‘heidelbergensis’ form extends half way to Riversdale where it merges with the atrofuscoid/magnificoid versions of mirabilis. But this population is further south down the Duiwenhoks growing in close proximity to retusa ‘nigra’ and the putative parent for that viz H. turgida. There is now so much evidence for the unifying of formal names and especially considering Essie’s exhaustive exploration and opinion, independently validating what I have to say.

Still further down the Duiwenhoks are these two populations for which there are sadly not enough pictures to remotely demonstrate the variation. What is perhaps worth noting is leaf shape, colour and translucency. Still further down the river is a major conundrum in the form of a population of dark unvariegated could be variegate. This is a problem of distribution because along the coast there are no haworthia between Stilbay in the east and the Duiwenhoks and then to Infanta in the east, apart from paradoxa. It is a major issue as there is no continuum as for the inland strip.

A few kilometers SE Heidelberg is this population of ‘heidelbergensis’ and I will follow with a further one nearby that demonstrates the transitions from retusoid to mirabiloid. This population is nevertheless already mirabiloid as is evident by the darker colours and the summer flowering time. In the last post I mentioned an aberrant population of presumably variegata. While trying to explain the complexity of species generally, I have avoided some serious issues with populations that simply do not fit anywhere, while they, only in their oddity. can they be claimed to fit a rational species definition. I hope to come to those.

Moving about 1km further SE is this population that also demonstrates variability and the nature of population differences. This population is more magnificoid in that there is a tendency for retusion of the leaf, for more textured surfaces and for less translucency coupled with more blackish-green coloration.

Before continuing with ‘heidelbergensis‘ just look at these magnifcoid/atrofuscoids from west of Riversdale where there is no turgida. Different habitats and also it is the retusoid version that does occur just here but infrequently.

These non-descript mirabiloids also from west of Riversdale can only roughly be distinguished from what I refer to as Heidelbergensoids. They occur from well east of Riversdale all the way to north of the Potberg, Bredasdorp, Riviersonderend, into the Worcester/Robertson area to south of Worcester, to Ashton, Robertson and then into the Little Karoo at Montagu and Barrydale. If you scrambled different quantities of varied colour thick paints on a board and scratched them around with a stick, heidelbergensis would appear as a thin web within the middle of the outcome.

In my Revision I recognised 4 varieties of heidelbergensis and this was really because I did not know what to do with these troublesome variables. Twenty five years later and with a lot more information, I can only consider a sensible explanation to be that they are part and parcel of a single system viz. H. mirabilis (and of course intimately linked to retusa/turgida). Here are two pictures of my first var. toonensis from SW Heidelberg. Unfortunately only two cultivated specimens and no idea of variation, but there is a lot better to come. Essie may have more because these are down the Slangriver that he knows well. What is so valuable about Essies conclusions is that he has arrived there on the basis of a wholly different set of populations and individual plants – and without herbarium support or influence.

These are a bit west of the Slangriver and taken in July during the wet season. So the plants are green, turgid and about as much translucency as they can show. It occurs to me now that I always thought of words rather than pictures, but in this digital age this is about how many pictures needed and of better quality.

Essie Esterhuizen commented “It was in this area where I realised in the early 90’s that I don’t have the ability to sensibly distinguish between the then maraisii, magnifica, mirabilis and heidelbergensis.”

To which Bruce Bayer replied. “Well Essie, its not a question of your ability. It simply cannot be done. Those names just existed because of the history of the subject and even now there is surely some doubt that you and I are stating the fact that a true classification does not meet the perceptions or needs of enthusiasts and collectors. Their need are real and it is time amateur journals started assuming some responsibility for a way of using names that is useful, while being accurate.”

This is also west of the Slangriver and not just making a mockery of any attempt to subdivide H. mirabilis into heidelbergensis, atrofusca, jakubii, magnifica, maraisii, shuldteniana and many more, you can clearly see that floribunda is involved too. I do not know of any floribunda southwest of Heidelberg except the evidence that appears in these small mirabiloids that this is where it has gone. Soumen Aditya is a good observer and it will be nice to know how his opinions of the names that are developing.

Soumen Aditya replies “Sir, true…one population and its species concepts no value at this moment. Horrible variations each single species and definitely link each others. Floribunda is a simple variation under many mirabiloid included maraisii. Thank you.”

This is a few km downstream on the Slangriver from toonensis and a great opportunity for someone to explore the prediction that more populations will exhibit the similarities that follow geographical order. There are some unique plants in this post and the Slangriver is no stranger to oddity. It is not easily accessible. It follows a strange path – first southward and then westwards to flow into the Breede River between Malgas and Infanta where there are many mirabilis and some turgida variants.

Some way to the northwest and the plants seem to have lost translucency while the leaves are shorter stouter and perhaps darker green. But there are MANY populations of this ilk varying in one direction or another, supporting the truth of Essie’s comment that categorizing them by any of the more relevant names is just not possible.

While considering what to post next on Mirabilis Travels I realized that I have to break mirabilis up and organize my pictures better. There are practically speaking 12 geographical zones and I only have pictures for about 200 populations of at least 400 I have visited that fall into the category I recognize as H. mirabilis. I cannot believe that anyone can dispute the statement that these populations and the internal variation form a continuum and any taxonomic subdivision will be purely artificial. In fact I go further and state that the species H. turgida, retusa, pygmaea, mutica and emelyae are more probably than not, included. When the involvement of H. floribunda with it’s associations, and then the probability that H. emelyae has further connections are considered as well, the enormity and complexity of the situation can be imagined. But here are pictures from two further populations to go with the previous SW Heidelberg.

Each of those mirabilis zones has something going on pointing to the complexity I refer to, but right now I am posting about the heidelbergensis mirage. This is the Swellendam zone and the population one from S Buffeljagsrivier. The home of groenewaldii, that is a variant that for me puts the whole retusa/mirabilis complexity into perspective.

This, further south from Buffeljagsriver. The plants were not in good shape and this is just one of the sad facts of life. It raises the issue of how to decide what it is. But the situation is that there is no clear idea of just what constitutes heidelbergensis. West of Buffeljagsrivier there are 6 known populations of mirabilis that I cannot state why exactly they are the same. There is an obvious similarity in respect of size, colour and flower but confoundment with the floribunda/variegata variant as well. Can it be understood and can it be explained? So lets rather continue southwards down the Breede River for the moment.

Further south along the Breede River is a “farm” where we found 8 populations of mirabilis and I will just show plants from 3 localities to show how they vary. See the floribunda influence too – floribunda as a recognizable population is absent south of the Bontebok Park but reappears as such down along the western end of the Potberg. But remember we are actually on a heidelbergensis mission to show that it is fiction just as magnifica and maraisii (and others) are.

.
.

Eureka – have I got it? For years I have been pondering the problem of how and why the form of rossouwii appears in mirabilis ’heidelbergensis’, mirabilis ’sublineata’, and emelyae ‘multifolia’. Essie has independently indicated that he observes this too. I should actually be travelling, in my series of posts about heidelbergensis, from west to east to show this. Sadly I do not have pictures of the heidelbergensis I saw further east of Heidelberg, nor any of Essie’s for the eastern Hessequas area. DNA does not create form. It only provides the materials to do so. Form depends on the creative power of the metaphysical element (causal mind – consciousness) of each life form? I will put a question mark there again, because this is a serious issue we need to contemplate. The photos I add are multifolia variants from 4 populations from Little Karoo. The last one touches on Warwick asking why I fingered a single variant from S of “heidelbergesis var. toonensis” along the Slangriver.

So we are still south of Buffeljags along the Breede River. Looking for heidelbergensis evidence. Not much in this lot but I did also interject with a post referring to rossouwii. Very interesting because now one has to consider all populations, related species AND their distribution. Because this is a whole and species cannot be summarily treated like cherries or plums and picked by whim and fancy. Haworthia rossouwii is very odd because it occurs at two centers about 30-40km apart and then appears to reoccur in a smaller form just south of Heidelberg.

Bit of a dilemma. Rather than go south I will for the moment go west from Swellendam towards Stormsvlei. The end will shock as it really demonstrates the impossibility of arriving at a solution on the mechanistic platform.

The predictability of a working hypothesis? Essie sent me this lovely picture of a plant from the mountains between Heidelberg and the Little Karoo. But like me Essie has only just awaken to the fact that you cannot get people to believe what you say unless you can give them a greater visual experience than you had yourself? Essie has a LOT to tell us. This is a link between the turgidoides (i.e. retusa and mirabilis together therein) and H. emelyae. It will eventually be recognized to also include rossouwii?

A little further west in the Luiperdsberg area west of Swellendam.

Now between Luiperdsberg and Bromberg on the way to Stormsvlei. There is not a smooth transition and a huge number of populations nevertheless supports the contention that this is a single system viz H. mirabilis.

I don’t have pictures of mirabilis on the south side of the Bromberg so these are from north. Also I am going to skip going still further north into the Worcerster Robertson Karoo and beyond the Riviersonderend River. I wrote all about mirabilis in Updates about 2006 and don’t have pictures for the complexity in that area and westwards, which would help to ease the oddity of transformation.

Tony Brook asked if mirabilis is the oldest name. Bruce answered “A very uncomfortable question. I have no doubt that it is atrovirens – but it is a tricky path and I am no fan of the nomenclatural code or the fancy footwork that goes with it. Goes back to Stearn and typification of old names.”

These Bromberg variants will surely excite the splitters but really when one sees so many of the populations from McGregor to Swellendam as well as to the north; and even the ecological semi-desert to the south, you have no option but to abandon an impossible point of view. Unfortunately the population of ‘heidelbergensis v. scabra’ (i.e. before my even conservative view was proved hopeless) at Leeurivier east of Drew, was ploughed out despite the non-arable nature of the shale. But there are pictures in Revisited and also in early volume of Updates re: H. mirabilis.

Authoritarianism is a real problem in present society and this comes up in plant classification too. So it is not compulsory to see these plants as H. mirabilis too. It is just my personal opinion. The plants on the north slopes of the Bromberg are in sandstone and these now are along the Riviersonderend River in manganic tertiary deposit. This is very unusual habitat. Taken late February as they flower in the heat of summer.

These were taken 3 years earlier and In January when cooler wetter conditions prevail. There are four microhabitats in a very small area.

Now this is a real freak for which I have no explanation. A single small population in a weird habitat on the south side of the Bromberg. Sandstone slabs and very wet. About 200m away is shale and also what one expect of H. mirabilis. But this flowers in November and has in-turning leaves that one finds in, say herbacea. Although there are individuals, as you can see, in which the leaves are erect-spreading. The flowers are nothing unusual except the buds are flattened top to bottom and the ends also turned down slightly like-in mirabiloids. So it throws everything out of the window or into disarray. It does not alone meet the requirements of a species and there is zero indication of what its real affinities may be. My deeper feelings are that it mirrors a dark side of (at least Haworthia) classification and that the real species is the entire “genus”. For the present I have speculated, as I have suggested in this series, that floribunda/variegata/chlorocantha are intimately connected to the turgidoids, and suspect that rossouwii is as well. I have treated this H. eliseae as a variant of rossouwii, as I have done with ‘calcarea’, ‘petrophila’ and my own ‘heidelbergensis var. minor”. Let me add something. Classification is not some thing confined to part of the group so that you can cherry-pick where and to what you want to apply a name. A classification has to be ALL inclusive and the specimens remaining in the incertae folders are only there because they are too poor to permit any decision.

About 200m away is H. mirabilis.

The myth of heidelbergensis has dissipated to the west and north and morphed into others, so let us go back down the Breede River and see what mirabilis holds for us in that direction. There’s a gap that I haven’t explored, the 20km stretch of the Breede River as it twists and turns south of Swellendam. The next two populations are south of Malgas, east of the river.

Here are more from the east bank of the Breede River south of Malgas. Note that the populations on the west bank all along the northern lower reaches of the Potberg are incredibly complex with names like joleneae, bobii confounded with atrofusca, floribunda, variegata, turgida, and mutica.

As an after thought … did anyone see this? Some how or other these two forms of H. eliseae came about, and I suspect it may have something to do with Rupert Sheldrake’s suggestion of morphic fields. Is this an overlap of the Worcester/Robertson herbaceoids and the more southerly mirabioids? I settled for a rossouwii/mirabilis interaction that is not entirely convincing. These two pictures were taken at the same time and place in one small population

I am not sure how to deal with the west bank populations. Essentially I am talking about H. mirabilis and have been exploring the heidelbergensis phenomenon. On the lower west bank of the Breede River we have a long conjunction of shale and sandstone mixed with tertiary deposits. The vegetation is either fynbos or renosterveld and there are vast differences over short distances. H. turgida in its best form is present (along the river itself) and so is the floribunda/variegata element. The result is two contrasting versions of mirabiloids. I will post the 6 or so turgidoid influenced populations first (fynbos) and then the more dark mirabiloids (Renosterveld) after. Note that there are no absolutes. On the decaying shales where there is kaolinite, we could be back west of Riversdale. I will start with what deviously and/or distastefully, acquired the name ‘bobii’.

This is just a few km west of bobii and again a link in a chain that makes a mockery of populations as species. I don’t think the Potberg is well explored at all. While this is on the eastern and north facing flank, there are two other populations known on the southern slopes travelling westwards.

On the southern face of the Potberg a few km further west and well off the beaten track …

Again travelling west along the south side of the Potberg is a similar population of plants. Unfortunately I only have two pictures of plants in cultivation that are not very useful except to show how growing conditions affect appearances. We arrived at site late afternoon in rainy weather. It was a long hike through thick brush and the light was also too poor for film photography.

Leaving the Potberg, returning to bobii at Infanta and then searching along the lower slopes between the Potberg and the Breede River. Just in what respect and by what margin do the next three populations differ from the previous 4? This was March 2006.

It is a bit of overdose because these are pictures of the same population as previously but taken a year later in September.

…