Where are we? Geography plays such a critical role in my perception of the species that it is important to try to understand why and how it touches on the issue. The Langeberg Mountains is a Table Mountain Sandstone about 10km wide and 1500m high running east-west and separating the inland Little Karoo from the Lowland Renosterveld of the southern Cape. There are five main travel routes through those mountains of which two are via river gorges viz. Cogmanskloof, Tradouw, and then there is also the Gouritz River Gorge where there is no road. Haworthia is not generally considered a sandstone and high mountain species because, firstly, there are few records to suggest that and, secondly, because they are averse to the higher moisture levels. However, there are many records in the low and close foothills. H. retusa ‘turgida’ is recorded in the higher areas and in the vegetation associated with the sandstone viz. the Fynbos. Fig. 1 is a view taken looking eastwards from Kleindoorn (Kleindoornrivier). This is about 16km east of Barrydale that sits at the northern end of the Tradouw Pass. The next farm is Brandrivier (B) and beyond that is Springfontein. Muiskraal is marked with an “M” and this is at the northern end of Garcia Pass from Riversdale to Ladismith. The “O” marks Oskop which is beyond Zandkraal and about 15km beyond Muiskraal. Another 10km will take you past Waterval and bring you to Aasvoelvallei at the confluence of the Grootriver with the Gouritz. The last stretch is another 10km over the Cloete’s Pass to Herbertsdale from where one travels back westward 7km to Towerlands.

What is there? I speak of species complexes because the classification and naming of Haworthia seems to defy any agreement. I do not think the nomenclatural system allows it either. South of the Langeberg are the H. retusa and H. mirabilis complexes as well as H. minima and H. pumila, as main role players. North of the mountains is H. arachnoidea and H. mucronata. There are incursions and crossing of the mountains by three of these and my contention that they do not all do so is simply because we do not see beyond the narrow confines of our classification system. North of the mountain there is enough evidence to suggest that H. mucronata and H. arachnoidea is very probably the same one complex despite the few situations where they manifest in near proximity. South of the mountains H. retusa and H. mirabilis have an even more complex relationship. It is difficult to begin to realize that there is a very small step between their manifestations south of the mountain to an integration of them both in H. emelyae that occurs north. I have great sympathy for those who contest the issue. I hope that the figures I present here, as a very small set of images from what may be a very small sample of existing populations, will help bridge the difficulty.

My field experience, however much field experience is a measure of any persons competence to process the product, suggests that there is a predictable coherence in the variation that one observes as one moves from discrete population to discrete population. This progressive variation is also driven by surface geology and habitat. It becomes no problem at all but to realize that as another set of samples, the Aasvoelvallei H. emelyae (see MBB 7850 figs 2a-l) is nothing more than the Towerlands H. retusa ‘turgida’ (MBB7849 figs 3a-b) in another guise. It is the habitats that are so vastly different. Towerlands is south facing mountain slope in sandstone and Fynbos vegetation next to a perennial mountain stream. It was no chance that the rain that caught us there stood in such sharp contrast to the warm sunshine that greeted us an hour later at Aasvoelvallei and H. emelyae a mere 10km away as the crow could fly. That habitat is Bokkeveld shale with a heavy overlay of the very white quartz that veins this formation, and the vegetation is Karoid.

H. mirabilis ‘maraisii’ is present north of the mountains at Montagu and towards Montagu from there to Barrydale. It is also present at Barrydale itself. At Kleindoorn we have an enigma (yet another) where the plants (MBB7847 figs. 4a-s) are very similar to the Barrydale plants and they also seem to flower in summer. But these same plants can be likened to H. emelyae ‘major’ at Muiskraal that flowers in the spring as does H. emelyae (and H. retusa). Figs. 5a-c MBB7844 shows plants at Brandrivier. The dramatic thing is that these are recent photographs and very unlike the plants observed there by me 30 years ago and recognized as H. emelyae ‘multifolia’.

Figs 8 are of JDV97-44 (originally GDM429) plants from beyond Muiskraal (where both ‘multifolia’ and ‘major’ occur in disparate habitats) at Zandkraal, where there are three populations to which an appellation ‘H. breueri’ has been fabricated. I downsize that to H. emelyae ‘breueri’ to bring some perspective back to Haworthia nomenclature. I do not include field photographs for Oskop and Waterval, but add two from south of Vanwyksdorp (see figs 9a-c ADH2966) where H. emelyae originated according to the collector (this was a Mrs. le Roux and the plants were sent to Smith via Mrs. Emely Ferguson). The photographs were taken just south of Vanwyksdorp by Adam Harrower and I do hope that readers will realize from this article that it is not at all strange that the von Poellnitz published illustration and type of H. emelyae (figs 10a-b) are so often said to be of H. retusa ‘turgida’. Curiously, H. Fourcade illustrated a plant (his no.144, see fig.11) of H. retusa ‘turgida’ and another (his no. 82, see fig.13) of H. reticulata as H. emelyae. This says a lot for subsequent attempts to interpret and argue vehemently about von Poellnitz’ H. correcta  (see fig.12).

What really struck me on this recent excursion was just how much inviting territory is unexplored. In contrast to that is Kleindoorn where we photographed those plants actually found by Gerhard on a previous visit there with Bob Kent and Kobus Venter many years ago. The plants are on the south side of a sandstone ridge. We could not find them elsewhere on our recent visit and it seems that suitable habitat is just not there for them. There is a most remarkable divide of shale and sandstone on Kleindoorn and while there is/was a form of H. arachnoidea in the sandstone (repeated at Zandkraal), only H. arachnoidea ‘nigricans’ is in the shale. At Brandrivier, E. Springfontein and Muiskraal the ‘multifolia’ is in a sandstone pebble layer that overlays the shale in the manner of the Tertiary terraces south of the mountain. West of Springfontein the plants are again in shale, but without the white quartz of other karoid sites mentioned.

Acknowledgement
It is not possible to do this fieldwork without the goodwill of landowners that is threatened by a glut of collectors who often trespass. Therefore I particularly want to thank Bertus le Roux of Kleindoorn, George and Jaco Nel of Brandrivier, Louis Botha of Springfontein, Anton Bredell of Aasvoelvallei, and Gregory Nicholson of Towerlands for access to their property. Gerhard Marx was contributory as was Kobus Venter on a previous excursion. Adam Harrower kindly sent me pictures from his records. ♦

There used to be a regular bus service between Riversdale and Ladismith and J. Dekenah made use of this for his excursions into the Little Karoo to find plants for G.G. Smith. He thus discovered H. emelyae ‘major’ at the northern mouth of Garcia Pass. He also submitted a single specimen of a plant collected from the karoid veld a little further north. It is that record that suggested to me that ‘major’ was linked to H. emelyae further east and north as well as to ‘multifolia’ to the west. Etwin Aslander found what I regarded as the equivalent of ‘multifolia’ (see figs 2a-i) on a low plateau north of Garcia pass on the farm Muiskraal that is at the foot of the mountains north of Garcia Pass.

Johan Meyer of Barrydale once showed me a cultivated plant that he claimed he collected on the slopes to the north of Etwin’s locality. This was a robust and rather tuberculate plant that was very reminiscent of the forms of H. emelyae at Zandkraal about 12km east. In an earlier chapter I wrote more widely about the problem of land ownership and access and eventually felt that I should look into the situation at Garcia Pass in respect of conservation status of ‘major’, when an opportunity to break the impasse arose.

Fortunately I have quite a long association with Muiskaal and its owners, since I was at the Karoo garden and Mr. Henry Chamberlain, owner of Muiskraal, finally allowed me access.  We visited both Etwin Aslander’s population (MBB7858 see figs 2a-i) as well as locating a population on the northern slopes as indicated by Johan Meyer (MBB7859 see figs 3a-i).

In addition we also found a population immediately west of the farmstead (MBB7857 see figs. 4a-f). Particularly interesting is the dramatic switches in dominances of shrubby plant species that must be a factor of differences in the very skeletal soils as well as how plants compete with one another (allelopathy). The plants in all three populations are of course similar to the east of Springfontein plants in as much as one generates a mental common denominator for a set of variable plants. This similarity must surely be seen in the similar geologies of the three sites which are on tertiary depositional terraces. We also explored extensively along the foot of the mountain between Muiskraal and Zandkraal without any success.

However, we also made contact with the owner of the upper areas around Garcia Pass itself and were able to explore both sides of the road at the ‘major’ locality (see MBB7861 see figs 5a-l) as well as the range eastwards.

We did find a new population some 3km east of the farm Onverwacht (see MBB7860 see figs. 6a-i) and Gerhard Marx reports seeing it immediately north of the same farm.  We did on a subsequent visit briefly get there, but did not find plants. Some of the plants were very ordinary and it is not surprising that J.R. Brown, in commenting on his collection, felt that he had two species viz. H. ‘major’ and H. maraisii as he might have named them. Definitely the plants I saw could have been mistaken for individuals in several populations in the sandstones south of MacGregor or even nearer Villiersdorp (Moddergat).  Unfortunately my camera failed me and I did not get many pictures, but one can see that the plants can have relatively spineless leaf surfaces.

There is something curious about the flowering times. The ‘major’ plants had evidence of recent flower stalks on a few plants while some plants were also beginning to flower. The lower populations at Muiskraal and Zandkraal were already shedding seed. I think it is significant to point out the geological features that I think are the main drivers of ‘difference’. Garcia Pass is a rift through the Langeberg Mountains that are uplifted sand stones. The northern foothills are an eroded remnant of tertiary deposits that overly Bokkeveld shale. The Zandkraal populations are associated with quartz outcrops or variants of such as are common throughout the Little Karoo. The vegetation differences between the sandstones, terrace gravels and shales are astounding and illustrated in the fact that botanists, following their original intention that a biome was a major association of plant species driven by massive and extensive landforms and climate, regard the Karoo vegetation and Fynbos vegetation as belonging to two different biomes. The terrace vegetation is primarily Renosterveld, but it has elements of what is now regarded as the Thicket biome. It would not be a serious overstatement of fact to say that no two plant species occur in each vegetation type. The farmer on Onverwacht (Fynbos) is a flower farmer picking from the field, while the present farmer on Muiskraal (generally Karoo) grows olives and grapes. Muiskraal used to be primarily a sheep farm and Onverwach barely.

Despite this observation on compositional differences of the vegetation, or because of it, my contention is that ‘major’ is a variant in the sandstones while ‘multifolia’ is an expression of the same species, H. emelyae, on the tertiary debris. They both go on to varied forms in quartzitic patches in the shales. In my opinion H. emelyae is in a continuum that is formed out of the H. retusa and H. mirabilis complexes south of the Langeberg Mountains, where broadly speaking H. emelyae moves into sandstone as ‘major’ and the other two species occupy that terrain as ‘turgida’.

Acknowledgement:
Sincere thanks to Mr. Henry Chamberlain of Muiskraal and Mr. Pierre Vermaak of Onverwacht. Gerhard Marx also motivated with interest and information. ♦