Volume 3, Chapter 4:- Classification with purpose

This essay was published in Alsterworthia International in a special issue No.3, June 2003.

This is a two-part essay. The first is to discuss a problem in the small Madagascan Aloe species, and the second to discuss Poellnitzia. The latter is now be listed in the genera of Southern African plants as Astroloba rubriflora. I think it is necessary to point out that while this may now be perceived to be ‘authoritative’, the taxonomic treatments are in fact not so.

A. The small Madagascan aloes:- In my pursuit of a communication framework for Haworthia, I have experienced many problems. Not least, is the response I get from botanists. From this experience I thus find the article in Taxon authored by no less than four botanists, really most instructive. The article (Taxon 44:513, 1995) deals with the differences within the genus Aloe and the re-instatement and transfers by P.V. Heath of some of the Madagascan species (A. haworthioides, A. boiteaui and A. parvula to Lemeea, and A. bakeri, A. bellatula, A. calcairophylla, A. descoingsii and A. rauhii to Guillauminia together with A. albiflora. As will be shown below, the information that Heath’s ‘classification’ either provides or obscures is virtually nil, and the reversion restores a heterogeneity which also says very little. The Taxon 5‑page article starts the discussion with a textbook quotation from Jeffrey (1968):‑ “There is no single criterion which by itself can be regarded as unfailing for recognising a genus. For practical purposes the genus should be regarded as an inclusive category whose species have more characteristics in common with each other than with species of other genera of the same family. There should be a clear morphological break between the members of the particular genus and the members of other related genera. Past and current practise are also important, since a major change in classification and naming can create communication problems and result in loss of information”. If one considers that Taxon is a prestigious scientific journal, this quote is surely unnecessary as the Asphodeloid genera have defied logic since Linnaean times.

Smith et al claim to have undertaken an investigation of Aloe to show that it is polymorphic. The basis of any genus is polymorphism. The evidence of the investigation is firstly, a table listing the ‘characters’ used by Lemee to distinguish Aloinella (= Heath’s Lemeea), against Aloe species which are suggested by Smith et al to have the same “characters”. Secondly, is a similar table juxtaposing Guillauminia with Aloe species which suggestively also have the same “characters”.

Why I have put “characters” in inverted commas is because it is well known in classification, that recognition of categories may be intuitive rather than rational. I claim that this is the case here as well, and that the analyses made by Smith et al are just fallacious. Even were the facts of “character” true, it would still be a fairly pointless exercise to try and simplify a multivariate state the way they do.

To illustrate this:- Take Lemeea in Table 1 in the Taxon article:-

Rosette small and dense. Smith et al cite A. aristata and A. humilis as having such rosettes as well, but Haworthia also has species which have small dense rosettes. The fact is that these small dense rosettes can readily be seen to be different, and recognition of species and even genera can be made on the basis of seeing only those rosettes. So an informed observer can distinguish many of the species on the basis of the rosettes only. This is in contradiction to a situation where experienced herbarium workers have confused members of wholly different genera eg. Aloe aristata and Haworthia, H. pumila(maxima) and H. limifolia, Aloe humilis and Haworthia, Haworthia venosa and Astroloba muricata (aspera/rugosa/corrugata).
Leaves narrowly linear. Smith et al cite Sect. Graminaloe. This is another misleading comparison. Any average observer in Aloe can recognise that section on the basis of its leaves. To suggest that one could not separate any of these Graminaloe from the small aloes in the group proposed by Heath on the basis of leaf character is an indictment of the skills of ordinary people.
Pedicels negligible. Chortolirion too, has a negligible pedicel as do several species of the sub‑genus Haworthia. But A. bowieae too is cited, and it could be claimed that, as Chamaealoe, it communicated something which is lost when it is treated as Aloe.
Filaments broad. No comment by Smith et al. Aloe haworthioides has the most remarkable filament structure and the broad upper part of the filaments can be mistaken for petals and an extended perianth tube. In fact it is quite ironic, and likely, that the same experienced amateurs who can see differences in the rosettes which Smith et al cannot, may mistake the filaments for petals. These filaments cannot emerge from the tube before the flower opens. To suggest that this condition in A. haworthioides is the same as the extended filaments in say Aloe ferox where both anthers and stigma are presented to pollinators outside of the tube, is misleading.
Anthers exserted. Many aloes do indeed have exserted filaments and it is hardly necessary to cite something as spectacularly different as the huge-stemmed A. ferox in a discussion of plants with small compact rosettes.
Ovary acute. This is cited as erroneous, but there is so little evidence anywhere for the shape of ovary, style and stigma in Aloe species, that this character is hardly worth noting.

Now take Fig.2 for Guillauminia of Bertrand:‑

Inflorescence lax. Smith et al would have to give a very precise definition of this term and say exactly how they would use it in this particular context. In Haworthia there are inflorescences which can be described as sparse and lax, and inflorescences which can be described as dense and firm (sometimes within the same species!). Hardly any of them are quite as firm as is the case in A. tenuior for example ‑ which has a variant on the Zuurberg having a lax inflorescence compared with its more robust relatives on the karoid plains where the inflorescence could by comparison be described as firm.
Perianth segments free. Smith et al cite A. myriacantha and A. albida. The actual situation is one which does not seem to have been explored in the genus Aloe, nor in related genera. The flower in the Aloaceae comprises three outer tepals and three inner. There are many different states within these genera. The outer tepals may be free or partially free or form a separate tube; they may be fused to greater or lesser degree with the inner petals to form a more convincing tube. More to follow.
Anthers exserted. This is a fairly meaningless statement in view of the degree to which anthers are exserted or not exserted in the genera as a whole.

The facts thus are that these two tables are just a fallacious assembly of observations used by Smith et al to negate an unsubstantiated opinion put forward by Heath. They do not comprise a series of facts which face the scrutiny suggested by Smith et al, and which are therefore just as weak as the rationale for Heath’s genera.

Tables 3 and 4 of the Taxon article, which supposedly tabulate the “characters” of the species which Heath placed in Lemmea and Guillauminia are technically no stronger and have no more substance than the previous two tables.

Smith et al conclude… “The species transferred by Heath to either genus do not have any features that might even remotely warrant their segregation from Aloe“. It is really regrettable that this kind of a statement is a feature of taxonomy. It can be liked to Shakespeare’s “Me thinkest thou dost protest too much!” It is obvious that there are characters which suggest difference of quite a major degree. If there were two more species that had those exserted flattened coloured filaments of A. haworthioides, they WOULD in all probability be taken to be a separate genus. The weakness of the protest is underscored by the next statement. Despite now claiming proof that the species can only be classified as Aloe, Smith et al go on to say “Should Heath’s views be followed, many more Aloe segregates deserve recognition”. It is a curious co-incidence that a paper by Forster in Bothalia 30:53 (2000) follows one in which Poellnitzia is included in Astroloba. Forster writes of the “characters” by which species of Lomatophyllum are easily distinguished. In fact there are two characters and these are the fleshy fruit and wingless seed. Just how much difference there is between the fleshy fruit of Lomatophyllum and the said non-fleshy fruit of Aloe is, is conjectural. The fruit of Aloe is initially also fleshy. Winged versus winglessness is a character set which has not been explored within Aloe. Thus even Lomatophyllum, as genus or subgenus, rests on flimsy ground. What is disturbing is this feature of scientific writing. Forster cites the cladistic analysis of the Alooidae by Smith and Vanwyk published in Taxon 40:557 (1991) giving it a credibility which can be proven to be totally misplaced. That analysis rests on a set of characters which are just as fallacious as similar allusions I make in this paper.

Actually therefore the statement by Smith et al is the only substance in the entire 5‑page article by the four botanists which is of note. The article has actually demonstrated that classification is not science (as this writer has claimed it to be), and neither is it art. It is simply the arrangement of groups of plants in an ill‑defined way where they are said to be species related together in genera. It is done in a way in which we can communicate with one another about them in the hopes that we can make general statements about specific things.

My complaint now is that the papers by both Heath and Smith et al fiddle with the classification and nomenclature, without adding one iota of information to the subject. The facts are:-

A. haworthioides and A. albiflora do in fact have characters which make them most unusual in the genus Aloe. There is no Aloe remotely like A. haworthioides which has the extraordinarily protruding flattened filaments. These form a reddish tube nearly as large and conspicuous as that of the perianth itself, which is almost hidden by the large floral bract. In the case of A. albiflora, the florets are widely campanulate and sparsely arranged on the slender peduncle. Reynolds placed 12 species together in his Group 1 of Madagascan Aloes and stated most plainly that they were not closely allied. Now Heath has ignored three of those 12 species and made two groups out of the disparate remnant. This is the problem that Smith et al should have addressed. The curious thing about this is that there appears to be a similarity between A. bellatula and A. perrieri (ignored and abandoned in Aloe by Heath, and not even considered by Smith et al), which in terms of variability could be extended to A. albiflora and perhaps further to A. parvula. There is no such similarity between A. haworthioides, and A. bouteaui. A. descoingsii and A. calcairophylla do seem to have unique florets. There is nothing odd about the florets in A. rauhii or A. bakeri that could necessarily separate them from, A. versicolor and A. parallelifolia.

In my own observation of Aloe, which I have had to consider when trying to find solutions for Haworthia, I have been struck by the artificiality of the whole classification by Reynolds. Not for a moment do I suggest anybody else could have done it better, but it is in fact just an eye-balled solution. Looking at the Madagascan aloes, it seems to me that species like Aloe conifera and cryptoflora are probably synonymous. In the South African aloes, there is a problem with Aloe striatula. in which the floret is quite unique (the inversion of the arrangement of outer and inner tepals). This species is also oddly placed with species like A. ciliaris and A. tidmarshii in a Subsection and Series (categories which Reynolds abandoned for his second volume for the non-South African species).

In once trying to identify a small Graminaloe from McClear in the Transkei, I could not separate it from neither A. myriacantha nor A. albida. Reynolds in his first volume had a very restricted distribution for A. myriacantha, but in the second volume it is shown to have the second widest distribution after A. buetneri. Thus it is obvious to me that it is in fact synonymous with A. albida, and may include even A. minima and A. saundersiae (the question of actinomorphy vs zygomorphy intrudes here!).

The Smith et al comment, that if Heath was followed many more segregates would have to be re-instated or created, is irrelevant. It suggests that these authors could authoritatively identify such segregates, and from what they have done in Tables 1 and 2 only, I have to conclude that they could not. The issue is what benefit would there be. Heath’s hypothesis, which is a flattering term, suggests affinities which are not there. The Smith et al annulment is appropriate, but poor in rationalisation.

An important issue is that of ‘characters’. The details of floral morphology are really not understood in the Aloaceae. This statement is underscored by the historical lack of attention that has been paid to the differences in the flowers of Haworthia and its three sub-genera, also of Astroloba and of Poellnitzia. In fact a recent treatment of Poellnitzia places it in Astroloba, which is so incongruous one has to ask why recognise genera in the Aloaceae? Several attempts have been undertaken to unravel the relationships of the genera and produce a structured phylogenetic tree (cladistics). None of these have considered the diversity within Aloe, nor the actual structure of the florets.

The quote from Jeffrey that Smith et al use, is probably useful at undergraduate level to suggest some kind of ideal. Why they have used it in this article in the Aloaceae where the genus Gasteria is probably the only genus which has more than 10 indisputably related species, is curious. My observation will be lost on the reader unless I state that subgenus Haworthia could be allied with the Graminialoe, the subgenus Hexangulares with Chortolirion, the subgenus Robustipedunculares with Astroloba, the genus Poellnitzia (now Astroloba) with Gasteria. Jeffrey’s requirements are no doubt met to some degree by these genera, but the genus Aloe is by far the greater body of species, which remains relatively intact by default rather than as a rational considered portrayal of morphological or genetic difference. As I would plead for Haworthia – do not make changes unless they contribute substantially to knowledge and understanding. Any changes should facilitate communication and make that communication meaningful. It may in fact be very useful to recognise some of those segregates of Smith et al which we are left to guess at. There are substantial segregates such as the group of Leptaloe and the maculate aloes. These could profitably be separated out simply to direct better attention to the extent of diversity in the family and the genera and make them more manageable.

The problem in Aloe may primarily be, one could suspect, because the revisionary work has been done by an amateur. This is G.W. Reynolds. No matter how much effort and diligence and intellectual skill has been applied, there is no way that the work can represent a real contribution to the essence of classification of biological systems. Smith et al are professional botanists and their observations here are no better directed. Reynolds, in his preface to “The aloes of South Africa” writes “The beginner will learn …” and goes on to the superficial comment found in every popular work about mans ignorance, and the refusal of nature to conform with mans wish for precision. But neither he nor other botanists really seem to make much progress beyond the beginner stage. The fact is that a classification is for people to use. The vast assemblage of species in the single genus Aloe need to be teased out for practical purposes and the weak rationalisation described above is a blow to any hope of clarity. The fact is that a species may be “seen” to be different but there may be no “characters” by which the degree of difference can be demonstrated.

B. Poellnitzia rubriflora:- A problem similar to that of Lemeea and Guillauminia, occurs with the genus Poellnitzia. This species P. rubriflora (L.Bolus) Uitewaal is a very distinctive species of plant which cannot be confused with any other. So it is really very curious that it has such a chequered taxonomic history. At least, this would be curious if this systematics and of its related genera and species had not suffered such cavalier and unbelievable treatment by botanists and would be botanists. Smith and Manning (Bothalia 11:53, 2000) now place P. rubriflora in Astroloba. They say.. “Vegetatively the species shows close affinity to some species of Astroloba and accords completely with the genus in its tubular actinomorphic flowers with included stamens.” This is in contrast to Rowley’s suggestion (in Smith, 1994) that Poellnitzia shared more characteristics with the genus Aloe than with Haworthia. Or with Parr (Bull.Afr.Succ.Plant Soc., 1971) who placed Poellnitzia in Haworthia.

The facts are that Poellnitzia is similar to all seven Astroloba species in that the plants have the leaves carried on extended stems in clear 5-farious tiers. In any Haworthia which have such extended stems, the leaves are trifarious or multifarious. Trifarious and even 5-farious arrangements do occur in some stemless species but just as dictated by spiral phyllotaxis. Poellnitzia is characterised from Astroloba on the basis of a very waxy epidermis.

The statement about the accordance of the flowers is a most extraordinary statement. In Poellnitzia the perianth tube is apparently entire as the margins of the outer petals are fused to the mid-rib of the inner petals for their greater length. The midribs of the inner petals are clearly visible between the ‘margins’ of the outer as they are in Astroloba, as they are in two of the three Haworthia subgenera and as they are in Aloe. The most obvious similarity of the Poellnitzia flower is probably to that of Gasteria, except that the curvature of the tube is down as it is in Haworthia sub-genus Haworthia, and opposed to up as in Gasteria.

This statement raises the question of actinomorphy. The flowers definitely are not actinomorphic in Astroloba – this is a persistent myth. There are always two distinguishable upper outer tepals and the single outer lower tepal. This is the same as in the Robustipedunculares of Haworthia. The Poellnitzia flower is only actinomorphic in a superficial sense and I would not call it so in an analytical and scientific sense. Just the curvature of the tube alone, denies any bilateral symmetry except in one plane. I frankly do not understand why this perception is so strongly held, but then I do not understand why formal botany has always treated the Liliaceae in such a cavalier way. The tepal tips may be actinomorphic but the tube is generally simply asymmetric and there is no way that one can bisect it in any plane but the vertical one. In Haworthia, and in most species of Alooidiae that I have examined, the upper inner petal overlaps the upper margins of both the lower petals. There is no similarity in the flower of Astroloba and Poellnitzia at all in the context of the smaller Aloid genera. On the other hand if flower structure is so important as to suggest one genus, what does one do with Aloe striatula, which must have a flower unique in the Aloaceae, if not in the Order. There is one upper outer tepal and two lower outer! – the flower is by the way, zygomorphic too. It needs to be seen what the arrangement is of the inner petals is to each other – they are truly turbiniform imbricate.

The authors close with a statement of pollination saying that Poellnitzia is pollinated by the Lesser Double-collared Sunbird. It is well known that Haworthia and Astroloba, with their whitish flowers, are also visited by the same Sunbird. The thrust of the Smith/Manning note is that a specialised pollination strategy is not sufficient grounds for separation of Poellnitzia as a genus. Thus the paper directly contradicts Smith, who in a 13-page article (Taxon 41:437, 1992) postulated Poellnitzia to have a mite-dominated pollination syndrome. It seems to have been then overlooked that the mites are possibly even more difficult to move between flowers than the smaller pollen grains. It is just curious logic, because a specialised pollination strategy surely implies any degree of associated structural adaptations relating to separation of even families. Smith in both that paper, and in S.Afr.J.Bot. 58,90, 1992, claimed either that Poellnitzia had a unique floral morphology or that “strong macromorphological evidence” separated it from other Aloid genera. Obermeyer-Mauve (Bothalia 11:119, 1973) wrote a short paper on the small flowered Aloineae which is also quite erroneous (as evident in Bayer, Nat.Cact.Succ.J. 27:78, 1972, and refuted in Bayer, 1976). She tacitly supported a suggestion made by Parr (Bull.Afr.Succ.Plant Soc. 6:145, 1971) that Astroloba and Poellnitzia be included in Haworthia. She fallaciously argued for the inclusion of Chortolirion in Haworthia as well. Rowley (Cact.Succ.J.Gt.Br. 43,2, 1981) actually validated his proposal that Poellnitzia belonged in Aloe. Smith wrote another paper (Hazeltonia 2:74, 1994 – in which no less than 43 literature references, but excluding the text-book by Jeffrey, are cited) saying…”morphological differences .. still readily serve to distinguish..” Poellnitzia, and that based on these it should form a distinct entity as a separate unispecific genus.

These articles by Smith et al, Smith and Manning, and Obermeyer-Mauve just demonstrate the ease with which the whole classification system can be disrupted, and the massive effort which is then required to restore a status quo. Neither protagonist is required to display any particular degree of expertise, and the botanist claims higher moral ground on the basis of science which has no more substance than that of religion. The actual unstated problem these botanists have is the embarrassment of monotypic (euphemistically now “unispecific”) genera. The word genus implies more than one anyway, and it would be more honest just to acknowledge the impracticality of unispecific genera, than parade a host of fallacious arguments to obscure them. The other even more unfortunate aspect is that all this fault-ridden argument passes over the heads of the unsuspecting community that is interested in the plants and wants to use the names. All that actually happens is that credibility becomes a very rare commodity in plant classification and nomenclature. If this is what professionals do, what can be expected from amateurs? The foundation is secured for the most uninformed and ignorant to parade as taxonomists – and they do.

The question of pollination comes into question because of the latest indignity. This is the transfer, described above, of the species to Astroloba by G.F.Smith and J.C.Manning published in Bothalia 30.1:53 (2000). There are many in congruencies in this half-page of text, and many more pages have to be filled to clarify this issue than should ever have been necessary.

The following claims are made in the Bothalia article:-

“Vegetatively Poellnitzia shows…

a. – close affinity to some species of Astroloba“.
Experienced and skilled botanists have confused Astroloba and Haworthia (at the Pretoria herbarium Astroloba muricata was identified for Mrs Bohnen as Haworthia venosa), and Aloe and Haworthia (in the Compton Herbarium there is a description by a very experienced botanist, of a new species Haworthia natalensis, attached to a specimen of Aloe aristata). P.V. Bruyns also brought me the first specimen of Haworthia pungens as an Astroloba. Many years ago I visited Miss Grace Britten at the Albany Museum and showed me a plant from the Longkloof. My reaction was..”But that looks like Poellnitzia!” The discovery of Haworthia pungens demonstrates that there is this similarity of an Haworthia to Poellnitzia as well.

b. – and “accords completely with the genus in its tubular actinomorphic flower with included stamens”.

“Actinomorphic” and “zygomorphic” are terms used to distinguish structural differences in flowers. “Actinomorphic” means the flower is radially symmetrical and can be bisected into two equal symmetric halves in more than one plane. “Zygomrphic” means that the flower can only be so bisected in one plane – usually the vertical. I have made this point for Haworthia and Astroloba before. The flowers in these genera are equally zygomorphic. There is no basic structural difference in respect of symmetry. What then is the case in Poellnitzia? The fact is that Poellnitzia has the zygomorphic floral structure of most of the Alooideae. There are two upper outer petals and one lower outer petal. There is one upper inner petal and two lower inner petals. The margins of the outer petals are fused to the midribs of the inner petals, but the lines of fusion are clearly visible on the outside of the tube so-formed. The raceme is borne horizontally and the individual flowers are borne secundly erect. Smith in Bothalia 25.1:35 (1995) has properly described this, and states that the upper-third of the tube is decurved. This is absolutely correct and also contributes to the fact that the flower cannot then be actinomorphic!. If the flower is properly examined it can be seen that it is in fact zygomorphic. Only the tip of the flower with the connivent petals, is actinomorphic… see point 3.

Curiously the zygomorphy of the white flowered Aloe albida is occasionally introduced as a red herring in the discussion of the classification of Haworthia. This zygomorphy is actually not in the least structurally different to that in the flower of many other Aloe species, including the four Reynold’s placed with A. albida in the Graminaloe and which he described as “regular” (actinomorphic!). The tube in A. albida is curved upward.
“A close relationship (between Poellnitzia and Astroloba) is supported by a preliminary survey of… aglycones in the roots of .. Alooideae”.

How close? What is the position of Astroloba to the three Haworthia subgenera? Or to Chortolirion or Gasteria? This is not evidence as it is presented here.
…”careful examination (sic!) of the flowers, shows that the aestivation is in fact imbricate and the species thus differs from Astroloba only in the more or less horizontal racemes…”.

This is not what careful examination will show. Careful examination will show that in Astroloba the upper inner petal includes the two lower slightly revolute inner petals. For this reason alone the flower can never be actinomorphic. In Poellnitzia, the margins of the inner petals are separate for their entire length. Sectioning the tube at any point in its length will show assymetry and the fact of zygomorphy as in Astroloba. The cross-section is that of an inverted triangle with the base shorter than the two sides. If the base of the flower is examined from the rear of the flower, the perigon can also seen to be slightly egg-shaped with the thin end of the “egg” below. The perigon is not round and thus cannot be actinomorphic. This is true for Astroloba, and also for Haworthia subgenus Robustipedunculares. To suggest that there is actinomorphic accord as described in point 1, is thus fallacious. Further, the stated similarity of tubular structure is absurd given the tubular nature of the flowers in the Alooidea generally.

Rowley (1976) refers to my claim that “the flower in Astroloba is positively not actinomorphic”, writing that to say so is.. “to imply that no Aloineae at all are 100% actinomorphic”. I implied and stated no such thing. I explained the position with regard to structural symmetry or assymetry as above. The only implication is that which the facts of structure have for the truth and practicality of classification. The fact that Aloe albiflora is actinomorphic (the flower is not even really tubular) is a problem of its own as is Poellnitzia and many other aloid oddities.
“These floral adaptations are now recognised as part of the syndrome of sunbird pollination…”

The authors are getting to the point where they eventually will claim that Poellnitzia is best treated as a species of Astroloba addapted to pollination by sunbirds. The fact is that Smith and Manning fail to cite a long paper in Taxon 41:437 (1992) by Smith, where Smith virtually excludes birds as pollinators. A curious point too is that the Poellnitzia flower (like Microloma) in contrast to most Aloe, Gasteria and the general red-flowered, bird-pollinated species has secund (erect) flowers. The others have pendulous flowers. Is Aloe albida then just a anthophorine bee-pollinated Haworthia.
… “and the resemblance between the flowers of Poellnitzia and the bird-pollinated genus Microloma (Asclepiadaceae) is particularly striking”.

This is a remarkable statement. It is to my knowledge only Microloma tenuifolium which has a very bright shiny red flower, and it is not always this red. I would not personally in my wildest dreams have considered the flowers of these two genera from vastly different families as comparable even on the most careless examination. One could say that Microloma is best treated as belonging in the Alooidae because it is also adapted to pollination by sunbirds. But, worse. Is Microloma pollinated by birds? The Asclepiads have a complicated pollination mechanism generally and most are insect pollinated. The flowers of Microloma sagittatum are not even red and the tube is not comparable to that of Poellnitzia. They are pinkish and even partly white. Microloma massonii has a small greenish brown obscure flower.
“In the wild the species is visited by Lesser Double Collared sunbirds…”

This statement is followed by an unlikely description of how the birds insert their beaks and tongues into the flower. Smith gave a very comprehensive account of the possible pollination of Poellnitzia in his Taxon (1992) account where he concluded that bird pollination was improbable. In my own experience I have often seen Poellnitzia in flower in the wild. It flowers in March-April. These are hot dry months when the veld offers very little nutriment to anything and flowers of any kind are few. In response to Smith and Manning I have made a point of examining Poellnitzia in the field and my observations suggest that Smith should stay with his 1992 more careful thoughts. The only prospective pollinator or mite “phoretor” (carrier) was the large Camponotus ant as observed by Smith. The flowers were tended by individual ants which walked over the tops of the flowers and carefully “palpated” the flowers with their feelers. The ants would often pay particular attention to the base of the tube and we (my wife and I) saw that occasionally a single flower would bear a small redder abraded spot to which the ants returned. The would place their mandibles on this point while palpating the flower with their feelers. The tube was never penetrated and more significantly over the period of two weeks, we never found any nectar in the flower tubes. This is not like Haworthia which can literally ooze nectar.

There are two complications. One may favour bird pollination and the other ant/mite pollination. In my field observation (early April when it is still hot and it was very dry too) seed-set was intense. But it was irregular. There was a syndrome of racemes with complete seed-set of the many consecutively maturing flowers from base to apex. There were racemes where most of the lower flowers had set seed, with the upper approximate third not doing so. There were racemes with seed set on the lower flower points, a long gap of 8-10 flowers points which were sterile and the end ones still bearing flowers and buds. There were also a few racemes in full flower and bud. Something had thus interfered with seed set over a period of about two weeks. The other interesting point is that there is barely any opening at all between the connivent tips of both sets of petals. The stigma is borne 3-4mm below the tip of the tube, surrounded by the anthers. Pollen can occasionally be seen outside the barely open slits between the connivent petal tips. Smith reasonably suggests that mites are responsible for this. Are they also responsible for carrying pollen back into the concealed stigma? I brought flowers back with me and placed them in water. A few days later I observed some yellow mites on the outside of the flowers. They were very active and appeared to be seeking re-entrance into the flowers which they could not achieve. I had to pry apart the petal tips fro the mites to enter. Yet in the flowers there were many mites. How did they get there? The ants are thought to need a nitrogen source (feeding on bird droppings for example). Do they use the mites as a food source and do they transport eggs or young mites between flowers? If this flower is adapted to a bird-pollination syndrome it is sending a strong message to birds that they are not welcome. One might as well say that the damage starlings and weaver birds do to Aloe flowers is strong evidence that the flower is adapted to bird pollination!
“The sugar composition of the nectar of Poellnitzia also refelects a shift from insect- to bird pollination…”

How does when deal with this rationale? One could use the same statement to separate the species from Astroloba. But I do not think when treating genera, one can exclude other genera from consideration in this manner and with this kind of argument.
“Adaptations for specialised pollination strategies alone are insufficient grounds for recognition of genera..”

The grounds for recognition of genera are as cited from Smith et al from Jeffrey. Smith and Manning have produced no facts by which Poellnitzia can be more closely related to Astroloba than to any of the other Aloid genera. All the facts have been distorted in order to reach an end, which is obviously to exterminate unispecific genera. Any structural character can, if so desired, be construed to have some or other adaptive significance. It is absurd to suggest that if these adaptations are directed at pollination, they are less significant for classification purposes. There is no connection of this kind whatsoever. A point is missed which I cannot adequately explain. Poellnitzia rubriflora demonstrates a uniformity which could even suggest that it is vegetatively propagated. I have yet to see two clones which I think I could consistently separate. I have even collected what I thought was a smaller form, only to find it grow out to equate others in cultivation.

It is actually very difficult to refute argumentation of this class against the general background of the treatment of the Alooideae by systematists and would-be systematists. I say this because Rowley (1976) also contributes to the debate with an odd statement.. :”Very often one finds that minute floral differences are linked with supporting vegetative characters…”. He is in fact referring to my recognition of the subgenera of Haworthia on the basis of floral characters. He is actually quite wrong and is repeating the attitude which Dr L.E. Codd expressed in private communication to me when questioned about the amalgamation of the smaller Aloid subgenera. Dr Codd said… “We are ignoring the small differences”. This is not true. Because, firstly, these writers are confusing the small size of the flowers with the significance of the differences. Were the flowers large, the differences would be more unquestionably seen to be large. The irony is that here in Smith and Manning, we have the case that more obvious large differences are discounted. Secondly, when a judgement of similarity or dissimilarity is made, the observer in fact makes a decision and then seeks a rationalisation. What are currently seen as Haworthia are three subgenera which have historically been judged to be similar and the rationalisation has never been properly constructed or questioned. The floral differences have been minimised. Capsule and seed structure has been ignored, and no attention has been paid to anatomy or things like root morphology. I have not done so either. The reason for this is that it is already obvious to me that the rationale already exists for the classification that I adopt viz. 3 subgenera, and one that I would propose viz. 3 genera. It is thus pointless to pursue other goals when proper attention is not paid to evidence already presented and which alone constitute grounds for the recognition of genera. But I do not think genera are real taxa in the sense that I see species to be. I feel that a classification arrived at over a period of 200 years has an accrued value. It is pointless to make changes which are not properly weighed against that value.

Rowley ends by sensibly making the point that the question of genera is artificial. I am very sceptical of his sanguine hopes for some future ideal solution when the … “time will be right”. To assume that botanists using sophisticated technology (DNA, molecular, cellular etc) or statistical methodology (cladistics and multivariate analysis) are going to produce better classifications will be unwise. Botanists have not effectively used the available simplest of evidence to classify or discuss the classification of the Alooideae, and amateurs are exacerbating the situation. Botanists are playing with technology, and with classification to display the results of this game. Amateur and impostor taxonomists are playing with classification to display their grasp of technology and knowledge. The motives are questionable, and the results prove it.