9. Haworthia cymbiformis (Haw.) Duv., Pl.Succ.Hort.Alenc. :7(1809).  Bayer :110(1976).  Bayer :35(1982).  Scott :91(1985).  H. concava Haw. Revis. :58(1821).  Aloe cymbiformis Haw., Trans.Linn.Soc. 7:8(1804).  Sims, Bot.Mag. 1:21,t.802(1805).  Salm-Dyck, Monogr. 11:t.1(1840). Type: Not preserved.  Neotype (designated here): icon, t.802, Bot.Mag..  Epitype (ex B&M): Walmer, Port Elizabeth, Smith 2844 (NBG):  H. planifolia Haw., Phil.Mag. 44:282(1825).  H. cymbiformis var. planifolia (Haw.) Baker JLinn.Soc. 18:209(1880).  Aloe planifolia (Haw.) Salm-Dyck, Monogr. 11:t.2(1840).  Type: icon. t2, Salm-Dyck, Monogr. 1840:  H. cymbiformis var. angustata V.Poelln., Feddes Repert.Spec.Nov. 45:166(1938).  Type:  Discovery region unknown, T.Foster.  Not preserved:  H. cymbiformis var. angustata fa subarmata idem. 45:166(1938).  Type: Rocklands, Adelaide, W.E.Armstrong in Triebn. 1187.  Not preserved.  Neotype (designated here): CAPE-3226 (Fort Beaufort): Rocklands, W.E.Armstrong in Smith 2801 (NBG):  H. cymbiformis var. compacta Triebn. idem. Type: Cape, west of Peddie, Mrs G. McLaren in Triebn. 1148.  Not preserved.  Neotype (designated here): CAPE-3327 (Peddie): W Woolridge, Peddie (-AB), Bayer 4648 (NBG).  H. planifolia var. exulata V.Poelln., Feddes Repert.Spec.Nov. 43:93(1938).  idem. 45:162(1938).  Type: Cape, Ubi?, C.H.Woolley in Long 392.  Not preserved.  Neotype (designated here): Photogr. H.G.Fourcade of Long 392 (NBG):  H. planifolia var. planifolia fa agavoides Triebn. et V.Poelln., Feddes Repert.Spec.Nov. 45:162(1938).  Type: Cape, Fort Beaufort, W.E.Armstrong in Triebn. 1169.  Not preserved.  Neotype (designated here): CAPE-3226 (Fort Beaufort): Sulphur Baths (-DC), Bayer 4655 (NBG):  et fa alta ibid. Type: Cape, Grahamstown, Mrs Helm in Triebn. 851.  Not preserved:  et fa olivacea ibid.  Type: Cape, Quagga West, Mrs Helm in Triebn. 853.  Not preserved:  et fa robusta ibid.  Type: Cape, Baakens Valley, Mrs I.King 100 in Triebn. 1066.  Not preserved.  Neotype (designated here): CAPE-3325(Port Elizabeth): Baakens Valley (-DC), Smith 3894 (NBG):  et var. incrassata V.Poell. idem. 45:163(1938).  Type: Cape, Kowie River, Mrs Archibald 335 in Long 446.  Not preserved:  et var. sublaevis V.Poelln., Kakteenk. 6:67(1938).  idem., Feddes Repert.Spec.Nov. 45:163(1938).  Type: Albany district, Mrs Britten in Triebn. 940.  Not preserved:  et var. longifolia Triebn. et V.Poelln. idem.  Type: Cape, Grahamstown, Mrs Helm in Triebn. 864.  Not preserved:  et var. longifolia fa calochlora ibid.  Type: Cape, Port Elizabeth, Mrs Helm in Triebn. 941.  Not preserved:  H. planifolia var. poellnitziana Res. ibid. 48:133(1940).  Type: ex Hort., Hamburg.  Not preserved:  H. lepida Smith JS.Afr.Bot. 10:21(1944).  Type: Cape, Albany district, between Carlisle Bridge and Fort Brown, Smith 5066 (NBG).

cymbiformis: boat-shaped.

Rosette to 130mm φ, partially stemmed, proliferous.  Leaves broad ovate to lanceolate, flat to slightly concave, generally <1/3 as thick as wide, usually opaque, green turning yellowish to pink hued on exposure.  Inflorescence to 250mm, 10-15 flowers, lax.  Flowers white.

1982 – This name has been used interchangeably with H. planifolia for a very widespread and variable species growing in the Eastern Cape.  This species has practically the same distribution range as H. cooperi but unlike that species is an opaque, truer green colour.  It forms dense mats with a fairly superficial root system and grows on rocky slopes and krantzes along rivers and streams.  H. planifolia Haw. was only described in 1825 and was stated to be less proliferous and with distinctly flatter leaves.  Baker placed the two species together and a study of Von Poellnitz’s and Smith’s records plainly show that neither could clearly distinguish these two species.  Von Poellnitz in fact described H. planifolia var. transiens and H. cymbiformis var. translucens from the same locality, and even stated that he did not know how to clearly separate the two species.  Von Poellnitz in Feddes Repert. Spec. Nov. 41:199(1937) records H. cymbiformis from Graaff Reinett (Triebner 861) but in 1938 (ibid. 45:163) describes H. planifolia var. incrassata from the same collection.  Berger, in a masterly piece of irrationality, justified separate sections for the two species.  Smith transferred H. incurvula V.Poelln. from the section Muticae to the section Obtusatae, on the basis of observed continuity with H. cymbiformis.  In Flowering Plants of Southern Africa (pl. 356, 1929) a plant identical to H. incurvula is illustrated under the name H. cymbiformis var. planifolia (Haw.) Baker.  It can only be concluded that one name be retained for the entire complex and this is H. cymbiformis.

Uitewaal (1948) put forward the view that H. obtusa was not related to H. cymbiformis at all, but that is was an earlier name for H. pilifera (here a synonym of H. cooperi).  He based his observation on a colour plate in the Kew herbarium, and goes as far as to say that Haworth’s original description is faulty.  The strongest argument in favour of Uitewaal’s contention is the historical one.  None of Haworth’s species can be referred to Baker’s H. cooperi and this is hard to concede knowing how widespread and variable that species is.  The name H. obtusa is here considered to be probably synonymous with H. cymbiformis and should probably be rejected as superfluous and as a source of confusion.  The distribution range extends from the Bashee River in Transkei, to Prince Alfred’s Pass in the west, and northwards to Fort Beaufort.  Variation within individual populations is small but no two populations are quite the same.  Thus there is scope for a good many more varieties than previously recognised.  However, to stay within manageable limits, only the one really different form viz. fa ramosa with the elongated stem, and varieties not limited to single populations, are upheld.  The var. incurvula is apparently restricted to Plutosvale which is a contradiction, but, because Smith discusses continuity with H. cymbiformis (atypical for Smith and not borne out by his field records), because of the possibility of a relationship with H. translucens subsp. tenera, and because of similar forms occurring in the Humansdorp area, it is maintained.  The var. transiens is really only more translucent, and at the type locality also larger, than incurvula, otherwise they may have been considered synonymous.  The var. umbraticola is a distinctive variety from the Swartwaterpoort west of Alicedale and northeastwards to Fort Beaufort. The leaves are very obtuse and round in cross‑section.  The fenestrate blunt tips with shining pellucid areas separated by dark green lines, make it a most attractive variety.  H. cymbiformis occurs primarily in the summer rainfall area and is very easy to grow in cultivation.  It proliferates to an rapidly and should be exposed to at least some direct sunlight to prevent bloating and excessive softening of the plants.  This species is also particularly prone to losing its roots with overwatering so it is also essentially a winter‑growing species which likes at least some resting period during the summer.

1999 – Sims in 1805 commented on the name H. cymbiformis, saying ‘Its name (with too much latitude by the way) is taken from its leaves’, which prompted Haworth to change the name to H. concava.  Von Poellnitz repeatedly exclaimed at the poor coloration of these plants in Europe as opposed to the colours which the plants develop under good light, when they are indeed very attractive.  The two illustrations in Salm Dyck’s Monograph of Aloe cymbiformis and Aloe planifolia are really very similar in relation to the variation within this one species.  One would think thus that a more sensible and conservative approach to species would have been achieved a great deal earlier than it has.  It is not obvious to which species H. cymbiformis is most closely allied, but it does seem to be closely associated with H. cooperi.  Bayer and Pilbeam may have been in error in their treatment of Uitewaal’s re-appraisal of H. obtusa Haw. as it is not easy to find the obvious field counterpart.  The solution suggested below may not be the most appropriate.  Col Scott also regarded obtusa as a variety of H. cymbiformis but unfortunately seems to have illustrated the H. cooperi variant.  A population on the Kat River near Fort Beaufort sampled by Scott, supports his argument and the decision taken here.  The explanation regarding the var. incurvula in the 1982 Handbook is poorly constructed.  What was meant, was the recognition of the variety despite its limited distribution, because of the evidence claimed by G.G. Smith, and because of its possible transitional nature towards H. gracilis.  H. lepida is regarded as a variant of H. cymbiformis because it could not be re-located despite a detailed description of the one locality where it was recorded.  A collection further to the east does not appear to differ dramatically from an already wide range of forms.

Breuer and Metzing nominate a specimen as a neotype when the early illustrations are excellent and are really the basis for the correct and historical application of the name. Haworth cited and accepted the illustrations in Botanical Magazine as well.

a. var. cymbiformis.
The typical species is considered to comprise the main body of the species which occurs from Port Elizabeth, eastwards to East London and inland to Adelaide and Committees on the Fish River.  Plants in this area tend to have broad, flat smooth leaves without spines.

Distribution: 3226 (Fort Beaufort): Near Alice (-DB), Smith 5635 (NBG); Kat River, W. Alice (-DB), Smith 105 (NBG); 13km S. Fort Beaufort (-DC), Smith 5617, 5617a (NBG); Rocklands (-DC), W.E.Armstrong in Smith 2801 (NBG); Sulphur Baths (-DC), Smith 2795, 3826, 7371 (NBG), Bayer 4655 (NBG); W. Sulphur Baths (-DC), Bayer & Bruyns 6593 (NBG).  3227 (Kingwilliamstown): Debe Nek (-CC), Britten in PRE 39472; Fort Murray Bridge (-CD), Smith 3111, 3317, 3576 (NBG); Bridal Drift (-DC), Smith 2806 (NBG); Umdanzini (-DD), Smith 5336, 5337 (NBG).  3325 (Port Elizabeth): Kranspoort, W. Patterson (-BC), Bayer 4549 (NBG); Slagboom Dam (-BC), Branch 37 (NBG); Below old Fort (-DC), Smith 5040 (NBG); Walmer (-DC), Smith 2790, 2844 (NBG); Baakens Valley (-DC), Smith 3894 (NBG), Paterson 155 (BOL).  3326 (Grahamstown): Fish River valley (-AA), Dyer 4549 (PRE); Carlisle Bridge (-AA), Smith 5567, 5597 (NBG); Cloudlands (-AB), Britten in BOL71308; Howiesonspoort (-AD), Smith 105a, 439, 909, 2843b, 5302, 5305 (NBG); The Fort (-BA), Courtenay-Latimer in Smith 5066 (BOL, PRE), Smith 5066 (NBG, PRE); Between Carlisle Bridge and Fort Brown (-BA), Smith 5066 (NBG); E. Fort Brown (-BA), Bayer 1620 (NBG); Ballinafad (-BB), Smith 3365, 5404a, 5405a (NBG), Bayer 4652 (NBG); Horseshoe (-BB), Smith 2765, 3124, 5311, 5312 (NBG); Committees (-BB), Compton 17834 (NBG), Smith 5071, 5404 (NBG); 3km W. Committees (-BB), Smith 3382 (NBG); S. Committees (-BB), Smith in NBG322/40; Giffords Bush (-BB), Smith 1997 (NBG); 30km E. Peddie (-BB), Smith 3515 (NBG); Grahamstown (-BC), Britten 218 (PRE); Grahamstown (-BC), Britten in PRE 34903; Grahamstown (-BC), Britten in PRE 39480; Grahamstown (-BC), Curator PRE Bot. Garden in PRE 26300; 24km S. Grahamstown (-BC); Stayner in KG254/70; Blaaukranz (-BC), Smith 5544 (NBG); Fernkloof (-BC), Smith 5629 (NBG); Mt. Drive (-BC), Britten (NBG), Dyer 6 (BOL); Kowie (-BC), Dyer 7 (BOL); Kariega (-DA), Branch 43 (NBG); Bussock Farm (-DA), Smith 768 (NBG).  3327 (East London): Peddie (-AA), Smith 3112 (NBG); Paradise, Wooldridge (-AA), Smith 2800, 5602, 5671 (NBG); Gqora (-AA), Smith 5777 (NBG); W Woolridge, Peddie (-AB), Smith 3113, 3115 (NBG), Bayer 4648 (NBG); Kapp-Fish confluence (-AC), Bayer 4654 (NBG); Kaffirdrift (-AC), Smith 655, 5255, 5261, 5262, 5263, 5274 (NBG); E. Fish River (-AC), Smith 5256 (NBG); Wesley to Falloden (-AD), Smith 3157 (NBG); Chalumna (-BB), Smith 571, 2786, 3089, 3391, 3392, 5131, 5313, 5400, 6199 (NBG).

Inadequately located: Cape, Marloth 6284 (PRE), Long in Smith 3892 (NBG), Smith 398, 2804, 3112 (NBG), Stellenbosch 3889, 5560; ex hort, Ross-Frames in NBG76/48; Zaysdorp, NBG101825, Warden (BOL); ex hort, Whitehill (NBG); Albany, Dyer in NBG1806/30, Britten in NBG 734/31, Luyt in NBG309/45.

b. var. incurvula (V.Poelln.) Bayer
:124(1976).  Bayer :36(1982).  Scott :94(1985).  H. incurvula V.Poelln., Feddes Repert.Spec.Nov. 31:85(1932).  H. cymbiformis var. planifolia Flow.Pl.S.Afr. 9:t356(1929).  Type: Grahamstown, Plutosvale, Mrs E. Ferguson.  Not preserved.  Neotype (B&M): Plutosvale, Britten (BOL71307).

incurvula: curved inwards.

As stated above, there is not very much substance to this variety either.  It is smaller than the norm and relatively narrow-leaved.  It has been collected many times from the same locality.  The flower is very similar to that of H. gracilis var. minima, but Smith nevertheless maintained that it is continuous with H. cymbiformis.  What he said was this “At the type locality near the top of the slope of a very deep valley, this plant is hardly variable, but as one descends, the plant changes, and at a point in the valley about a mile from the type locality, they are in appearance approaching H. cymbiformis.” The correct way to establish this is with physical evidence, and this is absent.

Distribution: 3326 (Grahamstown): Fish River Ridge (-AB), Britten in PRE 34959; Plutosvale (‑BA), Smith in NBG340/35 (BOL), Smith 5402 (NBG), Britten 11 (BOL), Britten in PRE 34909, in PRE 39477, Long 1029 (PRE), Dyer 3 (BOL), Dyer 2082 (PRE), Fourcade 99 (NBG), Smith 915 (NBG); Road to Plutosvale (-BA), Reynolds 2948 (PRE); 16km from Grahamstown (-BA), Erens in PRE 34910; S. Plutosvale (-BA), Smith 5402, 5403 (NBG); S. Hunts Drift (-BB), Smith 5741, 6508 (NBG).

Inadequately located: Albany, Dyer in NBG1803/30 (NBG), Smith 1123 (NBG), Luckhoff in NBG 404/34, Smith in NBG340/35, Britten in NBG740/31 (BOL); ex hort, NBG705/30, NBG1110/36, Luyt in NBG302/45, in NBG306/45, Whitehill (NBG).

c. var. obtusa (Haw.) Baker
JLinn.Soc. 18:209(1880).  Bayer and Pilbeam, Cact.Succ.J(U.S.)46:166(1974).  Scott idem. 48:260(1976).  Scott :93(1985).  H. obtusa Haw., Phil.Mag. 46:282(1825).  Type: Cape ex hort Kew.  Not preserved.  Lectotype (designated here): Icon Kew library:  H. umbraticola V.Poelln., Kakteenkunde 9:134(1937).  H. cymbiformis  var. umbraticola (V.Poelln.) Bayer :164(1976).  Bayer :36(1982).  Type: Swartwaterpoort, near Adelaide, W.E. Armstrong.  Not preserved.  Lectotype (B&M): icon (B):  H. hilliana V.Poelln., Desert Pl.Life 9:103(1937).  H. umbraticola var. hilliana  V.Poelln., Feddes Repert.Spec.Nov. 44:234(1938).  Type: Cape, ex hort Kew.  Not preserved:  H. obtusa var. pilifera fa truncata Jacobs., Handb.Succ.Pl. 2:574(1960).  Type: ?.  Not preserved.

obtusa: with obtuse leaves.

As discussed under H. cooperi, there is a real problem in understanding the situation concerning the blunt leaved forms of that species and of H. cymbiformis.  I have seen very dark green forms with brown venation in cultivation which I would not relate to H. cooperi.  The highly translucent forms from Fort Beaufort are greener and seem to be continuous with H. cymbiformis through the Swartwaterpoort.  Certainly the plant mentioned above from the Kat River, are very close indeed in appearance to H. cooperi var. obtusata except for the brownish-green coloration.  It has been noted that plants of that variety from Inverbolo (Upper Kei River) tend to become greener under lower light intensities.  The herbarium record shows that the typical form of H. cymbiformis also occurs along the Kat River.  Very robust forms in fact occur 10km S. Fort Beaufort.  H. cooperi var. pilifera is also present in very close proximity.  At Kagasmond and probably at Olifantsbeen nearby, the plants are very similar to the rather longer leaved forms, still with obtuse tips, which are found in Swartwaterpoort near Alicedale.  A Kagasmond collection is also cited under var. dielsiana, which is indicative of the difficulty in using dry herbarium material to make indisputable identifications.  G. Marx has made a collection from Swartwaterpoort in which the plants have the coloration of H. cooperi but the boat-shaped leaves of H. cymbiformis.

Distribution: 3226 (Fort Beaufort): Kagasmond (-CD), Bayer & Bruyns 6562 (NBG); Olifantsbeen (-CD), Krynauw in NBG268/43 (NBG); S. Adelaide (‑CD), Krynauw in NBG 67996; Blinkwater (-DA), Smith 6195 (NBG). Scott 600 (PRE), Bayer 4651 (NBG); Kat River, 10km SE. Fort Beaufort (-DC), Scott 1065 (PRE); Rocklands, Adelaide (-DC), Smith 2801 (NBG).  3325(Port Elizabeth (-BB), Bayer 4653 (NBG).  3326(Grahamstown): Thornkloof (-AA), Bayliss in KG382/76; S. Alicedale (-AC), Bayer 4650 (NBG); Alicedale (-AC), Britten in PRE 34905.

d. var. ramosa (Smith) Bayer comb.nov. 
H. cymbiformis fa ramosa (Smith) Bayer :149(1976).  Bayer :34(1982).  H. ramosa Smith, JS.Afr.Bot. 10:22(1940).  Type: CAPE-3427 (Peddie): Wooldridge (-AB), Smith 3168 (NBG).

ramosa: branched.

Consistency is a difficult ideal and this variety does not  conform well with the principle of substance.  It is only known from a long crescent-like rock-face north of Woolridge where plants vary from the normal stemless to increasingly stemmed plants on a gradient from west to east.

Distribution: 3327 (Peddie): Wooldridge (-AB), Smith 3168 (NBG, PRE), Smith 3105 (NBG), Bayer 4648 (NBG); NW. Wooldridge (-AB), Smith 3168, 3169 (NBG).

e. var. reddii (Scott) Bayer comb.nov. 
H. reddii Scott, Cactus Succ.J(US) 66:182(1994).  Type: CAPE-3226( Fort Beaufort): Waterdown Dam, Cathcart(-BB), Scott 8968 (PRE).

reddii: for Dr V.B. Reddi.

Plants from this population at Waterdown Dam have been known for a long time and identified (Bayer,1982) as possibly intermediate between H. cymbiformis and H. batesiana.  Col Scott similarly mentions both species names in his discussion and the matter appears rather problematic.  The population was portrayed as a depauperate one, at least in my perception, with quite considerable variation between the few odd plants at the site.  The few clones sampled in 1982 were not as robust as the one described by Scott, who also thought the population to consist of but a few individuals.  A re-visit to the site by Bayer and Bruyns in 1996 revealed that the south-facing cliff alongside the dam is clothed with huge numbers of plants.  The huge clumps are just like those of H. cymbiformis.  Some of the plants have very distinctive translucent dots and lines, others unmarked and uniformly opaque.  The floral characters mentioned by Scott are not definitive but the flowers do appear to have strongly colored veins.  At the time (Bayer :30, 1982) the species batesiana was still upheld, although not positively associated with field populations.  Since then the range of H. marumiana  has been shown to extend to at least Queenstown and further north.  The Cathcart population does not seem to belong there although the block patterning in the leaves does suggest this.

At the same time the known range of H. cymbiformis has been extended by P.V. Bruyns to the upper reaches of the Black Kei much nearer to Queenstown.  This is on a south-facing cliff at the farm Turnstream.   Here there are small forms which distantly resemble H. lepida as described and illustrated by Smith, and also suggesting the same possible link with H. marumiana and certainly with reddii.  This is particularly so because the Waterdown Dam is on the upper reach of the Black Kei.  A little to the southeast is also a population of larger plants on a very high west-facing cliff which are however with spined margins, and thus apparently belonging to the var. setulifera.  The block-patterning in the leaves of these two populations is not as marked as the Waterdown plants.  There is an old Galpin collection from even nearer to Queenstown.  Col. Scott was not correct in his opinions about H. marumiana var. batesiana and it is evident from his book that H. marumiana was unfamiliar to him too.  The population at the Waterdown Dam does not seem substantial enough in terms of viability nor range, to justify species rank.  What is needed is cognitive exploration of potential new localities to substantiate already expressed opinions.  In this case there are new records for var. batesiana as well as for H. marumiana and H. cymbiformis.

Distribution: 3226 (Fort Beaufort): Waterdown Dam, Cathcart(-BB), Scott 8968 (PRE); Klipplaat Dam (-BB), Bayer 4649 (NBG).  3227 (Kingwilliamstown): Gwatyn farm (-AB), Galpin 8280 (PRE); Turnstream (-AB), Bayer & Bruyns 6572 (NBG).

f. var. setulifera (V.Poelln.) Bayer comb.nov. 
H. planifolia var. setulifera V.Poelln., Kakteenkunde 5:54(1938). idem., Feddes Repert.Spec.Nov. 45:163(1938).  Type: East London, Stellenbosch 3332.  Not preserved.  Neotype (designated here): CAPE-3228(East London): Kwelegha bridge (-CC), Smith 5257 (NBG):  H. cymbiformis var. obesa V.Poelln. idem. 45:166(1938).  Type: Idutywa, Bashee River (-BA), G.W. Reynolds.  Not preserved.  Neotype (designated here): CAPE-3228(Butterworth): Xobo River, E. Idutywa (-BA), Smith 7796 (NBG).

setulifera: bearing small bristles.

In the earlier handbooks, the representative specimen for the species was poorly chosen from a Nahoon River specimen. It is apparent from Haworth’s acceptance of the Bot. Mag. illustration that the more western forms such as those at Baakens River or Howiesons Poort, would have been more in keeping with the original circumscription of either the typical species or H. planifolia. East and north of East London, H. cymbiformis begins to develop a thicker and shorter, more deltoid leaf and the teeth become markedly spined and von Poellnitz’ name is re-instated for this variety.

Distribution: 3128 (Umtata): Mquanduli (-DC), Walker in NBG2271/27 (BOL).  3227 (Kingwilliamstown): Highclere (AB), Bayer & Bruyns 6573 (NBG);  Inverbolo (-BC), Bruyns (NBG); Bluewater (-DA), Smith 676 (NBG); Near Komgha (-DB), Marloth 6510 (PRE); Tangla River (-DC), Smith 3881, 3882a (NBG); Newlands Location (-DD), Smith 3510 (NBG); Pump Stn. (-DD), Smith 611, 2785, 3096 (NBG); Kings Farm (-DD), Smith 3071 (NBG); McCleantown (-DD), Smith 3126, 3127, 2883, 2884, 2885, 3882 (NBG); Slippery Drift (-DD), Smith 3122, 3125, 3134, 3134a (NBG); below Horseshoe (-DD), Smith 3097 (NBG); Fort Jackson (-DD), Smith 3133, 3388, 3389, 3389a (NBG).  3228 (Butterworth): Willowvale (-AD), Luyt in NBG180/42, in NBG17/46, in NBG57/46, NBG341/38; Xobo River, E. Idutywa (-BA), Smith 7796 (NBG), Smith in NBG341/35 (BOL), Reynolds 2850 (PRE), Reynolds 146 (BOL); Xobo (-BA), Reynolds in PRE 39470, in NBG660/38, Smith 2760, 2796 (NBG); S. Mooiplaas (-CC), Bayer 1706 (NBG); Gonubie (-CC), Smith 6826 (NBG); Kwelegha bridge (-CC), Smith 5257 (NBG); Kwelegha (-CC), Smith 5251, 7183 (NBG); Kei River (-CC), Holmes (BOL).

g. var. transiens (V.Poelln.) Bayer
:162(1976).  Bayer :36(1982).  H. planifolia var. transiens V.Poelln., Feddes Repert.Spec.Nov.45:163 (1938).  Type: Cape, Prince Alfred Pass, Archibald 327.  Not preserved.  Lectotype (B&M): icon (B):  H. cymbiformis var. translucens Triebn. et V.Poelln. idem. 45:166(1938).  Scott :94(1985).  Type: Cape, Prince Alfred Pass, Lategan in Triebn.1137.  Not preserved.  Neotype (designated here): CAPE-3323(Willowmore): Prince Alfred Pass, Smith 5709 (NBG):  H. cymbiformis var. multifolia Triebn. idem. 45:166(1938).  Type: Uitenhage, W.E. Armstrong.  Not preserved.  Neotype (designated here): CAPE-3325 (Port Elizabeth): Hellsgate, UItenhage (-CB), Smith 2794 (NBG):  H. cymbiformis var. brevifolia Triebn. et V.Poelln. idem.: 165(1938).  Type: Cape, Hellsgate, Uitenhage, Mrs I. King in Treibn. 1068.  Not preserved.  Neotype (designated here): CAPE-3325 (Port Elizabeth: Hellsgate, Mrs I. King in Smith 2756 (NBG).

transiens: changing into.

In the publication where Von Poellnitz published his two varieties, he actually expresses doubt about identifying three parent species. Yet he states that H. planifolia var transiens is between H. planifolia and H. cymbiformis, but nearer to the former ‘to look at’. Two pages further he describes a var. translucens of H. cymbiformis from the same locality. In this work the circumscription of var. transiens is widened to include the Uitenhage elements at Hellsgate.  There are variants in the Gamtoos Valley (eg. Andrieskraal) which are reminiscent of this highly translucent form, and it is not certain just how these variants relate to either H. cymbiformis or H. gracilis. The element H. gracilis var. picturata is applied to those forms of that species which are very similar to H. cymbiformis variants. The ‘clear’ way (a comment made by Dr M. Hayashi) in which var. transiens is related to H. mucronata is symptomatic of the alternative solutions available in classifying Haworthias, which are not always clear.  It is generally understood that the Little Karoo species are continuous with the Eastern Cape species and this is commonly expressed in Von Poellnitz identifications, and also a feature of Col. Scott’s species distributions.

Thus H. mucronata can be allied with equal facility to either H. cymbiformis or H. cooperi, when in fact in the field it is more intimately related to H. decipiens. The location of this note is a powerful reminder that distinctions between species are highly blurred and that alternative solutions are possible.

Distribution: 3323 (Willowmore): Prince Alfred Pass (-CC), Smith 5624a, 5709, 5624, 5624a (NBG), Taute (BOL), Fourcade 3490 (BOL); Oskloof (-DA), Bruyns 7077 (BOL); Luiskraal (-DA), Forrester 399 (NBG).  3324 (Steytlerville): Scholtzberg (-CA), Van Jaarsveld 7804 (NBG); Andrieskraal (-DA), Fourcade 176 (NBG).  3325 (Port Elizabeth): Hellsgate, Uitenhage (-CB), Smith 2794 (NBG); Hellsgate, Mrs I. King in Smith 2756 (NBG); Kemachs (-CB), Smith 905 (NBG); Near Port Elizabeth (-DC), Smith 3892 (NBG); Stayner in NBG46/56, Stayner in KG80/70 (NBG), Taute in NBG1283/36, in NBG468/37, Taute .

Inadequately located: ex hort, NBG486/30.

Haworthia cymbiformis (Haw.) Duval [as Aloe cymbiformis Haw.] Curtis’s Botanical Magazine, vol. 21: t. 802 (1805) [S.T. Edwards]	Haworthia cymbiformis (Haw.) Duval [as Haworthia cuspidata Haw.] Addisonia, vol. 23: t. 741 (1954-1959) [M.E. Eaton]