48. Haworthia koelmaniorum Oberm. & Hardy, Fl.Pl.Afr. :f1502(1967). Bayer :127(1976). Bayer :68(1982). Scott :33(1985). Type: TRANSVAAL‑2529 (Groblersdal): Groblersdal (‑AB), Hardy & Mauve 2267 (PRE).
koelmaniorum: of Mr and Mrs Koelman.
Rosette stemless, slowly proliferous, 5-7cm φ. Leaves 14-20, dark brownish-green, opaque, somewhat recurved, scabrid with small raised tubercles, margins and keel with small spines. Inflorescence slender, to 35cm. Flowers slender, tepals fused and with revolute tips.
1982 – This unusual species is the most northern record in the genus, at Groblersdal in the Transvaal. It is closely related to H. limifolia and is more prominently tuberculate. Although described as having erect leaves, in cultivation they can recurve as in H. limifolia. The spiral arrangement referred to in the original is not a distinctive feature. In H. limifolia and H. venosa the angle of leaf divergence is above 160 degrees and a twin leaf spiral is often prominent. The fact that leaves are vertically coincident in threes is purely fortuitous in such an arrangement where seven vertical parastichies will often also be detectable. It is reported to have been collected out at its original habitat, but unconfirmed reports indicate that it is known elsewhere. It is a slow-growing species which can be propagated from whole leaves.
1999 – Little further evidence is available concerning this species. Although it has obviously been collected many times by collectors since first reported, nothing has found its way into the formal record. This is a tragedy, and a repeat of the often mindless collecting frenzy in the genus Aloe during the period 1950 to about 1970. The end-product of the average collection is a small contribution to a compost heap, and good permanent record should be regarded as some tithe towards the cost of a depleted natural stock. Col. Scott also described H. mcmurtryi, which I regard as a smaller variant. The similarity to the retuse Haworthias is really fortuitous and seeking such a relationship is neither realistic nor necessary. Dr Hayashi’s suggestion (in ms.) that H. koelmaniorum is linked to H. venosa var. tessellata via a northwestern corridor is slightly more probable, but flawed for other reasons. It has a more minutely papillate and scabrous texture than H. limifolia, and also more so than H. venosa.
a. var. koelmaniorum.
Distribution: 2529 (Groblersdal): Groblersdal (‑AB), Hardy and Mauve 2267 (PRE), Hardy in KG275/70 (NBG).
Haworthia koelmaniorum var. koelmaniorum JDV88/4 south of Groblersdal. Easily grown from leaf and seed so there should be no need for reliance on field collected plants. 
Haworthia koelmaniorum var. koelmaniorum JDV92/3 Groblersdal. Selection is also necessary to get speciemens which are really attractive.
b. var. mcmurtryi (Scott) Bayer comb.nov. Haworthia mcmurtryi Scott, Cact.Succ.J(U.S.) 56:69(1984). Scott :140(1985). Type: TRANSVAAL-2529(Groblersdal): Loskop, SW. Dam (-AB), McMurtry 5247 (PRE).
mcmurtryi: for Mr D. McMurtry.
This is smaller and with more prominent surface markings than the typical species. It has recently been observed and photographed by J.M. Esterhuizen at a number of sites in the Globlersdal area, and he confirms continuity with H. koelmaniorum.
Distribution: 2529 (Groblersdal): Loskop, SW. Dam (-AB), McMurtry 5247 (PRE).
Haworthia koelmaniorum var. mcmurtryi JDV93/50 southwest of Groblersdal. A smaller variety with shorter, squatter leaves. 
Haworthia koelmaniorum var. mcmurtryi JDV91/7 Loskop Dam. Unlike H. limifolia the leaves can be obsurely windowed. 
Haworthia koelmaniorum var. mcmurtryi JDV91/7 Loskop Dam. Also quite variable regarding tubercles and venation.