New species in Petalacte (Compositae) and Struthiola (Thymelaeaceae)

Petalacte vlokii Hilliard species nova ab affini P. epaleata Hilliard & Burtt foliis primariis maximis 5-8 x 1.2-2mm (nee 8-15 x 2-4mm), persistenter lanatis (nee supra glabrescentibus), floribus masculis tubo corollae 2.3-2.7mm longo (nee 3-3.5mm), ovario abortivo 0.4-0.8mm longo (nee O.15-O.5mm) distinguitur. Perennial herb up to 60cm tall, loosely branched, young parts white-woolly, leafy. Leaves (largest primary ones) 5-8 x 1.2-2mm, narrowly panduriform to elliptic, apex somewhat tapered or rounded, shortly apiculate, base broad, half-clasping, margins strongly revolute, both surfaces persistently white-woolly. Capitula heterogamous, c.4 x 2mm, many in crowded corymbose panicles c. 15mm in diameter at tips of branchlets. Involucral bracts in c.4 series, outer 2 series reduced, brown, shaft cobwebby, tip scarious, inner 2 series with greenish shaft and suborbicular opaque snow-white tip c.l x lmm. Receptacle smooth, without paleae or teeth. Flowers 4-6,0l ( -2) female, 3-5 functionally male, corolla tube in female flowers 2mm long, in functionally male flowers 2.3-2.7mm. Ovary 1.2-1.5mm long, aborted ovaries 0.4-0.8mm, sparsely hairy. Achenes not seen. Pappus bristles very briefly fused in a ring, contorted, barbellate below, barbs longer at the tips, acute, opaque white. Type: South Africa, Cape, Montagu division, 3320 CA, on^crest of Waboomsberg, on farm Pypsteelfontein, 4200ft, 29 ix 1990, Vlok 2416 (holo. E; iso. BOL, K).

cited Esterhuysen 24503 (above) as Petalacte epaleata, but when Mr Vlok, in 1991, sent to Edinburgh a specimen that matched it precisely, it became clear that an undescribed species was under examination. The two species differ in habit: P. vlokii is much more loosely branched than P. epaleata (Mr Vlok described it as a perennial reseeding herb, Miss Esterhuysen as a shrub, soft but becoming woody). Petalacte epaleata is a well branched shrublet. The leaves of P. vlokii are smaller than those of P. epaleata, less markedly apiculate, the margins more strongly revolute, the midrib scarcely or not visible below (readily seen in P. epaleata even though covered in wool) and persistently lanate on both surfaces (glabrescent above in P. epaleata). Vlok 2416 has fully developed flowers (these of Esterhuysen 24503 are only in bud); these prove to be smaller than those of P. epaleata (corolla tube of functionally male flowers 2.3-2.7mm long, not 3-3.5mm) and are seated on generally longer aborted ovaries (0.4-0.8mm, not 0.15-0.5mm). The compound inflorescences also appear to be smaller, about 15mm in diameter, not up to 35mm. Mr Vlok described the involucral bracts of P. vlokii as 'snowy-white, the florets yellow', but in a specimen of P. epaleata collected by him on Rooiberg Mountain {Vlok 2039) he noted 'flowers and involucral bracts pure white'. That there is a real difference in colour unrelated to age of the flowers needs confirmation. The flowers of P. coronata (L.) D. Don are said to be purple (and appear so when dried).
It is a pleasure to associate Mr Vlok's name with this third species of Petalacte; his deep interest in the Cape flora has led to many discoveries on the little-botanized Cape fold mountains. Petalacte epaleata is now known from three sites: the type locality on the Little Swartbergen nearLadismith (3321 AD) and two collections from the Rooiberg (3321 DA). The sites of/ 3 , vlokii lie further west, on the Waboomsberge (3320 CA) and the Langeberge near Swellendam (+ 3320 CD). Whether or not the two species are truly allopatric can only be resolved when the flora of these southern ranges is much better known.
I also want to thank Mrs P. Lorber, Curator of Bolus Herbarium, University of Cape Town, for gifting to Edinburgh a photocopy and fragment of the holotype of P. epaleata.

Struthiola anomala Hilliard, species nova a S. angustiloba Peterson & Hilliard habitu
(caulibus multis tenuibus simplicibus vel subsimplicibus e caudice stolonifero, nee suffrutice bene-ramoso), foliis juvenilibus magis pilosis, petalis rudimentariis pilis circumcingentibus carentibus (nee petalis c.0.75mm longis) distinguenda. Stock stoloniferous, many stems tufted from base, simple or subsimple, 1 5^0 c m tall, l-2mm in diameter near base, young parts with patent hairs c.lmm long, closely leafy, lower parts nude, rough with leaf scars. Leaves opposite, erect, imbricate, 9-11 x 2-2.5mm, lanceolate, acute, base slightly narrower, initially with white hairs c.lmm long on margins and midline below, eventually glabrescent. Flowers solitary in axils of leaves. Bracteoles c.4-5 x 0.5-0.8mm, linear-lanceolate, hairy as leaves. Calyx very pale creamy-or greenish-yellow, tube c.7-9mm long, narrowly cylindric, abruptly expanded in throat, lobes 4, c.2-3.8 x 0.5-1 mm, lanceolate-acuminate, glabrous or with a few hairs on swollen part of tube and near tips of the two larger calyx lobes. This interesting plant lends emphasis to the difficulty of generic delimitation within the ambit of Gnidia Linn. Dyer (1975, p. 393) adopted a broad generic concept for Gnidia, saying 'Consensus of opinion favours the reduction of the number of genera, mainly because of the variable nature of the differentiating characters', but gave no indication of the basis for that consensus. Moreover, neither Dyer's key nor his generic description of Gnidia make provision for Basutica E.P. Phillips and Pseudognidia E.P. Phillips, reduced by him under Gnidia. Dr B. Peterson (1982, p. 81) endorsed Dyer's view, and made several new combinations in Gnidia. Earlier, E. P. Phillips (1944, p. 61) had discussed the problem and concluded that' the number of stamens and the number of the calyx lobes, in conjunction with other considerations such as distribution and general similarity, must be taken as characters to separate the genera Gnidia , Lasiosiphon, Arthrosolen and Struthiola.' Phillips then went on to describe three new genera, Pseudognidia, Basutica and Struthiolopsis. Dyer and Peterson recognized only Struthiola and a heterogeneous Gnidia, and so harked back to Gilg's treatment in the Pflanzenfamilien (1895). The plant described here as Struthiola anomala bears so striking a resemblance to S. angustiloba (known only from the Natal Drakensberg between Royal Natal National Park and Giant's Castle) and S. tetralepis Schltr. (SW Cape) that the pragmatic course is to give the species a name in Struthiola despite its lack of well-developed petals encircled by stiff hairs, which will lead to difficulty in using currently available keys to genera. In Phillip's key (1944, p. 67 and 1951, p. 527), Struthiola keys out under ' 2 . Stamens as many as the calyx lobes' and then (lead 3) 'Petals surrounded by stiff hairs.' Struthiola anomala, lacking hairs, would then have to be distinguished (at lead 4) from Pseudognidia as they share the characters of opposite leaves and eight petals (only rudimentary in Struthiola anomala); the inflorescence in Struthiola is a spike, in Pseudognidia a pseudohead.
In Dyer's key (1975, p. 392), under the lead 'Petals free ', Struthiola is distinguished from Gnidia by four versus eight stamens; under the lead 'Petals absent' (which might be taken to apply to S. anomala ) , choice is restricted to Gnidia.
I sent the late Dr Peterson a specimen of Struthiola anomala in 1985, when the plant first came into my hands; he suggested that the petals and hairs had either been eaten by beetles or that the flowers were abnormal (note that he accepted it as a species of Struthiola). Neither alternative satisfied me, and I therefore asked Mr Tony Abbott to try to re-collect the plant during the course of his botanical expeditions to Ngele Mountain. After considerable effort over several seasons he succeeded; I am very grateful to him. He sent both dried and spirit material as well as photographs showing great stands of it on rocky grassy slopes. The material shows conclusively that the petals are reduced to rudiments similar to the fleshy bases from which the petals of S. angustiloba arise, and that the bristles are aborted, which characters, together with differences in habit and indumentum, mark it as a species distinct from S. angustiloba.