a aR OTR Males (iui THN BAC at vaya Baht ce Nahata it ~ a sh V ‘ Ceeey es ih , ¥ n a “se ptt) at ee a t f ; 4 H 5 } j ; : fi F : eae | Digitized by the Internet Archive in 2016 with funding from South African National Biodiversity Institute Libraries https://archive.org/details/journalofsouthaf22unse A ie ied i PLAN: é Catalorue » j 31-12. 19 ce | ~Bip 959 | Agee Ng = LANTIcUND | me TRIE TEP CONDE VOLUME XXII. | 1956. THE JOURNAL tf oes SOUTH AFRICAN BOTANY j gH PUBLISHED UNDER THE AUTHORITY OF THE TRUSTEES OF THE NATIONAL BOTANIC GARDENS OF SOUTH AFRICA KIRSTENBOSCH, NEWLANDS CAPE PROVINCE } s EDITOR H. B. RYCROFT, M.Sc. (S. Afr.), B.Sc. For. (Stell.), Ph.D. (Cape Town). DIRECTOR OF THE NATIONAL BOTANIC GARDENS, HAROLD PEARSON PROFESSOR OF BOTANY IN THE UNIVERSITY OF CAPE TOWN. THE JOURNAL OF SOUTH AFRICAN BOTANY. VoLtuME XXII. 1956 CONTENTS A New CycaD FROM THE CAPE Province. By R. A. Dyer, D.Sc. (8.Af.), F.R.S.8.Af. (With Plates I-III.) ee We a at i DISTRIBUTIONAL, ECOLOGICAL AND PHYTOGEOGRAPHICAL OBSERVATIONS ON THE FERNS oF SoutH West ArFrica. By EH. A. C. L. E. ee M.Sc. (S.Af.), D.Phil. (Oxon.), F.L.S. Two New SpPEcIES AND Two NEw Vana ¢ OF Anon FROM TmACARGLR, By G. W. Reynolds, Hon. D.Sc. (Cape a F.L.S. (With Plates IV.-X.) A New ALOE FROM Keene Conon. By G. Ww. Remnolds Bion: D. Se, (Cape Town), F.L.S. (With Plates XI and XII.) A New Mars Erica FROM WEMMERSHOEK. By Captain T. M. Soliton R. Ne Hon. D.Sc. (Cape Town). ae Be STUDIES ON THE EMBRYOLOGY AND Ram AnmONEnEEES OF Soune Asmara GENERA OF THE HAEMODORACEAE: DILATRIS Berg. AND WACHENDORFIA Burm. By Miriam P. de Vos, D.Sc.(Stel.) Notes on SoutH AFRICAN MARINE ALGAE. IV. By G.F. paper tass Ph. D. CRINUM MACOWANII. By Miss I. C. Verdoorn. (With Plates XIII-XVI.) A New ALOE FROM THE EASTERN TRANSVAAL. By G. W. Reynolds, Hon. D.Sc. (Cape Town), F.L.S. (With Plates XVII and XVIII.) Tue SoutH AFRICAN SPECIES OF A1zoaceaE. III. GaLenta. By R.S. Micon, M.A. (Cantab.), D.Sc. (Edin.), F.R.S.S.Af. THE IDENTITY OF ALOE PARVULA BERGER, WITH A NEW SHRNONIT Ne Naw AND A NEW SPECIES FROM Mapacascar. By G. W. Reynolds, Hon. D.Sc. (Cape Town), F.L.S. (With Plates XVII-XXITI.) THREE NEWSPECIES AND ONE NEw VARIETY OF ALOEFROM KARAMOJA, Tannin. By G. W. Reynolds, Hon. D.Sc. ae ae F.LS. pe Plates X XIII- XXVIIL ) ‘ : te a a Book REVIEW: Methods of Surveying and ers poem By Dorothy Brown (Professor H. B. Rycroft) THE FLORAL STRUCTURE OF NAGELOCARPUS SERRATUS (THN, ) Barocre By Margaret R. Levyns. é ‘ : : THE IDENTITY OF ALOE ABYSSINICA Larue AN, ELEGANS TODA AND A CAMPERI SCHWEINFURTH. By G. W. Reynolds, Hon. D.Sc. ns Town), F.L.S. (With Plates XXIX-XXXV.) : MaRINE ALGAE OF INHACA ISLAND AND OF THE INHACA /PRRTRAST A. I, By Win, Edwyn Isaac, B.Sc., Ph.D. (Wales), F.R.S.S.Af. (With Plates XX XVI- xX) : 6 oe oe o a fe INDEX OF PLANT Nearest Vor. XXII PAGE 65 161 195 JOURNAL OF SOUTH AFRICAN BOTANY VOL. XXII. Published: 31st DECEMBER, 1955. A NEW CYCAD FROM THE CAPE PROVINCE.* By R. A. DyEr. (National Herbarium, Pretoria.) (With Piatrs I—IITI.) Encephalartos arenarius 2. A. Dyer sp. nov., affinis #. latifrondi Lehm. plantis minoribus foliolis laxioribus tenuioribus glabris differt. Planta subacaulis vel usque | m. alta, e basi frequenter ramosa, truncis eylindricis usque 25 cm. diam. Folia 1—1-25 m. longa, juniora plus minusve glauca, circiter 30-juga; foliola mediana leviter imbricata, oblongo-lanceolata, 12—16 cm. longa, 2-5—4 cm. lata, acuta sessilia, basi 1—1-5 cm. lata, margine supra plano vel per rariter 1-lobato, margine infra lobis 3 rariter 4 subtriangularibus pungentibus armato, nervis parallelis satis tenuibus. Strobilus femineus 35—45 cm. longus, circiter 20 cm. latus, glaucus; facies squamarum 5 cm. late, 4 cm. alte, 2 em. prominentes, umbilico truncato 10—12 mm. lato, semina scarlatina, 5 em. longa, 2—2-5 em. diam. Squamae strobili masculini 3—3-5 cm. longae, 2-25—2-5 cm. latae, umbilico 1 cm. prominente. CAPE PROVINCE. Alexandria Div., Kaba Valley, Archibald 5420, Dyer 5439, 5441 in Nat. Herb. Pretoria (type), 5443: cultivated at N.B.G., Kirstenbosch, Henderson 2017; Cape Town Municipal Garden, Lewis. Stem simple or more often branched from the base into clumps, up to 1 m. long and at this stage usually reclinate. Leaves 90—125 cm. long, usually somewhat glaucous when young, apex generally recurved; petiole 15—20 cm. pulvinus 3—4 em. broad, with brown wool; leaflets in about 30 pairs on mature plants, sometimes remaining slightly glaucous, flat *Communicated at the meeting of the South African Association for the Advancement of Science in Grahamstown, July, 1955. 1 2 The Journal of South African Botany. or somewhat twisted, pungent, reduced in size towards apex and to petiole to one prickle, those in lower half more spaced and spreading and less lobed, those in the upper half closer and forming a V; medium ones 12—16 em. long, 2:5—4 em. broad, 1—1-5 cm. broad at attachment, closely veined on lower surface; upper margin entire or only very rarely with a tooth; lower margin usually 3-, rarely 4-lobed, reduced to 2—1 on lower leaflets; the lobes in the same plane as the body of the leaflet or only slightly twisted, pungent; leaflets of seedlings deeply tridentate or trilobed at the apex with or without an extra tooth on the upper margin and with 1—2 extra teeth or lobes on the lower margin. Female cone solitary, with stout peduncle 4—8 cm. long, mainly oblong and narrowed abruptly to apex, 35—45 em. long, about 20 cm. broad, more or less glaucous; median scales with exposed face mainly 4-sided with occasionally the top and bottom angles blunted, 5 cm. broad, 4 cm. high, protruding about 2 cm., slightly ridged from the angles to the central umbilicus and with lesser irregular ridges between, otherwise smooth and not tuberculate, with or without extremely minute papillae; umbilicus slightly concave, 10—12 mm. diam.; surface smooth: stipe pale yellow, angled, 4—5 cm. long, with sinus arms up to about 2 cm. long; seed coral-red, up to about 5 cm. long, 2—2-5 cm. broad. Male cone not seen intact, apparently smaller than female; median scales 3—3-5 cm. long, 2-25—2-5 em. broad, beak projecting | cm. beyond sporangial surface, with its face 7—10 mm. broad, slightly angled, nearly flat, glabrous. In 1852 Karl Zeyhert compiled a note on Encephalartos species of the eastern Cape in which he referred to L. altensteinii being near the Bush- man’s River not far from Olifantshoek and “‘E. tridentatus also, but sparingly, in that neighbourhood”. It is evident that the name Z. tri- dentatus is not correctly applied and one can only guess that Zeyher had in mind a plant related or idential to the species described above. There is, however, a bewildering collection of living plants, some with tridentate leafiets, in the Alexandria division on the right bank of the Bushman River and only a specimen could prove exactly what Zeyher referred to. Apparently he left no such record. Relatively little further field work was done on the genus in the Cape until the building up of the living collection of Encephalartos at the National Botanic Gardens, Kirstenbosch, early in this century. But even when Hutchinson and Rattray? monographed the Cycadaceae for the Flora Capensis in 1933, the present species was not recognised. In 1945, however, Henderson*® drew attention to the distinctiveness of the plant (Henderson 2017) in cultivation at Kirstenbosch and a specimen Tidmarsh sn. which Hutchinson and Rattray had identified as ZL. latifrons without citing it. A New Cycad from the Cape Province. 3 In 1953 Dr. E. E. A. Gledhill reported the discovery of a colony of unidentified Encephalartos on the inland sanddunes of Alexandria in the Kaba Valley. An inspection was carried out with her in 1954, which eventually led to the correlation of the above cited records. Specimens of H. arenarius were reported by the owners of the farms to be fairly widely distributed in the scrub bush and it was also stated that a large number had been removed for cultivation within recent times. The habitat in coastal bush on stabilised sand-dunes is strongly reminiscent of the con- ditions under which ZL. ferox (= EH. kosiensis) grows in its native haunts in Portuguese East Africa and Zululand. The closest affinity of #. arenarius is certainly with EH. latifrons and it might not always be possible to distinguish detached leaves or leaflets of each. On the other hand in the presence of full records there should be no doubt. E. latifrons produces robust erect trunks up to 6 feet and more tall, the leaves are rigidly curved, the leaflets are finely pubescent when young, densely overlapping and strongly veined below; female cones are large with the scales coarsely papillate and scaly pubescent on the surface. In £. arenarius the plants are less robust, the decumbent erect stems rarely attain a length above the ground of 3 feet, the leaves are comparatively long, the leaflets are glabrous and moderately densely spaced, the female cones are smaller, the surface of the scales is slightly wrinkled, glabrous and glaucous. The Kaba Valley population of #. arenarius, judging from an inspection of up to 100 specimens in all stages of maturity, is uniform. No other species is known from the immediate vicinity and there is no indication of hybridisation. To return to the subject of the variation among specimens on the banks of the Bushman’s River, that locality has been investigated by several botanists, including Miss I. C. Verdoorn in 1947. Typical EZ. altensteinii is the only recognised species so far found in the immediate vicinity and, amongst the nondescript population, similarities were noted with it and E. lehmannvi Lehm. forma, EH. horridus Lehm. and E. arenarius which are known from within some miles’ radius. Two suggestions to explain the different combinations of characters have been put forward; (a) that the specimens constitute unstable parent stock from which one or more of the species referred to have evolved by processes of segregation and isolation and (b) that the variations are the result of hybridisation between specifically distinct groups. I am of the opinion that some at least of the specimens seen on the right bank of Bushmans River, about 2 miles from the mouth, are the result of hybrid- isation. On the other hand the variable stock of HZ. lehmannii in the Fish River Valley further east may well have been the source of E. horridus 4 The Journal of South African Botany. and H. Iehmannii as segregated out near Uitenhage and Steytlerville, respectively. The field work within recent years has brought to light many interesting facts but has not made the taxonomist’s task easier, as, theoretically, it should have done. We are confronted by more problems than ever before of a similar nature to those referred to above. REFERENCES. 1. ZeEyHER: Per President: Phytol. 4 (1852), 613. 2. HuTcHINSON AND Ratrray: Flora Capensis, 5 Sect. 2 suppl. (1933), 24-44. 3. Henperson, M. R.: Journ. 8.A. Bot., 11 (1945), 26. Puate I. Encephalartos arenarius. Cape Province, Alexandria Division, Kaba Valley, 5 miles inland from Woody Cape, F. Muller’s farm. Northern facing. Photo R. A. Dyer, 232/2. 23 February, 1954. = specimen No. 5439. = Gledhill 5420. Puate II. Encephalartos arenarius. Cape Province, Alexandria Division. Kaba Valley, 4 miles inland from Woody Cape on “Old” Muller’s farm, eastern facing (cone glaucous). Photo R. A. Dyer 233/2. = Specimen No. 544laQ. 23 February, 1954. 36, aT ae HOT: BA S5 best Gains bed na | T on Cs : CTU Puatre II. Encephalartos arenarius. Alexandria Division, Cape Province. About 5 miles from coast, steep slope facing north-east on margin of bush clumps, cone glaucous, leaflets green, not very rigid. E. E. A. Gledhill 5420 (Archibald). Dec., 1953. Photo H. King. “a DISTRIBUTIONAL, ECOLOGICAL AND PHYTOGEOGRAPHICAL OBSERVATIONS ON THE FERNS OF SOUTH WEST AFRICA. By E. A. C. L. E. ScHELPE. (University of Cape Town.) South West Africa is one of the arid and semi-arid regions of the world regarded by Christ (1910) as being almost without ferns. This impression can be regarded as accurate if one considers the territory as a whole, ranging from coastal desert, through vast areas of thorn savannah to oceasional, shaded, watered kloofs on mountainsides. However, it became evident to the author on a collecting excursion to the territory in July, 1954, that the majority of species of the fern flora of the region occurred among rocky terrain of scattered rock outcrops or of mountain ranges. The minority of the ferns of South West Africa occur around such terrain in comparatively level ground e.g. Ophioglossum polyphyllum, or in seasonally wet areas, e.g. Marsilea spp. Another feature is the apparent absence of “fern allies” such as Lycopodium, Selaginella, Isoetes and Equisetum from the flora of South West Africa. Nevertheless, the possibility of the occurrence of Isoetes in seasonally wet areas in this territory cannot be excluded in view of the drought resistance of some South African species of the genus and the fact that the author’s visit took place during the dry winter season. A significant number of species of the ferns of South West Africa belong to genera with a noticeable proportion of xerophytic species, e.g. Actiniopteris, Ceterach, Cheilanthes and Pellaea. (Christ, 1910). The occurrence of these genera in xeric habitats in Africa has also been noted in Ethiopia (Schelpe, 1953) and in the Sudan (Macleay, 1953). Species of Pellaea and Cheilanthes are also known to occur in xeric habitats in North America (Pickett, 1923; Pickett & Manuel, 1925) as well as species of Notholaena (Gams, 1938). The rainfall of South West Africa is markedly seasonal in the northern and north-eastern parts of the territory (summer rainfall) and erratic though largely seasonal in winter in the southern parts. Active growth of the fern sporophytes takes place during the rainy season or subsequent to erratic rainy spells. As could be expected, the xerophytic ferns remain dormant during the dry seasons. The purpose of this paper is, firstly, to record the available data on the known distribution of the South West African ferns to date, secondly, to o 6 The Journal of South African Botany. discuss their habitat range in the light of known facets on the autecology of xerophytic and other ferns, and, thirdly, to discuss the distribution of the ferns of this territory in relation to their distribution in Southern Africa and elsewhere. The nomenclature used in this paper is according to the “Annotated Check-list of the Pteridophyta of Southern Africa” (Alston & Schelpe, 1952) except where more recent taxonomic work is cited. Specimens collected on the author’s 1954 expedition to South West Africa are housed in the Bolus Herbarium, University of Cape Town and a second set is at the British Museum (Nat. Hist.), London. DISTRIBUTION OF THE SOUTH WEST AFRICAN FERNS. Although it is realised that the distribution ranges of some ferns in South West Africa may be extended by further collecting in the more remote areas of the territory, it seems likely that the distribution patterns of such species would not be altered seriously. During the author’s travels in South West Africa, which were limited by time, an attempt was made to collect and record all the species observed in the areas visited. These observations have been supplemented with records from herbarium speci- mens prepared by earlier collectors represented in South African herbaria. The areas investigated by the author were along a route from Good- house through Warmbad, Karasberg, Keetmanshoop, Windhoek, Okahandja, Usakos, “Ameib”, Omaruru, Otjiwarongo, Waterberg, Otjiwarongo, Otavi, Grootfontein, Namutoni, Okakuejo, Outjo, Okahandja, Windhoek, Karasberg to Viools Drift. The South West African ferns can be segregated into four convenient habitat groups. 1. Species occurring in rock crevices or around boulder bases on rock outcrops. 2. Species not essentially associated with a rocky environment, either in open semi-desert, or under the shade of bush. _ 3. Seasonally aquatic or seasonally hygrophilous species. 4. Floating aquatic species. Since the majority of the ferns of South West Africa occur around more or less isolated mountains or ranges of hills, the positions of these hilly or mountainous areas considered in this paper are indicated in Fig. 1. The known distribution of the majority of the non-aquatic ferns are given in Table I. Distributional, Ecological and Phytogeographical observations 7 on the Ferns of South West Africa. TABLE I. DISTRIBUTION OF NoN-AQuatTic SoutH West AFRICAN FERNS. Pellaea robusta . . wo las| a | 213 lol 3 s|sele]/3\e #23 2 eA a Sion ieee 3 C ist) >) n 2 a/§|23|$|2| 8 \28)2) 2] 2 Bl/M iss] s]o}/8 (85/9/54) 2 a B 3] Ss 5p S i 8 S mH See Se ate Bnei Si)! Be 1 8 aloleeql¢)/S6ieiIS8| sisal 8 5 a i} = aS mH 2 ise! |) ~~ AWS evs © | 15) |) © |ieeidz S| S |) Approx. annual rainfall (mm.) | 90 | 150 | 350 | 380 | 230 | 470 | 500 | 500 | 530 | 520 Ceterach cordatum + y)+]t+ + + + Pellaea calomelanos +] +yt+ a ce eo ed Cheilanthes dinteri +)/+y/4+s)+yH+ Notholaena marlothii =e | a ee ee ed Actiniopteris australis + ar Se il ate ||l ate Cheilanthes hirta aF sr Cheilanthes parviloba + | + | Notholaena eckloniana + Doryopteris concolor var. kirkii | + + Cheilanthes multifida | || ae Adiantum caudatum Ophioglossum reticulatum =F Pellaea viridis + + Pellaea goudotii + Dryopteris thelypteris var. squamigera + Dryopteris sp. aff. D. pseudoguentziana . . + Microlepia speluncae ap Notholaena rawsoni + Pellaea deltoidea F 8 The Journal of South African Botany. Etosha Pan {> OTAVI HILLS “@ GUCHAB MTNS ad WATERBERG @ OKAHANDIA HILLs ERONGO MTNS. Windhoeke 6 AUAS MTNS. P MTNS » Walvis Bay Luderitz AUS HILLS a KARASBERG Fie. 1. Outline map of South West Africa showing the position of mountainous or hilly areas considered in this paper. 1. Species occurring in rock crevices or around boulder bases on rock outcrops. (a) Low rainfall areas (less than 400 mm. average annual rainfall). Cheilanthes hirta Pellaea deltoidea Cheilanthes parviloba Pellaea robusta Notholaena eckloniana Notholaena rawsoni (b) Higher rainfall areas (more than 400 mm. average annual rainfall). Adiantum caudatum Pellaea goudotii Cheilanthes multifida Pellaea viridis Distributional, Ecological and Phytogeographical observations 9 on the Ferns of South West Africa. (c) Low and higher rainfall areas (less and more than 400 mm. average annual rainfall). Actimiopteris australis Doryopteris concolor var. kirkii Ceterach cordatum Notholaena marlothii Cheilanthes dinteri Pellaea calomelanos As it is well known that the microclimates of small areas in hilly or mountainous regions are extremely variable, especially in predominantly seasonally arid areas, the preceding grouping of species can only be regarded as approximations of the general climate the species cited are probably able to endure. 2. Species not essentially associated with a rocky environment. In the drier open parts of the territory only Ophioglossum polyphyllum (as construed by Pichi Sermolli, 1954) is known to occur in this type of habitat. The production of its fronds in this habitat appears to be sub- sequent to the occurrence of erratic but short rainy spells. The perennially moist shaded habitats below the south-eastern cliffs of the Great Waterberg present a totally different picture. In this shaded forest habitat in an area with an annual average rainfall of nearly 520 mm., a few species were seen which were not recorded by the author nor any other collector, elsewhere in South West Africa. These species include Microlepia speluncae, Dryopteris thelypteris var. squamigera and Dryopteris sp. aff. D. pseudoguentziana. 3. Seasonally aquatic or seasonally hygrophilous species. The specimens and data gathered by previous collectors in the higher rainfall areas in the northern and north-eastern parts of South West Africa during summer rainy seasons, indicate that species of Marsilea are not infrequent in this region. They appear to occur in areas which are either wet or inundated during summer rains. However, they are not evident during the dry winter season. Five species are known to occur in the northern and north-eastern parts of South West Africa, but the identity of three of the taxa is still in doubt. Marsilea nubica M. villifolia. M. pubescens Dinter M.S. (? M. macrocarpa var.). M. profundelaciniata Dinter M.S. (? M. biloba var.). M. sp. aff. M. subterranaea. The absence of records of Marsileas from the banks of the Etosha Pan was confirmed by the author in the course of observations along its southern limit between Namutoni and Okakuejo. It seems probable that the high salinity of the water near the banks of the pan would be beyond the tolerance of Marsileas. 10 The Juornal of South African Botany. 4. Floating aquatic species. Azolla pinnata was collected by Miss Stephens in the Chobe swamps in the Caprivi Zipfel. ECOLOGICAL ASPECTS OF FERN DISTRIBUTION IN SOUTH WEST AFRICA. The alternation of generations in the Pteridophyta render it imperative that the separate tolerances of the gametophytic and sporophytic plants towards environmental conditions be considered when discussing the reproduction and persistence, respectively, of fern species in a certain habitat. Holttum (1938) summarised this view effectively: “The existence of fern plants is ultimately dependant on the occurrence, at least seasonally, of conditions in which the gametophyte generation can live and develop and the youngest stages of sporophyte growth become established.” Climatic factors. - Seasonally arid habitats abound in South West Africa. For a species to occur in such habitats it is essential that the sporophyte is able to endure drought and that the gametophyte either is also capable of drought resistance or capable of rapid development to the stage of producing a young drought resistant sporophyte during short and erratic, favourably moist conditions. The capability of drought resistance of the rhizomes of sporophytes of the xerophytic species is not in question, considering the presence of large and extensive rhizome systems in the field. More significant is the ability of such rhizomes collected during the dry season to produce fronds on being transplanted into moist soil in culture. Hence, it would appear that the drought resistance of the gametophyte is a most important factor in the possible extension of fern communities in seasonally arid habitats, if such gametophytes are not capable of extremely rapid development during short favourably moist periods. The limited work done on the drought resistance of prothalli of xero- phytic ferns is relevant to the discussion at this stage. Pickett (1923) observed that the gametophytes of Cheilanthes gracillima survived nearly five months’ air-drying. Pickett & Manuel (1925) found that 75 per cent of the gametophytes of Pellaea atropurpurea and Pellaea glabella in cultures survived air-drying of five months’ duration. In these ferns it was also observed that the gametophytes and the young sporophytes survived drought equally well. The effect of light intensity on the drought resistance of fern prothalli was also studied by Pickett (1914) on gameto- phytes of Camptosorus rhizophyllus; the results indicated that the mature Distributional, Ecological and Phytogeographical observations 11 on the Ferns of South West Africa. prothalli could withstand “practically unlimited interrupted drought as found in nature, if not exposed to direct sunlight, producing sporophytes through fertilisation at the time of occasional showers.” In the drier areas of South West Africa the most obvious feature of the local distribution of ferns on rock outcrops is their occurrence, limited almost entirely to the south aspect slopes. This is significant in relation to Pickett’s findings of greater drought resistance of the prothalli of Campto- sorus in lower light intensities. Also, it can be expected that the more shaded south aspect habitats would remain moist for a longer period after rain as compared with the less shaded, warmer, north aspect slopes. Such moist conditions expected to continue for a comparatively longer period on south aspect slopes would allow the formation of young sporophytes, whether the gametophytes be apogamous or not. Another adaptive feature of a xerophytic fern observed by Pickett (1923) in Cheilanthes gracillima, namely the retention of spores during the dry season and their release in the subsequent more moist season, has not yet been investigated among the xerophytic species of the Southern African ferns. Irregular germination of spores was observed by Pickett (1914) in Camptosorus rhizophyllus. Such a feature might be of importance if it occurs widely among xerophytic ferns in a habitat liable to receive short, erratic showers, occasionally followed by conditions dry enough to damage very young prothalli. The two species of Ophioglossum (O. polyphyllum, O. reticulatum) known to occur in South West Africa are presumably also able to perennate through dry seasons by means of their rhizomes. Whether these popula- tions have attained their present size and range largely by vegetative means or by a drought-resistant though long-lived gametophyte remains to be determined. The confinement of the majority of the ferns of South West Africa to rocky outcrops on hills and mountain slopes requires some explanation. Vast areas of thorn savannah in which no ferns were seen, occur in this territory. Judging from the frequently loose sandy substratum in such vegetation it seems that this substratum is too unstable and too rapidly drained or dried to offer a firm and moist surface of sufficient duration for a polypodiaceous fern prothallus and subsequent sporophyte to become established. Comprehensive temperature data are not yet available for South West Africa. However, it was gathered from local information that from the Waterberg northwards frosts do not occur and that in the vicinity of Windhoek frosts are infrequent. The vegetation on the mountains in the south of the territory, especially at higher altitudes, can be expected to 12 The Journal of South African Botany. experience frost during the winter. The occurrence of tropical species such as Microlepia speluncae, in the Waterberg can be correlated partly with the non-occurrence of frost and partly with ample supplies of water in this locality. Conversely, the higher temperatures prevalent in the north-eastern parts of the territory might exclude certain species with more temperate requirements. Edaphic factors Gams (1938) listed some fern species with apparent edaphic preferences for soils ranging from acidic to alkaline in reaction. Since ferns occur on rock outcrops in South West Africa ranging from sandstone and granite to dolomite, data is given in Table II on the presence of various species occurring about outcrops of different rocks. As it is known that species of angiosperms and bryophytes exhibit different tolerances to chemical edaphic conditions in different rainfall areas, approximate data on the range of precipitation in which the species occur in South West Africa is given. TABLE It. SUBSTRATA OF SOME SouTH WeEsT AFRICAN FERNS. D 2) : é z 5 2 Approximate mean ~ 18 a £ | annual rainfall range 3 | 9 = | < | of known habitats in So |S R Qa South West Africa. Ceterach cordatum oe “Se Lk 90—510 mm. Pellaea calomelanos _— 150—510 mm. Cheilanthes dinteri ne sie + 150—470 mm. Notholaena marlothii a = > 350—530 mm. Actiniopteris australis’ .. 3 — 350—530 mm. Adiantum caudatum Fe +530 mm. Cheilanthes hirta .. Par a + 150—230 mm. Cheilanthes multifida os +530 mm. Cheilanthes parviloba ie a + +350 mm. Doryopteris concolor var. kirkii . . - 230—510 mm. Notholaena eckloniana .. a + 90—150 mm. Pellaea goudotii +510 mm. Pellaea viridis 500—510 mm. It appears that Actiniopteris australis and Notholaena marlothii are able to tolerate both acid and alkaline substrata over a wide precipitation range. In South West Africa, Adiantum caudatum and Cheilanthes multifida have only been found on dolomite outcrops; however, the former has been seen growing as an epiphyte and forest-floor inhabitant in presumably Distributional, Ecological and Phytogeographical observations 13 on the Ferns of South West Africa. acid humus at the Victoria Falls and even the northern deltoid fronded form of Cheilanthes multifida (Schelpe, 1954) occurs on non-alkaline sub- strata in the Transvaal. The possibility of local ecotypes cannot be excluded, but it would seem that micro-climatic conditions are more important than edaphic factors in the limitation of the distribution of these two species. The rest of the species listed in Table II do not appear to tolerate substrata derived from dolomite, a view borne out by their known edaphic tolerances elsewhere in Southern Africa. Factors affecting seasonally aquatic species. The seasonally aquatic or seasonally hygrophilous Marsileas form an ecologically important group of the fern flora of South West Africa. Their ability to perennate through seasonally arid periods by virtue of their sporocarps has been recognised by Gams (1938) and others. Although it has been observed by the author that sporocarp formation can be induced by lowering the water level below ground level in cultures of Marsilea macrocarpa, data is lacking on the rapidity of formation of new sporo- phytes after extrusion of sporangia from the sporocarps at the onset of the rainy season. PHYTOGEOGRAPHICAL ASPECTS OF FERN DISTRIBUTION IN SOUTH WEST AFRICA. From the floristic point of view, the most striking feature of the pteridophyte flora of South West Africa is the apparent absence of members of the Psilotales, Lycopodiales, Selaginellales, Isoetales, Equisetales and Marattiales, even from the occasional moist and subtropical localities in the north-eastern ranges. Considering the range of habitat exhibited by members of these groups elsewhere in Southern Africa, the only locality in South West Africa seen by the author where they could possibly be expected to occur is on the south-eastern slopes of the Great Waterberg. It would appear that members of these groups failed to reach such isolated favourable areas in their more recent migrations through the African continent. Nevertheless the possibility of the occurrence of seasonal drought resistant species of Isoetes in South West Africa cannot be excluded. The distribution patterns in Southern Africa and elsewhere of ferns occurring in South West Africa fall into four more or less distinct groups. 1. Endemic species. 2. Species with a predominantly north-western Cape distribution. 3. Species with a predominantly southern African distribution. 4. Temperate and tropical outlier species. 14 The Journal of South African Botany. ! ‘ I a ‘ 1 Be 1 1 ¢ , t J 1 1 ? 2. 1 q ¢ i Ors = 4 1 i, eat} asl od cA 7A 1 U of iv ‘ i \ ¢ Fic. 2. Distribution map of Cheilanthes dinteri (+) and Notholaena marlothiu (@) in Southern Africa. 1. Endemic species. The only two clearly distinct, endemic, non-aquatic species occurring in South West Africa are Cheilanthes dinteri (Schelpe 1954) and Notholaena marlothu (Fig. 2). It seems possible that these two species originated in this area from populations of Cheilanthes multifida and Notholaena eckloniana, respectively, which apparently now have a very restricted distribution range in South West Africa. 2. Species with a predominantly north-western Cape distribution. This group includes Notholaena rawsoni, Pellaea deltoidea and Pellaea robusta. (Fig. 3). The limitation of the distribution of these three species to Namaqualand in the north-western Cape Province and to the southern parts of South West Africa poses a problem which is not easily explicable. They are undoubtedly ferns of arid terrain usually associated with rocky habitats. Superficially it would appear from mean monthly rainfall data Distributional, Ecological and Phytogeographical observations 15 on the Ferns of South West Africa. that the region occupied by these species is liable to a sparse winter rain- fall. An examination of the rainfall data for Namaqualand shows that such means give a false picture as short and erratic rainy spells are hable to occur at any time of the year, though the majority of these rainy spells occur during the winter months. 3. Species with a predominantly southern African distribution. Species included in this group are Ceterach cordatum (sens. lat.), Cheilanthes parviloba and Notholaena eckloniana. ‘The specimens of Ceterach cordatum seen in South West Africa all belong to the more dissected variety namaquensis (P. & R.) which is widely distributed in the drier areas of the north-western, western and southern Cape Province. The infrequent populations of Cheilanthes parviloba and Notholaena eckloniana in South West Africa seem to be much isolated outliers of predominantly southern Cape Province and eastern Southern African Feat] ¢ bene . ‘ ‘ { \ 4 i] x ‘ ~ 5 t Sh Us ' ‘ d ) ty 4 1 Aes, Af 1 [a ! ‘ 4 ‘ ’ 5 1 Ss56 oa acco . ¢ i (SS , 1 aoe ” 1 4 x , ‘ { oP 1 { ¢ ’ ' i] ? ry 1 ¢ ' 4 oc t ree ‘ ' , ; yo ' 1 1 ; ) “seers o bt ale -” va es 14 oy a , “7” teers ai if 7 ‘ ' U eect é ! 1 (ee s EO, 1 fers } BY a <e. 1 J Fic. 3. Distribution map of Nothdlaena rawsonii (+-) and Pellaea deltoidea (@) in Southern Africa. 16 The Journal of South African Botany. species respectively. If the increasing aridity of the central depression of Southern Africa during recent geological time is accepted, these out- | liers could be regarded as relics of far wider distribution patterns of these species. This view might be substantiated by the wide distribution of Ceterach cordatum and Pellaea calomelanos (Fig. 4) in Southern Africa, species which seem to have a wider tolerance towards extremes of temperature and water supply in the environment. oe ? \- eK mene em Fie. 4. Distribution map of Pellaea calomelanos in Southern Africa. 4. Temperate and tropical outlier species. Among the seasonally aquatic ferns of South West Africa, at least Marsilea nubica falls into this group. The non-aquatic species classified in this section (see Fig. 5) form a large section of the fern flora of the territory and include the following: Actiniopteris australis (sens. lat.). Adiantum caudatum (sens. lat.). Cheilanthes hirta. Distributional, Ecological and Phytogeographical observations 17 on the Ferns of South West Africa. Cheilanthes multifida. Doryopteris concolor var. kirki. Dryopteris thelypteris var. squamigera. Microlepia speluncae. Ophioglossum reticulatum. Pellaea calomelanos. Pellaea goudoti. Pellaea viridis (sens. lat.). As observed in examples of the previous section these outliers could be regarded as relicts left in limited congenial habitats despite the supposed general drying of the southern parts of Africa. The roughly reversed C-shaped distribution pattern of Adiantum caudatum and other ferns in Southern Africa appears to have no parallels among angiosperms in this region with the exception of the distribution pattern drawn by Weimarck (1940) for Aristea angolensis. In this case, e ‘ oe a! ew We c ian / + : i et tt bbe ooo ¢ % i oa wmse, -- omer tow x ~ 4 om Sy ) + . O & U sy 2 \ fe * 2 See, - CS ’ e ij po @.ceee © te er, ¢ od e c e 2 f t ; bo-- wn Se ‘ Liu ee — ene - == 4 = Fic. 5. Distribution map of Doryopteris concolor (@) and Microlepia speluncae (+) in Southern Africa. 18 The Journal of South African Botany. Weimarck considers the populations of this Aristea in the Katanga as relics of an old connecting link between east and west African populations in this latitude. A similar view could be taken of the populations of Adiantum caudatum and Microlepia speluncae at the Victoria Falls. It will be noted from Table III that Cheilanthes multifida is recorded from Southern and south-central Africa and also from north-eastern Africa. Comparable interrupted distribution patterns are exhibited by Pyrrosia schimperiana (Schelpe, 1952) and other ferns (MacLeay, 1953). Such distribution patterns could be held to indicate that the north-south continuity of older populations has been broken by adverse climatic changes developing at the middle of the distribution range. The present distribution patterns of other members of this group seem to indicate that the South West African populations have been isolated from the main ramifications of the distribution of these species in Africa by the advent of locally unfavourable conditions over the territory as a whole. This is considered more probable than the recent arrival of widely dispersed propagules of these species in extremely limited favourable habitats from sporulating populations thousands of miles to the east and north. In relation to this discussion it must be admitted that very little is is known about fern distribution or climatic conditions prevailing in the northern parts of Bechuanaland. Only Pellaea calomelanos among the non-aquatic ferns has been recorded to the author’s knowledge from the Kwebe Hills south of Lake Ngami. King (1942) records the widespread occurrence of inselberge in Bechuanaland. Such inselberge could possibly provide a habitat for xerophytic ferns and might provide “stepping stones” for the colonisation of rocky outcrops in South West Africa by eastern species in Southern Africa, assuming that the intervening distances of uncolonisable ground were less than the maximum distributional trans- port capacities of propagules from sporulating populations. However, in addition to incomplete distributional data, an inadequate knowledge of the sequence and nature of past climatic changes in the sub-continent prevent further substantiated argument on the possible past migrations and foci of spread of ferns in Southern Africa. In this phytogeographical discussion reference can be made to Cope- land’s (1939) rather dogmatic comments on the place of origin of the Marsileaceae and the xerophilous and other ferns. He regards these as having had a southern origin apparently entirely on the dubious grounds of species concentrations, with little or no reference to the distribution of fossil ferns or to the availability of suitable habitats in the regions con- sidered. From the evidence available at present, it seems to the author Distributional, Ecological and Phytogeographical observations 19 on the Ferns of South West Africa. that at least some of the genera of ferns occurring in Southern Africa including South West Africa are tropical outliers with tropical and not Antarctic areas of origin. ACKNOWLEDGEMENTS. The author wishes to thank the Council for Scientific and Industrial Research for a grant in aid of travelling expenses; also those residents in South West Africa who generously afforded their advice and hospitality to the author during his visit. SUMMARY. The known distribution of ferns in South West Africa is summarised and discussed in relation to the climatic and edaphic aspects of the habitats occupied by these species. The relative drought resistance of the sporophyte and gametophyte in xerophytic ferns is considered as a limiting factor in their distribution. The phytogeography of the ferns of South West Africa is discussed in relation to the known distribution of these species in Southern Africa and elsewhere. REFERENCES. Auston, A. H. G. (1934): Mr. John Gossweiler’s plants from Angola and Portuguese Congo: Pteridophyta. Journ. Bot., 1934, Suppl. Auston, A. H. G. and Scuetpr, HE. A. C. L. E. (1952): An annotated check-list of the Pteridophyta of Southern Africa, Journ. S.Afr. Bot., 18, 153—176. Curist, H. (1910): Der Geographie der Farne. Jena. CHRISTENSEN, C.: (1906): Index Filicum. Hafniae. (1932): The Pteridophyta of Madagascar. Dansk. bot. Ark., 7, 1—242. CoPELAND, EH. B. (1939): Fern Evolution in Antarctica. Philippine Journ. Sct., 70, 157—188. DeEmareEtT, F. (1949): Les pteridophytes du Congo Belge et du Ruanda-Urundi. I. Bull. Jard. bot. Etat, Brux., 19, 251—269. Drets, L. (1902): Die naturlichen Pflanzenfamilien; Polypodiaceae. Leipzig. Gams, H. (1938): Okologie der extratropischen Pteridophyten in Manual of Pteri- dology. The Hague. Hourrum, R. E. (1938): The ecology of tropical pteridophytes in Manual of Pteri- dology. The Hague. Kine, L. C. (1942): South African Scenery. Edinburgh. MacLeay, K. N. G. (1953): The ferns and fern allies of the Sudan. Sudan Notes and Records, 34, 286—298. Picur SERMOLLI, R. (1954): Adumbratio Florae Aethiopicae 3. Ophioglossaceae, Osmundaceae, Schizaeaceae. Webbia, 9, 623—660. PickerrT, H. L.: (1914): Some ecological adaptations of certain fern prothallia—Camptosorus rhizophyllus Link, Asplenium platyneuron Oakes. Amer. Journ. Bot., 1, 477—498. (1923): An ecological study of Cheilanthes gracillima. Bull. Torrey bot. Cl., 50, 329—338. 20 The Journal of South African Botany. Pickett, F. L. and Manuet, M. E. (1925): An ecological study of certain ferns, Pellaea atropurpurea (L.) Link and Pellaea glabella Mett. Bull. Torrey bot. Cl., 53, 1—5. ScHepr, E. A. C. L. E.: (1952): The genus Pyrrosia (Polypodiaceae) in Africa. Journ. S.Afr. Bot., 18, 123—143. (1953): Ferns in Central Ethiopia. Brit. Fern Gaz., 8, 61—64. (1954): The Cheilanthes multifida complex in Southern Africa. Journ. S.Afr. Bot., 20, 127—136. Tarpievu-Btiot, M. L.: (1951): Ophioglossacées in Flore de Madagascar et des Comores. Paris. (1953): Les Pteridophytes de l'Afrique intertropicale Francaise. Mem. Inst. Franc. d Afr. Notre, 28. Tryon, R. M. (1942): A revision of the genus Doryopteris. Contrib. Gray Herb., 148, 3—80. Wermarck, H. (1940): Monograph of the genus Aristea. Lunds Univ. Arsskr., N.F., Avd. 2, 36. Unton or SoutH Arrica: Department of Transport (1953): Average Rainfall, South West Africa. Pretoria. 2] Distributional, Ecological and Phytogeographical observations on the Ferns of South West Africa. ‘(Ee6[ ‘Avooe) uepng — (PEG ‘edjeyog) BIsopoyy uISeYyZION “UIEISe Ty pue uroyynog epylq[nur soyquepeyy ‘(C61 “JOlA Norpsey,) weyg ueqa ‘suoozoureg ‘4yseog AIOAT “eoumnsy youosy ‘(ZG “UodrZ) (ZP6l ‘uod1y) vipuy ‘uophog | seosesepeyy “eAuoy ‘epuesg ‘eyrAuesuey, “UIOISB A IYI “IVA IOTOOUOD siieydoAIOG — ‘(9061 ‘uesue4sTIyO) UOTUNey “UIO4SB AL (qe ‘Sues) ByITYy soyyURTIOYH ‘(ge61 ‘Avorjoey) uepng ‘(FE6T “UO¥STY) Blosuy (9061 ‘UesUe4sTIy) BISW [eordory, ‘(eg6 1 ‘odjoyog) eruissAqy UIOYSVGT (487 “suos) stpeaqgsne staioydormow ‘(ec6 ‘Avorjory) uepng “(2861 ‘uasue4stI4)) ‘reosesepey, “Boljy 4Sei (PEGI “UOISTY) BlosuYy ‘(egg 1 ‘edjoyog) vruisshqy ‘ST ogUoIBOSe], (9061 ‘ussuoysliyg) uvuun x ‘eAvleuIrpy (9061 ‘uesueystayQ) BUETET “49 ES “‘ROLIpY UIOYINOG qnoysnory} epi “BOLIpY UIoyyNoG ynoysnoryy opi, souR[euLO[eo voRT[eg (qe] ‘suos) UMyZRpz100 YoRIqo— “ULI NOG * BUBIUO[09 BUSB[OYO NT eqoplAred soyquepioyy ee “ec ‘goutAorg edeg UI04S9 AA. WQION, “+ eAsnqor BoR][eq Beplojyep voel[od TUOSMBI BUOB[OYION TyyopLVUL BUSBTOYION ToFUTp) SOF UB [IOUy) “BdlIyV Opisyno uor4Nqrasy4stq ‘spuR|sy URolIpy pue ‘@ollgy JO 4Sor UI UOTYNGIT4SIC "BOLIZY “MS Optsyno BOLIFY ULOYYNOG ‘[e1zU90 puB UIEISeT RIE OT LG STG ‘VOIULW ISHM HLAOG AdISLAO SNUGY NVOIUaW ISHAA HLOOG FO NOILAGIULSIGQG AHL NO VLVC ‘THE ATAVL The Journal of South African Botany. “(€G61 “F0lG Norpaey,) [eordoaqueg (ZEG “UosuoysIIyD) awosesepeyy ‘(g¢61 ‘Avorpory) uepng (C61 “JOl Netpaey,) uoogery, ‘osuoy uskoW ‘tawyg dureqg ‘suoo1owmley “4sReog AOA, “eournk) Yyouoasy “(FE6T “UO¥STY) vlosuy “UIO4SR OL ovounjods erdoposoryy (EG61 “J01@ Horpaey) erpuy yynog ‘(E61 “JOE nolpsey,) Suocsd9uIR_ ‘OSUO_ URIS [Og ‘efosuy “UIO4SE HL pur uzeyynog | esosturenbs area staeqydAjoyy staoydokaq (ZE6T ‘uosueqstay) ‘S] oUoLBOSRp, “S~T Oaou0g ‘aeoseSepey\ (ZEGL “UoSUo4sIAyO) BIqeay ‘(€C6 1 “JOld Norpawy) Uooqey ‘odu0D ueIsjog ULOSBOT ‘-1yopnos voeyljog ‘(E661 ‘Avorjory) wepng (G06 ‘S[PIC) volpy yseOT (9061 ‘Uosuo4sTayy)) ‘S[ oUodBosBy, ‘eruIssAqy ‘opaoA odep “ULOISBGT , pue usoygnog | - (qB] ‘Suos) SIplatA voeTog ‘(9061 “Uesue4ystsyO) “ST ‘(C61 ‘AvorTory_) wepng ‘(TS61 “401 Norpzey) ‘sniquine yy, ‘UoruneYy “ST O1OUTOL ‘reoseSepLI| : ‘(ES61 “F0Te Norpawy) suoosourey ‘osuo, UoAOT ‘taeyg ruRq~ “‘qsvoy AIOAy ‘Rournsy) YOu ‘UBpNg Youodyy (6P6T ‘oreutecy) OSuoOH uLIsloq eurjoreyD “Boureg ‘eormoury yeordoxy, ‘(PEG “UO4S[y) Blosuy “UIOISVHT winye]noyer wuNsso[sorydg ‘(€G61 ‘Avorporyy) uepng (906T (€G61 “GO[_ Notpaey,) oSu0g udOW ‘uesuoystIyy)) SoplIqof]T MONT “BISW ‘(9061 Suesue4sIayD) Botapy [eordouy, ULOISV HL (ye, ‘Suos) UNgepneo vUNqUeIp ? “BOLIPY OPIsyNoO UOIyNGAqSIG “‘SpuRy|sy uBolpy pure BOlIPY JO JSor UL UOTYNGLAYSICT "ROLUEW “AA'S Opisgno Bolapy ULOYYNOG UL UOLNLAySI a a A ae Mee Gea ee Seamer Boal setae “VOIUTY LSA AA HLAOG @aISENO SNUAY NVOLMIY OSa\\ HLOAOG AO NOILOGIMESIGG HHO NO VLVCT PEO TN GUTSY AT TWO NEW SPECIES AND TWO NEW VARIETIES OF ALOE FROM MADAGASCAR. By G. W. REYNOLDs. (With PLarEes I[V—X.) At the invitation of Professor J. Millot, Director of the Institut Scientifique de Madagascar, I flew to Tananarive on 8th June, 1955, to study the Aloes of Madagascar. A Land Rover with driver was provided, and together with Mr. B. M. Descoings, the Institute’s botanist, I travelled over 4,000 miles through the island, including the extreme south and the far north, investigating the Aloes. Several species were found at their recorded type localities, and many more elsewhere. Distribution was recorded, descriptions were drawn up in the field, a large number of photographs was secured in colour and black and white, while much herbarium material was carefully prepared. This material has been contributed to Tananarive and Paris, with duplicates for Pretoria and Kew, together with complete sets of photographs. As a result of this expedition which, fortunately, was very successful, and at Professor Millot’s invitation, I hope to attempt an illustrated review of the Aloes of Madagascar soon. In the meantime, two new species and two distinctive varieties await description, and they are now described with the full concurrence of Professor H. Humbert, Director of the Natural History Museum, Paris. Aloe Millotii Reynolds, species nova in subsect. Prolongatae, ser. macrifoliae Berger; A. antandroy H. Perrier et A. Bakeri Scott Elliot affinis. Planta succulenta. Caules decumbentes vel adscendentes, graciles, elongati, 7—9 mm. diam., 20—25 em. longi, e basi ramosi, apice per 5—7 cm. sublaxe foliati. Folia 8—10, patentia, recurvula, carnosula, 8—10 em. longa, 7—9 mm. lata; swpra viridia, rubescentia, saepe immacu- lata, canaliculata; subtws convexa, maculata; dentibus marginalibus deltoideis, albidis, -5—1l mm. longis, 5—10 mm. inter se distantibus armata. Inflorescentia simplex, 12—15 em. longa. Pedunculus gracilis, superne bracteatus, bracteae vacuae ovato-acuminatae, scariosae, albidae, 8 mm. longae, 5 mm. latae. Racemi laxe 6—8-floribus, 4—5 cm. diam. Bracteae ovato-acutae, infimae 7 mm. longae, 4 mm. latae, scariosae, 3—5d-nervatae. Pedicelli 6—8 mm. longi. Perigoniwm coccineum, 22 mm. longum, circa ovarium 7 mm. diametro, hine levissime constrictum et decurvulum, fauces versus levissime ampliatum; segmenta exteriora libera. 23 24. The Journal of South African Botany. Antherae 1 mm. exsertae. Stigma demum 1—2 mm. exserta. Ovarium 5 mm. longum, 2 mm. diametro. (Plates I[V—VI.) Habitat: Madagascar: Cap Ste. Marie, xerophytic bush and lme- stone plateaux, alt. 100—150 met., fl. 22 Feb., 1947, Humbert 20342! (P); on limestone at Cap Ste. Marie, ¢ 23° 35’ S., 49° 05’ E., alt. 100 met., fl. 5 July, 1955, Reynolds 7840, holotype (TAN), isotype (P, K, PRE). A. Millotii was collected by Professor H. Humbert (n. 20342) at Cap Ste. Marie, the southernmost point of Madagascar in 1946, and was provisionally named after Professor J. Millot. Plants were sent to the gardens at Tzimbazaza, Tananarive, for cultivation, but a description could not be drawn up without adequate flowering material. “Accompanied by Mr. B. M. Descoings, the Institute’s botanist, I visited Cap Ste. Marie on 5 July, 1955, and had the good fortune to find a con- siderable number of plants growing there on Eocene limestone, many being in flower. A description was drawn up, photographs were secured, and sufficient herbarium material was prepared. This very distinctive species can now be described and I am happy to adopt the name under which it has been known for some years. A. Millotii was found only on limestone, in exposed places or in partial shade, under intensely arid conditions. Plants form very small low undershrubs which are much-branched at the base, the outer stems being mostly decumbent with the centre ones erect to sub-erect. When partly supported by low twiggy bush, stems may reach 50 cm. in length with 16—20 leaves laxly grouped along the apical 20 cm. of stems. In exposed positions, stems average 20—25 cm. long with the apical 5 em. sub-densely foliate. Leaves are at first distichous in young stems, but become spirally twisted to rosulate with age. Marginal teeth vary from smaller to larger and are sometimes obsolescent near apex. Racemes are few-flowered, and only 2—4 simple inflorescences were noticed in a much-branched plant. The inflorescences vary from oblique to arcuate-ascending and suberect. Outer segments are free to base, but cohere in lower half. In pedicels, and shape and size of flowers. A. Millotii is scarcely separable from A. antandroy H. Perrier—which I found between Cap Ste. Marie and Tsihombe, also west of Tsihombe and near La Table, Tulear— but the latter is a much larger plant with rigid erect stems 50 cm.—l m. long, and very different leaves. A. Bakeri Scott Elliot, which I found near Fort Dauphin, is a near ally, but differs in forming large very dense groups of 50—100 plants, and in Two New Species and Two New Varieties of Aloe 25 from Madagascar. having somewhat ventricose flowers which shade from apricot-orange at base through orange to yellow at the green-tipped mouth. A. Millotw is a very distinctive species with a rather sprawling habit of growth, and could hardly be confused with any other species. Derscrrption: A small low undershrub, much branched at base. Stems decumbent, divergent or ascending, averaging 20—25 cm. long in exposed positions, 7—9 mm. diam. near base, slightly thicker upwards, 10—20-branched from base or higher, the apical 5—7 cm. sublaxly to sub-densely foliate with the internodes 5—10 mm. apart. Leaves 8—10, spreading to recurved, distichous in young shoots, becoming spirally twisted to rosulate with age, basally sheathing, 8—10 cm. long, 7—9 mm. broad at base, gradually tapering to the apex which is obtusely rounded and armed with about 5 rather soft white cartilaginous prickles about -5—1 mm. long; wpper surface dull grey-green with reddish tinge when in full sun, greener in partial shade, canaliculate, with or without a few small dull white spots low down; lower surface convex, usually greener than upper surface and with many scattered dull white spots throughout, the spots about 2 mm. long, 1—1-5 mm. broad at the middle, sometimes minutely subspinulescent; margins armed with small, white, cartilaginous, deltoid, isolated teeth -5—1 mm. long, 5—10 mm. ~ distant, the inter-spaces straight and the colour of the leaf, sometimes obsolescent near apex. Inflorescence always simple, averaging 12—15 em. long. Peduncle plano-convex and 3—5 mm. broad at base, terete upwards, about 3 mm. diam. below the raceme, with about 5 sterile-bracts in upper half, the lowest ovate-acuminate, thin, scarious, white, about 8 mm. long, 5 mm. broad at base, many-nerved, smaller upwards. Racemes laxly 6—8-flowered, 4—5 cm. diam., the pedicellate portion 3—5 cm. long, the scarlet buds grey-green tipped, spreading, open flowers nutant to sub-pendulous. Bracts ovate-acute, the lowest 7 mm. long, 4 mm. broad at base, reflexed, thin, scarious, 3—5-nerved. Pedicels spreading, averaging 6—8 mm. long. Perianth scarlet, paler at mouth, averaging 22 mm. long, basally flat to very slightly rounded, 7 mm. diam. across the ovary, constricted to 6 mm. above the ovary, thence slightly decurved and enlarging to the throat with the mouth wide open; outer segments free to base but cohering to the middle, obscurely 3-nerved, the nerves green at apex, the apices subacute, slightly spreading; inner segments free, thin and white at the edges, with a scarlet keel turning greenish at apex, the apices more obtuse and more spreading than the outer. 26 The Journal of South African Botany. Filaments almost white, filiform-flattened, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn exserted 1 mm. Style pale lemon, with stigma at length exserted 1—2 mm. Ovary pale orange-yellow, 5 mm. long, 2 mm. diam., truncate at apex. Capsule round-oval, 15 mm. long, 12 mm. diam. Native Name: Vaotsay. The natives make a decoction from the leaves and drink it for stomach troubles. A. bulbillifera H. Perrier var. Paulianae Reynolds. Varietas nova, a forma typica bulbillis epipedunculis, infra ramos non epiramis geruntibus differt. (Plates VII, VIII.) Habitat: Madagascar: Majunga Province, ex 85 km. south-west of Antsohihy, c. 15° 03'S., 47° 44’ E., alt. 100 met., coll. Madame L. Paulian, cult. Tzimbazaza, fl. 9 June, 1955, Reynolds 7656 holotype (TAN), isotype (P, K, PRE): 85 km. south-west of Antsohihy, 20 July, 1955 (bulbils only), Reynolds 7904 (TAN, P, K, PRE). This distinctive variety was first collected by Madame L. Paulian, wife of Dr. R. Paulian, Deputy Director of the Institut Scientifique de Madagascar. Plants were cultivated in the Institute’s gardens at Tzimbazaza, Tananarive, where I saw some in flower on 9 June, 1955, together with the typical form that was also flowering. Subsequently, on 20 July, 1955, while travelling northwards to Diego Suarez, I found considerable numbers of the variety on rocky slopes, on a ridge above forest, at a point 24 km. south of the Analalava road junction and 85 km. south-west of Antsohihy. Capsules had dehisced and were then very dry, but the bulbils were well developed on the peduncle below the branches. In the typical form of A. bulbillifera (Plate VII, Fig. 1), the inflorescence is taller, considerably more branched from below the middle, the branches are very much longer, while the peduncle is more slender, and unable to support the weight of the inflorescence which bends over sideways. Bulbils form only on the slender branches below the racemes and develop after the flowers are over. In the var. Paulianae (Plate VII, Fig. 2), the peduncle is stouter, erect - or suberect, rigid, and is branched only in the upper quarter, the branches being considerably shorter and much more compact. Bulbils form only on the peduncle from about the middle to the lowest branch or a little higher, and not on the branches. These sessile bulbils commence developing when the flowers open. When the capsules dehisce and become very dry, the bulbils comprise rosettes of 4—6 leaves which are up to 5 cm. Two New Species and Two New Varieties of Aloe 27 from Madagascar. long, 15 mm. broad, with dentate margins. When the peduncle dries out, the bulbils drop and root. There is little difference in leaf characters, bracts, pedicels and flowers. Description: Plant succulent, acaulescent, solitary. Leaves about 24, densely rosulate, lanceolate-attenuate to ensiform, spreading, about 60 em. long, 9 cm. broad at base; upper surface dull green with reddish tinge, without spots or markings, slightly canaliculate; lower surface convex, greener, without spots or markings: margins sinuate- dentate, armed with firm, compressed-deltoid, dull white teeth 3 mm. long, 12 mm. distant, the teeth slightly forward-uncinnate. Sap dries yellow. Inflorescence a branched panicle up to 2 m. high, 2—4 simultaneously. Peduncle basally plano-convex and 15—20 mm. broad, brown with a bloom, branched in upper quarter, with about 12 compact arcuate- ascending branches, the lowest 2—3 branches with 1—3 branchlets; the peduncle developing sessile bulbils below the first branch and not on the branches, the bulbils at length comprising rosettes of about 5 leaves which are thick, fleshy, spreading to recurved, 5 cm. long, 15 mm. broad, with dentate margins. Racemes cylindric-acuminate, sublaxly flowered, the terminal 25—30 cm. long, 6 cm. diam., branch racemes 15—20 em. long, buds spreading to nutant, open flowers subpendulous. Bracts ovate-acute, 4 mm. long, 3 mm. broad, thin, scarious, 3—5- nerved. Pedicels slender, reddish-brown, 8—10 mm. long, suberect to spreading. Perianth scarlet, paler at mouth, sub-cylindric, 27 mm. long, basally flat, 7 mm. diam. across the ovary, thence very slightly narrowed on the underside and slightly decurved, enlarging towards the throat; outer segments free to base but cohering to the middle, apices subacute, slightly spreading; inner segments free, broader than the outer, with 3 crowded nerves forming a scarlet keel, apices more obtuse and more spreading than the outer. Filaments filiform-flattened, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn exserted 2 mm. Style pale lemon, with stigma at length exserted 3—4 mm. Ovary pale olive, 5 mm. long, 2 mm. diam., obtuse at apex. A. acutissima H. Perrier var. antanimorensis Reynolds var. nov. in Sect. Prolongatae ser. macrifoliae Berger. A forma typica plantis humili- oribus, caulibus brevioribus gracilioribus, folii angustioribus brevioribus, pedicellis et floribus brevioribus differt; A. Decaryi Guillaumin et A. antandroy H. Perr. affinis. (Plate V.) 28 The Journal of South African Botany. Habitat: Madagascar: Tulear Province, 8 km. north-west of Antanimora, c. 24° 44'S., 45° 36’ E., alt. 300 met., fl. 28 June, 1955, Reynolds 7792 holotype (TAN), isotypes (P, K, PRE). Large numbers of flowering plants were found on 28 June, 1955, on flat rock surfaces among thorn bushes and tall stemmed Pachypodiums, 8 km. north-west of Antanimora, at km. 313 on the main road from the north to Ambovombe and Fort Dauphin. Plants are of low shrubby erowth and form dense patches 1—2 met. and more across. The var. antanimorensis differs from typical A. acutissima in having much shorter stems which are only 6—9 mm. in diameter, against 20—30 mm. The leaves are shorter and only 15—20 mm. broad at base, while pedicels average 10 mm. long against 15—20 mm. The flowers are 25 mm. long against 30—35 mm. in the typical form. A noteworthy character of the variety is that the inflorescence is as tall as the stem and rosette combined, sometimes taller. ; A species to be considered is A. Decaryi Guillaumin—which I did not succeed in finding. This species is described as having leaves only 8—9 mm. broad, pure white marginal teeth, scape as long as the leaves (i.e., 15—19 cm.), and perianth outer segments free to base. In our new variety, apart from other differences, the lower half of the perianth is a definite tube. A. antandroy H. Perr. is another ally, but this species has much longer subscandent stems, considerably narrower spotted leaves, and very differently shaped flowers. Description: Plant succulent, of low broad shrubby growth. Stems 30—50 cm. long, 6—9 mm. diam., branched at base, the apical 10—15 cm. sublaxly about 10-foliate, the internodes 10 mm. distant, striate. Leaves about 10, basally sheathing, narrowly lanceolate-attenuate, spreading, 15—20 em. long, 18—20 mm. broad at base; wpper surface dull grey-green, unicoloured, flat to slightly canaliculate; lower surface convex, similar to upper surface; margins armed with deltoid, firm, pale brownish teeth 2 mm. long, averaging 8 mm. distant. Inflorescence simple or sometimes with a short branch, 40—50 em. tall. Peduncle slender, basally plano-convex and 5—7 mm. broad, terete upwards, brown, with about 6 sterile-bracts, the bracts deltoid, thin, scarious, up to 12 mm. long, 5 mm. broad, many-nerved. Racemes cylindric-conical, 8—10 cm. long, sub-densely flowered, youngest buds suberect and partly obscured by their bracts, older buds spreading, open flowers subpendulous. Bracts narrowly deltoid, the lowest 12 mm. long, 3 mm. broad at base, thin, scarious, about 7-nerved. Two New Species and Two New Varieties of Aloe 29 from Madagascar. Pedicels 10 mm. long. Perianth reddish-scarlet, 25 mm. long, obtusely tapering at base to the pedicel, 5 mm. diam. across the ovary, very slightly constricted above the ovary, thence slightly decurved and enlarging towards the throat, the mouth open, trigonous; outer segments free for half their length, obscurely 5-nerved, apices sub-acute slightly spreading; inner segments free but dorsally adnate to the outer for half their length, broader than the outer, with thin pale borders, keeled in upper half, the apices more spreading than the outer. Filaments filiform-flattened, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn exserted 1 mm. Stigma at length, exserted, 2 mm. Ovary olive-brown, 4 mm. long, 2 mm. diam., the apex truncate. Aloe itremensis Reynolds, spec. nov., nulli alii arcte affinis, forsitan sectionem propriam constituere debet. Planta succulenta, acaulescens vel brevissime caulescens. Folia 12—16, rosulata, patentia, usque ad 30 cm. longa, 4-5 cm. lata, sensim attenuata; supra viridia, rubescentia, immaculata, leviter canaliculata; subtus convexa; ad margines dentibus deltoideis, parvis, 1—1-5 mm. longis, 5—8 mm. distantibus armata. Inflorescentia gracilis, 2-3-ramosa, 1—1-20 met. alta. Racemi cylindrico- acuminati, 15—20 cm. longi. Bracteae anguste deltoideae, 6 mm. longae, 2 mm. latae, subscariosae, 3-nervatae. Pedicelli 8—9 mm. longi. Peri- gonium rubrum, anguste cylindrico-clavatum, 25 mm. longum, circa ovarium 5 mm. diametro, hine levissime ampliatum; segymenta exteriora libera, 5-nervata, interiora angustiora. Genitalia demum 1 mm. exserta. Ovarium 4:5 mm. longum, 2 mm. diametro. (Plate X.) Habitat: Madagascar, on the Itremo range, 54 km. west of Ambato- finandrahana, c. 20° 37'S., 46° 34’ E., alt. 1,700 met., fl. 20 June, 1955, Reynolds 7706, holotype (TAN), isotype (P, K, PRE). A. itremensis was found on the western side of the plateau along the top of the Itremo range at a point on the roadside 54 km. west of Ambato- finandrahana, and 94 km. west of Ivato, on the road to Amborampotsy and Morondava, at an altitude of 1,700 met., and at approximately 20° 37’ S., 46° 34’ E. Plants were noticed only on steep sandstone slopes and outcrops, usually in exposed positions, but sometimes in shade. 30 The Journal of South African Botany. Professor Millot confirms that plants did not fit any published description, and I am naming the species after the only locality where I found it. A. itremensis occurs as solitary plants not forming groups. It has little or no stem, and is characterised by its very small marginal teeth, slender tall inflorescence which is compactly only 2—3-branched from the middle or higher, and by its very narrowly deltoid bracts. Descrietion: Plant succulent, solitary, not forming groups. Stem 0—20 cm. long, up to 3 cm. diam. Leaves 12—16, rosulate, arcuate-ascending, spreading to recurved in upper quarter, up to 30 cm. long, 4-5 cm. broad at base, gradually tapering to an acute apex; wpper surface basally flat, slightly canaliculate upwards, dull-green with reddish tinge, without spots or markings; lower swrface convex, similar in colour to upper surface; margins armed with broadly compressed-deltoid brownish teeth which are paler at apex, 1—1-5 mm. long, 5—8 mm. distant, larger near base, smaller upwards, obsolescent towards apex. Inflorescence a slender 2—3-branched panicle 1—1-20 met. high. Peduncle plano-convex and 14 mm. broad at base, terete upwards, brown with a bloom, compactly 2—3-branched from the middle or higher. Racemes cylindric-acuminate, the terminal averaging 15—20 cm. long, the lateral a little shorter, the buds suberect, older buds spreading, open flowers subpendulous, all unicoloured reddish-scarlet. Bracts narrowly deltoid, 6 mm. long, 2 mm. broad at base, reflexed at the middle, thin, subscarious, with 3 crowded nerves sometimes appearing as |-nerved. Pedicels reddish-brown, 8—9 mm. long, slender. - Perianth narrowly cylindric-clavate, averaging 25 mm. long, tapering at base into the pedicel articulation, 5 mm. diam. across the ovary, enlarging a little to the throat and slightly contracted at the mouth; outer segments free to base, with pale thin edges, obscurely 5-nerved throughout, the apices sub-acute, slightly spreading; inner segments free, slightly narrower than the outer, 5 mm. broad in upper two-thirds with | thin white 2 mm. broad border, prominently scarlet-keeled, the apices more obtuse than the outer. Filaments filiform-flattened, the 3 inner narrower and lengthening before the 3 outer, with their anthers in turn exserted 1 mm. Stigma at length exserted 1 mm. Ovary pale olive, 4:5 mm. long, 2 mm. diam., obtusely tapering into the style and faintly 6-striped longitudinally. Two New Species and Two New Varieties of Aloe 31 from Madagascar. ACKNOWLEDGMENTS. I am greatly indebted to Professor J. Millot, Director of the Institut Scientifique de Madagascar, Tananarive, for his invitation to study the Aloes in Madagascar, and for providing every possible facility and assistance during my visit. To Dr. R. Paulian, the Deputy Director, I am obliged for arranging transport throughout the Island, and for considerable help in many other directions. Iam grateful to the Institute’s botanist, Mr. B. M. Descoings for much considerate help and many kindnesses during seven weeks of continuous travelling together. I am also indebted to the South African Council for Scientific and Industrial Research for a travelling grant which enabled me to fly to Madagascar to investigate the Aloes. “uD OZ FysIeEH “ecér ‘AINE ¢ “Ty “Rose Depey ‘oe “03g deg ye ouojsouny ud ‘spjouAoy, WOT 27h AT TOVIg *xordde 4 "ECG | TANG © “fof *xoudde ¢ UIO4S FUMOMO 4] ‘avosesepeyy Jo yutod ysouruseyynos oy ‘ore ‘09g deg ye opeys persed ut yUe[ ‘sploudoy wo 2077 A VIG OTS] “Oi +o ae aD Iie, 2. Prate VI. Aloe Millotii Reynolds. Fic. 1. Flowers 1/1 from bud to fruit. Fic, 2. Various flowers 1/1 showing variation. “RZRZVCUNZ], YB CGR, “OUNL G "ToT SULIOY, “SplOUAoY epUDIYNYG “IBA Jolddod "HY vwafypeqing “Fo 7° ISI i upMevads ayy SoAleuvur ff, AY ‘doLiog “HY pwafypeqng v)P "| “DIyl "[ “DIyy ‘spjoudoyy apuniynyg “aera soloed “PF viefypqing “VW. “TTT A Givitd “T Dlyy wm Fg ~' | ty GCG] CY ULL ae suopaes oud ut gueyd i I % “OTT AION Sé A) ayy | "GCG, “‘oune BZ jo JSOM-U LOU "OOVIS PIMAF OF PNG UO | / | SHOMOL yp "ZL ‘la UUEy I “TR: ey G ISB. qeqIqey [VINgvU UL syuR[q “[ “Sly “SP[OUAOYT SUSUIMOWMUDIUD “ABA IOLLIO “EL VUMISSYNID 90] 7 *xoadde ¥ x OVPBIN ULIY NOY “BLOUTIUBR PUY 6 “wil “XI LV] ‘spjoudoyy sasuawade 00) 7 osugs uolyeurpod-qsod oy pnq woaz *[/ | SHOMOIop “Z “OTT “you G-] WSIOH “Ol PE OF WBOSRSRPLIY “RUBY RapuURUyoOIRguriy JO ysom IVT Og WO SRqrqey PRaMgeuw ur gunz op “Ploy ‘o “OTH ‘xX MIVId A NEW ALOE FROM KENYA COLONY. By G. W. Rrynoxps. (With Plates XI and XII.) Aloe ukambensis Reynolds, spec. nov. in sect. Tropicales Berger pertinet, A. abyssinica Lam. affinis. Planta succulenta, acaulis. Folia 30—40, dense rosulata, lanceolato- attenuata, usque ad 50 cm. longa, 10—12 em. lata, leviter canaliculata, striata, maculata vel immaculata, ad margines sinuato-dentata, dentibus deltoideis 3—4 mm. longis, 8—10 mm. inter se distantibus munita. Inflorescentia 50 cm. alta, 2—3-ramosa. Racemi capitati, 4—6 cm. longi, 6 cm. diam. Bracteae ovato-deltoideae, 5 mm. longae, 2:5 mm. latae, scariosae, 3-nervatae. Pedicelli 18—20 mm. longi. Perigoniwm rubrum, 40 mm. longum, trigono-cylindricum; segmenta exteriora per 20 mm. libera. Antherae 2—3 mm. exsertae. Stigma demum 4 mm. exserta. Ovarium 5 mm. longum, 2-5 mm. diametro. Habitat: Kenya Colony: Maboloni Hill, Yatta, coll. Benton, fl. 7 Dec., 1952, Bally 8382! (KA); Maboloni Hill, Ndaleni Ranch, 47 miles east of Thika, 73 miles north-east of Nairobi, alt. 4,500 ft., c. 1° 05’ S., 37° 28’ E., fl. 6 June, 1955, Reynolds 7651, holotype (EA), isotype (PRE, K). A. ukambensis was first discovered by Mr. P. R. O. Bally in January, 1952, on Mbwinzao Hill between Makindu and Kibwezi, on the Nairobi- Mombasa main road. (Alt. 4,500 ft., c. 2° 22’S., 37° 54’ E.). In 1943 Mr. Bally found the species again on the Yatta Plateau near Kitui. In January, 1952, Police Officer Benton found plants on Maboloni Hill, near the Thika River, about 47 miles east of Thika. The species has also been collected by Mr. David Sheldrick, Warden of the Tsavo National Park, on Yamanyani Hill, about 25 miles north of Voi. When Mr. Bally and I visited Maboloni Hill in August, 1953, no plants were found in flower, but early in June, 1955, when Mr. and Mrs. H. A. Delap drove me to the same hill, several flowering plants were found, although the maximum flowering time is December—January. A striking character of A. ukambensis is the leaf lineation, and the arrangement of the ““H’’-shaped spots. The leaves are conspicuously dark- veined longitudinally, the veins being 1—2 mm. wide. The ““H”’ spots are almost square and occur only in line with, and interrupting the veins. In some plants the spots disappear with age, while in others the basal half of the lower surface of leaves may be copiously “H’’-spotted. 33 34 The Juornal of South African Botany. The capitate racemes are usually only a little higher than the leaves, while in the flowers, the filaments and style lie along the lowest side of the perianth. Description: An acaulescent plant forming small to large dense groups on gneiss outcrops. Stem none or very short. Leaves 30—40, densely rosulate, lanceolate-attenuate, up to 50 cm. long, 10—12 em. broad at base; upper surface flat low down, slightly canaliculate upwards, grey-green with reddish tinge, conspicuously striate longitudinally, with or without a few scattered oval to lenticular white spots low down; lower surface convex, similar to upper surface in colour and lineation, varying from unspotted to spotted in lower half, the spots “H’’-shaped, in line with, and interrupting the veins; margins sinuate- dentate, armed with compressed-deltoid, reddish-brown teeth which are paler at apex, 3—4 mm. long, 8—10 mm. distant. Sap dries pale-orange. Inflorescence 2—3-branched, 50 cm. high, the racemes sometimes only slightly higher than the apices of leaves. Peduncle plano-convex and 15 mm. broad at base, brown with a bloom, with 2—3 branches (sometimes 4) from about the middle, the lowest branch subtended at base by a broadly ovate, thin, scarious, brownish many-nerved bract about 15 mm. long and broad. Racemes densely flowered, capitate or sub-capitate, 6 cm. diam., the pedicellate portion 4—6 cm. long, youngest buds bright red with greyish tips, suberect, older buds spreading, open flowers pendulous. Bracts ovate-deltoid, dull white, scarious, thin, 5 mm. long, 2-5 mm. broad at base with 3 crowded nerves in median line. Pedicels slender, suberect, 18—20 mm. long, deflexed at apex, reaching 30 mm. in the fruit. Perianth bright red, averaging 40 mm. long, cylindric-trigonous, basally tapering to the pedicel and shortly stipitate, cylindric and 7 mm. diam. across the ovary, above the ovary slightly narrowed on underside and slightly decurved, cylindric-trigonous upwards, the mouth wide open; outer segments free to the middle (tube 20 mm.), obscurely 3-nerved at apex, apices subacute, slightly spreading; inner segments free but dorsally adnate to the outer to the middle, broader than the outer, with 3 crowded . nerves forming a scarlet keel in upper half, the apices more obtuse and more spreading than the outer. Filaments pale-pink, filiform-flattened, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn exserted 2—3 mm. Stigma at length exserted 4—5 mm. Ovary orange-brown, 5 mm. long, 2-5 mm. diam. Capsule 18—20 mm. long, 8—9 mm. diam. A New Aloe from Kenya Colony. 35 ACKNOWLEDGMENTS. Iam indebted to the South African Council for Scientific and Industrial Research for a travelling grant which enabled me to investigate this and several other species of Aloe in East Africa. I am also indebted to Mr. P. R. O. Bally, and to Mr. and Mrs. H. A. Delap for driving me to Maboloni Hill, and for sending plants by air to Johannesburg. Puate XI. Aloe ukambensis Reynolds. Plants in natural habitat, Maboloni Hill, 37 miles east of Thika, Kenya Colony. Alt. 4,500 ft., c. 1° 05’ S. .37° 28’ E. FL. 6 June, 1955. Height 75 cm. “HOTFRLIBA BuLMoys [/ [| StOMOLyT “qmuayp 07 pnq utoay [/] SaoMO]sT ‘xorudde + 9oRfINS IOMOT Jeo] *xoadde - x oovyans aoddn yroryt ‘sploudoy susvlaquipyn ao) a DET © hy 6 Ol T Slt LWT on rc) JOURNAL OF SOUTH AFRICAN BOTANY VOL. XXII. Published: 3rp. APRIL, 1956 A NEW MARSH ERICA FROM WEMMERSHOEK. By Caprain T. M. Santer, R.N. Erica Bakeri Salter sp. noy. (Ericaceae—KHricoideae) § Leptodendron. Fruticulus paludosus ad 60 cm. altus, supra tortuose ramosus, partibus herbaceis omnino glabris. Rami ramulique graciles politi castanei saepe contorti. Folia anguste linearia plerumque patentia saepe recurva, 2-5—3 mm. longa, petiolis rubris. Flores, pedunculis et sepalis inclusis, rosei apice ramulorum brevium terminales, 1—3-nati ultimi interdum subumbellati. _Pedunculi 1—2 mm. longi vel leviter longiores: bracteae 3 remotae circa 0-5 mm. longae brevissime ciliatae. Sepala attenuato- ovata apicem versus carinata circa 1 mm. longa, breviter ciliata. Corolla ovoideo-urceolata 3—3:2 mm. longa, sed in siecitate in adventu brevior, tubo quam lobis erectis 3-plo longiore. Stamina inclusa filamentis supra leviter sigmoideis: antherae oblongo-cuneatae basifixae, 0-5 mm. longae, basin ampliatam versus scabridae, poro fere dimidio lobi: aristae dorsales scabridae, 0:3—0-4 mm. longae. Ovariwm sessile subturbinatum in parte superiore dense hispidum: stylus inclusus, stigmate capitato. CAPE PROVINCE: Paarl Div.; in marshy ground near the rail- way at Wemmershoek, Baker 420 holotype in the Bolus Herbarium, Flowers Aug.—Oct. A marsh shrublet up to 60 em. high, intricately branched above, the herbaceous parts entirely glabrous. Branches and branchlets slender, often irregularly contorted, shining chestnut brown. Leaves narrow- linear, mostly spreading, often recurved, 2:5—3 mm. long, usually much shorter than the internodes, bright green in the living state: petioles red. Flowers, including the peduncles and sepals, rose-pink, terminal on short branchlets, 1—3-nate, the ultimate sometimes subumbellate. Pedwncles 37 38 The Journal of South African Botany. 1—2 mm. or slightly more long: bracts 3, remote, about 0-5 mm. long, very shortly ciliate. Sepals ovate-attenuate, keel-tipped, about 1 mm. long, shortly ciliate. Corolla ovoid-urceolate, 3—3-2 mm. long, but appearing shorter when dried, the tube about three times as long as the erect lobes. Stamens included, the filaments somewhat sigmoid above: anthers oblong-cuneate, basifixed. about 0-5 mm. long, scabrid towards the widened base, the pore nearly half as long as the lobe: awns dorsal, scabrid, 0:-3—0-4 mm. long. Ovary sessile, subturbinate, densely hispid on the upper part: style included: stigma capitate. Though this glabrous marsh species seems to fall most nearly under the Section Leptodendron as defined, it is in no way related to any of the species in that rather ambiguous Section. It appears, although the corolla and all the herbaceous parts are glabrous and the leaves 3-nate, to be closely related to some of the marsh species of the § Ephebus which occur in the Cape Peninsula, viz. #. cyrillaeflora Salisb. E. paludicola L. Bol., #. limosa L. Bol., EB. Salteri L. Bol. and E. heleogena Salter. Very like these species in habit, it often depends on surrounding marsh vegetation for support, and the chestnut brown or reddish brown stem and branches are very similar. It is not, of course, known for certain whether Hrica Bakeri occurs in any other locality, but it is noteworthy that the last four of the five marsh heaths mentioned above are peculiar, in that each of them seems to prefer its own particular mountain marsh, never having been found elsewhere, and they do not grow in association. #. cyrillaeflora has the same habit in the Cape Peninsula, only occurring now in one marsh between Simonstown and Kommetje, but, according to the late Dr. N. E. Brown, it was formerly recorded from Rocklands. This may be a mistake for Brooklands which is near the marsh mentioned above for it is unlikely that it would grow at such a low altitude. Salisbury’s type was collected over 150 years ago in the rather vague locality ‘‘Stellen- bosch Division; Hottentots Holland’, but it has never been observed since, except in the Cape Peninsula. I have named this new species in honour of Colonel H. A. Baker, its discoverer. Colonel Baker has recently been carrying out an intensive field study of the Ericaceae of the Cape Peninsula and has discovered almost certain hybridization between Hrica empetrifolia L. and E. mollis Andr., hitherto unrecorded, besides bringing to light several facts which make modifications in existing descriptions necessary. I have to thank him for undertaking much of the necessary preliminary research in estab- lishing #. Bakeri as a hitherto undescribed species, ascertaining that no other specimens are lying amongst the Incertae in the three local herbaria and for preparing the accompanying Figure. A New Marsh Erica from Wemmershock. 39 Fie. 1. Hrica Bakeri Salter. (Natural size.) (Baker 420.) Del. H. A. Baker. —== ss —— _ 2 Fie. 2. 1. Flower x 6. 2. Corolla x 6. 3. Sepal x 6. 4. Gynaecium x 6. 5. Stamen X 6. 6. Anther, side view x 12. 7. Anther, back view x 12. (Baker 420.) Del. H. A. Baker. STUDIES ON THE EMBRYOLOGY AND RELATIONSHIPS OF SOUTH AFRICAN GENERA OF THE HAEMODORACEAE: DILATRIS Berg. AND WACHENDORFIA Burm. By Miriam P. DE Vos. (Department of Botany, University of Stellenbosch.) ABSTRACT. The literature on the embryology of genera placed in the Haemo- doraceae by various taxonomists is reviewed. In a number of cases a knowledge of the embryological characteristics has contributed to a better understanding of the relationships of these genera. Embryological investigations in Dilatris pillansii Barker have shown that a periplasmodium is formed in the anthers and the pollen-grains develop successively; the ovule is sessile, orthotropous, bitegmic and crassinucellate; a parietal cell and a row of four megaspores are formed, the embryo-sac development is of the normal type, and the endosperm development helobial. A small chalazal haustorium develops from the primary basal endosperm cell. The information of Dellert on the embryological development of Wachendorfia paniculata Burm. is supplemented. The embryological investigations indicate that (1) Dzlatris and Wachendorfia are very close in their relationships, (2) these two genera are also closely related to Xiphidium and Anigosanthus, and (3) that these genera stand closer to the tribe Hypoxideae than to any other group. Embryological results therefore support the placement of Dilatris and Wachendorfia together in one tribe and also of the four genera in one family as Bentham and Hooker and Hutchinson have done. It has been 1epeatedly suggested in the past that the Haemodoraceae is not a natural family and probably not of monophyletic origin. Baillon (1891), for example, states that it is a heterogeneous family which pro- bably ought to disappear for being composed of fragments from divers groups. Engler (1897) calls it an unnatural family, the genera of which should be moved partly to the Liliaceae and partly to the Amaryllidaceae. Lotsy (1911) agrees with Pax (1889) and thinks it to be “‘wenig ein- heitlich”’, and most probably of polyphyletic origin; ‘‘ja, sie ist méglicher- 41 42 The Journal of South African Botany. weise mit der Zeit ganz aufzulésen und unter die Liliaceae und Amarylli- daceae zu verteilen”. Pax (1930) considers it either an old family of which a few remaining groups have been left, or “eine nicht einheitliche gestaltete Familie polyphyletischen Ursprungs”’. On the other hand, Hutchinson (1944) considers it a natural family as he has delimited it. He states, “. . . I am firmly of the opinion that the two tribes” (i.e. Haemodoreae and Conostyleae which Pax separated into different families) “should be again associated to form the Haemo- doraceae. Thus constituted, but with the genus Aletris removed to the Liliaceae, the family is natural and homogeneous, not only in its facies, but in its general distribution which is predominantly austral, . . .” The delimitation of the family by various systematists has been very different. See the summary in Table I. REVIEW OF THE EMBRYOLOGICAL LITERATURE. In an attempt to discover the natural affinities of the Haemodoraceous genera, the embryological development of some has already been worked out. As could be expected, this information has lead to some valuable suggestions about the relationships of these genera. Stenar (1927, 1938) has shown on embryological evidence that a close relationship exists between Xiphidium and Aniyosanthus, and Dellert (1933) has shown that Wachendorfia stands close to Anigosanthus. Both Dellert and Stenar further find enough evidence to suggest that these respective genera are also related to the tribe Hypoxideae of the Amarylli- daceae, and that they can be classed together. Pauridia has been found on embryological (de Vos 1948, 1949) and other evidence (cf. Schulze, 1893; Brackett, 1923) to be closely allied to some South African members of the Hypoxideae. This genus should be placed in this group—as has been done by Bentham and Hooker, Hutchin- son and by Baker in Thiselton-Dyer’s “Flora Capensis’”’. Cave (1952) has demonstrated that the early stages of the embryology of Odontostomum agree with that of Cyanastrum (Fries, 1919; Nietsch, 1941), which in Engler and Prantl’s system is the only genus of the family Cyanastraceae, and which Hutchinson on the other hand places in the Tecophilaeaceae together with Odontostomum and the four other genera of the tribe Conanthereae. According to Cave these results sub- stantiate Hutchinson’s placement of Odontostomum and Cyanastrum together in one family. The seed anatomy of Cyanastrum is unique in the Liliiflorae, as it has no endosperm, the embryo is very large, and the cells in the chalaza proliferate after fertilization to form a large tissue containing food reserves (Fries, Nietsch). It is unfortunate that stages atter fertilization in Odontostomum were not available for investigation. Studies of the Embryology and Relationships of 43 South African Genera of the Haemodoraceae. Taste I. Tae DeLiirarion oF THE HAbMODORACEAE BY VARIOUS 'TAXONOMISTS. In Bentham Engler Baker in | Adam- Endlicher & & Hutchin- Flora son | Phillips} Dis- 1836-40, Hooker, Prantl, son, Capensis, & 1951. | tribu- 1842. 1883. 1930. 1944, 1896. Salter, tion. 1950. Haemodorum .. Hae Hae-Euhae Hae Hae-Haem Austr. Wachendorfia .. se 5 “A a5 Hae-Euhae}| Hae Hae S. Afr. Schiekia - s 2 Amer. Hagenbachia i 7 99 3 Amer. Dilatris. . 5 60 5 5 55 55 S. Afr. Lachnanthes % 5 a 55 Amer. Barberetta Fe % - Fa S. Afr. Xiphidium 0 90 90 99 Amer. Lanaria Bs a Ama-Cono > Ama |S. Afr. Phlebocarya 39 o 90 op Austr. Pauridia .. | ? Hypox |?Ama-Hyp Hae Hypox | Ama-Hyp | Ama Hae |S. Afr. Tribonanthes .. Hae Hae-Cono | Ama-Cono | Hae-Cono Austr. Conostylis 90 ; 50 61 Dn Blancoa ae 35 ; 35 ss bp Anigosanthus .. is 56 56 96 ” Macropidia a - a 33 Lophiola os 33 35 3 Amer. Aletris .. 3 * Lil-Ale Lil-Nar Amer., Asia Peliosanthes Smi Hae-Oph | Lil-Mon Lil-Pel Asia Ophiopogon i 5 5 Lil-Oph 59 Liriope .. ay ras Aa . a Sansevieria Lil ; Lil-Dra Agav Hae-Oph Lil Afr., Asia Conanthera x Hae-Con | Ama-Con Teco Amer. Cyanella 5 35 <5 = Hae-Con Hae Ama |S. Afr. Zephyra bs = is AA Amer. Tecophilaea Tr. aff. - a. a Py Odontostomum re Lil-Asph 3 ss Key ro ABBREVIATIONS.—Families: Agav(aceae), Ama(ryllidaceae), Hae(modoraceae), Ir(ideae), Hypox(idaceae), Lil(iaceae), Smi(laceae), Teco(philaeaceae). Subfamilies: Ale(troideae), Mon(doideae). Tribes: Asph(odeleae), Con(anthereae), Cono(stylideae) or Cono(styleae), Dra(caeneae), Euhae(modoreae) Haem(odoreae), Hyp(oxideae), Nar(thecieae), Oph(iopogoneae), Pel(iosantheae). Cyanella, the only other genus of Hutchinson’s Tecophilaeaceae which has been studied embryologically (de Vos, 1950), shows a somewhat similar early development: no pe:iplasmodium develops in the anthers and the walls in the p.m. cells form simultaneously!; the ovule shows a parietal cell, normal type of embryo-sac development, and nuclear endo- sperm development. It differs in the possession of many ovules in each chamber many of which ripen, and from Cyanastrum in the possession of a well developed endosperm. this group are needed. 1de Vos, unpublished. More embryological investigations in NATIONAL MARY GUNN LIBRAR} BOTANICAL PRIVATE BAG X 101 PRETORIA 0001 REPUBLIC OF SOUTH AFRICA INSTITUTE 44 The Journal of South African Botany. Sansevieria, a genus which Bentham and Hooker placed in the Haemo- doraceae, shows in its embryological development its close relationship with Dracaena (Stenar, 1942; Wunderlich, 1950) and not with the Ophiopo- goneae as Bentham and Hooker thought. This supports Engler and Prantl and Hutchinson who group these two genera together. Aletris, now usually considered to be a genus of the Liliaceae, has been studied by Ono (1928, 1929, cited by Schnarf, 1931). According to Schnarf’s review, Aletris has been found to possess a parietal cell, normal type of embryo-sac development, and helobial endosperm. Whether Ono was able to make any suggestions about the relationship of this genus is not known, as his papers have not been available. These are the only Haemodoraceous genera which have been examined embryologically up to now, as far as could be found out. From this short review it is’ clear that the Haemodoraceae as constituted by most systematists is probably not a homogeneous group. More embryological investigations, the term used in its broadest sense as defined by Cave (1953), to include both sporophytic and gametophytic structures that are concerned in sexual reproduction. are highly desirable in other genera of the Haemodoraceae, especially in the Australian and New World — groups. Such information will probably help greatly in discovering the natural affinities of the genera. In the present paper the development of the ovule, seed and pollen in Dilatris is described. As Dellert’s description of the embryology of Wachendorfia is not complete because of the scarcity of material she had available, and as I had already started investigating Wachendorfia before — Dellert’s paper came to my notice, a short revision of the embryology of this genus is also given. This leaves only Lanaria and Barberetta of the South African genera still to be studied. INVESTIGATION. Method.—Young flower buds and older isolated ovaries were fixed in FAA. In still older stages ovules were dissected out before fixation to get rid of the hairy ovary walls. After embedding in paraffin wax sections were cut 10—15 yw in thickness, and stained either with Heiden- hain’s iron-alum haematoxylin or with Delafield’s haematoxylin. DILATRIS PILLANSIT BARKER. Pollen. Before reduction division in the pollen-mother-cells the tapetal cells become binucleate and start migrating into the pollen-sacs to form a periplasmodium. This reaches its maximum development when the young pollen-grains are separating after the reduction division Studies of the Embryology and Relationships of 45 South African Genera of the Haemodoraceae. and forming an exine. Remains of the periplasmodium are still to be seen some time after the exine has developed. The cell-wall formation in the p.m. cells is successive, the first wall being formed after the first reduction division. The generative cell which is very small and more or less pear-shaped, is formed in one corner of the pollen-grain against the end of the distal wall (Fig. 25). The vegetative nucleus, which at this stage lies near the generative cell, is much larger and often shows two nucleoli. The nuclei of the two cells stain to about the same intensity. Position and form of ovule.—The ovary is inferior and three-chambered and has a single ovule on a large axile placenta situated in the upper part of each chamber. The ovule is sessile and is attached by means of its broad chalaza to the placenta. It is orthotropous but slightly bent and pendulous, the micropyle pointing downwards (Figs. 2—4). A sharp dent occurs in the integuments on one side near the micropyle. A broad vascular bundle containing a large number of scalariform tracheids runs from the placenta into the chalaza. Here it expands and lies in contact with the entire broad base of the nucellus (Fig. 4). It does not branch off into the integuments. The placenta enlarges rapidly and forms a thick cushion-like out- growth round the base of the ovule, and at the time of fertilization the micropyle nearly touches it (Figs. 3A, B). It can be compared with the aril found in various plant groups, but it does not enlarge to form a third integument. After fertilization the ovule overtakes it in growth. Fies. 1—6. Dilatris pillansii, sections through developing ovule and seed. Fia. very early stage with nucellus and primordia of integuments, x 80; 2, slightly older with megaspore-mother-cell, developing integuments, and placental outgrowth connected below with the outer integument, in median longitudinal section, x 80; 3A—B, mature embryo-sac stage; 3A, median longitudinal section with the lower part of the outer integument and placental outgrowth con- nected; 3B, near median section, two sections removed from 3A, the outer integument free from the placental outgrowth, « 50; 4, ovule after fertiliza- tion showing enlarging embryo-sac and disappearing nucellus, near median, x 50; 5, near median section of developing seed with endosperm tissue, small embryo, and degenerating chalazal haustorium, x 35; x denoting spaces not filled up with tissue; 6, longitudinal section of fertile ovary chamber showing nearly ripe seed and part of flange; dotted lines indicate where the carpel walls break at seed dispersal, x 8; 6A & 6B, cross-section through ovule and placental outgrowth to show connection: 6A, through lower part of ovule, 6B, through micropylar region, x 11; a, placental outgrowth (aril); en, endosperm; es, embryo-sac; fl, flange; 7, inner integument; m, micropyle; nu, nucellus; oz, outer integument; ov, ovary wall. Integuments—The two integuments develop from the base of the young nucellus, the inner one beginning its development slightly before the outer. Both consist of two cell layers, except at the base of the outer integument and the tip of the inner, where they are thicker. The micropyle is usually formed by the inner integument only, as the outer 46 The Journal of South African Botany. de Vos, Embryology of the Haemodoraceae Studies of the Embryology and Relationships of 47 South African Genera of the Haemodoraceae. is slightly shorter than the inner on the side away from the placenta (Figs. 3A, B). In a few cases the outer integument overtops the inner and forms a very short continuation of the micropylar canal (Fig. 4). Where the pendulous ovule touches the placenta, the outer integu- ment is attached to the placental outgrowth by means of a narrow strip of parenchymatous tissue in the median longitudinal line (Figs. 2—4, 6A, B). This begins to develop in an early stage and is still to be seen in the nearly ripe seed. (The seed shown in Fig. 5 still has it, two sec- tions removed from the section figured—on account of its narrowness a section not quite median does not show the connection.) This con- nection perhaps indicates that the orthotropous condition of the ovule is not primary, having been developed from an anatropous condition by reduction. In a median longitudinal section it looks somewhat like the raphe of an anatropous ovule, except that no vascular bundle runs through it. No reference in the literature could be found to any such state in other orthotropous ovules. Nucellus and embryo-sac.—-The nucellus begins its development as a hemispherical projection from the placenta. A subepidermal archesporial cell is differentiated which divides periclinally to form a megaspore- mother-cell and a parietal cell. A small cell-complex consisting of several layers originates from the latter by periclinal and anticlinal divisions. The epidermal cells at the tip of the nucellus divide once periclinally. Ultimately the megaspore-mother-cell therefore lies embedded under five or six layers of nucellar tissue (Figs. 7—10). Fires. 7—14. Dilatris pillansii, development of female gametophyte. Hic. 7, subepidermal archesporial cell; 8, megaspore-mother-cell and parietal cell; 9, slightly older megaspore-mother-cell and developing nucellar cap formed by divisions in the parietal cell and nucellar epidermis; 10, still older stage with megaspore-mother-cell in Prophase I; 11, four megaspores formed after reduc- tion division; note the palisade-shaped nucellar epidermal cells lower down; 12, mature embryo-sac, the egg-cell not shown; nucellar cells beginning to dis- integrate; 13A & B, base of embryo-sac in two adjacent sections showing enlarging antipodal cells, with primary endosperm nucleus in 13B; 14, base of embryo-sac with two basal endosperm cells (developing chalazal haustorium), two hypertrophied antipodal cells on a small postament, and one free nucleus of the upper endosperm chamber; all x 350; an, antipodal cells; be, basal endosperm cells (chalazal haustorium); p, postament. The megaspore-mother-cell elongates and undergoes reduction divi- sion. Four megaspores are formed in a row, and the chalazal one develops into a normal 8-nucleate embryo-sac (Figs. 11—13) which at first has a narrow pointed base. The three large antipodal cells are inversely pear- shaped and lie in one level with their pointed bases directed towards the chalaza. The two polar nuclei lie just above them and sometimes in contact with them. At the micropylar end the egg-apparatus shows no extraordinary features. Before fertilization the embryo-sac does not The Journal of South African Botany. 48 a Embryology of the Haemodoraceae Vos, Studies of the Embryology and Relationships of 49 South African Genera of the Haemodoraceae. enlarge much and takes up only a small space in the upper part of the nucellus. Fertilization was not seen, but remains of what probably was the pollen-tube were found in the micropyle of one ovule. The nucellus is large and crassinucellate. At first 1t is ovoid but later becomes somewhat pear-shaped with a massive base and a pointed tip. The widening of the base is brought about by an elongation ot the epidermal cells in the anticlinal direction in the lower half of the nucellus (Figs. 3A, 4, 12). These cells elongate up to six or seven times their width, thus becoming palisade-shaped. In early stages the base of the nucellus consists of small meristematic cells rich in protoplasmic contents. It is sharply defined from the chalaza, as the vascular elements in the latter end abruptly against the meristematic base of the nucellus. The developing embryo-sac absorbs the nucellar cells along its sides as far as the nucellar epidermis. After fertilization this process is accelerated and somewhat later only the nucellar cap at the micropyle consisting of the nucellar epidermis, and the palisade-shaped epidermal cells near the base are left. Still later they too are absorbed and no nucellar tissue remains in the ripening seed. The antipodal cells increase in size at about the time of fertilization and then disappear gradually. Before their disappearance the adjacent nucellar cells are absorbed by the embryo-sac, leaving the antipodals on a small pedestal of nucellar cells (the “postament’”’, Schnarf, 1929, p. 59) (Fig. 14). On account of the faster absorption of nucellar cells towards the chalaza a small chalazal endosperm haustorium is formed, and the postament with the antipodal cells is left on the farther side where they degenerate slowly. At the same time the longitudinal axis of the embryo-sac changes so that the haustorium which at first is some- what lateral, becomes basal. Endosperm. —The two polar nuclei fuse at about the time of fertiliza- tion and after fertilization the primary endosperm nucleus moves from its position near the antipodal cells to that side of the embryo-sac where the small haustorium is being formed towards the placenta (Fig. 13B). The endosperm formation is helobial. Although the first division of the primary endosperm nucleus was not observed, a small basal endo- sperm cell is formed after this mitosis. Two, and shortly afterwards four very large basal cells are formed from the first one by vertical divisions (Fig. 14, 15). They contain much protoplasm and each has at least two large hypertophied nuclei. They form the small chalazal haustorium which projects towards the placenta. After the basal nucellar tissue has been absorbed they nearly touch the chalazal vascular bundle and 50 The Journal of South African Botany. probably help to convey food materials to the developing endosperm. Later, when the seed has reached its full size but is not yet quite ripe, these four cells are absorbed. In the large upper cell formed after the first division of the primary endosperm nucleus, repeated nuclear divisions take place and a large number of nuclei are formed in the thin layer of cytoplasm which lies against the wall of the embryo-sac (Fig. 15). In one case one of these nuclei and some cytoplasm was seen to intrude into one of the degenerating antipodal cells. Cell-wall formation is progressive from the base towards the micropyle. Later the embryo-sac becomes filled with endosperm tissue by repeated nuclear and cell-divisions. Fras. 15—25, Dilatris pillansii, development of embryo and seed, and gametogenesis in anther. Fic. 15, submedian longitudinal section of ovule after fertilization showing two of the four cells of the chalazal haustorium and numerous endo- sperm nuclei, x 350; 16, zygote attached to nucellar cap, x 350; 17, embryo in quadrant stage, x 350; 18, embryo in octant stage, with only two of the four terminal cells shown, x 350; 19, embryo in a later stage where the deriva- tives of the first apical and first basal cell can still be identified, x 350; 20, longitudinal section of still undifferentiated embryo of nearly ripe seed, with short, bent suspensor and elongated central cells, x 250; 21, embryo, same as in Fig. 20, showing its position in a small sac or diverticulum, with sur- rounding endosperm and integuments, < 50; 22, section through seed-coat of nearly ripe seed, x 250; 23, section through outer layers of endosperm and adjacent seed-coat, x 250; 24, transverse section through pollen-sac showing periplasmodium and pollen-mother-cells in interphasé of reduction division, x 350; 25, young pollen-grain with generative and vegetative cell, x 400; cu, cuticular layers; en, endosperm; 7, inner integument; oz, outer integument. Embryo.—After the formation of cell-walls in the endosperm the fertilized egg-cell begins to divide. ‘The first division is transverse and this is followed by a vertical division in the apical cell and a transverse one in the basal cell (Fig. 17). The octant stage is reached after another vertical division in both apical cells at right angles to the first, a vertical division in the middle cell, and a transverse one in the basal cell (Fig. 18, where two of the four terminal cells are shown). The cell-divisions after this stage were not followed. The derivatives of the apical cell of the two-celled proembryo are richer in protoplasmic contents than those of the basal cell; they divide repeatedly and form the larger part of the embryo. The cells derived from the middle cell of the four-celled pro- embryo also take part in the embryo formation, which therefore con- forms to the Asterad or to the Crucifer (Onagrad) type. The suspensor which is derived from the basal cell of the four-celled proembryo, remains short and consists of three or four cells. In the ripe seed the embryo is oval in longitudinal section and un- differentiated except for a slight elongation of the central cells. It lies in a small sac or diverticulum formed by the micropylar tip of the embryo- sac which is folded flat against the enlarging endosperm. It is covered by a thin layer of endosperm on the outside and the suspensor is bent (Rigs. 20, 21). Studies of the Embryology and Relationships of 51 South African Genera of the Haemodoraceae. de Vos, Embryology of the Haemodoraceae St tbo The Journal of South African Botany. Seed.—Usually only one seed matures in an ovary, the other two ovules aborting. Degeneration of these ovules usually takes place after the embryo-sac has developed but may sometimes occur earlier. One ovule, for example, was found with four disintegrating megaspores. When degeneration occurs in the mature embryo-sac, the egg-apparatus usually disintegrates while the antipodal cells become hypertrophied and may increase up to 70 p» in length, with nuclei 35 yw in diameter. After fertilization the shape of the viable ovule changes completely from a more or less spherical to a round inverted watchglass-shaped structure with a concave lower (chalazal) and a convex upper surface. This is caused by an enormous growth in width so that its transverse diameter is later about four times its height. It fills the entire ovary chamber completely, and the micropylar tip becomes flattened against its convex upper surface, thus forming the small sac in which the embryo is situated (Figs. 5,6, 21). The large placenta pushes against the concave lower surface. The chalaza and hilum are in the centre of this concave surface, and the axis of the seed from chalaza to micropyle is oblique. Barker (1940) has observed that the ovule “.. . is clasped by a flange which projects inwards round the entire edge of each carpel. The carpel-wall containing the ripe seed thus held in place by the flange begins to separate from the receptacle, starting at the base and con- tinuing upwards, until it finally breaks away completely with the perianth attached to the apex, thus forming a mechanism suitable for dispersal by wind.” The present developmental studies show that the flange is formed by an enlargement and lignification of a zone of cells of the septa separat- ing the fertile ovary chamber from the others, and of the axis above and below the placenta (Fig. 6). This flange then holds the seed in place when the fertile carpel with its seed breaks away from the receptacle by a tear which develops in the thin part of the septum next to the flange (the tear is indicated by dotted lines in Fig. 6). The placenta is left behind attached to the receptacle. The seed-coat is formed mainly by the outer epidermis of the outer integument (Fig. 22). These cells become palisade-shaped and contain a brown substance. The inner layer of cells of this integument and both layers of the inner integument become crushed and disorganized (Fig. 23). ° Three layers of cuticle are present: one on the outside, the second between two of the crushed layers (i.e. between the two integuments), and the third just inside the seed-coat. The endosperm in the ripe seed is copious and contains starch, protein and oil. The cell-walls are thin, except for the outer layer of endosperm cells (Fig. 23) which have very thick outer walls which give a positive reaction with Schultze’s solution. These cells contain small, round Studies of the Embryology and Relationships of 53 South African Genera of the Haemodoraceae. globoid-like bodies which give neither a starch, protein, nor oil reaction. It is uncertain what their nature is. Germination.—Seeds collected between January and March 1949 from the mountain slopes of Kleinmond, Caledon Division, were kept in wet vermiculite under ordinary laboratory temperatures from the end of April. Germination was very slow and at the end of July of that year only 4 per cent of the seeds had germinated. The experiment was then stopped on account of loss of most of the seeds through fungal attack. The undifferentiated condition of the embryo in the mature seed probably requires this long germination period. WACHENDORFIA PANICULATA BURM. Before Dellert’s paper on the embryology of Wachendorfia paniculata came to my notice, I sectioned and stained preparations to study the embryological development of this plant. As this series is complete showing stages not seen by Dellert, some supplementary information about the embryology of this plant is given. The present investigation confirms Dellert’s observations, including some points about which she was not quite sure, e.g. that the pollen-grains are formed successively, that the parietal cell-complex she observed in the young ovule is formed from one primary parietal cell which is cut off from a subepidermal archesporial cell (Figs. 26—28), that the embryo-sac development is of the normal type, and the development of the endosperm helobial. Fies. 26—44, Wachendorfia paniculata, development of ovule and seed. Fic. 26, subepidermal archesporial cell; 27, megaspore-mother-cell and parietal cell; 28, enlarging megaspore-mother-cell and parietal cell-complex; 29, second metaphase of reduction division; nucellar epidermis two-layered at tip; 30, 31, rows of four megaspores; in 31 the upper megaspore lies upon the second one; 32, four megaspores, three of them with two nuclei; chalazal megaspore is functional; 33A & B, young embryo-sac with four nuclei: 33A, two nuclei in upper half, with disintegrating megaspore above them; 33B, lower half; 34, egg-apparatus; 35, lower half of mature embryo-sac with antipodal cells and polar nuclei; 36, embryo-sac after fertilization with a large upper and a small basal endosperm cell; 37, basal endosperm cell; 38, embryo-sac with a number of free endosperm nuclei in the upper chamber; 39, stage in cytokinesis, with cells in the lower half of the upper endosperm chamber and free nuclei in the top half; 40, chalazal haustorium in cross-section; 41, young embryo consisting of about twelve cells; 42, slightly older embryo with suspensor consisting of two cell rows; 43, section through seed-coat and outer layer of endosperm of nearly ripe seed; 44, young pollen-grain with generative and vegetative cell. Fries. 26—35, 37, 41, 42, x 350; 36, 38, 39, x 80; 40, x 200; 43, x 250; 44, x 400; cw, cuticular layer; en, endosperm; 7, inner integument; oi, outer integument. The following describes some of the embryological characteristics not observed by Dellert: Pollen.—In the young pollen-grains the generative cell which is small and more or less pear-shaped, is formed against the distal wall nearer one corner (Fig. 44). The Journal of South African Botany. 54. ad Heke de Vos, Embryology of the Haemodoraceae Studies of the Embryology and Relationships of South African Genera of the Haemodoraceae. Or on Ovule.—The ovule is hemitropous and sessile. Four megaspores are formed in a row (Figs. 30, 31). The upper one, however, often lies slightly out of line in an oblique position (Fig. 31) on account of the second reduction division spindle in the upper interphase cell being oblique (Fig. 29). This is probably the reason why Dellert could not find more than three megaspores in her material. Often more than one megaspore starts developing, e.g. in Fig. 32, where three of the mega- spores have two nuclei. The chalazal one usually gets the upper hand and the other megaspores degenerate very slowly, so that in some ovules where the embryo-sac is in the four-nucleate stage, remains of them are still visible (Figs. 33A and B). Dellert’s suggestion that the development of the endosperm is most probably helobial, is correct (Figs. 36, 37). In the upper endosperm chamber which is formed after the first division of the primary endo- sperm nucleus, a large number of free nuclei are formed (Fig. 38). Cyto- kinesis starts from the base upwards, so that a few cases were seen with the lower half of the embryo-sac already in a cellulai state, but the wider upper part still nuclear (Fig. 39). The primary basal endosperm cell divides to form four large cells with hypertrophied nuclei and much cytoplasm. In a transverse section through the ovule these cells lie in a row (Fig. 40). They form a chalazal haustorium which gradually absorbs the nucellar cells that separate them from the vascular bundle in the chalaza, and later only some crushed cell-walls lie between them and the vascular bundle. The anti- podal cells degenerate early, and no postament is formed. Dellert has suggested that the endosperm cells above the four cells of the chalazal haustorium, which in one stage look different from those near the micropyle, probably also arise from the primary basal cell. This, however, has not been confirmed. Although these endosperm cells at first look different from the upper ones, they also seem to originate from the free nuclei of the upper endosperm chamber (cf. Figs. 38, 39). The primary basal endosperm cell forms only the four cells of the chalazal haustorium. The embryo-sac widens considerably as the endosperm develops, and later becomes reniform or sometimes oval in longitudinal section, with the embryo and micropyle at the lower end and the hilum in the middle of one side, the concave side when reniform. The chalazal haustorium later is pulled away from the chalazal region and lies in a space which is still to be seen in some of the mature seeds. Later the haustorial cells are absorbed together with some of the neighbouring endosperm cells. The endosperm of the mature seed does not show the two regions seen earlier. Like Dilatris it is, however, differentiated into a peripheral 56 The Journal of South African Botany. l-celled layer with thick outer walls which give a cellulose reaction, and an inner mass of cells which contain starch, protein and oil as reserves. The outer layer of endosperm contains protein but hardly any starch. The earliest stages of the division of the zygote were not found. A later stage shows a young undifferentiated embryo with a short suspensor which consists of one, and a little later two rows of cells (Figs. 41, 42). In the ripe seed the embryo is differentiated and shows a slender terminal cotyledon and a shoot growing-point about half-way down its length. The suspensor remains short, stout and straight. In these points the embryo therefore differs considerably from that of Dilatris. The seed is covered with brown bristly hairs or papillae which are formed by the cells of the outer integument growing out to form long tubular projections. The inner layer of this integument is very thin and the cells can hardly be recognized. The inner integument still consists of two cell-layers which have turned brown. Only two cuticular layers are present: between the two integuments and inside the inner one. The outer cuticle has disappeared on account of the external papillose pro- trusions from the outer epidermis. The nucellus has disappeared completely. One, two, or all three seeds of the ovary may ripen, most frequently two. DISCUSSION. Relationship between Dilatris and Wachendorfia.—Tables II and III summarize the embryological characteristics of those genera of the tribes Taste I]. Tue most Important EMBRYOLOGICAL CHARACTERISTICS OF GENERA OF THE ‘TRIBES HAEMODOREAE, CONOSTYLEAE AND HyPOXIDEAE. Dila- | Wachen-] Xiphi- |Anigos-| Hy- | Cureu- | For- Pauri- tris. dorfia. | dium. | anthus.| poxis. | ligo. | besia.4 |Ianthe.*} dia. Periplasmodium in pol- len sacs at So at + + + + + +3 +3 +8 Development of pollen | suce suce suce suce succ suce prob? | prob*® | prob® suce suce succ Integuments .. se 2 2 2 2 2 2 2 2 2 Nucellus crassinucellat or tenuinucellate .. cras eras cras cras ten. ten ten ten ten Parietal cell .. “ie + + —- + 0 0 0 0 “0 Megaspores, number and arrangement .. 4, lin 4, lin 45 lin 45 Por) 4.2 3 Anal A ae lin Embryo-sac develop- ment . - ae ie NT NT NT NT NT NT |Allium} NT NT Endosperm development | He He He He He, Nu Nu Chalazal haustorium .. -- -- + + &O 0 1 Now Empodium Salish. 2? Now Spiloxene Salisb. 3de Vos, unpublished. Studies of the Embryology and Relationships of 57 South African Genera of the Haemodoraceae. Haemodoreae, Conostylideae and Hypoxideae, where these characteristics have been worked out. In Table If such embryological characteristics are tabulated which have often been considered of importance to show relationships (cf., e.g. Schnarf and Wunderlich, 1939, and Cave, 1953), and in Table III some other points of similarity and difference are tabulated which, taken together with the points in Table IT, also indicate relationships (Maheshwari, 1950, p. 357). From these tables it is clear that the embryological characteristics of Dilatris and Wachendorfia are remarkably similar, except for the shape of the ovule and the structure of the embryo and seed-coat (not tabulated). Embryological evidence therefore supports the placing of these two genera in one family, as systematists have always done in the past, or even one tribe, as Bentham and Hooker and Hutchinson have done— although the ovary is superior in Wachendorfia and inferior in Dilatris. TaBLE II]. Somer orHER EMBRYOLOGICAL CHARACTERISTICS OF GENERA OF THE TRIBES HAEMODOREAE, CoNOSTYLEAE AND HyPoOXIDEAE. Dila- |Wachen-| Xiphi- |Anigos-| | Hy- | Curcu-| F[or- Pauri- tris dorfia | dium. | anthus.| poxis. | ligo. besia. | Ianthe.| dia. Position of generative cell in pollen (distal or proximal) Be dist dist Funicle .. 0 0 short. | short + ak + Ovule .. sas £0 orth hemi orth orth ana ana hemi ana hemi Micropyle directed downwards .. + on on oe Integuments largely 2- layered ag + + + more + + Inner integument wider at tip ar ale + + Integuments which form micropyle , inner inner both both both both Parietal. _ cell forms several layers + + a 0 0 0 0 0 Nucellar epidermis 2 or more-layered at tip.. - + = az + + = Nucellar epidermal cells palisade-like towards base .. a8 ae aL ae = = = Antipodals large, at same level ee + =e + Antipodals degenerate late early at fert | early +late | late or| late early Number of cells in chalazal haustorium 4 4 1 ] 1 or 0 0 Embryo small in ripe seed sf + — + Lf) atte ae 4 Suspensor bent 2 cell short | short? short | short | short rows 5 = yo —, = ~- wee! Authority present] Dellert | Stenar | Stenar Stenar de Vos investi-| 1933 & | 1938 1927 1925 1948, 1949 gation | present investi- gation 58 The Journal of South African Botany. Taste 1V. Some MorrpHonocicAL CHARACTERISTICS OF GENERA OF THE TRIBES HAEMODOREAE, CoNOSTYLEAE AND HyPoOxXIDEAE. Dila- | Wachen-| Xiphi- | Anigos-| Hy- | Cureu- | For- Pauri- tris. dorfia. | dium. | anthus.} poxis. | ligo. | besia. | Ianthe.| dia. Stamens a ak: 3 3 3 6 6 6 6 6 3 Ovary .. a of inf sup sup inf inf inf inf inf inf Ovules per chamber .. 1 ] (oo) io) oa) 6—8 oO ca) fo-0) Placenta Be .. | large large large | large | small | small | small | small | small Of the two genera Dilatris shows more advanced characteristics, both morphological and embryological, viz., the inferior ovary, the very large placenta to which the ovule is attached by an extra strip of parenchymat- ous tissue, the flattened watchglass-shaped seed, the very pronounced palisade-like cells of the nucellar epidermis, and perhaps also the un- differentiated embryo. ; Relationship between Dilatris, and Wachendorfia, Xiphidium and Anigosanthus.—Dellert has shown on embryological grounds that Wachendorfia stands close to Anigosanthus, and Stenar (1938) similarly shows that Xiphidium and Anigosanthus are closely related. In Tables II and III their results are summarized. Although the embryological development of Xiphidium and Anigosanthus has not been worked out fully, there seems to be enough evidence to agree with their views and, as shown above, to add Dilatris to this group of related genera. The anatomical investigations of Schulze (1893) also showed the two Cape genera, Wachendorfia and Dilatris, to be related to the American genera Xiphidium and Schiekia. The most important differences in the ovary and anthers known at present in the four above-mentioned genera are not embryological but morphological: (1) difference in the number of ovules per chamber. This need not be considered a character of great taxonomic importance and may only indicate which of the genera are more advanced in this respect. (2) Difference in the number of stamens. It has become evident that this also need not be considered of very great taxonomic importance in certain groups, as similar reductions have been found to occur within several other homogeneous groups of the Liliiflorae, e.g. in the Hypoxideae where Pauridia with three stamens is in all other respects (general morphology, anatomy, embryology, seed development and pollen morphology, cf. Schulze 1893, Brackett 1923, de Vos 1949, Erdtman 1952) similar to certain Cape Spiloxene (Ianthe) species. (3) Difference in the pollen morphology. Wachendorfia, Xiphidium and Dilatris have monosuleate grains with the sulcus sometimes operculate, and Anigosanthus has Studies of the Embryology and Relalionships of 59 South African Genera of lhe Haemodoraceac. “jsopolar(?) to subisopolar 2-aperturate” grains (Erdtman 1952, pp. 46, 198). KErdtman does not state whether he considers these characters to be related and only refers to Hutchinson who places the Conostyleae (to which Anigosanthus belongs) in the Haemodoraceae “as a very distinct tribe’. The similarity in the embryological characteristics of these four genera is such that they must be closely related and it warrants their being placed in one family at least. This is therefore against Engler and Prantl’s system (1930) in which Wachendorfia, Dilatris and Xiphidiwm are placed in the Haemodoraceae and Anigosanthus in the Amar yllidaceae- Hypoxidoideae, but supports workers like Bentham and Hooker and Hutchinson, who place all four genera in one family. Unfortunately, these four genera are the only members of the tribes Conostyleae and Haemodoreae that have been examined embryologically. It is quite probable that some of the other genera of these two tribes will show somewhat similar embryological characteristics. In this respect it is interesting to note that Conostylis also has hemianatropous ovules according to Schnizlein (1860), and that Haemodorum has two hemiana- tropous ovules on thick placentae in each chamber, with a small embryo at a distance from the hilum (B. & H. III, p. 673). On the other hand, there is Lanaria which has been placed by Bentham and Hooker and by Hutchinson in the (Ku)haemodoreae and by Pax and Hoffmann in the Conostylideae. Preliminary investigations show that the embryology of this monotypic genus differs from that of the four genera discussed above and also from the tribe Hypoxideae. It has, for example, no periplas- modium and the pollen-grains develop simultaneously. More embryo- logical investigations in these tribes are therefore necessary. Relationship of the four genera with the Hypoxideae.—Dellert has also found embryological evidence to suggest that Wachendorfia, Anigosanthus, and the members of the tribe Hypoxideae can belong to one family. Stenar (1938) agrees and adds Xiphidium to the group, as mentioned above. From Tables II and III some important points of similarity are evident between the Hypoxideae and the four genera of the Conostyleae- Haemodoreae group: (1) the presence of a periplasmodium which, of all the Liliiflorae, has been found in these genera only; (?\ successive cell- wall formation in the p.m. cells; (3) the presence of two integuments (which is of such common occurrence in the Lilliflorae that not much importance can here be attached to it); (4) normal type of embryo-sac development; (5) helobial type of endosperm development, except for a few South African Hypoxideae. Dellert also mentions the antipodals 60 The Journal of South African Botany. which degenerate early. But in Dilatris they degenerate late and there- fore no similarity in this characteristic exists. On the other hand some rather important differences between the embryological characteristics of the Hypoxideae and the Conostyleae- Haemodoreae group are evident, the most important being the absence of a parietal cell in the ovules of the Hypoxideae, the smaller size of the antipodals, and the smaller placentae. The pollen morphology also differs (Erdtman, 1952). Dellert discusses the first two points and con- cludes that they are not of such importance as are the points of similarity, and Schnarf (1931, p. 14) states that the nature of the antipodals has only a limited systematic value. Pax and Hoffman (1930) place the four tribes, Alstroemerieae, Hypo- xideae, Conanthereae and Conostylideae, in the subfamily Hypoxidoideae of the Amaryllidaceae. But the problem whether this subfamily is homogeneous, still needs further investigation. Comparative embryo- logical and pollen investigations indicate a possible polyphyletie origin: the development of the pollen and ovule in the Alstroemerieae differs considerably fiom that of both the Hypoxideae and Conostylideae (Stenar, 1925, 1938 pp. 227-278). In anatomy the Alstroemerieae also differs from the Hypoxideae (Schulze, 1893, p. 382; Schnarf, 1892, p. 325). The embryology of the Conanthereae has not been worked out satisfactorily, the only genera investigated so far being Odontostomum (Cave, 1952) and Cyanella (de Vos, 1950) These investigations indicate that their develop- ment follows another pattern (or patterns). Further investigation is therefore still needed. The anatomy (Schulze, 1893) and the pollen morphology of the Conanthereae also differ: Erdtman (1952, p. 46) has stated that the pollen morphology of the Conanthereae supports the establishment of a separate family, the Tecophilaeaceae. The embryological investigations- so far done indicate that the three genera, Wachendorfia, Dilatris and Xiphidium, of the Haemodoreae and the one genus, Anigosanthus, of the Conostyleae stand closer to the Hypoxideae than to either the Alstroemerieae or the Conanthereae. According to the comparative embryological evidence available, the four genera and the Hypoxideae can possibly be included in one family. OPSOMMING. ’n Oorsig word gegee van die afgrensing van die Haemodoraceae deur verskillende taksonome, asook van die literatuur oor die embriologiese ontwikkeling by genera van die familie. In sekere gevalle het hierdie kennis bygedra tot ’n beter begrip van die verwantskappe van die betrokke genera. Embriologiese ondersoek by Dilatris pillansii toon dat die tapetum van die helmknoppe ’n periplasmodium vorm wat tussen die jong stuif- Studies of the Embryology and Relationships of 61 South African Genera of the Haemodoraceae. meelkorrels indring, en dat die selwande van die jong stuifmeelkorrels suksessief neergelé word na die reduksiedeling. Die generatiewe sel van die stuifmeelkorrels word in een hoek teenaan die distale wand gevorm. In die vrugbeginsel is daar in elke hok een sittende, ortotrope, kras- sinusellate saadknop met twee integumente aan ’n groot plasenta. Uit die subepidermale archespoorsel ontstaan ‘n makrospoormoedersel en ’n dek-sel. L.g. deel om ’n klein dek-selle-kompleks te vorm bo die makro- spoormoedersel en die makrospore. Die epidermis van die nucellus aan die top deel ook periklinaal. Na die reduksiedeling van die makrospoor- moedersel word vier makrospore in ‘n ry gevorm, en die chalazale een is funksionerend. Die kiemsakontwikkeling is volgens die normale tipe. Die anti- podeselle is groot en lank en degenereer eers na die bevrugting. Die endospermontwikkeling is volgens die helobiale tipe. ’n Klein basale sel word afgesny en daaruit ontstaan vier selle wat hipertrofie ondergaan en ’n chalazale haustorium vorm. In die boonste endospermsel ontwikkel die egte endospermweefsel, eers deur die vorming van ’n groot aantal vrye kerne wat later met selwande omgewe word. Die embrio ontwikkel volgens of die Asterad- of die Crucifer-tipe. In die ryp saad is dit ongedifferensieer en die kort suspensor is gebuig. Die embrio lé in ‘n klein uitstulping of sakkie van die kiemsak wat platgedruk is teen die endosperm. Die saadhuid is hoofsaaklik uit die buitenste opperhuid van die buitenste integument opgebou. Meer informasie word gegee oor die embriologie van Wachendorfia paniculata wat alreeds voorheen ondersoek is (Dellert, 1933). Die embrio- logiese ontwikkeling van hierdie plant. is in ooreenstemming met die van Dilatris, behalwe dat die saadknop hemitroop is, en die kiem in die ryp saad gedifferensieer is, met ’n reguit suspensor wat uit twee rye selle bestaan; die bou van die saadhuid verskil ook effens. Hierdie embriologiese studie toon dat (a) Dilatris en Wachendorfia baie na verwant is, (6) hierdie twee genera ook na verwant is aan Xiphi- dium en Anigosanthus, en (c) dat die vier genera nader verwant is aan die tribus Hypoxideae as aan enige ander groep waarvan die embriologie’ uitgewerk is. Die plasing van Dilatris en Wachendorfia in een tribus en van die vier genera in een familie, soos deur Bentham en Hooker en Hutchinson gedoen, word dus deur die vergelykende embriologie bevestig. Volgens die embriologie te oordele sou die vier genera en die Hypoxideae moontlik in een familie geplaas kan word. Opregte dank is verskuldig aan Prof. P. G. Jordaan en Dr. K. R. Sporne vir die oorlees van die manuskrip. 62 The Journal of South African Botany. REFERENCES. Apamson, R. S., and Satrer, T. M. 1950: Flora of the Cape Peninsula. Cape Town. Baruton, M. H., 1891: Dictiounusaire de Botanique, III: 4. Baker, J. G., 1896-97: Haemodoraceae. In Thiselton-Dyer, Flora Capensis, VI: 1. Barker, W. F., 1940: The genus Dilatris Berg. with the description of a new species. Journ. S. A. Botany, 6: 147. BEenTHAM, G., and Hooker, J. D., 1883: Genera Plantarum, III: 671. London. Brackett, A., 1923: Some genera closely related to Hypoxis. Contrib. Gray Herb., 69: 155. Cave, M. S., 1952: Sporogenesis and gametogenesis in Odontostomum hartwegii Torr. Phytomorph., 2: 210. 1953: Cytology and embryology in the delimitation of genera. Chron. Bot., 14: 140. DELLERT, R., 1933: Zur systematischen Stellung von Wachendorfia. Osterreich. Bot. Zeitschr., 82: 335. bE Vos, M. P., 1948: The development of the ovule and the seed in the Hypoxideae. I, Ianthe Salisb. Jowrn. S.A. Bot., 14: 159. —— 1949: Do. II, Pauridia Harv. and Forbesia Eckl. Journ. S.A. Bot., 15: 13. ————— 1950: Die ontwikkeling van die saadknop en saad by Cyanella capensis L.: n Geval van polyembryonie. S.A. Journ. Sci., 46: 220. ENpDLIcHER, S., 1836-40: Genera Plantarum secundum ordines naturales disposita, I. Vindobonae. — 1842: Mantissa Botanica Supplementum Secundum. Vindobonae. Eneter, A., 1897: Ubersicht tber die Unterabteilungen, u.z.w. In Engler & Prantl, Die natirlichen Pflanzenfamilien, Nachtrage zum II-IV. Teil: 344. ENGLER, A., and PRANTL, K., 1930: Die naturlichen Pflanzenfamilien, 2te Aufl. 15a. ErptMan, G., 1952: Pollen morphology and plant taxonomy. Angiosperms. Stockholm. Friss, T. C. E., 1919: Der Samenbau bei Cyanastrum Oliv. Swensk. Bot. Tidskr., 13: 295: Hurcuinson, J., 1944: The families of Flowering Plants. Il, Monocotyledons. London. Lorsy, J. P., 1911: Vortrage tuber botanische Stammgeschichte, III: 800. Jena. MauveEsHwaRt, P., 1950: An introduction to the embryology of Angiosperms. New York. Nietscu, H., 1941: Zur systematischen Stellung von Cyanastrum. Osterr. bot. Zeitschr., 90: 31. ; Ono, F., 1928: Endospermbildung von Liliaceen. Bot. Mag. Tokyo, 42: 445. ———— 1929: Embryologie der Liliaceae, mit besonderer HRiuicksicht auf die Endospermbildung. I, Melanthioideae und Aletroideae. Sci. Rep. Tohoku Univ. Sendai, 4, Ser. 4: 381. Pax, F., 1889: Haemodoraceae. In Engler & Prantl, Die natirlichen Pflanzen- familien, 1. Aufl. IT, 5: 92. ———— 1930: Haemodoraceae. In Engler & Prantl, Die natiirlichen Pflanzen- familien, 2. Aufl. 15a: 386. and Horrmann, K., 1930: Amaryllidaceae. In Engler & Prantl, Die natiirlichen Pflanzenfamilien, 2te Aufl. 15a: 391. Puinires, E. P., 1951: The genera of South African Flowering Plants. Dept. of Agric. Bot. Survey Mem. 25. Pretoria. Scuarr, W., 1892: Anatomie der Hypoxideen und einiger verwandter Pflanzen. Bot. Centralbl., 52: 145—327. Scunarr, K., 1929: Embryologie der Angiospermen. In Linshauer, Hdb. der Pflanzenanat., 11, 2, X/2. Berlin. 1931: Vergleichende Embryologie der Angiospermen. Berlin. ———— and WuNDERLICH, R., 1939: Zur vergleichenden Embryologie der Liliaceae- Asphodeloideae. Flora, N.F. 33: 297. ScHnizLEIn, A., 1860: Emige merkwiirdige Formen von Ovula bei Monocotylen. Flora, 43: 529. ‘ Scnuuze, R., 1893: Beitrage zur vergleichenden Anatomie der Liliaceen, Haemo- doraceen, Hypoxidoideen und Velloziaceen. Engler bot. Jahrb. 17: 295. Stenar, H., 1925: Embryologische Studien. II, Die Embryologie der Amaryllideen. Studies of the Embryology and Relationships of 63 South African Genera of the Haemodoraceae. Akad. Abhandl. Uppsala. 1927: Zur Engwicklungsgeschichte der Gattung Anigosanthus Labill. Bot. Not., 1927: 104. ———— 1938: Die systematische Stellung der Gattung Xiphidium Loefl. Svensk Bot. Tidskr., 32: 274. 1942: Zur Embryologie der Dracaena-Gruppe. Heimbygdas Tidskr., 8: 183. SuEssENGUTH, K., 1921: Beitrage zur Frage des systematischen Schlusses der Mono- cotylen. Beth. bot. Centralbl., 38, 2: 1. WonveruicyH, R., 1950: Die Agavaceae Hutchinsons im Lichte ihrer Embryologie, ihres Gyn6ozeum-, Staubblatt- und Blattbaues. Osterr. Bot. Zeitschr. 97: 437. NOTES ON SOUTH AFRICAN MARINE ALGAE. IV. By Grorce F. PAPENFUSS. (Department of Botany, University of California.) The present paper, like the earlier ones in this series (Papenfuss, 1940, 1943, 1952), is primarily concerned with questions relating to the identity, taxonomy and nomenclature of some of the South African marine algae. Grateful acknowledgment is made to the Carnegie Corporation of New York for a Grant-in-Aid, received through the University of Cape Town (1939-1940), the Guggenheim Foundation for a Fellowship (1949- 1950) and the National Science Foundation for a Grant-in-Aid (1954-) which enabled me to search for (in European herbaria) and to study the types and other published specimens referred to in this paper. The funds granted by the National Science Foundation have also made it possible to secure the services of research assistants to aid in the study of the South African marine algae. I wish to thank Dr. Lois Egerod, Mr. Max Hommersand and Miss Shirley Sparling for valuable assistance. For the loan of specimens I am indebted to the keepers of the herbaria at Birming- ham, Hamburg, Kiel and Stockholm. CHLOROPHYCOPHYTA. Caulerpa mexicana Sonder ex Kiitzing. Caulerpa mexicana Sonder ex Kiitzing, 1849, p. 496. C. Harveyana Kitzing, 1857, p. 3, pl. 5, fig. ii. C. crassifolia (C. Agardh) J. Agardh, 1872, p. 13. C. taxifolia (Vahl) C. Agardh £ crassifolia C. Agardh, 1822, p. 436. C. pinnata sensu Weber-van Bosse, 1898, p. 289 (non Fucus pinnatus Linnaeus fil., 1781, p. 452: cf. Howe, 1905, p. 574). This is the first record of the occurrence of this widely distributed species in South Africa. The only South African specimens at hand were collected by Dr. Mary A. Pocock and me at Durban (The Bluff) and Umhlanga Rocks, Natal, in July 1938. These plants are only | to 2 cm. high and are smaller than some of those from other parts of the world in which the fronds may attain a height of 10 em. Dwarf individuals also occur, however, in other regions (cf. Taylor, 1928, p. 96; Bérgesen, 1952, p. 9). 65 66 The Journal of South African Botany. Since the time that Howe (1905, pp. 574-576) pointed out that the plant that Linnaeus (1781) described as Fucus pinnatus is not the same as the taxon described by C. Agardh (1822) as Caulerpa taxifolia B crassi- folia, this species has invariably been reported under the name C. crassi- folia (C. Agardh) J. Agardh. The taxon was first described as a species, however, by Sonder (ex Kiitzing. 1849) under the name C. mexicana; and it is this name rather than C. crassifolia that must be used (cf. Art. 70 of the Code). Weber-van Bosse (1898) established four formae under Caulerpa pinnata, two of which, typica and mexicana, are of concern here. (Both were transferred to C. crassifolia by Bdérgesen, 1907, p. 363.) Forma typica is representative of the type of C. crassifolia and f. mexicana of the type of C. mexicana. In view of the restoration of the specific epithet mexicana and in conformity with the current Code (Art. 35), the form representative of the type of C. mexicana would continue to bear the name mexicana and, if it were recognized, f. typica would have to receive another name. In consideration of the somewhat doubtful taxonomic value of all the formae recognized under C. pinnata or C. crassifolia, it seems best, however, to refrain from giving any of them formal recogni- tion under C’. mexicana at this time. (Bdrgesen, loc. cit., has remarked that intermediates often occur between f. typica and f. mexicana.) As regards the synonymy of this species, it should be pointed out that J. Agardh (1872, p. 14) included (with a query) Caulerpa obtusa Lamouroux (18094, p. 332; 18096, p. 143), of uncertain source but perhaps from the Antilles, in the synonymy of C. crassifolia var. mexicana. To my knowledge, the type of C. obtusa has never been re-examined. Lamour- oux’s illustration of the species (1809). pl. 6, fig. 3) suggests a plant rather unlike conventional C. mexicana but some of the specimens of the latter species from Bermuda in the herbarium of the University of California do show some resemblance to Lamouroux’s illustration. It is thus not inconceivable that C. obtusa and C. mexicana are conspecific and that the former binomial will have to replace the latter as the correct name of this species. PHAEOPHYCOPHYTA. Halopteris funicularis (Montagne) Sauvageau. Halopteris funicularis (Montagne) Sauvageau, 1904, p. 393, fig. 80. Sphacelaria funicularis Montagne, 1842a, p. 13. Synonyms or misidentifications based wholly or in part on South African material. Stypocaulon funiculare (Montagne) Kiitzing, Reinke, 1891, p. 22; Barton, 1893, p. 111; De Toni, 1895, p. 517. (These three records are based on Areschoug’s Sphacelaria globifera.) Notes on South African Marine Algae LV. 67 Anisocladus congestus Reinke, 1890, p. 213; 1891, p. 32 (only as to South African material—not as to lectotype from New Zealand): De Toni, 1895, p. 520 (cited from Reinke); Barton, 1896, p. 197 (cited from Reinke); Delf and Michell, 1921, p. 97. Sphacelaria globifera Areschoug, 1851, p. 20; 1854, p. 364. Sphacelaria paniculata sensu Krauss, 1846, p. 213; sensu J. Agardh, 1848, p. 36; sensu Areschoug, 1851, p. 19. (Non Sphacelaria panic- ulata Suhr, 1840, p. 278.) Stypocaulon paniculatum sensu Hohenacker, 1854, No. 154; sensu Magnus, 1873, p. 142, pl. 2, fig. 42; sensu Barton, 1893, p. 111 (cited from Krauss); sensu De Toni, 1895, p. 516 (cited from J. Agardh). (Non Sphacelaria paniculata Suhr.) Sphacelaria scoparia sensu Harvey, 1838. p. 396; 1846, text to pl. 37 (p.p.); sensu Drége, 1843, pp. 111, 223 (on p. 111 in error as Sphaero- coccus scoparius); sensu Areschoug, 1851, p. 21 (cited from Drege and Harvey). (Non Halopteris scoparia (L.) Sauvageau.) Stypocaulon scoparium sensu Kiitzing, 1849, p. 466 (cited from Harvey); sensu Barton, 1893, p. 111; sensu De Toni, 1895, p. 518 (cited from Kitzing). (Non Halopteris scoparia (L.) Sauvageau, cf. Sauvageau, 1904, p. 376.) As will be evident from the somewhat lengthy synonymy given above, the South African material of Halopteris funicularis has been the source of a good deal of confusion. Until the time of Sauvageau (1904), who was the first to clarify the situation as regards the identity of the South African plant, published accounts (e.g., De Toni, 1895) conveyed the impression that the Stypocaulaceae were represented in South Africa by some five species: H. funicularis, H. congesta, H. scoparia, Stypocaulon paniculatum (=?H. hordacea) and Phloiocaulon squamulosum. Actually, as far as is known, only two species oceur in South Africa: H. funicularis and P. squamulosum. I have examined representative specimens of all the published records based on South African material, as listed in the synonymy above, with the exception of that of Drege (1848, pp. I11, 223) and have found them to be of H. funicularis. A brief review of a few of the points relating to the history of Halopteris fumcularis, especially as regards its synonyms that are based wholly or in part on South African material, may be useful. The species, as Sphacelaria funicularis, was first described by Montagne (1842a) from material from Auckland Island. The type material is in Montagne’s collection in the cryptogamic herbarium of the Muséum National d’Histoire Naturelle, Paris, where I have seen it. Sauvageau (1904, p. 394) has pointed out that it is not possible to determine with certainty whether the type material is representative of Halopteris 68 The Journal of South African Botany. funicularis (as circumscribed by him) or of H. congesta and he (p. 395) has accordingly designated sporophytie material from Chile, reported upon by Montagne in 1852 (p. 267), as typical of his concept of H. funicu- laris. Since the time of Sauvageau’s monograph of the Sphacelariales, investigators have followed his circumscription of H. funicularis and I have done the same as regards the South African material of Halopteris. On the basis of material from South Africa, Areschoug (1851) described this species as new to science under the name Sphacelaria globifera. Reinke (1891) reduced S. globifera to the synonymy of Stypocaulon funiculare (H. funicularis). In more recent times (ef. Delf and Michell, 1921, p. 97) this taxon has passed in South Africa as Anisocladus congestus Reinke. Reinke (1890) described the monotypic genus Anisocladus from material from South Africa and New Zealand, without designating a type. Sauvageau (1900-1914, pp. 293, 411-415) reduced Anisocladus to the synonymy of Halopteris but retained Reinke’s species as an autonomous taxon, H. congesta. This species shows a striking resemblance to H. funicularis— in fact, can be distinguished from it with certainty only on the basis of a character presented by fertile sporophytic plants: in H. funicularis the clusters of unilocular sporangia contain intermingled, paraphyses whereas in H. congesta they lack paraphyses. Among the specimens of Halopteris from South Africa available to him, Sauvageau could find none that was referable to H. congesta—all were of H. funicularis. He, however, did not see Reinke’s material. Representative specimens of Halopteris collected by me in South Africa were sent to Miss Moore several years ago, when she was working on the New Zealand species of this genus (Moore, 1951, 1953), but the fertile sporophytic and hence identifiable specimens all proved to be of only one species—H. funicularis. During the summer of 1954 I had an opportunity to visit the Uni- versity of Kiel, where Reinke’s herbarium is conserved, and to arrange for a loan of his material of Anisocladus congestus. Reinke had three specimens from South Africa: one from Port Natal named ‘‘Sphacelaria paniculata Hering”, one labelled “Cap. bon. sp. (ex herb. Jessen.)”, and the third one labelled “Port Natal 1839 leg. Krauss’’. Both the specimens from “Port Natal’’ are gametophytic whereas the one from “Cap. bon. sp.” is sporophytic. The last-named specimen is definitely representative of Halopteris funicularis. Miss Moore (1951, p. 277), who has had a great deal of experience with both H. funicularis and H. congesta, is not certain that these two species can be distinguished on the basis of sexual plants. Since, to judge from the sporophytic plants, only one species of Halopteris, H. funicularis, occurs in South Africa, it seems reasonable to conclude Notes on South African Marine Algae IV. 69 that the gametophytic plants among Reinke’s South African material, as well as those in my own and other collections, are also representative of H. funicularis. As is true of the various species of Halopteris that Miss Moore has studied, the gametophytic plants of H. funicularis are oogamous. On the basis of my examination of Reinke’s material, Halopteris congesta may thus with reasonable certainty be excluded from the South African flora. New Zealand therefore constitutes the type region of this species and I propose as lectotype a sporophytic specimen that Reinke received from von Mueller and that is labelled “North Island, New Zealand 1889”. Finally, it is necessary to say a few words about the geographic dis- tribution of Halopteris funicularis in South Africa. On the basis of material collected by Krauss (1846, p. 213), this species was reported from Port Natal (now Durban) under one or another of the names listed in the synonymy above by J. Agardh (1848, p. 36), Areschoug (1851, p- 19), Magnus (1873, p. 142), Barton (1893, p. 111) and De Toni (1895, p. 516). In spite of the fairly extensive and intensive collecting that has been done in the Durban region in recent times, H. fwnicularis has never again been obtained in this part of South Africa. It is extremely doubtful that the specimens in question were collected at Port Natal. More likely they were secured in the Cape Peninsula or in the region of Cape Agulhas whence Hohenacker (1854) had material for distribution and where I obtained it in quantity in the drift at Struis Bay. As in the New Zealand- Australia region (cf. Moore, 1953, map 9), H. funicularis in South Africa is a plant of the colder waters—along the south and west coasts of the Cape Province. Chnoospora minima (Hering) Papenfuss, comb. nov. Fucus minimus Hering, 1841, p. 92. Chnoospora pacifica J. Agardh, 1847, p. 7. Chnoospora fastigiata J. Agardh var. pacifica (J. Agardh) J. Agardh, 1848, p. 172. On the basis of material collected by Krauss (1846, p. 213) at Port Natal, Hering in 1841 described a plant from South Africa under the name Fucus minimus. Hering’s diagnosis was very brief and did not permit recognition of the species by subsequent phycologists. Endlicher (1843, p. 29) associated the species with the marine lichen known for a long time by the name Fucus pygmaeus, and he was followed in this disposition of the taxon by De-Toni (1895, p. 209). Kiitzing (1849, p. 590) included Hering’s species in his list of species of Fucus but he did not see material of it. Barton in 1896 (p. 195) reduced the species to the synonymy 70 The Journal of South African Botany. of the South African plant now known as Carpomitra filiformis (Suhr) Papenfuss (1943, p. 82). It was evident that Barton had seen some of Hering’s original material and it seemed likely that a specimen would be on file either in the her- barium of the British Museum or in that of Kew. I searched for such a specimen at these institutions in 1939 and 1950 but to no avail. Search in many European herbaria was equally fruitless until the summer of 1954 when an authentic specimen was finally found in the herbarium of the Staatsinstitut fiir allgemeine Botanik at Hamburg, and arrangements were made for the loan of it. The label accompanying this specimen contains the inscription “Fucus minimus Hering nov. species Port Natal Coll. Dr. Krauss Cap. b. sp. e herbaria Heringiana”’ and also the diagnosis as published by Hering. Furthermore, this is the specimen that Barton saw as is evident from the annotation “‘Carpo- mitra chytraphora Kiitz.” in her hand. Hering’s specimen (which I have designated lectotype) is a small fragment but examination of it leaves no question that it is not repre- sentative of Carpomitra filiformis but is identical with the widely dis- tributed plant that was described by J. Agardh (1847) as Chnoospora pacifica from material from “St. Augustin” on the Pacific Coast of Mexico (type, No. 50087 in Herb. Agardh!). Since Fucus minimus Hering ante- dates Chnoospora pacifica J. Agardh, the latter name must yield to the new combination Chnoospora minima (Hering) Papenfuss. Chnoospora minima is a reasonably common alga along the east (warmer) coast of South Africa. The species was reported from South Africa by Barton (1896, p. 195, as C. fastigiata) and by Levring (1938, p- 21, as C. pacifica), who also gave some good illustrations of its habit and anatomy. RHODOPHYCOPHYTA. Galaxaura magna Kjellman. Galaxaura magna Kjellman, 1900, p. 82, pl. 15, figs. 1-10, pl. 29, fig. 46. Corallopsis dichotoma Suhr, 1839, p. 70, pl. 3, fig. 44. (Non Galaxaura dichotoma Lamouroux ex Decaisne, 1842, p. 116.) The identity of Corallopsis dichotoma Suhr (1839) has been uncertain since the species was founded. Harvey (1846, text to pl. 69) reduced the taxon to the synonymy of the plant now known as Scinaia furcellata (Turner) Bivona (a species which does not occur in South Africa) and he was followed in this by Kitzing (1849, p. 715). J. Agardh (1852, p- 423) placed the species, with a query, in the synonymy of Scinaia salicornioides (Kiitzing) J. Agardh and he was followed by De Toni (1897, p. 106), who, however, later (1900, p. 461) also listed it under species inquirendae. Notes on South African Marine Algae LV. 7A Setchell (1914, p. 108) was apparently the first person to have examined some of Suhr’s material of Corallopsis dichotoma. He obtained a fragment (part of which is still in the herbarium of the University of California) of a specimen that is on file in the National Herbarium of Victoria and from a study of it concluded that the species agrees well with the description and figures of the South African plant described as Galaxaura magna by Kjellman (1900). By implication Setchell designated the specimen in the National Herbarium of Victoria as lectotype of C. dichotoma. During the summer of 1954, I found another of Suhr’s specimens in the herbarium of the University of Kiel, filed at the end of the genus Galaxaura. This specimen is labelled “‘Corallopsis dichotoma” in Suhr’s hand and with it are notes by him. The collector and locality (merely given as “Von der stidafrikanischen Kite” in his paper) are given as “Ecklon in der Algoa Bay”. Anatomical examination of the Kiel speci- men as well as of the fragment of the Melbourne specimen has shown that both are representative of G. magna. Both specimens are of sexual plants. This generation is less distinctive than the tetrasporangial genera- tion and may be confused with the sexual generation of the South African plant that Zanardini (1846; 1862, pl. 31, figs. BI—5) described as G. Diesingiana. However, on the basis of our present knowledge of these two species—thanks to the magnificent paper on the South African species of Galaxaura by Svedelius (1942)—and by comparison with the material of both species in my collection, there is little question of the identity of Corallopsis dichotoma with G. magna. In G. magna the cortex of sexual plants is thinner and the medulla is thicker than in sexual plants of G. Diesingiana. Since Corallopsis dichotoma Suhr (1839) antedates Galaxaura magna Kjellman (1900), it would have been necessary to replace the latter specific epithet with the former, were it not for the fact that the binomial G. dichotoma already exists. In 1842 (p. 116) Decaisne described under this name a manuscript species of Lamouroux based on material from Guadeloupe and Cuba. (The identity of G. dichotoma Lamouroux ex Decaisne still remains to be clarified as far as I have been able to determine. ) Rhodophyllis reptans (Suhr) Papenfuss, comb. nov. Halymenia reptans Suhr, 1834, p. 735, pl. 2, figs. 17, S; Kiitzing, 1866, p. 32, pl. 91, figs. c-f. Rhodophyllis capensis Kitzing, 1849, p. 786; 1869, p. 19, pl. 50, figs. g—1. Rhodymenia nigricans Harvey, 1847, pl. 46 (nomen nudum). Ectoclinium kowiense Holmes, 1896, p. 349. Suhr in 1834 described a species from Algoa Bay as Halymenia reptans and in 1841 a species from “Cap” as Phyllophora reptans. A good deal 72 The Journal of South African Botany. of uncertainty has existed as regards the identity of these two species and at times they were confused with each other. J. Agardh (1852a, p. 480), Grunow (1867, p. 82) and Schmitz (1894, p. 201, footnote) thought Phyllophora reptans was merely a small form of Gelidium pristoides. Kylin (1938) regarded it as an autonomous species of Gelidiwm, G. reptans (Suhr) Kylin, an opinion with which I concur. From a study of one of Suhr’s specimens in Reinbold’s herbarium, Schmitz (loc. cit.) concluded that Halymenia reptans is a species of Rhodo- phyllis which might be identical with FR. capensis Kiitzing, the type of which is a specimen collected by Pappe at Kalk Bay. During the summer of 1954, I found in the herbarium of the University of Kiel another of Suhr’s original specimens which I borrowed for study. It is labelled in Suhr’s hand “‘Halymenia reptans $ Am Fuss der Halymenia fur- cellata 8. Algoa-Bay.” I have designated this specimen as lectotype. Study of it has revealed that this taxon actually is conspecific with R. capensis Kiitzing, and since Suhr’s binomial is the older, it becomes necessary to make the new combination Rhodophyllis reptans (Suhr) Papenfuss. Harvey (1847) in his day had illustrated and distributed specimens of the species under the name Rhodymenia nigricans but in accordance with Art. 44 of the Code this name must be treated as a nomen nudum. In 1950 I had an opportunity of searching for and of arranging for the loan of various specimens of South African marine algae in the herbarium of the University of Birmingham that were published by Holmes. (These and other specimens have since been deposited in the herbarium of the British Museum.) Among these were the type and two other specimens of the plant described by Holmes (1896) as Hetocliniwm kowiense on the basis of material received from Becker, who collected it at “The Kowie” (Port Alfred). Examination of this material has shown that H. kowiense is a species of Rhodophyllis, and there appears to be little question that it is identical with R. reptans. In fact, one of Holmes’ specimens is a plant that Becker distributed under the name “Halymenia reptans Suhr’. Holmes compared his species with Rhodo- phyllis but states (p. 349) “ . . . the nemathecioid sorus of tetraspores at once distinguishes the plant from the genus Rhodophyllis”. The tetra- sporangium bearing tips of this plant agree well with those of R. reptans and other species of this genus and are quite unlike those of H. dentatum (the type of this genus), which, as illustrated by J. Agardh (1879, pl. 30), contain paraphyses intermingled with the sporangia. Plants of Rhodophyllis reptans from Port Alfred and other localities of the south-eastern coast of South Africa are generally slenderer than those occurring in False Bay and adjacent regions, which correspond to Notes on South African Marine Algae IV. 73 the type of R. capensis, but anatomically the two kinds of thalli appear to be identical. Nienburgia serrata (Suhr) Papenfuss, comb. nov. Nitophyllum serratum Suhr, 1836, p. 342, pl. 4, fig. 29. The identity of Nitophyllum serratum has been uncertain ever since the species was established by Suhr in 1836. Kutzing (1849, p. 869) included it in his list of Species dubiae, J. Agardh (18526, p. 672) listed it under Species inquirendae and thought it might be the juvenile stage of a Plocamium, and De Toni (1900, p. 665) placed it under Species minus notae aut quoad sectionem incertae. During the summer of 1954 I found in the herbarium of the University of Kiel the only specimens of Suhr’s material of Nitophyllum serratum that I have seen. The specimens consist of a few small bits on mica in an envelope and are labelled as collected at ““Algoa Bay” and with them are notes on the species by Suhr, his original pencil drawings of it and the printed description and figures. I have designated this material as lectotype of NV. serratum. Examination of Suhr’s material revealed that it is representative of a small serrated member of the Delesseriaceae that has been collected by Dr. Mary A. Pocock and me at a number of localities along the south- east and east coasts of South Africa. Some years ago this plant was identified as representative of Nienburgia Kylin (1935), which genus was originally described by Kylin (1924, p. 46) under the name Heteronema. It was thought that the South African plant probably represented a new species and it is gratifying to have found that it is the long-lost Nitophyllum serratum of Suhr. The species should therefore be known as Nienburgia serrata (Suhr) Papenfuss. The occurrence in South Africa of a species of Nienburgia constitutes an interesting extension of the geographic distribution of this genus. Heretofore, as far as I am aware, Nienburgia has been credited with only three species, N. Andersoniana (J. Agardh) Kylin, N. borealis (Kylin) Kylin, and N. japonica (Yamada) Kylin—all from the North Pacific. Placophora monocarpa (Montagne) Papenfuss, comb. nov. Polysiphonia monocarpa Montagne, 1842b, p. 254; 1846, p. 89, pl. 143, fig. 3; 1856, p. 424; Harvey, 1847, p. 56; Kiitzing, 1849, p. 807; 1863, p. 13, pl. 38, figs. d-h; J. Agardh, 1863, p. 941; Barton, 1893, p. 176. Herposiphonia? monocarpa (Montagne) De Toni, 1903, p. L061. Placophora Eckloniae Pocock ex Scagel, 1953, pp. 36-39; Pocock, 1953, 19) ly 14S, A, Tle Tc 74 The Journal of South African Botany. On the basis of material obtained in South Africa by Guadichaud, during the voyage of the Bonite, the polysiphonous reproductive filaments of this species were described by Montagne (18426) as a new species of Polysiphonia—P. monocarpa. These filaments were later illustrated by Montagne (1846) and Kiitzing (1863). Duplicates of the material collected by Guadichaud were available to several later phycologists (Harvey, 1847; Kiitzing, 1849, 1863; Barton, 1893) but none of them contributed to a better understanding of the species. De Toni in 1903 transferred Polysiphonia monocarpa to Herpo- siphonia, with a query. In associating it with Herposiphonia he probably was influenced by Montagne’s comment that the species formed a velvety layer on the larger brown algae. I have seen specimens of the original material of Polysiphonia mono- carpa in the collections of Thuret (Paris), Kiitzing (Leiden) and Harvey (Dublin) and in the herbaria of Kew and the British Museum of Natural History. Microscopic examination of fragments of the material in the collections of Thuret and Kiitzing revealed that the polysiphonous filaments spring from a pseudoparenchymatous prostrate system (which apparently escaped the notice of Montagne, Harvey, Kiitzing and Barton) and that the species is representative of the genus Placophora. Comparison of Polysiphonia monocarpa with the two known species of Placophora, P. Binderi and P. Eckloniae—both based on material from South Africa, has shown that Montagne’s species is identical with P. Eckloniae, which has been well described by Scagel (1953) and Pocock (1953). It thus becomes necessary to make the new combination Placophora monocarpa (Montagne) Papenfuss and to reduce P. Eckloniae to the synonymy of this binomial. Since the time that Guadichaud collected the original material of Placophora monocarpa, over a hundred years ago, the species escaped the notice of collectors in South Africa until a few years ago when Dr. Pocock rediscovered it. She has obtained it at a number of localities and always as an epiphyte on the stipe of Ecklonia maxima (Pocock, 1953, p. 39). Montagne mentioned only “ad Fucos majores” as the substratum, but it seems very likely that his material likewise grew on EH. maxima. The characteristic formation of only one cystocarp on each of the poly- siphonous filaments of female plants, which suggested the specific epithet monocarpa, is clearly shown in plate 7 (figure g) of Dr. Pocock’s paper. Notes on South African Marine Algae IV. 75 NcAnpH, ©. A. REFERENCES. 1822. Species algarum ... Vol. 1 (2). [8] + pp. 169-531. Lund. AGaARDH, J. G. a 1847. Nya alger fran Mexico. Ofv. K. Sv. Vet.-Akad. Férhandl. 4: 5-17. 1848. Species genera et ordines algarum, ... Vol. 1. viii + 363 pp. Lund. 1851. Species genera et ordines algarum,... Vol. 2 (1). xii + 336 pp. + [addenda and index, pp. 337-351]. Lund. 1852a. Species genera et ordines algarum, ... Vol. 2 (2:1). Pp. 337- 504. Lund. 18526. Species genera et ordines algarum, ... Vol. 2 (2:2). Pp. 505- 700 + [addenda and index, pp. 701-720]. Lund. , 1863. Species genera et ordines algarum, ... Vol. 2 (3:2). Pp. 787-1291. Lund. 3 1872. Till algernes systematik. [afd. 1]. Lunds Univ. Arsskr. 9 (afd. 2, nr 8). 71 pp. 1879. Florideernes morphologi. K. Sv. Vet.-Akad. Handl. 15 (6). 199 pp., 33 pls. ARESCHOUG, J. E. 1851. Phyceae capenses. 32 pp. Uppsala (Thesis). 1854. Phyceae novae et minus cognitae in maribus extraeuropaeis collectae quas_ descriptionibus observationibusque illustravit. Nova Acta R. Soc. Se. Upsal., ser. 3, 1: 329-372. Barton, ETHEL S. 1893. 1896. BdRGESEN, F. 1907. 1952. DECAISNE, J. A provisional list of the marine algae of the Cape of Good Hope. Journ. Bot. 31: 53-56, 81-84, 110-114, 138-144, 171-177, 202-210. Cape algae. Journ. Bot. 34: 193-198, 458-461, 5 figs. An ecological and systematic account of the Caulerpas of the Danish West Indies. K. Danske Vidensk. Selsk. Skrifter, 7 Raekke, Naturvidensk. Math. Afd., 4: 337-392, 31 figs. Some marine algae from Mauritius. Additions to the parts pre- viously published, IV. K. Danske Vidensk. Selsk. Biol. Medd. 18 (19). 72 pp., 33 figs., pls. 1-5. 1842. Mémoire sur les corallines ou polypiers calciféres. Ann. Sc. Nat. Bot., sér. 2, 18: 96-128. Der, ELLEN M. and Marcaret R. MICHELL. 1921. The Tyson collection of marine algae. Ann. Bolus Herb. 3: 89- 119. De Tont, J. B. 1895. Sylloge algarum... Vol. 3. Fucoideae. xvi + 638 pp. Padua. 1897. Sylloge algarum ... Vol. 4. Florideae, Sect. 1. xx + lxi + 386 + [2] pp. Padua. 1900. Sylloge algarum... Vol. 4. Florideae, Sect. 2. Pp. 387-— 773. Padua. 1903. Sylloge algarum ... Vol. 4. Florideae, Sect. 3. Pp. 775-1521 + 1523 [bis]—1525 [bis]. Padua. DrieGe, J. F. 1843. Zwei pflanzengeographische Documente. Flora 26 (Besondere Beigabe). 230 pp., 1 map. ENDLICHER, 8. L. 1843. Mantissa botanica altera. Sistens generum plantarum, suppl. 3. {6] + 111 pp. Vienna. Grunow, A. 1867. Algae, in Reise der Gsterreichischen Fregatte Novara um die Erde ... Bot. Theil 1. Pp. 1-104, pls. 1, 1a, 2-11. Vienna. Harvey, W. H. 1838. The genera of South African plants, ... Ixvi + 429 pp. Cape Town. 1846. Phycologia britannica: ... Vol. 1. xv pp., 120 plates and accompanying text of unnumbered pages + vili pp. London. 1847. Nereis australis, ... viii + 124 pp., 50 pls. London. 76 The Journal of South African Botany. Herine, K. : : 1841. Diagnoses algarum novarum a cl. Dre. Ferdinand Krauss in Africa australi lectarum. Ann. Mag. Nat. Hist. 8: 90-92. HouENACKER, R. F. 1854. Algae marinae siccatae, Lief. 4: Nos. 151—200. Hormes, E. M. 1896. New marine algae. Journ. Bot. 34: 349-351. Howe, M. A. 1905. Phycological studies—II. New Chlorophyceae, new Khodophyceae and miscellaneous notes. Bull. Torrey Bot. Club 32: 563-586, pls. 23—29. KJELLMAN, F. BR. 1900. Om floridé-slagtet Galaxaura dess organografi och systematik. K. Sv. Vet.-Akad. Handl. 33 (1). 109 pp., 20 pls. Krauss, F. 1846. Pflanzen des Cap- und Natal-Landes, Algae. Flora 29: 209-215. Kurzine, F. T. 1849. Species algarum. vi- 922 pp. Leipzig. : 1857. Tabulae phycologicae ... Vol. 7. ii + [1] + 40 pp., 100 pls. Nordhausen. 1863. Tabulae phycologicae... Vol. 13. (1] + 31 pp., 100 pls. Nordhausen. 1866. Tabulae phycologicae ... Vol. 16. [iv] + 35 pp., 100 pls. Nordhausen. 1869. Tabulae phycologicae... Vol. 19. iv. + 36 pp., 100 pls. Nordhausen. Kyiin, H. = 1924. Studien ber die Delesseriaceen. Lunds Univ. Arsskr. N. F. Avd. 2, 20 (6). 111 pp., 80 figs. Z 1935. Zur Nomenklatur emiger Delesseriaceen. K. Fysiogr. Sallskap. Lund Forhandl. 5 (23). 5 pp. 1938. Verzeichnis einiger Rhodophyceen von Stdafrika. Lunds Univ. Arsskr. N. EF. Avd. 2, 34 (8). 26 pp:, 0 figs:, 8 pls: LamMouroux, J. V. F. 1809a. Observations sur la physiologie des Algues marines, et description de cinq nouveaux genres de cette famille. Nouv. Bull. Se. Soe. Philom., sér. 2, 1: 330-333, pl. 6, fig. 2. 18096. Mémoire sur les Caulerpes, nouveau genre de la famille des Algues - marines. Journ. Bot. [Desvaux] 2: 136-146, pls. 5-7. LEVRING, T. 1938. Verzeichnis emiger Chlorophyceen und Phaeophyceen von Sud- afrika. Lunds Univ. Arsskr. N. F. Avd. 2, 34 (9). 25 pp., 10 figs., 4 pls. LINNAEUS, C. 1781. Supplementum plantarum .. . [16] + 467 pp. Brunswick. Maenus, P. 1873. Zur Morphologie der Sphacelarieen nebst Bemerkungen tiber die Ablenkung des Vegetationspunktes der Hauptachse durch den nahe am Scheitel angelegt werdenden Tochterspross. Festschr. Feier Hundertjahr. Ges. Naturforsch. Freunde Berlin. Pp. 129-156, 4 pls. Montaene, [J. F.] C. 1842a. Prodromus generum specierumque phycearum novarum, in itinere ad polum antarcticum ... 16 pp. Paris. 1842b. Troisieme centurie de plantes cellulaires exotiques nouvelles. Ann. Se. Nat. Bot., sér. 2, 18:241—282. 1846. Cryptogames cellulaires: Algues, in A. N. Vaillant, Voyage autour de monde exécuté pendant les années 1836 et 1837 sur la corvette la Bonite, Botanique (by Gaudichaud), Vol. 3. Pp. i-xi, 1-112. Paris. 1852. Algas, in C. Gay, Historia fisica y politica de Chile . . . , Botanico, Vol. 8. Pp. 228-393. Paris. Notes on South African Marine Algae IV. 77 1856. Sylloge generum specierumque cryptogamarum ... xxiv + 498 pp. Paris. Moorr, Lucy B. 1951. Reproduction in Halopteris (Sphacelariales). Ann. Bot. N. S8., 15: 265-278, 54 figs. 1953. Some distribution problems illustrated from brown algae of the genus Halopteris. Proc. Seventh Pacifie Sc. Congress, Vol. 5. Pp. 13-18, 10 maps. PapPenruss, G. F. 1940. Notes on South African marine algae. I. Bot. Not. 1940: 200— 226, 16 figs. 1943. Notes on South African marine algae. II. Journ. 8. Afr. Bot. 9: 79-92. 1952. Notes on South African marine algae. III. Journ. 8S. Afr. Bot. 17: 167-188. Pocock, Mary A. 1953. South African parasitic Florideae and their hosts. 1. Four members of the Rhodomelaceae which act as hosts for parasitic Florideae. Journ. Linn. Soc. London Bot. 55: 34-47, 9 figs., pls. 5-9. REINKE, J. 1890. Uebersicht der bisher bekannten Sphacelariaceen. Ber. Deutsch. Bot. Ges. 8: 201-215, 3 figs., 2 charts. 1891. Beitrage zur vergleichenden Anatomie und Morphologie der Sphace- lariaceen. Bibl. Bot. 5 (Heft 23). 40 pp., 13 pls. SAUVAGEAU, C. 1900-1914. Remarques sur les Sphacélariacées. xii + 634 pp., 128 figs. (Pp. 1-348 appeared in Journ. de Bot. 14-18.) Bordeaux. ScaGEL, R. F. 1953. A morphological study of some dorsiventral Rhodomelaceae. Univ. Calif. Publ. Bot. 27: i-iv, 1-108, 20 figs. Scumitz, [C. J.] F. 1894. Neue japanische Florideen von K. Okamura. Hedwigia 33: 190- 201, pl. 10. SETCHELL, W. A. 1914. The Scinaia assemblage. Univ. Calif. Publ. Bot. 6: 79-153, incl. pls. 10-16. Sune, J. N. 1834. Uebersicht der Algen, welche von Hrn. Ecklon an der siidafri- kanischen Kuste gefunden worden sind. Flora 17: 721-735, 737-— 743, pls. 1, 2. 1836. Beitrage zur Algenkunde. Flora 19: 337-350, pls. 3, 4. 1839. Beitrage zur Algenkunde. Flora 22: 65-75, pls. 1-4. 1840. Beitrage zur Algenkunde. Flora 23: 257-265, 273-282, 289-298. 1841. Beitrage zur Algenkunde. Verhandl. K. Leopold.-Carol. Akad. Naturf. 18, Suppl. 1: 273-288, 3 pls. SVEDELIus, N. 1942. Zytologisch-entwicklungsgeschichtliche Studien tiber Galaxaura eine diplobiontische Nemalionales-Gattung. Nova Acta Reg. Soc. Se. Upsal. ser. 4, 13 (4). 154 pp., 80 figs. TAYLOR, W. R. 1928. The marine algae of Florida with special reference to the Dry Tortugas. Papers Tortugas Lab. Carnegie Inst. Wash. 25. (Car- negie Inst. Wash. Publ. No. 379.) [4] + 219 pp., 3 figs., 7 tables, 37 pis. WEBER-VAN BosskE, ANNA. 1898. Monographie des Caulerpes. Ann. Jardin Bot. Buitenzorg 15: 243-401, pls. 20-34. ZANARDINI, G. 1846. Del vero posto che alle Galaxawra si compete nella serie dei vege- tabili marini. Gniorn. Bot. Ital. 2: 48-52. 1862. Scelta di ficee nuove o pit rare del mare Adriatico figurate, descritte ed illustrate (decade terza). Mem. Ist. Veneto Se. Lett. 10: 447— 484, pls. 26-33. CRINUM MACOWANII. By I. C. VerRDoorNn. (National Herbarium, Pretoria) (With Plates XIII to XVI) Crinum macowanii Bkr. emend; ('. macowani Bkr. in Gard. Chron. 9: 298 (1878) et in Bot. Mag. sub. t. 6381 (1878) pro parte. Lectotype: MacOwan 2122 (K). Bulbus ovoideus vel sub-globosus, 6—18 cm. diam., in collum plus minusve distinctum 11—29 cm. longum attenuatus. Folia c. 17, e collo breve, usque 18 cm. longo, enata, viridula, nonnunquam glauca, patentia, arcuata, longitudine plicata, usque 80 cm. longa, plerumque 4—9 cm. lata, marginibus plus minusve undulatis, pilis cartilaginosis sparse ciliatis. Inflorescentia usque 90 cm. alta. Pedunculus viridulus parce glaucus, robustus, sub-compressus, usque 70 cm. longus, 2-7 cm. latus, 1-7 cm. crassus. Umbella cum 10—25 floris. Bracteae involucrates 2, plus minusve deltoideae, base 5-5 cm. latae, usque 10 cm. longae; bracteae inter floribus nonnullae, lineares, c. 10 cm. longae. Pedicelli breves vel usque 4 cm. longi, germen 2 cm. longum. Perianthii tubus viridis vel subfuscus, curvatus, c. 10 em. longus; lobi saepe albidi, late roseo-carinati, ¢. 10 em. longi, exteriores 2:5 cm. lati, interiores 3 cm. lati, semper formam cam- panulatam conniventes apicem versus recurvi; cacumen viride, apex reflexus minute calcariformis. Filamenta albida; antherae atrae. Capsula falcata. CAPE. Noman’s Land, south-western frontiers of Natal, MacOwan 2122, lectotype (K); near Butterworth, MacOwan 508 (GRA, BOL). NATAL. Pietermaritzburg, Dyer 4869; Lower Tugela; Umblali, Meeuse 9658. TRANSVAAL. Pilgrimsrest: 17 m. S.W. Pilgrimsrest, Codd & Ver- doorn 7603; Sabie, Codd 5652. Lydenburg: 4 m. 8. of Lydenburg Codd & Verdoorn 7619. In the Gardener’s Chronicle 1878, when drawing up the description of Crinum macowanu (written ““Macowani’’) Baker confused two entities. He intended describing the species represented by the specimen MacOwan 2122, but believing it to be the same as two plants flowering at Kew, his description was based mainly on these living plants. In the Botanical 79 80 The Journal of South African Botany. Magazine of the same year, on plate 6381, under the name CU. macowani is a painting of one of these living plants and it is obviously C’. moore: Hook. f., a species quite distinct from that represented by MacOwan’s specimen. This error, we now learn, was noted by William Watson of Kew in 1900, and more recently by Mrs. Isaac of the Bolus Herbarium but to the best of my knowledge, nothing has been published on the subject. The specimen MacOwan 2122 came originally from “Noman’s Land” the south-western frontier of Natal. Unfortunately attempts to rediscover it in this particular area have so far failed, but certain specimens from Pietermaritzburg and Umbhlali in Natal and the Pilgrimsrest and Lyden- burg districts in the Transvaal are an excellent match of MacOwan 2122 (K) and MacOwan 508 (GRA, BOL). (The latter specimen, from near Butterworth, MacOwan himself identified as being the same species as 2122.) The amended description given above is therefore based only on these eastern examples although the name is applied to a much wider range of specimens. The reason for not citing all the specimens is because the broad view includes some forms which future workers may consider worthy of varietal rank. Descriptions of the forms and the reasons for considering them as normal variants of a single entity are given below. In the broad view of the species the distribution extends from Komgha in the eastern Cape, northwards through Natal and the Orange Free State to the Transvaal where it occurs quite extensively. The accom- panying distribution map shows the districts in South Africa in which specimens have so far been collected. An intensive survey might prove that it occurs in some other areas as well. The species is characterised by the following features: Leaves cedar green or glaucous, longitudinally folded to form a trough, arcuate spread- ing from a short stem-like column, usually broadly truncate (the apex easily dying off), but if intact, long and acuminate to the apex, fairly sparsely ciliate and undulate on the margin; perianth tube 5—12 cm. long, lobes large, more or less 10O—25 em. long and 2:5—3 em. broad, conniving in a campanulate shape with only the apical portion recurved at maturity, white or pale pink, always with a broad rose-pink dorsal keel or band; anthers black; fruit with part of the perianth persisting as a beak. oa The typical form is usually found near streams or in vleis. It has an inflorescence up to 90 cm. tall and the leaves are usually green not glau- cous. (Plates xiii and xiv.) A form found in the eastern Cape, although with a fairly tall inflo- rescence, often as tall as the typical form, has the leaves usually glaucous and the length of the pedicel and perianth tube varying considerably. ‘(VlIOJOIg JW poxBATy[No) [eIeN ‘YRTYUTE, UroaZ ATPEUIBIAO ‘uLTOF [RoIdAT—Z “Sy yRIBN ‘FainqzzaeVuLsoyzotq ‘oesmMoo1oyem ‘uIofF pRoidA[—| “Sty TIEX GLWTd ‘yRBVAsURL, “SanquopATT Jo YANOS Survoaqgs Teou ULOF PROIdA T—Z “Blof ‘[BBASURAT, OIGRG ‘opis URES ye UTTOZ [ROIdAT—] “S1yp “AIX ULV Td ae Soe RL PLATH XV. Fig. 1—Form with glaucous leaves and shorter perianth tube; grassland, Komgha, Cape. Fig. 2—Low plant with glaucous leaves and perianth segments tinged pink with a deeper rose pink keel; growing in dolerite outcrop, Estcourt, Natal. Fig. 3, 4—Plants medium height, leaves glaucous; growing in grassveld, Doorn- kloof, Irene. Fig. 3 was taken on a cool rainy day and Fig. 4 in the heat of a sunny day. Fig Fig Fig. Fig. PLATH) XVI. . 1—Low plant, glaucous leaves; in grassveld 6 miles 8S.W. of Lydenburg (note flowers drooping in hot sun). . 2—_Low plant, glaucous leaves; in rocky grassveld 15 miles 8S.W. of Lydenburg. (Note beaked fruits on right hand from neighbouring plant). 3—Low plant, glaucous leaves; in grassveld east of Pretoria. 4—Low plant, glaucous leaves; originally from Marico, Western Transvaal (cultivated at Pretoria). Crinum macowann 8] Some of these specimens have the pedicels up to 5 cm. long, as opposed to from about 1—3 cm. long, and the perianth tube only 5 cm. long as opposed to about 9 cm. This is the case in the plant on Plate xv, Fig. 1, but it is not clear since the position of the ovary is not easily seen. (Plate xaN, lie, 13) From a dolerite outcrop near Estcourt is recorded a form which appears very distinct having glaucous leaves, a short peduncle and the flowers more pink than white as the dorsal keel is broad and a very deep colour and the rest of the segment is tinted with pink. (Plate xv, Fig. 2.) Distribution records of Crinum macowanit Bkr. emend. (sensu lata). These differences may seem very significant when one compares such plants with the typical form, but over the wide range of its distribution intermediate stages can be found in all these features. The leaf colour seems to depend on the soil for in the drier regions and when near stones or rocks the leaves are distinctly glaucous, while in rich moist soil they are cedar green. This was clearly illustrated in the Lydenburg district when a special investigation of the Crinums of that region was made. 82 The Journal of South African Botany. For some miles to the south west of Lydenburg in red sandy loam on gentle rocky slopes there were hundreds of plants of this Crmum growing. They were all low plants with glaucous leaves. (Plate xvi, Figs. 1 and 2.) Not far from these. beside a stream, grew the specimen which has been included in the typical form having a tall inflorescence and green leaves. (Plate xiv, Fig. 2.) Many plants similar to the last mentioned, that is the more or less typical form, were found in vlei ground north of Lydenburg. and on some slopes intermediate forms were seen. While the habitat seems to influence the colour of the leaves and average length of the peduncle it has not yet been established if the deeper colour of the flowers in some plants, such as those at Estcourt, is connected with the conditions under which they grow, but it is unlikely that a difference in colour intensity can be considered of taxonomic value. In the central and western Transvaal the specimens collected have been mostly the form with a short peduncle and glaucous leaves. Such plants grow around Pretoria (Plate xvi, Fig. 3) and at Groot Marico, the most westerly record to date (Plate xvi. Fig. 4). Near Warmbaths in the Waterberg district and at Irene a form has been seen that is of medium height and the leaves usually sub-glaucous. These intermediate speci- mens support the view that the plants are best treated as a taxon varying within specific limits. Photographs of the plants at Irene illustrate an interesting feature that has been observed repeatedly (Plate xv, Figs. 3 and 4). The photo- graph of Plate xv, Fig. 3 was taken on a cool rainy day and it will be observed that in this photo a number of the open flowers are spreading. In Plate xv, Fig. 4 taken on a hot sunny day all the open flowers are drooping. Without knowing the reason this difference in the appearance of the flowers might be considered to be a distinguishing feature for a form or variety. The following key distinguishes C’. macowani in the broad application of the name from C. moorei Hook. f. with which Baker confused it, and two closely related species in the Transvaal C. bulbispermum (Burm. f.) Milne Redhead & Schweickerdt and C. graminicola Verdoorn. Perianth segments white or pink, not banded dorsally, sub-spread- ding at maturity; inflorescence up to 6 ft. tall; leaves more or less flat, shortly narrowing to base and apex .. C. moore Perianth segments with a distinct dorsal rose coloured reel (banded), conniving in a campanulate or funnel shape with only the apical portion recurving at maturity; inflorescence up to 3 ft. tall; leaves usually broadly truncate, the apical portion dying off easily, if intact then long and tapering to the apex: Crinum macowanii Leaves more or less flat, never glaucous, spreading at ee level; peduncle always short ae Leaves arcuate spreading from a short or ome stem- ix column, peduncle usually over 1 ft. tall if short then leaves glaucous: Stem-lke column of clasping leaf bases usually over 1 ft. tall; leaves always glaucous, several narrow ones in the centre; perianth funnel-shaped; anthers white to fuscus, not black; fruit not beaked : : 6 Stem-like column usually under | ft. tall; gave s green if glaucous plants small, only | or 2 narrow leaves in centre; perianth campanulate; anthers black; fruit beaked 83 C. graminicola C. bulbispermum C. macowania A NEW ALOE FROM THE EASTERN TRANSVAAL By G. W. REYNOLDS. (With Plates XVII and XVIII.) Aloe modesta Reynolds. Spec. nov., A. linearifolia Berger (Leptoaloe) affinis; floribus subsessilibus facile distinguitur. Planta parva, succulenta, acaulescens, sub terra bulbiformis. Folia 4—6, rosulata, 15--20 cm. longa, 8—9 mm. lata, linearia, sensim attenuata, supra viridia, leviter canaliculata, subtus convexa, ad margines edentata. Inflorescentia sim- plex, 25—30 em. alta. Pedunculus 3—4 mm. diam., levissime incrassatus, bracteis vacuis ovato-acuminatis 15 mm. longis, 6—7 mm. latis munitus. Bracteae ovato-acuminatae, 10 mm. longae, 6 mm. latae, subscariosae, 3—5 nervatae. Pedicelli | mm. longi. Perigoniwm luteo-viridulum, 13 mm. longum, cylindrico-trigonum, 4 mm. diam.; segmenta exteriora libera. Antherae 2—3 mm. exsertae. Stigma demum 3—4 mm. exserta. Ovarium 4 mm. longum, 2:5 mm. diametro. [Plates XVII, XVIII] SOUTH AFRICA. Transvaal, near Dullstroom, 6,900 ft., coll. J. H. de Wet, cult. Johannesburg, fl. 15 Jan. 1956, Reynolds 7626 holotype (PRE). Our remarkable new species was discovered by Mr. J. F. de Wet in February 1955, near Nederhorst, about four miles north-east of Dull- stroom in the Eastern Transvaal at an altitude of 6,900 ft. When two plants—sent by Mr. de Wet—flowered in Johannesburg in January 1956, I went to Dullstroom and was conducted by Mr. de Wet to a locality where several plants were growing wild in stony ground with grass. A. modesta provides a link between species in the Section LEPTALOE and some species in the Section LeproaLor. Having flowers 13 mm. long, and with perianth outer segments free to base, A. modesta seems nearest allied to A. minima Bak., and the var. blyderivierensis, but in size, leaf characters and habit of growth it is nearest to A. linearifolia Berger. The leaves of A. modesta dilate and thicken below ground level, forming an ovoid bulb-like swelling (25 mm. long, 20 mm. diam.) which is a charac- ter of A. kniphofioides Bak., but the latter has very different leaves, and 35 mim. red flowers. A striking character of A. modesta is that the flowers are densely congested in a subcapitate raceme, and are almost sessile, pedicels being at most only 1 mm. long. This unique character separates A. modesta from all other species in both LepraLor and LEPTOALOE. 85 86 A New Aloe from the Eastern Transvaal Description: Plant acaulescent, succulent, small, solitary. Roots thick, fleshy, tapering, not fusiform. Stem none, with the leaves dilating and thickening below ground level and forming an ovoid bulb-like swelling about 25 mm. long, 20 mm. diam. Leaves 4—6, rosulate, compactly ascending, 15—20 cm. long, 8—9 mm. broad at ground level, linear in lower half, gradually tapering to an acute apex; wpper surface dull deep green, slightly canaliculate, without spots or markings; lower surface rounded, the colour of the upper surface, copiously white-spotted near base only, the spots crowded, dull pale green, lenticular; margins with exceedingly narrow cartilaginous edge, straight, edentate. Inflorescence simple, 25—30 em. high. ‘Peduncle 3—4 mm. diam. low down, slightly thickening upwards, with several ovate-acuminate subscarious, sterile-bracts, the lowest 15 mm. long, 6—7 mm. broad, 5—7-nerved, gradually smaller upwards. Raceme subcapitate, slightly conical, 35—40 mm. long, 30—35 mm. diam., very densely flowered, the buds suberect, open flowers sub-hori- zontally spreading. . Bracts lowest ovate-acuminate, 10 mm. long, 6 mm. broad, thin, subscarious, 3—5-nerved. Pedicels 1 mm. long. Perianth yellowish-green, 13 mm. long, cylindric-trigonous, basally obtuse, cylindric and 4 mm. diam. across the ovary, thence trigonal with the mouth slightly upturned; outer segments free to base, obscurely 3-nerved, apices subacute; imner segments spathulate, with broad thin white borders and with three crowded nerves forming a greenish keel throughout, the apices more obtuse than the outer. Filaments pale rose, filiform-flattened, the three inner narrower and lengthening before the three outer with their anthers in turn exserted 23) Tdiao, Stigma at length exserted 3—4 mm. Ovary green, 4 mm. long, 2-5 mm. diam., obtusely tapering into the style. Puate XVII. Aloe modesta Reynolds. Plants x } approx., ex near Dullstroom, Eastern Transvaal, 6,900 ft., cult. Johannes- burg, fl. 7 Jan. 1/956: Fie. |. Eire, 2: Puiarr XVIII. Aloe modesta Reynolds. Fre. 1.—Roots, underground swelling. and base of leaves 1/1. Fre. 2.—Raceme 1/1. Fre. 3.— Flowers 1/1 from bud to post-pollination stage; also bract 1/1. Iie, 33, JOURNAL OF SOUTH AFRICAN BOTANY VOL. XXII. Published: Juny Ist, 1956 THE SOUTH AFRICAN SPECIES OF AIZOACEAE. Til. GALENIA L. By R. 8. ADAMSON. The genus Galenia is confined to southern Africa where it is especially characteristic of the drier regions of the west and south-west. It is absent from the main summer-rainfall areas. The genus is a distinctive one characterised by exstipulate leaves, small flowers, a superior ovary with a single pendulous ovule in each carpel, and stamens twice as many as the segments of the perianth and arranged in pairs alternating with the perianth lobes. Along with the endemic South African genus Plinthus which has isomerous stamens, it constitutes a group in the tribe Aizoineae. The genus was founded by Linnaeus in 1753 for the species G. africana. Shortly afterwards his son added a second species. The connotation of the genus was extended by Fenzl in 1839 when he merged Kolleria Presl. with it. In 1843 Walpers also combined Sialodes E. and Z. It was further enlarged in 1862 when Fenzl transferred to it a number of species previously assigned to Aizoon. The only comprehensive account of the species is that by Sonder, Fl. Cap. 2. 473 1862, which was largely based upon the work of Fenzl. In it Sonder recognised 18 species. Since that time a number of additional species have been described, especially from Namaqualand and 8.W. Africa, but no general account has been undertaken. In collections there has been some confusion in the identifi- cation of specimens. In the present account not all the described species have been upheld, some are reduced to varietal rank, some to synonymy. In all 27 species are recognised of which 6 are described for the first time. Several new varieties are also described. 87 88 The Journal of South African Botany. In the course of this revision, in addition to fresh specimens, material in the following collections has been examined:— Bolus Herbarium, Cape Town.. ne os BOL University of Cape Town 4a JA ie CT South African Museum, Cape Town .. te SAM National Bot. Gardens, Kirstenbosch re NBG National Herbarium, Pretoria. . a: ey PRE Albany Museum, Grahamstown af a GRA Royal Bot. Gardens, Kew a 2 Ss K British Museum (Nat. Hist.) London a BM Botany School, Cambridge... ay oe CGE Naturhist. Riksmuseum, Stockholm .. ee 8 _The specimens in the Linnean herbarium, London have also been examined. To the curators and staff of these institutions, thanks are returned for facilities for examination of the specimens. GALENIA LL Sp: P3859 mii53. (Gens Plvedhoml6on iat: Kolleria Presl Symb. Bot. 2. 24. t14. 1832. Sialodes Eck. & Zeyh. Enum. 329. 1837. Shrublets, less often herbaceous. Leaves opposite or alternate, simple, usually small, exstipulate. Flowers small, in cymes, each with a bract and two bracteoles: bracts and bracteoles often adnate to the perianth. Perianth segments 4 or 5, rarely 3, united at the base, the lobes coloured inside. Stamens twice as many as the perianth segments, in pairs alter- nating with the segments: anthers versatile. Ovary superior, carpels 2, 3, 4 or 5, each with | pendulous ovule: styles free, as many as the carpels: stigma all along the inner side of the style. Fruit dry, leathery, splitting at the top. Seeds compressed or flattened, reniform or orbicular. The type species is G. africana L. Perennials, most commonly with shoots that die back to the base in the dry season. Erect, or more commonly diffuse with secund branches. In most species there is variation in leaf size, the lower larger; several produce unusually large leaves at the beginning of the growing season. The larger leaves often drop in the later stages. In many species the leaves have enlarged water-storage cells in the epidermis. In the fresh state these project as glistening papillae, in dried material they appear as rounded or polygonal depressions on the surface. The flowers are small, either sessile or nearly so. While the structure of the flower is very uni- form, the form of the inflorescence and the size of the flower are characte- ristic for the species. The flowers are protandrous though the stigmas ripen before the shedding of pollen is complete. The stamens ripen in The South African Species of Aizoaceae. 89 succession, the filament elongates to 3 times its length at dehiscence. One of each pair elongates at a time. Most abundant in the drier regions in the west and south-west, especially along the margins of the Karoo. Also occurring on the southern coastal belt. All are lowland plants, none are found on the higher moun- tains. Fenzl (Ann. Wien. Mus. 2. 288. 1839) was the first to divide the genus into subgenera. He recognised two:—(1) Hugalenia perianth 4: styles 2: fruit tardily dehiscent; and (2) Kolleria perianth 4 or 5: styles 3, 4 or 5: fruit dehiscent. His scheme has been generally followed by later authors (e.g. Sonder Fl. Cap. 2. 474. 1862; Pax in Engl. & Prantl Nat. Pflanzf. 3.1.b. 43. 1889; Della Torre & Harms Gen. Siphonog. 1. 154. 1904; Pax & Hoffmann in Engl. & Prantl Nat. Pflanzf. ed.2. 16.c. 224-5. 1934). However, an examination of a number of specimens has thrown doubt upon whether this division is a natural one. There are included in the subgenus Hugalenia a number of species which in appearance, habit and general structure are so much like species in the subgenus Kolleria that subgeneric separation on the number of styles seems unnatural. Several of these species have either 4 or 5 perianth segments, and in a few there are examples with 3 styles. It is to be noted that several of these species were originally ascribed to Aizoon and transferred by Fenzl. In addition to the species referred to, there is in the subgenus Hugalenia a distinctive group of species characterised by opposite narrow leaves, symmetrical inflorescences, 4 perianth segments and two carpels. The species in this group are summer-flowering, Dec.—Feb., whereas all the others flower earlier, Aug.—Oct. It is proposed to separate this group as a subgenus. As it contains the type species it will carry the name Galenia. As so separated the subgenus has the same connotation as the original genus. All the other species, which are characterised by secund branches at least in the inflorescence and most commonly by obovate leaves, are placed in the subgenus, Kolleria with an enlarged connotation. This sub- division groups the species into easily distinguishable groups which are based on habit, leaf-form and arrangement, inflorescence and flower struc- ture. The two subgenera are defined as:— Subgen. 1. Galenia. Eugalenia Fenzl Ann. Wien. Mus. 2. 288. 1839. p. pte. Erect or suberect shrublets. Leaves opposite, linear or linear-oblong, usually glabrous or nearly so. Flowers in terminal cymes, usually panicu- late, not secund. Perianth segments 4, in two dissimilar pairs, membra- nous at the edges and tips. Styles 2, ovary flattened. Fruit not or tardily dehiscent. Plants green or yellow, not gray or white. 90 The Journal of South African Botany. The type species is G. africana L. Subgen. 2. Kolleria Fenzl 1.c. emend. Presl Symb. Bot. 2. 24. t.14. 1832. pro gen. Prostrate or diffuse shrublets. Leaves usualiy secund, opposite or alternate, most commonly obovate or spatulate, very often gray or white. Flowers in secund cymose inflorescences. Perianth segments 4 or 5, usually alike. Styles 2, 3, 4 or 5. Fruit dehiscent. The type species is G. herniariaefolia (Presl) Fenzl. Subgen. 1. Galenia. Key TO THE SPECIES. to 1. Stems densely covered with stout spreading hairs. . 1. Stems glabrous or with short appressed hairs. 2. Leaves subulate, connate at the base: perianth Pee with a stout mucronate tip 6. Ecklonis 2. Leaves not subulate, not connate, perianth without a mucronate point. 3. Stem with internodes as long as the leaves: ileesras at least 2 em. long. 4. Leaves soft, drooping when dry, soon deciduous. 5. Leaves linear or linear-lanceolate: flowers yellow- green, 1—1-5 mm. long. 6. Leaves flat, often over 2 em. long: inflorescence hispidissima aspreading panicle .. 1. africana . Leaves filiform: inflorescence with long secund branches i ee ie ae a 1. africana var. secundata 5. Leaves oblong or obovate: flowers 1-5—2 mm. long ae si ee ri << a 1. africana var. pentandra 4. Leaves rigid, distant, less than 2-5 cm. long: flowers white, in a flat- topped umbel 3 3. rigida Stems with internodes not longer than the leaves: leaves squarrose, less than 2 em. long. 7. Leaves up to 2 em. long, usually widened to the tip: inflorescences many-flowered, exposed: ovary not indented at the top... 4. namaensis . Leaves rarely up to I em. long, not widened to the tip: inflorescences few-flowered, surrounded by leafy branches: ovary indented at the top NS) ~~ Or . procumbens 1. G. africana L. Sp. Pl. 359. 1753. G. linearis Thunb. Prod. Pl. Cap. 77. 1794. An erect aromatic green or yellow-green softly woody shrublet, 0-5— — 1-5m. high. Stems pale-coloured, glabrous except for very short appressed hairs on the youngest parts. Leaves opposite, bright green but the older yellow or orange, soft, drooping when dry, flat, linear, narrowly oblong or very narrowly spatulate, 1-5—5 em. long, 0-2—0-5 cm. wide, acute, the upper shorter. Inflorescence a terminal panicle, pyramidal, corymbose or almost umbellate, 3—10 cm. long, 2—12 em. wide: flowers numerous, yellow-green. Bracts at the base of the inflorescence leaf-like, The South African Species of Aizoaceae. Il in the distal parts very small. Bracteoles not longer than the flowers. Perianth segments 4, 1—1-5 mm. long, the outer cucullate at the tip, the inner flatter, with a cylindrical mucronate pcint, generally sickle- shaped or bent back. Anthers yellow or orange. less often red. Styles thick. The type is in herb. Linn. (507: 1.) Common especially at the western and southern margins of the Karoo. Not eaten by stock and spreading extensively in overgrazed or disturbed vegetation, also along roadsides. Certainly much more abundant and widespread than in the past. Very variable in size, amount of branching, and in colour. Occasionally covered all over with short hairs. Easily recognised by the long soft leaves and many-flowered inflorescences with small yellow-green flowers. The inflorescences may be solitary or several together and vary much in shape and density. At the base they are forked but often appear to be trichotomous, one division appearing as a lateral branch. The ultimate branches are usually divaricate. S.W. AFRICA. Windhoek Liebenberg 4479 (PRE); Fish Riv. Marloth 3734 (PRE); Sjambok Riv. Pillans 6464 (BOL. K.); Holsog Marloth 8984 (PRE); Scharfenstein Pearson 8575 (BOL. K); Schackalskuppe Pearson 4799 (K); Zoutpansdrift Pearson 5013 (GRA. K); KI. Karasberg Galpin 14129 (BOL. K.). NAMAQUALAND. Steinkopf Schlechter 31 (BOL, GRA, PRE), s.n. (BM, K); Okiep Pillans 4952 (BOL); Horeas Riv. Pillans (BOL); Walle- kraal Pillans (BOL): Garies Henrici 2158 (PRE). CALVINIA. Loeriesfontein Acocks 15209 (PRE); Nieuwoudtville Theiler 89 (PRE); Calvinia Theron 1622 (PRE): Botter Kloof Adamson 5235—6; Blaaukrantz Pass Adamson 4552. KENHARDT. Brandvlei Hsterhuysen 4001 (BOL, PRE); Tyson 4001 (NBG); Kenhardt Adamson 2033. VAN RHYNSDORP. Van Rhynsdorp Thode A1895 (K, PRE). CLANWILLIAM. Blaauberg Pillans (BOL); Clanwilliam. Henrici 3295 (PRE); Leipoldt 912 (BOL); Rogers 16803 (BM); Greys Pass Pearson 5146 (BM, K); Warm Baths Stephens 7260 (K). CERES. Leeuwfontein Pearson 3244 (BM, GRA); Karroo Poort Adamson 4728; Pearson 4814 (BM): Hottentots Kloof Pearson 4814 (BM, K). SUTHERLAND. Middelpost Adamson 5244—5, 5247: Roggeveld Worsdell (K): Verlaten Kloof Adamson 5239. LAINGSBURG. Rd. to Sutherland Pole Evans 19107H (PRE): Matjesfontein Galpin 4075 (GRA): Foley 4 (PRE): Marloth 702 (PRE): Whitehill Compton 12604 (NBG). 92 The Journal of South African Botany. PRINCE ALBERT. Prince Albert Stokoe 9323 (BOL), sn. (SAM). WORCESTER. Touws Riv. O. Kuntze (K): Hex Riv. “R.L.D.” (PRE): Rehmann 2469 (BM, K): Tyson 763 (GRA, SAM): Worcester Bolus 5055 (BM, BOL): s.n. (BM): Heklon & Zeyher (GRA, PRE). TULBAGH. Tulbagh Pappe (GRA). PAARL. Paarlberg Drege (S). STELLENBOSCH. Hottentots Holland Ecklon & Zeyher (86) (S): Zeyher (S): Somerset West Parker 3630 (BOL, K, NBG). CAPE. Blaauberg Pillans (BOL): Milnerton Adamson 2691 (CT): Koeberg Pappe (K): Zeyher (SAM): Cape Flats Rehmann 2010 (BM): Salt Riv. Schlechter 285 (BM, BOL, CGE, GRA), s.n. (PRE): Rosebank Page (CT): Pillans 3278 (BOL, PRE). CALEDON. Sonderende Schlechter 5632 (BM, GRA, K, PRE). BREDASDORP. Elim Zeyher (S). ROBERTSON. Robertson Britten 647 (GRA). MONTAGU. Koo Barnard 774 (SAM): Montagu Ecklon & Zeyher 309 (S): Page 115 (PRE): Kogmans Kloof O. Kuntze (K). SWELLENDAM. Swellendam Bowie (K): Bolus (BOL): Breede Riv. Zeyher 2638 (S, SAM). t RIVERSDALE. Garcias Pass Adamson 5146—8: Muiskraal Galpin 4074 (GRA): Albertinia Mwir 597 (PRE). MOSSEL BAY. Herbertsdale Burchell 6414 (K): Mossel Bay Moss 6815 (BM). GEORGE. George Bolus (GRA): Olifants. Riv. Bolus 2313 (SAM), 2315 (BOL, K, SAM): Montagu Pass de Castelnau 65 (PRE). UNIONDALE. Georgida v. Rensburg $1062 (PRE): Uniondale Fourcade 4286 (BOL, K). WILLOWMORE. Willowmore v. Z. Bakker 237 (PRE). ALBANY. Grahamstown Foley 1657 (K). Without locality:—Alexander (PRE): Drege (BM, K): Ecklon (8): Forsyth (K): Forster (Kj: “Hort. Chelsea” (BM): Sieber 65 (S), “Swartland”’ (SAM): Talbot (K): Thunberg (BM): Verreaux (B) Wallich (CGE, K): Zeyher (K, SAM). A specimen without collector’s name is marked “St. Helena” (CGE) For hybrids see under G. namaensis and G. procumbens. Var. secundata Adamson var. noy. A typo differt foliis angustissimis filiformibus, inflorescentia secunda ramis alternis secundis subdivaricatis, floribus minoribus, staminibus saepe 4—5 vel 4—6 solum perfectis. ; More slender and less erect. Stems not over 45 em. high, white. Leaves ‘very narrow, linear-filiform, soft, slightly glaucous. Inflorescence secund, The South African Species of Aizoaceae. 93 with few main branches but many short alternate divaricate lateral branchlets. Flowers about 1 mm. long. Inner perianth segments with an incurved flap at least half their length. Stamens 8, 6 or 4, commonly 4 complete and 2—4 with very small anthers. The type is Bryant 115 (K). S.W. AFRICA. Luderitz Kinges 2462 (PRE): KI. Karasberg Galpin 14129 (BOL, K, PRE). KENHARDT. Kakamas Wasserfall 1143 (PRE). PRIESKA. Prieska Bryant 115 (K), J115 (PRE): Pole Evans 2334 (K). VICTORIA WEST. Victoria West Reynolds 4186 (PRE). Var. pentandra Hiern Welw. Cat. Afr. Pl. 1.2. 411. 1898. Sprawling, rarely over 40 cm. high, and more woody. Branches often divaricate. Leaves oblong or obovate, 1-3—2-5 cm. long, 0-4—0-8 cm. wide, hooked or falcate at the tip, narrowed to a petiole at the base. Inflorescence corymbose with rather stout branches. Flowers 1-5—2 mm. long. Bracts and bracteoles adnate to the perianth and often shorter. Perianth segments 4, the inner bent inwards at the tips. Styles occasional- ly 3, about as long as the perianth. The type is Welwitsch 1093 (BM). Originally described as having 5-merous flowers with 5 stamens but an examination of the type material does not confirm this. It is possible that the bract or bracteoles may have been mistaken for part of the perianth, and in older flowers some of the stamens fall off and an appear- ance of isomery results. This variety seems to agree with the meagre description of var. halimifolia Sond. Fl. Cap. 2. 478. 1862 but the only specimen so named in Sonder’s herbarium, Drege G. africana L.a, consists of two parts, a bunch of leaves that agree with the description, and a flowering shoot which is typical G. africana. Var. pentandra was first recognised by Oliver Fl. Mrop: Air: 2. 585. 1871 but was not named by him. ANGOLA. Mossamedes Welwitsch 1092 (BM), 1093 (BM, K). NAMAQUALAND. Vogelklip Schlechter 11301 (BM, K, PRE). 2. G. hispidissima Fenz] in Sond. Fl. Cap. 2. 478. 1862. A low-growing many-stemmed plant up to 15 cm. high, in bunches. Stems decumbent, woody at the base only, rather thick, the younger densely covered with stout yellow or white bristle-like hairs 12 mm. long, standing out at right angles. Leaves opposite, linear, flat, 0-7— 2-5 cm. long, green, narrowed to the base, acute at the tip, with scattered bristle hairs on the edges and underside. Flowers in a crowded panicle. Bracts like the leaves but much shorter, those on the ultimate branches 94 The Journal of South African Botany. shorter than the flowers. Perianth segments 4, 2—3 mm. long, acute, with some bristle-hairs outside. Anthers orange or red. Styles 2, short. The type is Drege 7055. Sandy soils among bushes on the lowest mountain slopes. CLANWILLIAM. Warm Baths Pearson 7172 (SAM): Stephens 7172 (GRA, K, SAM). PIKETBERG. Piketberg Bolus 7545 (BM, BOL, GRA, K, NBG, PRE): Edwards (BOL): 170 (BOL): Guthrie 2607 (NBG). CERES. Gydouw Leipoldt 4012 (BOL). Without locality—Drege (8). 3. G. rigida Adamson sp. nov. : -Fruticosa glabra ramosa, ramis floriferis erectis eramulosis. Folia subrigida lineari-olbonga erecto-patentia superiores internodiis breviora inflorescentia umbellatocorymbosa compacta. Bracteae superiores reflexae. An erect rigid branching shrublet up to 50 cm. high, often yellowish. Stems pale, covered with wax, sometimes sticky. Leaves rigid, linear- oblong 1—1-25 em. long, 0-2—0-3 em. wide, flat or more usually concave above, ascending, glabrous and often sticky, the upper shorter than the internodes. Inflorescence terminal, cymose-umbellate, flat-topped, rather compact, on an erect unbranched shoot 5—20 cm. long. Bracts smaller than the leaves, the lower reflexed, all but the ultimate longer than the flowers. Bracteoles very small. Perianth segments 1-5—2 mm. long, white inside, distinctly connate, the outer flattish, cucullate at the tip, the inner with the acute or acuminate membranous tip sharply bent inwards. The type is Adamson 5240 (BOL). Rocky places especially on dolerite in the N.-W. Karoo. CLANWILLIAM. Olifants Riv. Rodin 5051 (BOL, K, PRE). SUTHERLAND. Btn. Sutherland & Middelpost Adamson 5240—1, 5243, 5246. 4. G. namaensis Schinz Bull. Herb. Boiss. 5. App.3. 76. 1897. Shrubby, erect, diffuse, or occasionally sprawling, up to 60 em. high? The young parts covered with short closely appressed hairs, the older . glabrescent. Leaves dull green or glaucous, yellow when old. Branches white, ascending or divaricate. Leaves usually crowded, rather rigid, linear but more or less widened towards the tip, 0-3—2 em. long, most commonly 0-5—1-5 cm., about 2 mm. wide, folded upwards, most often falcate with the tip hooked. Inflorescences exposed, terminal on the main shoot and branches, usually numerous, varying from laxly cymose to compact, umber-like or almost hemispherical, 2—3 om., rarely up to The South African Species of Aizoaceae. 95 5 em. across. Bracts like small leaves, the upper very small: bracteoles thick, concave above, about as long as the flowers. Perianth segments about 2 mm. long, white or greenish inside, the outer cucullate, the inner with a shortly mucronate tip. Anthers yellow, orange or less often pink. Ovary flat, not indented at the top. Seeds black, reniform. The type is Fenchal 171 (Z). Rocky places, especially in dry river beds, at the north and west margins of the Karroo. Rather variable in the length and closeness of the leaves and in the number and density of the inflorescences. The species has been much confused with G. africana and G. procumbens. From the former it is distinguished by the shorter more rigid leaves, the small inflorescences largely on lateral branches, and the white flowers. From the latter by the longer and more clavate leaves, the more compact and exposed inflorescences, and by the ovary not indented at the top. S.W. AFRICA. Rehoboth Rodin 2781 (BOL, PRE, K); Swakop Riv. Moss 17938 (BM); Windhoek Codd 5809 (PRE); Steyn 8964 (PRE); Walfish Bay Rodin 2139 (BOL, K, PRE); Nomtsas Pearson 9327 (BOL, K); Rossing Pole Evans 19334H (PRE); Putzonderwater Pole Hvans 19374H (PRE); Waterberg Bradfield 450 (PRE); Keetmanshoop Fenchal 171 (K); Gerstner 6397 (PRE); Anglo-German Kalahari Ireland (BOL). NAMAQUALAND. Kuboos Pillans 5428 (BOL); Okiep Pillans 4952 (BOL); s.l. Zeyher (K). KENHARDT. Kakamas Lewis (SAM). PRIESKA. Vogelstruisdrift Bryant J166 (PRE). HAY. Douglas Burchell 1741 (K). KIMBERLEY. De Beers Mine Wilman (McG. Mus. 427) (K, PRE); Shaws Drift Acocks 2581 (PRE). CARNARVON. Buffelshout Burchell 1613 (K). DE AAR. DE Aar Theron 533 (PRE). FAURESMITH. Knolfontein C. A. Smith 5393 (K, PRE); Berg Riv. C. A. Smith 5437 (K); Bokland C. A. Smith 4623A (PRE); Ruigtpoort Henrici 2610 (PRE); Groenvlei Rabie (PRE). PHILIPPOLIS. Spion Kop C. A. Smith 4593 (PRE), 4596 (K); Philippolis Sidey 403 (PRE). G. africana x namaensis. Specimens intermediate between the putative parents have been referred to this hybrid. These range from plants much like G. africana but with much shorter leaves and closer inflorescences, on the one hand, to plants like G. namaensis but with distant straight leaves on the other. S.W. AFRICA. Swakop Riv. Galpin & Pearson 7480 (K, PRE, SAM). 96 The Journal of South African Botany. GORDONIA. Narugas nr. Upington Moss 11559 (BM). PRIESKA. Marydale Thorne (SAM). 5. G. procumbens L.f. Suppl Pl. 227. 1781. An erect much branched shrublet 30 cm. to 1 m. high, less often semi- prostrate, often orange or yellow in colour, glabrous or with a few appressed hairs on the youngest parts. Branches opposite, ascending or divaricate, sometimes partly secund. Leaves opposite, deciduous on the older parts, linear, 0-3—0-6 cm. long, squarrose or falcate, hooked at the tip, usually deeply concave above, those on the lateral branches often crowded. Flowers usually few, in lax cymes on branches shorter than surrounding leafy shoots, the ultimate branches of the cyme monochasial. Bracts as long as the flowers: bracteoles short. Perianth segments 4, 1-5—2 mm. long, white or tinged with pink inside, the outer cucullate, the inner with an acute sharply incurved tip. Anthers orange, less often red. Styles 2 or rarely 3; ovary flattened, indented at the top with rounded lateral bulges. The type is Linn.f. 91 in herb. Linn. (507: 3). Rocky or stony places near the margins of the Karoo, especially in the south. Variable in habit. Though originally described as ““‘procumbent”’ the commonest forms are erect shrublets and agree completely with the type specimen. Prostrate or semiprostrate forms occur for the most part by roadsides and on overgrazed areas. There are all stages of transition. Owing to this and to the original description of the leaves as “‘ovate’’, there has been a good deal of confusion in collections. Specimens have been referred to a variety of species. Some of the larger erect forms have been grouped under the MS name “G. flavescens’’ along with specimens of G. namaensis. NAMAQUALAND. Kamiesberg Ecklon & Zeyher (GRA, S). CALVINIA. Middelpost Adamson 5237; Zak Riv. Burtt-Davy 17560 (GRA). KIMBERLEY. Riet Riv. Acocks 2600 (BOL, K, PRE); Boshof Rd. Acocks 8540 (K, PRE). ‘ HERBERT. Douglas Kotze 802 (PRE). BEAUFORT WEST. Beaufort West Henderson 11(K); Sunnyside Esterhuysen 2739 )BOL): rd. to Aberdeen Adamson 5248. HANOVER. Hanover Shaw 50 (K); Sim (Galpin 5977) (PRE). MURRAYSBURG. Murraysburg Tyson 146 (BM, SAM); Koudeberg Adamson 5207—8. GRAAF-REINET. New Bethesda Acocks 13224 (PRE): Graaf-Reinet Bolus 102 (BOL, GRA); Koudeberg Adamson 5206. The South African Species of Aizoaceae. 97 MIDDELBERG. Wapadsberg Adamson 5251; Grootfontein Gill 32, 82 (PRE); Theron 87 (PRE); Verdoorn 1522 (PRE), 1605 (BOL. PRE); Middelburg Rd. Flanagan 1382 (BM, PRE, SAM); Bangor Farm H.#H. Bolus (BOL); Rietpoort Comins 710 (K, PRE); Naauwpoort Dyer 1906 (PRE); Culmshoek Galpin 7680 (PRE). CRADOCK. Cradock Brynard 154 (PRE); Dyer 1027 (PRE). TARKA. Tarka Riv. Zeyher (116, 12) (8). SOMERSET EAST. Groot Riet Vlei Acocks 11959 (PRE); Zwaggers- hoek Theron 50 (BOL, K, PRE). UITENHAGE. Commonage Prior (K). FAURESMITH. Groenvlei Verdoorn 1028 (PRE); Brandfontein Henrici 7349 (PRE); Brakpoort Henrici 4392 (PRE); C. A. Smith 5174 (PRE) Luckhoff Sidey 342 (PRE). KOFFIEFONTEIN. Dreifontein Kies 293 (K, PRE). Without locality: —Thunberg (BM). G. africana x procumbens. Specimens collected at Uniondale (Adamson 5188—9) may be this hybrid. They exhibit leaves like those of G. procumbens but more widely spaced, white stems, and a lax terminal panicle of small greenish-yellow flowers. G. namaensis x procumbens Though this hybrid has not been clearly determined, there are speci- mens which are in many features intermediate between the two closely related species which probably represent it. 6. G. Ecklonis Walp. Rep. 2. 232. 1843. Sialodes glauca EK. & Z. Enum. 329. 1837. G. glauca Sond. Fl. Cap. 2. 478. 1862. non Walp. A procumbent green or yellow-green almost herbaceous plant with many stems from a short woody rootstock. Stems 10—40 cm. long, glabrous or nearly so, sometimes with some appressed hair-like scales on the youngest parts. Leaves opposite, linear-subulate, rolled upwards, 0-5—I cm. long, often falcate or reflexed, distinctly connate at the sheathing base. Flowers in a lax terminal dichasial cyme. Bracts like the leaves but smaller, straight, not longer than the flowers. Perianth segments 4, 3—4 mm. long, acute, white or red inside, green outside, with a short straight cylindrical mucro at the tip. Anthers pale yellow. Styles 2. The type is Ecklon 2149 (B), lectotype (K). Sand at low altitudes in the south-west. Often reddish in the older stages. A very distinctive species, easily recognised by the herbaceous habit, connate leaves and large flowers with 98 The Journal of South African Botany. a mucronate perianth. The name, G. glauca Sond., by which this has been known, is antedated by G. glauca Walp. a synonym of G. pubescens var. pallens. PIKETBERG. Piketberg Guthrie 2614 (NBG): Piketberg Rd. T'yson 2322 (BOL). MALMESBURY. Malmesbury Compton 13253 (NBG); Mamre Rd. Salter 4900 (BM, K). TULBAGH. Winterhoek Zeyher (S, SAM); Predikstool Zeyher (8); Tulbagh Kloof Bolus 7546 (BOL, GRA, K, PRE). WORCESTER. Villiersdorp Compton 22914 (NBG). PAARL. Hermon Esterhuysen 6054 (BOL, K). STELLENBOSCH. Stellenbosch Flats Duthie 1908 (BOL). Without locality:—Ecklon 2149 (K); Ecklon & Zeyher (CGE): “Herb. Steudel” (PRE). Subgen. 2. Kolleria cf. p. Key TO THE SPECIES. 1. Styles 3, 4 or 5, rarely 2 by reduction. 2. Perianth segments produced into a cylindrical point. 3. Leaves rhombic: perianth segments 4—5: styles 4: stems prostrate 15. prostrata 3. Leaves obovate, acute: perianth segments ip - styles 5: stems decumbent- ascending .. 16. acutifolia 2. Perianth segments not produced into a cylindrical point. 4. Plants prostrate, secund: inflorescence branches secund. 5 Glabrous except for a few hairs on the youngest parts. 6. Leaves linear or linear-lanceolate. 7. Perianth segments 5: styles 5. 8. Branches prostrate: leaves flat: flowers crowded on lateral branches... .. 7. filiformis §. Branches ascending: leaves concave above: flowers solitary, axillary . . : 9. exigua 7, Perianth segments 4: styles 4: leaves con- cave above .. 8. herniariaefolia 6. Leaves oblong, elliptical or spatulate, "mostly concave above. 9: Leaves small, succulent: flowers less than 1-5 mm. long: styles 5 3.6 11. subcarnosa 9. Leaves not succulent: flowers at least 1-5 mm. long. 10. Leaves acute: styles mostly 3. 11, Stems pale or yellow: leaves often sticky, green: flowers large 10. collina 11. Stems white, thickly covered with wax: leaves grayish. . Oc .. 14. pubescens var. cerosa 5. Hairy, sometimes only on the young parts. 12. Leaves linear or linear-lanceolate, usually acute. 13. Stems slender: leaves flat: aes glab- rous: styles usually 5 ae 7. filiformis 13. Stems woody: leaves concave "above: perianth hairy: styles usually 3. .. 12. affinis 1. Styles 2. The South African Species of Aizoaceae. 12 Leaves obovate or spatulate, usually obtuse 14, Plants green, glabrescent. 15. Leaves papulose: flowers on lateral branches: perianth 5: styles 5 15. Leaves finely papulose: flowers axil- lary: perianth 4: styles 4 14. Plants gray or white, hairy. 16. Leaves small, acute, covered with short appressed hairs: styles 3 16. Leaves larger, obtuse, covered with loosely appressed hairs: styles 5 4. Plants diffuse or ascending, often with secund flowering branches. 17. Glabrous or nearly so. 18. Leaves acute, green: stems yellow: styles 3 .. 18. Leaves obtuse, spatulate. 19. Leaves papulose, not succulent. 20. Flowers in a terminal forking in- florescence: styles 4 or 5 20. Flowers on secund lateral branches 19. ulose. 21. Leaves persistent: Leaves succulent, usually coarsely pap- inflorescence secund: stems herbaceous .. Bye 21. Leaves soon deciduous: inflorescence not secund: stems woody Be 17. Covered with hairs or scales, at least on the younger parts. 22. Leaves succulent 22. Leaves not succulent. 23. Stems woody: leaves acute, covered with short appressed hairs: styles 3 .. a 23, Stems herbaceous: leaves obtuse loosely covered with long hairs styles 3or5 .. 24. Leaves orbicular, thick: stem densely covered with large papillae go ole oe Me 24. Leaves not orbicular: stems not covered with papillae. 25. Covered all over with flat spreading scales: leaves succulent a 56 ee a6 O60 25. Glabrous or covered with appressed hairs or hair- like scales. 26. Succulent: leaves cylindrical-oblong: flowers in a dense terminal panicle 26. inflorescence mostly secund. 27. Leaves smooth, often white, folded up- wards: perianth segments 4 27. Leaves papulose or hairy. 28. Leaves not obviously papulose, hairy, gray inflorescence usually rigid 28. Leaves papulose. 29. Leaves with appressed hairs or hair- like scales Usually woody: leaves more or less spatulate: 29. Leaves glabrous or nearly ‘80. 30. Stems branches: with flowers ascending crowded: inflorescence pedunculate 30. Stems with divaricate bran- ches: flowers in a lax group 12. 19. 20. 23. 21. 22. 25. . pubescens . pubescens var pallens . affinis . secunda . collina - cymosa . pubescens var. Fourcadet - papulosa - pruinosa - sarcophylla affinis portulacacea . Dregeana squamulosa fallax . erystallina fruticosa crystallina . pruinosa hemisphaerica 99 100 The Journal of South African Botany. 7. G. filiformis (Thunb.) N.E. Br. Bothalia 1. 155. 1923. Mesembryanthemum filiforme Thunb. Prod. Pl. Cap. 89. 1789. Fi. Cap. ed Schult. 417. 1823. Prostrate, slender, green or reddish. Stems 20—80 cm. long, branched from the base, with small lateral branches in the flowering region. Young parts either glabrous, minutely papulose, or with few or occasionally many appressed hairs. Leaves alternate, linear or linear-oblong or oblanceolate, usually acute, 0-5—1 cm. long, 0-1—0-3 cm. wide, flat or concave above, glabrous or hairy. Flowers on alternate secund branches, 1—5 em. long, usually crowded near the base. Bracts leaf-like. Perianth segments 5, about 2 mm. long, glabrous or very thinly hairy. Anthers red, less often yellow. Styles 5 or rarely 3. The type is in herb. Thunberg, Uppsala. Sand or gravel, especially in river beds. Rather variable in the amount of hairs. The type specimen is glabrous but plants otherwise indistinguishable have appressed hairs at least on the young parts. CALVINIA. Nieuwoudtville Maguire 211 (BOL). HERBERT. Douglas Orpen 12, s.n. (SAM); St. Clair Orpen 128 (SAM). HOPETOWN. Hopetown Marais 6 (PRE). SUTHERLAND. Tanqua Riv. Hutchinson 674 (BOL). WORCESTER. Matroosberg Stn. Adamson 5096. LAINGSBURG. Tweedside Compton 3063 (BOL). ROBERTSON. McGregor Hsterhuysen 4455 (BOL); Ashton Adamson 3764. MONTAGU. Koo Adamson 5102; rd. to Barrydale Adamson 4644, 5109. HOPEFIELD. Darling Stokoe (SAM). CAPE. Paarden Is. Levyns 5666 (CT); Ndabeni Adamson 3133 (CT, SAM). CALEDON. Sonderende Adamson 4574—80. BREDASDORP. Cape Agulhas Hsterhuysen 4412 (BOL). SWELLENDAM. Storms Vlei Adamson 4583; Fries, N orlindh & Weimarck 1517 (BOL); Breede Riv. Zeyher 2637 (8, SAM); Slang Riv. Adamson 4609; Malagas Esterhuysen 4449 (BOL). RIVERSDALE. Albertinia Adamson 4848. MOSSEL BAY. Herbertsdale Burchell 2551 (K). UNIONDALE. Uniondale Paterson 3082 (GRA). KING WILLIAMSTOWN. K. Williamstown Tyson 2925 (SAM). Without locality: —Lehmann (K). The South African Species of Arzoaceae. 101 8. G. herniariaefolia (Presl) Fenzl in Sond. Fl. Cap. 2. 476. 1862. Kolleria herniariaefolia Pres] Sym. Bot. 1. 24. tl4. 1830. Aizoon microphyllum Bartl. Linnaea 7. 541. 1832. A. Herniaria Reichb. ex Steud. Nom. ed. 2. 1. 4. 1848. Prostrate, rather dark green or brownish. Stems woody, up to 75 cm. long, branched, in the upper parts with regular close alternate flowering branches, 1—6 em. long. which are pinnately or bipinnately branched. Young stems with scattered hairs, the older glabrescent. Leaves small, linear-oblanceolate. 0-3—0-6 cm. long, 0-1—0-4 cm. wide, concave above, those on the main stem obtuse, those on the branches acute. Flowers usually numerous on the secund lateral branches, sessile. Peri- anth segments 4, 2—3 mm. long, white inside, glabrous outside, spreading at anthesis but incurved later. Anthers pink. Styles 4 or rarely 3. The type is Sieber 164 (O) cotype (K). Sand at or near the coast in the south-west. Allied to G@. filiformis but more woody, with much closer and more regular flowering branches, narrow distinctly concave leaves, and 4- merous flowers. N. E. Brown (Bothalia 1. 155. 1923) says the plant is glabrous but all the specimens seen have hairs on the young parts. Brown (1.c.) quotes Aizoon microphyllum Bartl. as a synonym of G. filiformis. The type has not been seen but a specimen so named in Bartling’s hand- writing, Sieber 164 (S), is undoubtedly this species. CAPE. Cape Town Bolus 4631 (K); Raapenberg Guthrie 707 (CT); Black Riv. Page (CT); Roodebloom Ecklon & Zeyher 4821 (SAM); Papen- dorp Zeyher (SAM); Rondebosch Sieber 164 (K. 8); Constantia Hcklon & Zeyher 2142 (SAM); Muizenberg Bolus 4631 (BOL); St. James Wolley- Dod 2168 (BOL); Simonstown Compton (CT); Wolley-Dod 2062 (K); Wright (K); Paarden Is. Adamson 1448 (BOL); Milnerton Adamson 2655, 4756: L. Bolus (BOL); Compton 13056 (NBG); Moss 2407 (BM). CALEDON. Onrust Schlechter 9481 (GRA, K); v. Niekerk 344 (BOL): SWELLENDAM. Storms Vlei Adamson 4586A: Infanta Adamson 4867. RIVERSDALE. Botteliersfontein Mwir 1824 (BOL). GEORGE. Montagu Pass Adamson 5178. 9. G. exigua Adamson sp. nov. Subglabra procumbens ramulis ascendentibus. Folia lineari-oblanceol- ata supra canaliculata minute papulosa. Flores axillares solitarii. Peri- gonil segmenta 5. Style 5: ovarium latissimum apice depressum. A small darkish green almost glabrous plant with many procumbent stems and alternate ascending branches. Stems 10—20 cm. long. Leaves alternate, linear-oblanceolate, 0-8—1-5 cm. long, 0-1—0:3 cm. wide, 102 The Journal of South African Botany distinctly narrowed to the base, concave above, subobtuse, often horned at the tip, very finely papulose, glabrous except for a few scattered hairs on the youngest. Flowers solitary, axillary, much shorter than the leaf-like bracts. Perianth segments 5, 1-5—2 mm. long, usually erect, the outer distinctly cucullate, the inner incurved at the tip. Anthers yellow or pink. Styles 5: ovary broadly pentagonal, more than twice as wide as deep, slightly depressed on top. The type is Acocks 1900 (PRE). Open places in the north-western karroo. Allied to G. filiformis but differing in habit, the more concave and almost petiolate leaves, the cucullate perianth, and the broad 5-chambered ovary. HERBERT. Douglas Acocks 1900 (PRE); Salt Lake Hill Ester- huysen 4225 (BOL, K.) KIMBERLEY. Magersfontein Wilman (McG, Mus. 3417) (BOL, K); Mauritzfontein Acocks 64 (PRE). HOPETOWN. Hopetown Acocks 14347 (PRE). JAKOBSDAL. Jakobsdal Codd 3425 (PRE). 10. G. collina (EK. & Z.)Walp. Rep. 2. 232. 1843. Kolleria collina BE. & Z.Enum. 327. 1839. G. humifusa Fenzl in Sond. Fl. Cap. 2. 476. 1862. A green almost glabrous diffuse, decumbent or prostrate shrublet, often sticky. Stems from a woody base, much branched, up to 50 cm. long, the younger yellow or white. Leaves green, glabrous or the youngest with a few appressed hairs, often sticky, acute, rather variable in size and shape, the lower sub-opposite, oblanceolate or oblong-spatulate, more or less petiolate, 0-5—2-5 em. long, 0-3—0-6 cm. wide, flat or folded upwards, often partly opposite, the upper alternate, sessile, 0-4—1 cm. long, 0-2—0-4 cm. wide, usually concave above. Flowers on alternate secund lateral branches 1—3 cm. long, sometimes with short branchlets. Flowers in cymose groups, often crowded, the central flower sessile, the lateral on branchlets 1—3 mm. long. Bracts leaf-like, much longer than the flowers. Perianth segments 5, 2-5—3 mm. long, white or yellowish, more or less cucullate. Anthers yellow or pink. Styles 3, less often 5. The type is EH. & Z. 2143 (B), lectotype (S). Open ground, most often rocky. Readily distinguished by the acute glabrous green leaves, yellow stems and rather large flowers. The lower leaves fall off at or about flowering time. S.W. AFRICA. Aus Dinter 4154 (BM). NAMAQUALAND. Okiep Pillans 4767 (BOL); Orange Riv. Drege (S, SAM); Groot Derm Pillans 5595 (BOL, K); Lekkersing Taylor 1126 The South African Species of Aizoaceae. 103 (BOL); Steinkopf Schlechter, sn. (BM, BOL, K); Witbank Pillans 5550 (BOL, K); Bowesdorp Schlechter 11216 (BOL, PRE); Kamieskroon Thorne (SAM); Wallekraal Adamson 5023, 5048. Hondeklip Bay Adamson 5041; Pillans (BOL); Bitterfontein Zeyher 722 (BM, CGE, K, PRE, S, SAM). CALVINIA. Nieuwoudtville Lavis (BOL). VAN RHYNSDORP. Van Rhyns Pass Adamson 4545; Van Rhyns- dorp Adamson 4977; Knechtsvlakte Schlechter 8157 (BM, BOL, GRA, K, PRE); Nieuwerust Adamson 4986, 4989. KIMBERLEY. Alexandersfontein Power (PRE); Poljaspoort Acocks 524 (PRE). LAINGSBURG. Cabidu Barker 6784 (NBG). CALEDON. Swartberg Ecklon & Zeyher 2143 (S, SAM): Dasberg Stokoe (SA). RIVERSDALE. Garcias Pass Adamson 4858. GEORGE. Top of Montagu Pass Adamson 5179. Without locality:—Masson (BM). Var. Meziana Adamson comb. nov. G. Mezxiana K.Mill. Bot. Jahrb. 43. Beibl. 97. 68. 1909. G. Dinteri Schellenb. Bot. Jahrb. 48. 501. 1913. Usually quite prostrate, more slender, not sticky, with appressed hairs on the younger parts. Leaves narrower, linear-oblanceolate, not petiolate, flat, 1—2 em. long, not over 0-3 cm. wide. Flowers smaller, 1—1-5 mm. long, mostly in axillary groups. Perianth usually slightly hairy. Ovary wider than long: styles 3 or 4. The type is Schlechter s.n. (B); neotype Schlechter 34 (K). Dry open sand or gravel in Namaqualand and 8.W. Africa. Rather variable in the amount of hairs. G. Meziana was originally described as having 5 stamens only. This appears to be a rare condition. All the specimens seen have 10. G. Dinteri and G. Mizana are the same and cannot be separated. S.W. AFRICA. Aus Dinter 6094 (BM, BOL, K). NAMAQUALAND. Steinkopf Schlechter 34 (BOL, K, PRE); Wolfhoek Esterhuysen 1331 (BOL). 11. G. subcarnosa Adamson sp. nov. Prostrata. Folia parva subcarnosa acuta pruinoso-papulosa. Flores in ramis secundis aggregatis. Perigonii segmenta 5, styli 5. Prostrate, very commonly red or purple, forming circular mats or patches 10—25 cm. across. Young stems and leaves pruinose or finely papulose with a few appressed scales. Leaves small, 3—7 mm. long, 1—3 wm. wide, thick, acute or subacute, concave above, narrowed to 104 ~The Journal of South African Botany. the base, usually much crowded. Flowers on short alternate secund lateral branches. Bracts like small leaves: Perianth segments 5, 1-5 mm. long, finely papulose, incurved at the tip. Styles 5: ovary 5-angled, broader than deep. The type is Acocks 14358 (PRE). Open rocky places on the karoo. A distinctive species recognised by the mat habit, small thick folded leaves which are pruinose-papulose, and 5-merous flowers on small lateral branches. ; MIDDELBURG. Carlton Hills Acocks 15934 (PRE). VICTORIA WEST. N. of Victoria West Acocks 14358 (PRE). SUTHERLAND. Sutherland Acocks 16980 (PRE). BRITSTOWN. S. of Britstown L. H. Taylor 907 (BOL): Marloth 9813 (PRE). KIMBERLEY. Roodepan Acocks H867 (PRE); Orange River L. Gibbs (BM). S. RHODESIA. Matopos L. Gibbs 12 (BOL). 12. G. affinis Sond. Fl. Cap. 2. 476. 1862. Gray or reddish, rarely green. Stems woody, prostrate or ascending, rarely over 30 cm. high, with spreading secund branches. Stems pale- coloured, pruinose, the younger with short appressed hairs. Leaves mostly alternate, distant or crowded, obovate or spatulate, 0-5—1-2 em. long, 0-2—0-4 cm. wide, acute, folded upwards and concave above, covered either densely or sparsely with short closely appressed hairs. Flowers on short pinnate secund branches. Bracts like the leaves but smaller, longer than the flowers. Perianth segments 5, 1-5—2 mm. _ long, hairy outside. Anthers yellow. Styles 3, less often 5. The type is H. & Z. 2139 (S). Rather rare on rocky koppies on the karoo. Distinguished by the small acute hairy leaves. Sonder (l.c.) describes the plant as erect but most of the specimens seen are prostrate or spraw- ling. NAMAQUALAND. Bitterfontein Zeyher 724 (BM, BOL, CGE, K, S). WILLISTON. N. of Williston Acocks 14721 (PRE). - COLESBURG. Coleskop Thorne (SAM). MONTAGU. Baden Levyns 8016 (CT). SWELLENDAM. Hassequas Kloof of Ecklon & Zeyher 2139 (S, SAN). RIVERSDALE. Gouritz Riv. Burchell 6498 (K). Without locality:—Drege (Tephras pubescens E.M.) (K). The South African Species of Aizoaceae. 105 13. G. secunda (L.f.) Sond. Fl. Cap. 2. 474. 1862. Aizoon secundum L. f. Supp. Pl. 261. 1781. A. contaminatum E. & Z. Enum. 326. 1837. A. elongatum K. & Z. Le. A. propinquum EK. & Z. l.c. A prostrate many-stemmed plant covered with coarse villous loosely appressed hairs, usually white, gray or yellow. Stems 10—50 cm. long, woody at the base, secund, branched from the base, with in the distal part many short flowering branches 1—5 cm. long standing out at right angles. Leaves alternate, coarsely papulose, those on the main shoots broadly obovate, 0-7—2 em. long, 0:4—1-2 cm. wide, obtuse, narrowed to the base, flat or concave above, often partly or wholly glabrescent in the later stages, deciduous at or about flowering time. Leaves on the lateral branches smaller, more acute, usually folded upwards, and persistent. Flowers on the short lateral branches, solitary or in groups of 2—5 on short branchlets. Bracts leaf-like: bracteoles very short. Perianth segments 5, 2—3 mm. long, coarsely villous, yellow or white inside, the tips incurved. Anthers yellow, less often red. Styles 5 or occasionally 3. Fruit rounded, depressed on top. The type is in herb. T’hunberg, Uppsala. Open places on sand or gravel in the coastal belt and at the margins of the karroo. Wide-spread but certainly extending along roadsides and in disturbed ground. Distinguished by the coarsely villous rather broad leaves and large 5-merous flowers. Very variable in habit, in the number and closeness of the branches, the closeness of the leaves, and in the density of the hairy covering. Extreme forms look distinct but are connected by all stages or intermediate. While most commonly the shoots are separate, sometimes close mats are formed. Forms with very crowded leaves may resemble species of Plinthus. In general plants from the drier habitats or from sea-side stations have more numerous and more persistent hairs than those from moister inland stations. Var. strigulosa Sond. Lec. is merely an especially glabrescent. form. In the more hairy forms the papulose leaf surface is obscured. The various species described by Ecklon & Zeyher l.c. are not more than habitat or seasonal states and cannot be maintained. S.W. AFRICA. Gr. Kalk Thorne (SAM); Buntfeldschuh Dinter 3936 (BOL, PRE). KENHARDT. Nagbult Acocks 12650 (PRE). PRIESKA. Prieska Bryant 115la (K, PRE); Page (BOL). WILLISTON. Rietkokoos Acocks 9618 (PRE). CARNARVON. Carnarvon Henrici 25 (PRE). 106 The Journal of South African Botany. VAN RHYNSDORP. Van Rhynsdorp Thode A1896 (PRE); Knechts- vlakte Schlechter 8312 (BM, GRA, K, PR#). CLANWILLIAM. Matjes Riv. Wagener 146 (NBG). PIKETBERG. Sauer Compton 15094 (NBG). CERES. Laken Vlei Compton 12079 (NBG); Hottentots Kloof Adamson 5016 (PRE). WORCESTER. Brandvlei v. Breda 816 (PRE); Karroo Gd. Cross 48. (BOL, NBG). LAINGSBURG. Whitehill Compton 3604 (BOL, NBG), 15233 (NBG), 17407 (NBG) Pieter Meintjes Rogers (BOL). BEAUFORT WEST: Nieuwveld Mts. Adamson 5212, 5214: rd. to Aberdeen Pole Evans 11 (PRE). ABERDEEN. Aberdeen Rd. Pole Evans 2541 (K). MURRAYSBURG. Murraysburg Tyson 12 (BOL), 429 (K, SAM). GRAAF REINET. Graaf Reinet Adamson 5274; Maguire 704 (NBG). MIDDELBURG. Middelburg Acocks 15944 (PRE); Theron 850 (PRE); Middelburg Rd. Flanagan 1694 (PRE, SAM); Naauwpoort Rogers 12030 (PRE). CRADOCK. Cradock Rogers-Wolfe (BOL); Mortimer Bolus 9218 (BOL, K), 9268 (K). QUEENSTOWN. Klipplaat Riv. Thode A571 (K). KIMBERLEY. Alexandersfontein Brueckner 944 (BOL, K, PRE); Mosterts Hoek Hafstrom & Acocks H1002 (PRE). CAPE. Green Point Adamson 1827 (BOL, PRE); Dawson (BOL); Salter 7678 (CT); Stokoe (SAM); Zeyher (SAM); Mouille Pt. Levyns (CT); Three Anchor Bay Salter 6397 (CT); Sea Point Adamson 3096; Milnerton Adamson 4755; St. James Moss 18039 (BM); Robben Is. Adamson 3932: Rondebosch Adamson 4466. ROBERTSON. Robertson Adamson 4651: Levyns 4321 (CT); Zeyher (SAM); Nuy Adamson 4462; McGregor Esterhuysen 4454 (BOL). MONTAGU. Wagonboomsberg Adamson 5097; Concordia Barnard 759 (SAM). LADISMITH. Dobelaarskloof Levyns 6697 (CT). SWELLENDAM. Storms Vlei Adamson 4931 (PRE); Hcklon & Zeyher 2134 (8S); Levyns 3476 (CT); Swellendam Zeyher (SAM); Hassequas- kloof Ecklon & Zeyher 2134 setae Zeyher (SAM); rd. to Port Beaufort Adamson 4604. ria RIVERSDALE. Riversdale Muir 2737 (PRE); Schlechter 1807’ (BM, K, PRE); Oakdale Levyns 2657, 3162, 4182 (CT); Muir 106 (BM, SAM), 2399 (PRE). MOSSEL BAY. Mossel Bay Rogers 17796 (K), 26953 (K): Gouritz Riv. Burchell 4389 (K); Schlechter 5719 (BOL); Zeyher 1 (SAM). The South African Species of Aizoaceae. 107 GEORGE. Kromme Riv. Bolus 1866 (BOL). OUDTSHOORN. Cango Bolus 11915 (BOL); Oudtshoorn Rogers 4268a (K). UNIONDALE. Avontuur Fowrcade 1661 (GRA, PRE). UITENHAGE. Uitenhage Compton 20304 (NBG); Ecklon & Zeyher (GRA); Prior (K); Schmitz (PRE); Schlechter 2561 (BOL, K); Prior (K); Zeyher (SAM); Swartkops Riv. Ecklon (S); Ecklon & Zeyher 189 (BOL), 2131 (BOL); 2132 (GRA, K, SAM); Sveber 2132 (8); Zeyher 189 (BOL, K), 2633 (K, PRE, 8, SAM), s.n. (SAM): Paadepoort Bolus 1866 (BOL). PORT ELIZABETH. Port Elizabeth Cruden 471, 483, 484 (GRA); Paterson (CGE); Humewood Pillans (SAM); Redhouse Paterson 3232 (CGE); Perseverance Rodin 1283 (BOL, K, PRE). ALEXANDRIA. Nanaga Acocks (PRE); Story 1299 (K, PRE), 1300 (PRE). ALBANY. Alicedale Cruden 141 (GRA); Rogers 12030 (BOL, K); _Bushmans Riv. Maguire 528 (NBG); Grahamstown Dyer 1680 (GRA, K); MacOwan 445 (GRA); Bothas Hill Britten 5299 (GRA); West Hill Adamson 4762; Lombards Post Burchell 3913 (K); Cradock Rd. Adamson 5260; s.l. Bowker (K); Cooper (BM, K). BATHURST. Kowie Britten 1906 (GRA PRE); Tyson (BOL); s.1. Bowker (K); Cooper 7 (BM, K). EAST LONDON. E. London Galpin 5844 (GRA); Nahoon Nanni 129 (PRE). FAURESMITH. Liickoff C. A. Smith 485 (BM), 3516 (BOL); Rose Marie Kies (PRE). NATAL. s.l. Cooper 1281 (BOL, K) (NB. On the ticket of Pillans Humewood is written “Berea, Durban’’). Without locality:—Bowin (BOL); Burtt-Davy 11959 (PRE); Drege 7061a (K), ‘CBS’ (PRE); Ecklon (BOL); Ecklon & Zeyher 2134 (K): Prior (K); Sim 2664 (PRE); Zeyher 2132 (K), 2134 (K), s.m. (8). Also recorded as an alien from Australia (Bot. Gds. Sydney 26301 (K) and from U.S.A. (Curtis 6819, Sheldon 10101 (K)). 14. G. pubescens (EH. & Z.) Druce Rep. Bot. Ex. Cl. Brit. Is. 1916. 624. 1917. Aizoon pubescens EK. & Z. Enum. 326. 1837. G. spathulata Fenzl in Sond. Fl. Cap. 2. 475. 1862. Prostrate, in green or gray-green patches 1 m. or more across. Young parts with closely appressed rather coarse hairs, the older glabrescent. Stems woody at the base, 25 cm.—1-5 m. long, much branched, the branches often ascending. Leaves alternate, papulose, those on the main shoots broadly obovate-spatulate, 0-6—2-2 cm. long, 0:3—1-7 cm. wide, obtuse or subobtuse, narrowed at the base, those on the branches 108 The Journal of South African Botany. smaller, more acute, often concave above. Flowers on alternate secund lateral branches 2—8 cm. long, either distant or crowded, and often themselves branched. Flowers solitary or with a pair of branchlets from the base 1—3 mm. long, each bearing 1—3 flowers. Bracts like leaves but smaller, less narrowed at the base, and usually folded upwards. Perianth segments usually 5, 2—2-5 mm. long, slightly incurved at the tip, with few or many closely appressed hairs outside, pinkish inside. Anthers usually pink. Styles 5, less often 3. The type is Zeyher 2638 (S). Bushy places at low altitudes especially where sheltered. Drought resistant and common on roadsides and in disturbed areas. Has been utilised as a fodder plant in drought, certainly alien in some of its stations. Rather like G. secunda but less strictly prostrate, more branched, much less hairy, and with the hairs closely appressed. Distinguished by the broad papulose glabrescent leaves and the rather irregular arrange- ment of the flowering branches which commonly grow out as long shoots, or are formed at the base of long shoots. The leaves on the main stem fall at or about flowering time. VRYBURG. Taungs Parker (PRE). HAY. Floradale Esterhuysen 2319 (BOL, PRE). KIMBERLEY. Kimberley Hsterhuysen 821 (BOL); Farrer (GRA); Moran 56, 98 (GRA). BARKLY WEST. Daniels Kuil Hsterhuysen 812 (BOL). HERBERT. Valschfontein H'sterhuysen 5150 (BOL). CLANWILLIAM. Herculesfontein Zeyher 3 (GRA). CERES. Basson’s Gat Burchell 1251 (K). RICHMOND. Rhenosterfontein Acocks 15841 (PRE). MURRAYSBURG. Murraysburg Tyson 254 (SAM). CRADOCK. Cradock Brynard 275 (K); Convent 29 (GRA); Cooper 518 (BM, BOL, K). CAPE. Ndabeni Adamson 3159, 3212 (CT); Pinelands Adamson 3312 (CT, PRE); Salter 8621 (BOL); Camps Bay Adamson 3544 (BOL); Levyns. (CT); Salter 8783 (BOL). BREDASDORP. Bontebok Park Compton 22615 (NBG); rd. to Swellendam Hsterhuysen 19556 (BOL). ROBERTSON. Robertson Karroo Compton 5735 (NBG). MONTAGU. Montagu Page 41 (PRE). SWELLENDAM. Dipka Adamson 4606—7. HEIDELBERG. Barrydale Adamson 4616. RIVERSDALE. Riversdale Schlechter 1817 (GRA); Gouritz Riv. Schlechter 5719 (BMK, PRE); coast Muir 4709 (BOL). The South African Species of Aizoaceae. 109 MOSSEL BAY. Herbertsdale Burchell 6419 (K); Mossel Bay Guthrie 3981, 4317 (NBG). HUMANSDORP. Mind Plaas Fourcade 6061 (BOL); Gamtoos Riv. Fourcade 2578 (BOL). UNIONDALE. Uniondale Fourcade 4658 (BOL). UITENHAGE. Uitenhage Hcklon (S); Ecklon & Zeyher 188 (BOL, GRA, K), 2133 (BOL, K); Marais 142 (K, PRE); Pappe (K); Thode 4654 (K); Zeyher (S); Swartkops Riv. Zeyher 188 (SAM), 2635 (SAM), 2638 (S), sn. (SAM); Enon Thode A2663 (PRE); Hankey Fourcade 2578 (K); Coega Rogers 2275 (K). PORT ELIZABETH. Port Elizabeth Paterson (CGE); Redhouse Mogg 4764 (PRE); Swartkops Brock 16 (GRA); Addo rd. Hries, Morlindh & Weimarck 781 (BOL); Addo Reserve Brynard 331 (K), s.n. (PRE). ALBANY. Alicedale Cruden 93 (GRA); Rogers 12123 (BOL, K); Swartrivierspost Burchell 6419 (K); Grahamstown Adamson 5269, 5272, 5275; Bolton (S); Lindstedt 34 (PRE); Rogers 2606 (BOL), 3778 (GRA), 2736 (GRA, K, PRE); Highlands Turberville (PRE): Belmont valley Adamson 5253; Cradock rd. Adamson 5259, 5261—4; West Hill Adam- son 4713; Bushmans Riv. Baur 1098 (K); Assegai Riv. Compton 19062 (NBG); Fish Riv. Adamson 4929; s.1. Cooper 7 (BOL, K). BATHURST. Port Alfred Bolus (BOL); Britten 786 (GRA, PRE); Compton 21084 (NBG); Dyer 634 (PRE); Schonland 180 (GRA); Tyson (Marloth 8583) (PRE). PEDDIE. Keiskama mouth Mills (BOL). KING WILLIAMSTOWN. Debe Nek Rogers 4466 (BM, BOL, CGE, SAM); King Williamstown Tyson 1033 (BOL, SAM), 6091 (SAM). EAST LONDON. Selborne C. A. Smith 3694 (PRE), Galpin 1837; Flagstaff Hill Galpin 5844 (PRE); E. London Worsdell (K). KOMGHA. Kei Bridge Flanagan 119 (BOL, NBG, SAM). STUTTERHEIM. Stutterheim Acocks 9027 (PRE). VICTORIA EAST. Garfield Acocks 11093 (PRE). QUEENSTOWN. Essex Adamson 5271; Bolus (BOL). O.F.S. Kroonstadt Pont 320 (PRE); Bloemhof Burtt-Davy 1537 (K). Without locality: —Burtt-Davy 1495 (K); 7826 (PRE); Drege 57la & b (K), G. spathulata d (8); J. F. Drege 3011 (PRE); Ecklon & Zeyher 2133 (K, PRE); Flanagan (PRE): Fourcade 2782 (BOL); Rogers 17164 (PRE). G. pubescens is a variable species. In addition to variations it can be related to habit, season, etc.. There are several well-marked varieties. KeryY TO THE VARIETIES. 1. Stems and leaves with appressed hairs at least when young. 2. Flowers on secund branches: perianth 5: styles 5. 3. Leaves alternate: perianth hairy at ae .. var. pubescens 3. Leaves opposite: perianth glabrous .. A .. var. Fourcadet 2. Flowers axillary: perianth 4: styles 4 .. a0 -. var. pallens 110 The Journal of South African Botany. 1, Stems and leaves without hairs: stems white. 4. Stems densely papillose: leaves papules: pee 5: styles 5 ¢ .. var. lignosa 4, Stems thickly covered with Wax: ‘leaves waxy- prainose perianth 5: styles 3—4 .. . O° : -. var. cerosa Var. pallens Adamson comb. nov. G. pallens (KE. & Z.) Walp. Sep. 2. 232. 1843. Kolleria pallens EK. & Z. Enum. 328. 1837. K. glauca KK. & Z. 1.c. : G. glauca Walp. l.c. Glabrous except for scattered hairs on the youngest parts. Leaves gray or gray-green, thinner, usually folded upwards, very finely papulose. Flowers axillary, not on lateral branches, usually in groups of 3, the lateral on branchlets 2—5 mm. long. Bracts leaf-like. Perianth segments 4, papulose on the back. Styles 4. The type is Zeyher 2636 (8). Apparently rare in the Uitenhage district. Not collected within recent years. Usually quoted as “G. pallens Fenzl” (e.g. Fl. Cap.) but the combina- tion was first made by Walpers. UITENHAGE. Swartkops Riv. Hcklon & Zeyher 2144, 2145 (GRA, SAM), s.n. (8) Zeyher 1049 (BOL, GRA, K, SAM), 2636 (K, 8, SAM). Without locality:—Drege (Tephras humifusa K.M.a) (BM, CGE, K): Ecklon & Zeyher s.n. (CGE). Var. lignosa Adamson var. nov. A typo differt caulibus lignosis prostratis Junioribus dense papuloso- papillosis, foliis subglabris, floribus axillaribus solitariis 5-meris. Prostrate. Stems woody, the younger densely covered with white papulose papillae. Leaves glabrous except for a few scattered appressed hairs on the youngest, rather coarsely papulose. Flowers on the main stem and branches, not crowded. Perianth segments 5. Styles 5. The type is Mogg 8109 (PRE). VRYBURG. Armoeds Vlakte Mogg 3960, (PRE), 8109 (CGE, PRE). BARKLY WEST. Hartsrivier Marloth 976 (PRE). KIMBERLEY. Kenilworth Ferrer (K). Without locality:—Burtt-Davy 13108 (PRE). Var. cerosa Adamson var. nov. A typo differt caulibus lignosis ascendentibus albis densissime cerosis, foliis glabris junioribus ceroso-pruinosis veteribus ee floribus SouIE tariis, perigonii segmentis 5, stylis 3 raro. 4. Decumbent with ascending branches, glabrous but wax-covered. Stems white, woody, with a thick rough coating of wax. Leaves alternate, The South African Species of Aizoaceae. 111 pale-coloured, oblanceolate, 0-5—1-3 cm. long, 0-2—0-3 cm. wide, flat or often folded. Young leaves covered with wax, the older papulose. Flowers on lateral secund branches, solitary or each with a single short branchlet. Bracts leaf-like, longer than the flowers. Bracteoles as long as the flowers. Perianth segments 5, 1-5—2 mm. long, thin, papulose on the back. Anthers mostly pink. Styles 3, rarely 4 or 5. Ovary flattened. The type is Hsterhuysen 811 (BOL). Local in the Bechuanaland region. HAY. Postmasburg Hsterhuysen 811 (BOL); Matsap Acocks 2233 (BOL, K, PRE). Var. Fourcade Adamson var. nov. A typo differt caulibus diffusis, foliis oppositis subglabris, inflorescen- tia laxe secunda ramis distantibus. Suberect or diffuse but often with semiprostrate branches. Glabrous except for scattered hairs on the youngest parts. Leaves opposite, dark green, papulose, oblong-spatulate, 1—1-7 cm. long, 0-3—0-5 cm. wide, usually folded upwards. Inflorescence spreading, secund, 5—15 cm. long, with alternate divaricate branches 0-5—1-5 cm. long and about 1 em. apart. Flowers in lax groups of 3—5. Bracts leaf-like. Perianth segments 5, glabrous. Styles 5. The type is Fourcade 4660a (BOL). UNIONDALE. Uniondale Fouwrcade 4660a, 6460, (BOL). 15. G. prostrata Schellenb. Bot. Jahrb. 48. 506. 1913. A slender prostrate many-stemmed gray-green plant in circular patches 50—80 cm. across. Stems with opposite branches, the younger covered with appressed hairs, the older glabrescent. Leaves opposite, rhombic-spatulate, 0-6—1 em. long, 0-3—0-5 cm. wide, narrowed to a petiole at the base, acute or obtuse-mucronate at the tip, thinly covered with appressed hairs. Flowers erowded on secund lateral branches. Perianth segments 4—5, 2—3 mm. long, pinkish, thinly hairy outside, with a cylindrical acuminate projection behind the incurved tip. Anthers usually pink. Styles 4, less often 3 or 5. The type is Bolus 6815 (B), lectotype (BOL). Readily distinguished by the opposite rhombic leaves and the perianth with acuminate tip. Originally described as near to G. Ecklonis (glauca) but except for a similarity in the, perianth there does not seem any close affinity. BARKLY WEST. Barkly West Bolus 6815 (BOL). pone KIMBERLEY. Kimberley Acocks 1630 (PRE); Flanagan 1480 (BOL, SAM); Wilman (PRE). 112 The Journal of South African Botany. PRIESKA. Prieska Bryant 1151b (K). HANOVER. Hanover Sim (Galpin 5978, 5979) (PRE). MURRAYSBURG. Murraysburg Adamson 5219; Koudeberg Adamson 5210. MIDDELBURG. Gordonville Acocks 16559 (PRE). QUEENSTOWN. Queenstown Bolus (BOL); Lesseyton Mt. Galpin 1941 (PRE). PHILIPPOLIS. Philippolis C. A. Smith 4592 (PRE). FAURESMITH. Veld Res. Henrici 2890 (PRE); C. A. Smith 4607 (PRE); Verdoorn 979 (PRE); Fauresmith Brueckner 962 (PRE): C. A. Smith 4599 (PRE). BETHULIE. Orange Riv. Flanagan 1480 (PRE). BLOEMFONTEIN. Bloemfontein Potts 2830, 3363 (PRE). 16. G. acutifolia Adamson sp. nov. Suffruticosa, sparse pilosa vel glabrescens. Folia opposita, acuta obovata petiolata saepissime rubro-marginata. Flores axillares solitarii. Perigonii segmenta 5 in mucrone angusto producta. Styli 5. Stems many, decumbent or ascending, from a thick woody rootstock. Young parts thinly appressed hairy, the older glabrescent. Leaves gray-green, opposite, acute, obovate or rhombic-ovate, commonly red at the edge, 8—12 mm. long, 4—5 mm. wide, narrowed at the base to a petiole 2—3 mm. long. Flowers sessile, axillary, on branches 1—2 cm. long, often rather crowded. Bracts like the leaves but smaller and less petiolate: bracteoles slightly longer than the flowers, obtuse or emargi- nate. Perianth segments 5, 2—2-5 mm. long, pale cream-coloured or pinkish, slightly hairy outside, produced behind the incurved tip into a slender point about 1 mm. long. Anthers yellow. Styles 5, rather short. The type is Esterhuysen 813 (BOL). Allied to G. prostrata but more woody, less prostrate, less hairy, with narrower more acute leaves, and more crowded flowers. BARKLY WEST. Newlands Esterhuysen 813 (BOL); Caerwinning Esterhuysen 814 (BOL). 17. G. papulosa (E. & Z.) Sond. Fl. Cap. 2. 475. 1862. Aizoon papulosum E. & Z. Enum. 326. 1837. — A green or yellow-green glistening almost herbaceous plant, in bunches 10—40 cm. high, often with spreading or decumbent branches. Stems and leaves covered with large obtuse transparent papillae but without hairs or scales. Leaves succulent, alternate, obtuse, the lower broadly oblong or rotund, petiolate, 1—3 cm. long, 0-8—1-7 cm. wide, the upper smaller subsessile, flat or often folded and concave above. Flowers in lax, The South African Species of Arzoaceae. 113 more or less secund cymes, with bracts like the leaves but much smaller. Bracteoles thick, about as long as the flowers. Perianth segments 5, 1-5—2 mm. long, papulose on the outside, incurved at the tip. Anthers yellow, rarely red. Styles 3, less commonly 5. The type is H. & Z. 2137 (B), lectotype (SAM). River beds and sheltered places in the karoo, especially on shales. Readily recognised by the coarsely papulose succulent leaves without hairs or scales. The lower broader leaves fall off about flowering time. Allied to G. sarcophylla but less woody and without hairs. S:W: AFRICA. Swakopmund Pearson 531 (K, SAM); Windhoek Dinter 348 (BM, K, SAM); Rogers 29520 (PRE); Hereroland Balck 15 (K); Aus Dinter 6239 (BM). NAMAQUALAND. Steinkopf Schlechter 34 (GRA), 38 (GRA); Zywarts Pillans (SAM). CALVINIA. Calvinia Macpherson 3689/14 (BCL): de Bosch Adamson 4555. GORDONIA. Keimoes Barnard (SAM). PRIESKA. Prieska Lloyd (PRE). VAN RYHNSDORP. Klaver Stokoe (SAM). BRITSTOWN. Britstown Page (BOL). VICTORIA WEST. Hutchinson Acocks 9716 (PRE); Victoria West Thode A1897 (PRE); Pampoenpoort Henrici 4768 (PRE). FRASERBURG. Fraserburg. Hafstrom & Acocks 423 (PRE); rd. to Sutherland Hafstrom & Acocks 721 (PRE). SUTHERLAND. Tanqua Riv. Compton 7536 (NBG): Roggeveld Worsdell (K). BEAUFORT WEST. Krom Riv. Goatcher (BOL). ABERDEEN. Groot Riv. Hutchinson 3136 (K). MURRAYSBURG. Murraysburg Bolus 28 (BM, BOL, SAM); Tyson 28 (K). PRINCE ALBERT. Fraserburg Rd. Thorne (SAM); Prince Albert Bolus 11515 (BOL, K, PRE); Krige (BOL). LAINGSBURG. Koup Compton 10403 (NBG); Esterhuysen 4509 (BOL); Laingsburg Adamson 4693; Whitehill Compton 9296, 11269 (NBG). Matjesfontein Fries, Norlindh & Weimarck 736 (BOL). PIKETBERG. Piketberg Bolus (BOL). LADISMITH. Groot Riv. Adamson 4623—46. HEIDELBERG. Lemoenshoek Adamson 4590; Warmwaterberg Adamson 4622, 4628, 5129. - MOSSEL BAY. Gouritz Riv. Ecklon & Zeyher 2137 (SAM). GEORGE. Camfer Fourcade 5677 (BOL). 114 The Journal of South African Botany. UNIONDALE. Rd. to Willowmore Fourcade 4654 (BOL, K); Pole Evans 20 (PRE). JANSENVILLE. Klipplaat Thode A.572 (PRE). FORT BEAUFORT. Klipbank Rogers 17118 (PRE). Without locality:—Burtt-Davy 12648 (PRE): Drege (Tephras papulo- sa) (CGE): Rogers 200 (BOL). Var. microphylla Adamson var. nov. A typo differt caulibus fruticosis, foliis anguste oblongis, stylis saepis- sime 9. Stems more woody, thicker, and more laxly branched. Leaves usually crowded, narrow-oblong, about three-times as long as wide, usually slightly widened to the tip, concave above, 0-6—0-8 em. long, 0-2—0-4 em. wide. Styles most often 5. The type is Adamson 5249A (BOL). S.W. AFRICA. Rehoboth v. Rhyn (PRE); Usakos Marloth 1297 (BOL, PRE); Bullpoort Liebenberg 5105 (K, PRE). CALVINIA. De Bosch Adamson 4554, 4556. PRIESKA. Prieska Bryant 231 (K), J231 (PRE), s.n. (PRE). BEAUFORT WEST. Rd. to Aberdeen Adamson 5249A. NB.—A specimen from Richmond, van Wyks Opslag, Gill 54 (PRE) probably belongs to G. papulosa. It differs in having a reddish woody stem and gray leaves, very slightly succulent and finely papulose. 18. G. sarcophylla Fenzl in Sond. Fl. Cap. 2. 475. 1862. A low-growing more or less sprawling plant in clumps up to 30 cm. high and 50 em. across, the whole glistening, grayish-green or the older parts red or yellow. Stems many, decumbent with ascending branches, the oldest slightly woody, the younger coarsely papulose and covered with loosely appressed hair-like scales. Leaves succulent, coarsely papulose and with scales at least when young, alternate or the lowest opposite, obovate-spatulate, obtuse, the lower 0-6—1-6 em. long, 0-3— 0-8 cm. wide, narrowed to a petiole at the base, the upper leaves smaller, sessile. Flowers on close or distant secund lateral branches, usually numerous. Bracts thick, obtuse, longer than the flowers; bracteoles about about as thick as long, as long as the flowers. Perianth segments 5, about 2 mm. long, not succulent, covered with scales on the back. Anthers yellow, less often pink. Styles 3, occasionally 5, rarely 2 by reduction. The type is Zeyher 719 (B), lectotype (S). Sand or gravel especially in river beds in the karroo. Readily distinguished by the papulose succulent leaves with few or many hair-like scales on the younger parts, and the flowers crowded on lateral branches. The South African Species of Aizoaceae. 115 S.W. AFRICA. Swakopmund Moss 18014 (BM); Pearson 529 (SAM); Rogers (BOL), Wiss 965 (PRE); Luderitz Kinges 2458 (PRE); Ababies Pearson 9126 (BOL, K)); Noachabib Pearson 8673 (BOL); Namalong Marloth 1532 (PRE); Schinz 916 (K); Gt. Karasberg Pearson 8574 (BOL, K), 8673 (BOL, K); Kl. Karasberg Dinter 4853 (BOL, K, PRE): Bakhuis Pearson 5448 (K). NAMAQUALAND. Steinkopf Schlechter 38 (BOL, K, PRE), s.n. (BM); Nieuwfontein Pearson 3356 (BOL, K, SAM); Orange River Pearson 6027 (SAM); Concordia Schlechter 11327 (BOL, PRE); Okiep Diimmer (K); Springbok Lewis (SAM); Zeyher 719 (BM, BOL, CGH, GRA, K, 8, SAM); Sabatsfontein Lewis (SAM); Hendeklip Bay Pillans (BOL); Bitterfontein Adamson 4991; Zeyher (SAM). VAN RHYNSDORP. Nieuwerust Adamson 4895; Scharkraal Pearson 3074 (BOL, K); Klaver Lavis (BOL); Lewis (SAM); Widouw Riv. Stokoe (SAM). CLANWILLIAM. Pakhuis Adamson 4514; Bidouw Johnson 547 (NBG); Matjes Riv. Wagener 146 (NBG); Buchufontein Lewis (SAM). GORDONIA. Upington Webb 5 (K). KENHARDT. Kakamas Acocks (PRE); Lewis (SAM); Aughrabies Falls Wasserfall 1037 (PRE). CALVINIA. Loeriesfontein Adamson 2047; Granaatboskalk Comins 674 (PRE); Calvinia Schmidt 397 (PRE); Ekerdam L. H. Taylor 2734 (NBG); Botterkloof Adamson 5077; de Bosch Adamson 5084—5; Com- mando Doorns Adamson 5083. PRIESKA. Prieska Bryant (BOL); Page (BOL). WILLISTON. Rietkolkoos Acocks 9616 (PRE); Dwaar Riv. Burchell 1457 (K). FRASERBURG. Fraserburg Hafstrom & Acocks 422, 2296 (PRE). SUTHERLAND. Both Middelpost & Sutherland Hutchinson 717 (BOL, K, PRE). BEAUFORT WEST. Krom Riv. Goatcher (BOL); Klipbrak Rogers 17118 (PRE); Sunnyside Hsterhuysen 2741 (BOL, NBG, PRE); rd. to Aberdeen Adamson 5249. RICHMOND. Gt. Tafelberg Acocks 8734 (PRE). MURRAYSBURG. Murraysburg Tyson 431 (BM, BOL, K, SAM). MIDDELBURG. Middelburg Theron 221 (PRE). CRADOCK. Cradock Brynard 274 (K, PRE). GRAAF REINET. Graaf Reinet Bolus 2314 (BOL). ABERDEEN. Groot Riv. Hutchinson 3136a (K). LAINGSBURG. Whitehill Compton 11196 (NBG); Matjesfontein Foley 1657 (K). MONTAGU. Koo Adamson 5103. 116 The Journal of South African Botany. UNIONDALE. Rd. to Willowmore Fourcade 4658 (BOL, K); Pole Evans 2286 (PRE). KNYSNA. Vlugt Bolus 2314a (BOL). FAURESMITH. Liickhoff Pole Evans 1881—-2 (PRE); C. A. Smith 483, 485 (PRE); Veld Res. Henrici 4707 (PRE); Groenvlei Verdoorn 1034 (Pre) Van Wyksbrak OC. A. Smith (PRE); Pampoenpoort Henrici 4761 (PRE). KOFFIEFONTEIN. Poortjesdam Henrici 4503 (PRE). BLOEMFONTEIN. Bloemfontein Henrici 2736 (PRE). Without locality:—Drege 7060 (K); Rogers 205 (BOL). 19. G. portulacacea Fenzl in Sond. Fl. Cap. 2. 475. 1862. A gray-green diffuse or sprawling plant with the shoots ascending. Stems herbaceous or woody at the base only, 15—30 em. long, the younger covered with loosely appressed white hair-like scales. Leaves opposite except in the flowering region, oblong or oblong-spatulate, 1—3 em. long, 0:30-6 cm. wide, much narrowed to the base, subacute at the tip, flat, not succulent, papulose, covered with loose hair-like scales when young. Flowers in lax terminal cymes, in groups of 5—7, the lateral on branchlets 0-5—1-5 em. long. Bracts like the leaves but much narrower and usually folded upwards. Perianth segments 5, about 3 mm. long, erect, the tips incurved, green outside. Anthers yellow. Styles 3 or 5. The type is Drege (Dreikoppen) (B). On sand at or near the margins of the karoo. In appearance somewhat like G. sarcophylla but not succulent and with larger leaves. In collections it has been much confused with G. pube- scens but is distinguished by the opposite narrower leaves with loose hair-like scales and the lax inflorescence. Sonder l.c. uses three different spellings for the specific epithet, the others being “‘portulacea”’ and “‘portu- lacoides’’. KIMBERLEY. Alexandersfontein Power (PRE). CALVINIA. Nieuwoudtville Lavis (BOL). NAMAQUALAND. Hondeklip Bay Pillans (BOL). CLANWILLIAM. Graafwater Adamson 4402. CERES. Karroo Poort Adamson 5087: Compton 3502 (BOL). WORCESTER. Brandvlei v. Breda 316 (PRE). ROBERTSON. Robertson Karroo Adamson 5154; Levyns 4321 (CT). SWELLENDAM. Bonnievale Compton 3422 (BOL). MONTAGU. Kogmans Kloof Adamson 4645. RIVERSDALE. N. side Garcias Pass Adamson 4596: Riversdale Story 2878 (PRE); Muiskraal Adamson 5149. MOSSEL BAY. Gouritz Riv. Schlechter 5719 (PRE). GRAAF REINET. Graaf Reinet Bolus 116 (BOL, GRA). The South African Species of Arzoaceae. 117 20. G. squamulosum (E. & Z.) Fenzl in Sond. Fl. Cap. 2. 477. 1862. Aizoon squamulosum E. & Z. 326. 1837. A white or less commonly yellow decumbent or prostrate plant covered all over with flat spreading scales. Stems 20—50 ecm. long, slightly woody at the base, with erect or ascending branches. Leaves opposite or alternate in the flowering region, somewhat succulent, coarsely papu- lose, oblong or obovate, 1—2 cm. long, 0-3—0-7 cm. wide, obtuse, narrowed to the base, usually rolled upwards. Flowers sessile, solitary. Bracts leaf-like. Perianth segments 5, 2 mm. long, densely covered with scales outside, yellow inside. Anthers yellow. Styles 2. The type is Ecklon & Zeyher 2136 (S). In damp brackish hollows and stream beds in the western and north- western karoo. Distinguished by the dense covering of scales. Like G. sarcophylla but larger, prostrate, much more scale-covered, and with 2 carpels. S.W. AFRICA. Swakop Riv. Galpin 7467 (PRE); Galpin & Pearson 7466 (K, SAM); Swakopmund Pearson 5317 (K, SAM): Stomprietfontein Loeb X1 (PRE); Bucholtzkram Pearson 3385 (BOL, K); Windhoek Rogers 29768 (GRA, K); v. Niekerk 330 (BOL); Aus Dinter 6113 (BOL, K). NAMAQUALAND. Witbank Pillans 5555 (BOL); Orange Riv. Ecklon & Zeyher 8136 (S, SAM); Eenkokerboom Schlechter 11065 (BM, BOL, GRA, K, PRE); Garies Adamson 5015. GORDONIA. Upington Broom (GRA); Webb 5, 11 (PRE). VICTORIA WEST. Pampoenpoort Acocks 14678 (PRE). 21. G. fruticosa (L.f.) Sond. Fl. Cap. 2. 477. 1862. Aizoon fruticosum L.f. Supp. Pl. 261. 1781. A whitish diffuse woody shrublet with rigid usually secund branches up to 50 cm. long. Stems yellowish, thinly covered with appressed hairs, the internodes 1—4 em. long. Leaves opposite, gray or almost white, densely covered with silky appressed hairs, oblong or obovate, usually folded upwards, 0-5—1-5 em. long, 0:3—0-5 cm. wide, acute or subacute with a hooked point. Inflorescence rigid, forking at the base, above with divaricate alternate usually secund branches. Flowers on short branch- lets usually separated. Bracts leaf-like, longer than or as long as the flowers. Perianth segments 4, about 1 mm. long, erect, hairy outside either all over or at the tip only. Anthers red or yellow. Styles 2. The type is in herb. Thunb. Uppsala. Open Karoo on shales. Distinguished by the rigid woody habit, leaves covered with silky hairs and the small flowers. Variable in size and the extent and branching of the inflorescence which is either secund or almost erect. After flowering 118 The Journal of South African Botany. the inflorescence branches harden and may become spinous. The bracts vary from leaf-like to very small and not longer than the flowers. Forms with red and with yellow flowers are found side by side without any other difference. The leaves are finely papulose but this is hidden by the hairs. S.W. AFRICA. KI. Karasberg Dinter 4951 (BOL, K, PRE); Galpin 14200 (BOL, PRE). NAMAQUALAND. Port Nolloth Galpin & Pearson 1671 (BOL); Okiep Scully (SAM); Hondeklip Bay Adamson 5007, 50341 Pillans (BOL); Bitterfontein Hsterhuysen 1296 (PRE); Godman 48 (BM); Zeyher 723 (K, S). VAN RHYNSDORP. Nieuwerust Adamson 4978, 4987, 4990; Van Rhynsdorp Adamson 4537, 4540: Maguire 143 (NBG); Klaver Adamson 4532, 4534, 4968, 4970, 4521 Maguire 138 (NBG); Nardouw Stokoe (SAM); Knechtsvlakte Comptow 15095, 20870 (NBG); Ebenezer Drege (8). CALVINIA. Calvinia Henrici 3395—6 (PRE); Commando Doorns Adamson 5086 Bushmanland Pearson 6221 (K). CLANWILLIAM. Clanwilliam Levyns 1261 (CT); Olifants Riv. Zeyher (SAM). PIKETBERG. Eendekuil Howes 187 (K, PRE); Het Kruis Stephens & Glover 8726 (BM, BOL, K). CERES. Beukiesfontein Adamson 5086. WORCESTER. Hex Riv. Inebenberg 4198 (PRE); Marloth 3751 (BOL). LAINGSBURG. Koup Zsterhuysen 4508 (BOL); Whitehill Compten 2932 (BOL, K), 3540 (BM, BOL, K). ABERDEEN. Aberdeen Maguire 754 (NBG); van Wyksbos Theron 906 (K, PRE). BEAUFORT WEST. Beaufort West Acocks 14303 (PRE); Pole Evans 31 (PRE). MONTAGU. Kiesies Doorns Michell 262 (PRE). ROBERTSON. Robertson Karroo Adamson 4650, 4652—4, 4661, 4662, 4664. Without locality: —Drege (Tephras pubescens) (K): Masson (BM). Var. prostrata Adamson var. nov. A typo differt caulibus lignosis prostratis, foliis papulosis paucipilosis, floribus numerosis in ramis secundis divaricatis congestis. Woody, quite prostrate, the branches secund. Leaves papulose, rather thinly hairy, not over 1-2 em. long. Inflorescence with many secund alternate branches and numerous branchlets 0-3—0-7 em. long. Flowers crowded. Bracts longer than the flowers, partly reflexed. Perianth seg- ments 1-5 mm. long. Anthers orange-yellow. The South African Species of Aizoaceae. 119 The type is Salter 7012 (K). Distinctive in habit, papulose leaves, and very crowded flowers. CAPE. Vissershoek Salter 7012 (BOL, K). 22. G. erystallina (EH. & Z.) Fenzl in Sond. Fl. Cap..2..477. 1862. Aizoon crystalinum HK. & Z. Enum. 327. 1837. G. namaquensis Schellenb. & Schlitr. Bot. Jahrb. 48. 500. 1913. G. leucoclada Schellenb. & Schltr. l.c. (p.pte.). A gray diffuse shrublet with spreading secund branches, 20—40 cm. long. Stems woody at the base but not rigid. Young stems and leaves more or less densely covered with coarse hairs or hair-like scales, spreading or loosely appressed. ‘Leaves opposite or alternate in the flowering region, coarsely papulose, oblanceolate, oblong or spatulate, most often folded upwards, 0-5—1-7 em. long, 0-3—0-6 em. wide, usually obtuse. Inflores- cence secund, most often lax, forking, with many alternate lateral branches bearing the flowers in groups. Bracts leaf-like, much longer than the flowers: bracteoles about as long as the flowers... Perianth segments 4 or 5, about 2 mm. long, papulose and often ‘hairy outside, incurved at the tip, yellow or pink inside. Anthers orange-yellow or pink. Styles 2. The type is H. & Z. 2138 (W), lectotype (SAM). Rocky or stony soils in the western and southern parts of the Karoo. Like G. fruticosa buat less rigid, with prominently papulose leaves and a laxer more secund inflorescence with leaf-like bracts. Very variable in habit, in the amount of hair covering, in the shape of the leaves, and in flower colour. While extreme forms look very distinct and some have been described as species, they ail appear to be connected by intermediates. The type specimen is a semi-herbaceous form with rather thick semi- succulent leaves covered with narrow scales. This form appears to be rare and confined to Namaqualand. More woody forms are much more common: those with distinctly hairy stems and leaves have been described as G. namaquensis Schellenb. & Schitr. (type Schlechter 11290). Glabrescent woody forms have been attributed to G. leuwcoclada Schellenb. & Schltr. The type of this species, Schlechter 11279, belongs to G. pruinosa, but some of the quoted specimens, e.g. Bolus 9532, undoubtedly belong here. There seems no constant difference between forms with 4-merous and 5-merous flowers, indeed both have been found on the same plant. Red or yellow coloration of the perianth and. anthers also seems to have no diagnostic value. NAMAQUALAND. Port Nolloth Bolus 9532 (BOL), s.n. (BOL) Galpin & Pearson 7487 (K); Pillans 5198 (BOL); Okiep Bolus 9532 (K): L. Bolus (BOL); Scully (SAM); Messklip Schlechter 11290 (BM, GRA, K, PRE); Arakup Schlechter 11240 (BM, BOL, K, PRE); Kastel Poort 120 The Journal of South African Botany. Bolus 9532 (BOL, K); Copperberg Pillans 5158 (BOL); Lekkersing L. E. Taylor 1104 (BOL): Witbank Pillans 5555 (BOL); Wallekraal Adamson 5043, 5045, 5047, 5019: Pillans (BOL); Garies Adamson 5013: Kamiesberg Esterhuysen 1358 (BOL, PRE): s.1. Scully 258 (BM). CALVINIA. Loeriesfontein Pearson 4855 (BOL, K, PRE). VAN RHYNSDORP. Widouw Riv. Lewis (BOL): Magwre 135 (BOL): Stokoe (SAM) Klaver Lamb 3642 (SAM); Stokoe (SAM): Van Rhynsdorp Adamson 4544; Lewis (SAM); Thode A1896 (PRE); Knechts- vlakte Schlechter 8313 (BM, BOL, K). CLANWILLIAM. Zout Riv. Schlechter 8103 (BOL), 8107 (BM); Valleyfontein Ecklon & Zeyher 2138 (SAM). WORCESTER. Worcester v. Breda 112 (PRE). PIKETBERG. Piketberg Adamson 4478. HOPEFIELD. Hopefield Letty 130 (PRE). MONTAGU. Rd. to Barrydale Adamson 5109. UITENHAGE. Uitenhage Prior (K). Without locality:—Drege (G. salsoloides Fenzl) (S); Verreaux (S). Var. maritima Adamson var. nov. A typo differt caulibus prostratis ramulis erectis subherbaceis, foliis longioribus albis oblanceolatis, floribus in ramulis congestis. A white prostrate plant. Main stems woody, up to 60 cm. long, branched with numerous erect or ascending almost herbaceous branchlets. Leaves opposite, oblanceolate, 0-5—2 cm. long, flat or folded upwards, gradually narrowed to the base, covered with coarse appressed hairs. Flowers rather crowded on the lateral branches. Bracts leaf-like, longer than the flowers. Perianth segments 5, 1-5 mm. long, erect, hairy on the back. Styles 2, erect, as long as the perianth. The type is Wolley- Dod 3127 (BOL). Damp hollows in silt near the sea. CLANWILLIAM. Lamberts Bay Adamson 4485; Bidouwberg Lewis (BOL). CAPE. Paarden Is. Adamson 1943, 3117 (CT), 3400 (SAM); Wolley- Dod 3127 (BM, BOL, K); Levyns 5156 (CT). Without locality:—Drege (G. sericea a) (S); Verreaux (S). 23. G. fallax Pax in Engler Pfizw. Afrik. 3.1. 47. fig. 94 J-L. 1915. A gray-white erect rounded shrublet. Stems stout, branched from the base, 15—30 cm. high, pinkish, the younger papulose, the older pruinose. Leaves opposite, succulent, narrow oblong, rolled upwards, 1—2-5 cm. long, more or less hooked at the tip, papulose or slightly scaly when young, pruinose when old. Flowers in a crowded terminal The South African Species of Aizoaceae. 121 panicle. Bracts thick, slightly longer than the flowers: bracteoles as thick as long. Perianth segments 4, about 2 mm. long, the outer thick, cucullate, papulose on the back with hairs at the tip, the inner thinner, with hair-scales at the tip. Styles 2, erect. Sandy or rocky river beds in 8.W. Africa. Perianth segments very slightly joined at the base. The type of this species has not been seen. It is identified from the somewhat inade- quate description and excellent figures. S.W. AFRICA. Walfisch Bay MacOwan s.n. (BOL); Swakop Riv. Galpin & Pearson 7466 (SAM); Luderitz Hobart-Hampden (BM); Kinges 2054 (PRE); Moss 23 (BM), 17883 (BM); s.l. Marloth 4616 (PRE); Rogers s.n. (BOL). NAMAQUALAND. Port Nolloth Galpin & Pearson 7489 (SAM). 24. G. pruinosa Sond. Fl. Cap. 2. 477. 1862. G. leucoclada Schellenb. & Schltr. Bot. Jahrb. 48. 500. 1913 p.pte. A gray or white diffuse or suberect shrublet up to 50 cm. high. Stems white, papulose or with appressed scales when young, glabrescent when older, much branched, the branches opposite, ascending or divaricate. Leaves opposite, wedge-shaped or obovate-spatulate, obtuse, coarsely papulose, 0-5—1-5 em. long, 0-3—0-6 cm. wide, the youngest with appressed hair-like scales, the older glabrescent. Inflorescence forking, the branches with internodes 4—8 em. long in the lower part, in the distal part with alternate lateral branchlets, 2—5 em. long. Flowers lax or in groups, the groups often dense but separated 0:-5—1 cm. Bracts ovate, 2—4 mm. long; bracteoles very small. Perianth segments usually 5, about 2 mm. long, papulose and with appressed scales outside, the tips incurved. Anthers yellow. Styles 2 or occasionally 3, thick, spreading. The type is Drege (Gariep ) (W), lectotype (S). Rocky places in the N.W. Cape. A distinctive species readily known by the white woody stems and forking pedunculate inflorescences. The plants distributed by Dinter under the name “‘G. leucodermis” only differ in having more succulent leaves and shorter more compact inflorescences. They cannot be con- sidered as distinct. The type specimen of G. lewcoclada (Schlechter 11279) is this species. S.W. AFRICA. Angra Paquena Galpin & Pearson 7512 (K); Aus Dinter 6164 (BM, BOL, K); Achterfontein Pearson 9228 (BOL). NAMAQUALAND. Orange Riv. Drege (S): Port Nolloth Pearson 514 (K); Groot Derm Pillans 5332 (BOL. K), 5595 (BOL); Messklip Schlechter 11279 (BM, K, PRE). CLANWILLIAM. Wolverton Schlechter 11441 (BM, GRA). 122 The Journal of South African Botany. PIKETBERG. Eendekuil Levyns 1683 (CT). Without locality: —Drege (Tephras papulosa EM c) (CGE). 25. G. hemisphaerica Adamson sp. nov. Fruticulus ramosissimus hemisphaericus ramis subdivaricatis dichot- omis. Folia opposita vel superiores alterna, subcarnosa obtusa papulosa, vetera decidua. Flores laxe dichotomo-cymosa. Perigonii segmenta 5. Styli 2 raro 3. : A hemispherical shrublet, 20—25 cm. high, 30—50 em. diam. Stems woody, pale brown, much branched, the branches almost divaricate and forking. Young stems papulose, the older smooth, glabrous. Leaves opposite, soon deciduous, slightly succulent, flat, obtuse, wedge-shaped, papulose, 0-8—1-5 cm. long, 0-2—0-4 em. wide. Inflorescence terminal, dienasial, lax, elongated. Bracts like the leaves but much shorter; bracte- oles not longer than the flowers. Perianth segments 5, 1-5—2 mm. long, almost flat, papulose outside. Styles 2, rather thick, rarely 3. The type is Dinter 6440 (BOL). Dry open places on sand. The whole plant rolls in the wind when dry. Allied to G. pruinosa but distinct in habit and the inflorescence. S.W. AFRICA. Buchuberg Dinter 6440 (BOL); Aus Dinter 3681 (BM, BOL, K). NAMAQUALAND. Steinkopf #. P. Phillips 1637 (SAM). 26. G. cymosa Adamson sp. nov. Suffrutex diffusa subsecunda ramis oppositis subdivaricatis. Folia subglabra spatulata vel obovata petiolata papulosa. Panicula secunda laxe cymosa ramis subdivaricatis internodiis longis. Perigonii segmenta 4 vel 5. Styli 4 vel 5, raro 3, longi tenues. A diffuse branching shrublet up to 50 cm. high, the branches secund, opposite, often divaricate. Young parts with a few appressed hair-like scales, the older glabrescent. Leaves opposite or the upper alternate darkish green, broadly spatulate or obovate, obtuse, petiolate, 0-5—2 cm. long, 0:-4—0-6 cm. wide, finely papulose. Inflorescence a secund panicle, the lower internodes 1—2-5 cm. long, the branches widely forking, the ultimate alternate. Bracts leaf-like but sessile, not longer than the flowers, often falcate-recurved. Perianth segments 5, less often 4, 2—3 mm. long, green and papulose on the back. Anthers yellow, Styles 5 or 4, rarely 3, very slender as long as the perianth. The type is Adamson 5135 (BOL). Rather sheltered places among rocks in the southern parts of the karoo, especially near the margins. The South African Species of Aizoaceae. 123 Allied to G. pruinosa but less woody, more slender, the stems not white, the leaves thinner and petiolate, and the inflorescence secund and laxer, also with 4 or 5 styles. LAINGSBURG. Whitehill Compton 2923 (BOL, K), 3176 (BOL), 3596 (BOL, NBG, K), 3605 (BOL, NBG): Matjesfontein Foley 159 (PRE). PRINCE ALBERT. Prince Albert Bolus 11514 (BOL, K). LADISMITH. Warm Baths Adamson 4618: Warmwaterberg Adamson o25) D127, Dl3d3d. ollao: HEIDELBERG. Lemoenshoek Adamson 5142—3. MONTAGU. Rd. to Barrydale Levyns 424 (CT). RIVERSDALE. Waterval Muir 3753 (PRE). GEORGE. Olifants Riv. Bolus 2316 (BOL, K). ALBANY. Cradock Rd. Dyer 2124 (GRA, K). FT. BEAUFORT. Ft. Beaufort Ecklon & Zeyher (GRA). 27. G. Dregeana Fenzl in Sond. Fl. Cap. 2. 477. 1862. A stout rigid whitish woody shrublet up to 45 cm. high. Stems white, covered all over with large projecting papillae, on the young parts some- times flattened as scales. Branches opposite, ascending or divaricate. Leaves opposite, whitish, almost orbicular, thick petiolate, coarsely papulose, obtuse, the blades 0-7—1-5 em. long, 0:7—1-3 cm. wide, the petioles about 2 mm. long. Flowers in a terminal crowded flat-topped panicle. Bracts like the leaves but much smaller, the uppermost about as long as the flowers. Perianth segments 5, 3—4 mm. long, densely papulose outside, the three outer segments cucullate, the two inner flat, obtuse or emarginate, all membranous edged. Anther deep orange or pink. Styles 2, rather long. The type is Drege (Gariep) (B), lectotype (S). Dry river banks in Namaqualand. A very distinctive species easily recognised by the orbicular leaves, densely papillate stem, and large flowers. S.W. AFRICA. Kahamstal Dinter 8173 (K). NAMAQUALAND. Verleptpram Drege (S); Bushmans Riv. Pillans 5063 (BOL), 5387 (BOL); Koodoes Riv. Pillans 5529 (BOL); de Koei Leighton 2411 (BOL); Hells Kloof Compton 18862 (NBG). LIST OF NUMBERED SPECIMENS QUOTED. ACOCKS. 64 exigua; 524 collina; 1630 prostrata; 1900 exigua; 2233 pubescens v. cerosa; 2600 procumbens; 4867 subcarnosa; 8540 procumbens; 8734 sarcophylla; 9027a pubescens; 9616 sarcophylla; 9618 secunda; 9716 papulosa; 11093 pube- scens; 11959 procumbens; 12650 secunda; 12789 pubescens; 13209 africana; 13224 procumbens; 14303 fruticosa; 14347 exigua; 14358 subcarnosa; 14359 secunda; 14678 squamulosa; 14721 affinis; 15934 subcarnosa; 15841 pubescens; 15944 secunda,; 16559 prostrata; 16890 subcarnosa. 124 The Journal of South African Botany. ADAMSON. 1448 herniariaefolia; 1827 secunda,; 1943 crystallina v. maritima; 2047 sarcophylla; 2055 herniariaefolia ; 2083 africana; 2691, 3096 secunda; 3117 crystal- lina; 3133 filiformis; 3159 pubescens; 3212 pubescens; 3400 crystallina v. mari tima; 3544 pubescens; 3764 filiformis; 3932 secunda; 4402 portulacacea; 4412 filiformis; 4462, 4466 secunda; 4478, 4485 crystallina; 4514 sarcophylla; 4521 fruticosa; 4531 crystallina; 4532, 4534, 4537, 4540 fruticosa; 4543, 4544 crystal- lina; 4545 collina; 4552 africana; 4554, 4556 papulosa v. microphylla; 4555 papulosa; 4556 erystallina; 4574, 4575, 4576, 4577, 4579, 4580, 4885 filiformis; 4586a herniariaefolia; 4590 papulosa; 4598 portulacacea; 4604 secunda; 4606, 4607 pubescens; 4609 filiformis; 4616 pubescens; 4618 pruinosa; 4622, 4623, 4624, 4625, 4626, 4628 papulosa; 4640, 4642 cymosa; 4645 portulacacea,; 4650 fruticosa; 4651 secunda; 4652, 4654, 4658, 4661, 4662, 4664, 4665 fruticosa; 4728 africana; 4755 secunda; 4756 herniariaefolia; 4762 secunda; 4763 pubescens; 4817 herniariaefolia; 4848 filiformis,; 4836 collina; 4893 papulosa; 4929 pubescens ; 4931 secunda; 4968, 4970 fruticosa; 4977 collina; 4978 fruticosa; 4985 sarcophylla; 4986 collina; 4987 fruticosa; 4989 collina; 4990 fruticosa; 4991 sarcophylla; 5007 fruticosa; 5013 erystallina; 5014 secunda; 5019 erystallina; 5023 collina; 5034 fruticosa,; 5041 collina; 5043, 5045, 5046, 5047 crystallina; 5048 collina; 5049, 5051 erystallina; 5077 sarcophylla; 5076, 5080 crystallina; 5083, 5084, 5085 sarco- _ phylla; 5086 fruticosa; 5087 portulacacea; 5096 filiformis; 5097 secunda; 5102 filiformis,; 5103 sarcophylla; 5107 collina; 5109 erystallina; 5225, 5127 cymosa; 5129 papulosa; 5133, 5135, 5142, 5143 cymosa; 5146, 5147, 5148 africana; 5149 portulacacea,; 5154 portulacacea; 5178 herniariaefolia; 5179 collina; 5188, 5199 africana x procumbens; 5206, 5207, 5208 procumbens; 5209, 5210 prostrata; 5212, 5214 secunda; 5234, 5235, 5236 africana; 5227 procumbens; 5239 africana; 5240, 5241, 5243, 5246 rigida; 5244, 5245 africana; 5246 procumbens; 5249 sarco- phylla; 52494 papulosa v. microphylla; 5250, 5251 procumbens; 5252, 5259 pubescens; 5260, 5261, 5262, 5263, 5264, 5269,.5271, 5272 pubescens ; 5272 secun- da; 5275 pubescens. BAKKER. 237 africana. BARKER. 104 procumbens; 6748 collina. BARNARD. 759 secunda; 774 africana. BAUER. 1098 pubescens. BELCK. 15 papulosa. BOLUS. 28 papulosa; 116 portulacacea; 182 procumbens; 1866 secunda; 2314, 2314a sarcophylla; 2313, 2315 africana; 2316 cymosa; 2741 africana; 4631 herniariae- folia; 5055 africana; 6815 prostrata; 7545 hispidissima; 7546 Ecklonis; 9218 9268 secunda; 9532 crystallina; 11514 cymosa; 11515 papulosa; 11915 secunda. BRADFIELD. 450 namaensis. BRITTEN. 789 pubescens; 647 africana; 1906 secunda; 5499 secunda. BROCK. 16 pubescens. BRUECKNER. 944 secunda; 962 prostrata.- BRYANT. 115 africana v. secundata; J115, J116 namaensis; J128 erystallina; J231 -papulosa v. microphylla; 377 sarcophylla; 1151a secunda; 1151b prostrata; 1153 sarcophylla. BRYNARD. 154 procumbens; 274 sarcophylla; 275, 231 pubescens. BURCHELL. 1251 pubescens; 1457 sarcophylla; 1613 namaensis; 1764 namaensis ; 2551, 2594 filiformis; 3913 secunda; 6389 secunda; 6414 africana; 6417 filiformis; 6419 pubescens; 6498 affinis. BURTT-DAVY. 1495, 1537 pubescens; 7566 procumbens; 7862 pubescens; 11959 se- cunda; 12648 papulosa; 3108 pubescens v. lignosa; 17560 procumbens. CASTELNAU. 65 africana. CODD. 3425 exigua; 5809 namaensis. COMINS. 674 sarcophylla; 710 procumbens. COMPTON. 2923 cymosa; 2932 fruticosa; 3063 filiformis; 3176 cymosa; 3422 portu- lacacea ; 3540 fruticosa; 3596 cymosa; 3604 secunda; 3605 cymosa; 3802 portulaca- cea; 5735 pubescens; 7356, 9296, 10403 papulosa; 11196 sarcophylla; 12079 secunda,; 11239 papulosa; 12604 africana; 13056 herniariaefolia ; 13253 Ecklonis ; 15094 secunda; 15095 fruticosa; 15233, 17407 secunda; 18862 Dregeana; 19062 pubescens; 20304 secunda; 20870 fruticosa; 21084, 22615 pubescens; 22914 Eck- lonis. CONVENT. 29 pubescens. The South African Species of Arzoaceae. 125 COOPER. 7 pubescens; 518 pubescens; 1281 secunda. CROSS. 48 secunda. CRUDEN. 93 pubescens; 141, 471, 483, 484 secunda. DINTER. 348 papulosa; 3681 hemisphaerica; 3936 secunda; 4154 collina 4853 sarco- phylla; 4888 africana; 4951 fruticosa,; 6094 collina v. Meziana; 6113 squamulosa ; 6164 pruinosa; 6239 papulosa; 6440 hemisphaerica; 8173 Dregeana. DREGE. 571b pubescens; 7060 sarcophylla; 7061a secunda. DREGE, J. F. 3011 pubescens. DUTHIE. 1908 Ecklonis. DYER. 634 pubescens; 1027 procumbens; 1680 secunda; 1906 procumbens; 2124 cymosa. ECKLON. 2149 EHcklonis. ECKLON & ZEYHER. (83) africana; 188 pubescens; 189 secunda,; 309 africana; 2131, 2132 secunda; 2133 pubescens; 2134 secunda; 2136 squamulosa; 2137 papulosa; 2138 crystallina; 2139 affinis,; 2142 herniariaefolia; 2143 collina; 2144, 2145 pubescens v. pallens; 4821 herniariaefolia. EDWARDS. 171 hispidissima. ESTERHUYSEN. 811 pubescens v. cerosa; 812 pubescens; 813, 814 acutifolia; 821 pubescens; 1296 fruticosa; 1331 collina v. Meziana; 1356, 1358 crystallina; 2319 pubescens; 2739 procumbens; 2741 sarcophylla; 4001 africana; 4225 exiqua; 4412, 4449 filiformis; 4454 secunda; 4455 filiformis; 4518 fruticosa; 4509 papulosa; 5150 pubescens; 6054 Ecklonis; 19556 pubescens. FENCHAL. 171 namaensis. FLANAGAN. 119 pubescens; 1382, 1404 procumbens; 1480 prostrata; 1694 secunda; 1657 africana; 1661 secunda. FOLEY. 4 africana; 159 cymosa; 1657 sarcophylla. FOURCADE. 1661 secunda; 2578, 2782 pubescens; 4286 africana; 4654 papulosa; 4658, 4660a pubescens v. Fourcadei; 5677 papulosa; 6061 pubescens; 6460 pube- scens v. Fourcadet. FRIES, NORLINDH & WEIMARCK. 736 papulosa; 781 pubescens; 1517 filiformis. GALPIN. 1837 pubescens; 1941 prostrata; 4073 papulosa; 4074, 4075 africana; 5644 pubescens; 5476 prostrata; 5977 procumbens; 5978 prostrata; 14129 africana v. secundata,; 14200 fruticcsa; 7467 squamulosa; 7680 procumbens. GALPIN & PEARSON. 1671 fruticosa; 7466 squamulosa; 7512 pruinosa; 7466 fallax ; 7467 sarcophylla; 7480 africana x namaensis ; 7487 crystallina; 7489 fallax. GERSTNER. 6332 crystallina; 6397 namaensis. GIBBS. 12 subcarnosa. GILL. 32 procumbens; 54 papulosa; 82 procumbens. GODMAN. 618 fruticosa. GUTHRIE. 766 herniariaefolia; 2607 hispidissima,; 2614 Ecklonis; 3981, 4317 pube- scens. HAFSTROM & ACOCKS. 422 sarcophylla; 423 papulosa; 721 papulosa; H1002 secunda,; 2296 papulosa. HENDERSON. 11 procumbens. HENRICI. 25 secunda; 1898 papulosa; 2158 africana; 2610 namaensis; 2736 sarco- phylla; 2890 prostrata; 3295 africana; 33, 35, 36 fruticosa; 4392 procumbens; 4503 sarcophylla; 4761 sarcophylla; 4768 papulosa; 9349 procumbens. HOBART-HAMPDEN. 23 fallax. HOWEKS. 187 fruticosa. HUTCHINSON. 574 filiformis; 717 sarcophylla; 3136 papulosa; 3136a sarcophylla. JOHNSON. 547 sarcophylila. KIES. 293 procumbens. KINGES. 2054 fallax; 2256 pruinosa; 2458 sarcophylla; 2462 africana v. secundata. KOTZE. 802 procumbens. LAMB. 3642 crystallina. LEIGHTON. 2411 Dregeana. LEIPOLDT. 912 africana; 4012 hispidissima. LETTY. 130 crystallina. LEVYNS. 424 cymosa; 1261 fruticosa; 1363 pruinosa; 2657 cymosa; 3162, 3476 secunda; 4182 secunda; 4321 portulacacea; 5156 crystallina; 5166 filiformis; 6687 secunda; 8016 affinis. LIEBENBERG. 4198 fruticosa; 4479 africana; 5105 papulosa v. microphylla. LINDSTEDT. 34 pubescens. 126 The Journal of South African Botany. LOEB. X1 squamulosa. MACOWAN. 445 secunda. MACPHERSON. 3698/14 papulosa. MAGUIRE. 135, 138 crystallina; 143 fruticosa; 211 filiformis ; 528, 704 secunda ; 754 fruticosa. MARAIS. 6 filiformis; 142 pubescens. MARLOTH. 702 africana; 976 pubescens v. lignosa; 1297 papulosa; 1532 sarco- phylla; 3734 africana; 3701 fruticosa; 4316 fallax; 8247 secunda; 8583 pubescens; 8984 africana; 9813 subcarnosa. MICHELL. 262 fruticosa. MOGG. 3960 pubescens v. lignosa; 4764 pubescens; 8109 pubescens v. lignosa. MORAN. 36, 198 pubescens. MOSS. 2407 herniariaefolia; 6815 africana; 11559 africana x namaensis; 17487 africana; 17880 fallax; 17938 namaensis; 18039 secunda; 18014 sarcophylla. MUIR. 106 secunda; 597 africana; 1824 herniariaefolia; 2399, 2737 secunda; 3753 cymosa; 4709 pubescens. NANNI. 129 secunda. ORPEN. 12, 128 filiformis. PAGE. 115 africana; 41 pubescens. PARKER. 3630 africana. PATERSON. 3082 filiformis ; 3232 secunda. PAXON. 127 namaensis. PEARSON. 514 pruinosa; 529 sarcophylla; 530 papulosa; 3074 supbanenee 3244, 3278 africana; 3356 sarcophylla; 3355 squamulosa; 4799, 4814 africana; 4855 erystallina; 5013, 5146 africana; 5317 squamulosa; 5448, 6027 sarcophylla; 6221 fruticosa; 7172 hispidissima; 8573, 8574, sarcophylla; 8575 africana; 9126 sarco- phylla; 9228 crystallina; 9327 namaensis. PHILLIPS, E. P. 1637 hemisphaerica. PILLANS. 3278, 4952 africana; 4969 collina; 5013 Dregeana; 5158 pruinosa; 5198 crystallina; 5332 pruinosa; 5387 Dregeana; 5428 namaensis; 5529 Dregeana; 5550 collina; 5555 crystallina; 5595 pruinosa,; 6464 africana; 18048 collina; 18049 sarcophylla; 18052 fruticosa; 18053 collina; 18054 fruticosa. POLE EVANS. 11 secunda; 20 papulosa; 51 fruticosa; 1881, 1882 sarcophylla; 2286 sarcophylla; 2335 africana; 2541 secunda; 19107H africana; H19334, H19374 namaensis. PONT. 320 pubescens. POTTS. 2830, 2363 prostrata. REHMANN. 2010, 2469 africana. REYNOLDS. 4186 africana v. secundata. RODIN. 1283 secunda; 2139 namaensis; 2781 namaensis; 3051 rigida. ROGERS. 200 papulosa; 205 sarcophylla; 2275, 2606, 3778 pubescens; 4268 secunda; _ 4466 pubescens; 12030 secunda; 12123 pubescens; 16532 secunda,; 16803 africana; 17118 papulosa; 17164 pubescens; 17796 secunda; 26953 secunda; 27366 pube- scens; 29520 panulosa; 29768 squamulosa, SALTER. 4900 Ecklonis; 6397 secunda; 7012 fruticosa v. prostrata; 7678 secunda; 8621, 8783 pubescens. SCHLECHTER. 31 africana; 34 africana & collina v. Meziana & papulosa; 38 papulosa & sarcophylla; 285 africana; 1817 secunda; 2616 secunda; 5632 africana; 5719 pubescens; 8103 crystallina; 8107 crystallina; 8157, 8159 collina; 8312. secunda; 8313 crystallina; 9481 herniariaefolia; 11065 squamulosa; 11216 collina; 11240 erystallina; 11279 pruinosa; 11290 erystallina; 11301 africana v. pentandra; ~ 11327 sarcophylla; 11441 pruinosa. SCHINZ. 916 sarcophylla. SCHMIDT. 397 sarcophylla. SCHONLAND. 180 pubescens. SCULLY. 258 erystallina. SHAW. 50 procumbens. SIDEY. 342 procumbens; 403 namaensis. SIEBER. 164 herniariaefolia; 351 africana; 2132 secunda. SIM. 2664 secunda; Galpin 5976 prostrata; Galpin 5977 procumbens; Galpin 5978 prostrata. The South African Species of Aizoaceae. 127 SMITH, C. A. 483, 485 sarcophylla; 3516 secunda; 3694 pubescens; 4592 prostrata; 4593, 4596 namaensis; 4599, 4603, 4607 prostrata; 4623A namaensis; 5174 procumbens; 5393, 5437 namaensis. STEPHENS. 7172 hispidissima; 7260 africana. STEPHENS & GLOVER. 8762 fruticosa. STEYN. 9864 namaensis. STOKOE. 9223 africana. STORY. 1119 africana; 1299, 1300 secunda; 2878 portulacacea. TAYLOR, L. E. 907 subcarnosa; 1104 crystallina; 1126 collina; 2734 sarcophylla. THERON. 87 procumbens; 89 africana; 221 sarcophylla; 533 namaensis ; 850 secunda; 906 fruticosa; 1150 procumbens; 1622 africana. THODE. A571 secunda; A572 papulosa; A665 pubescens; A1895 africana; A1896 crystallina; A1897 papulosa; A2663 pubescens. TYSON. 12 secunda; 28 papulosa; 146 procumbens; 254 procumbens; 429 secunda; 431 sarcophylla; 763 africana; 1033 pubescens. 2322 Ecklonis; 2925 filiformis; 4001 africana. VAN BREDA. 112 crystallina; 316 portulacacea. VAN NIEKERK. 330 squamulosa; 344 herniariaefolia. VAN RENSBURG. 51062 africana. VERDOORN. 979 prostrata; 1034 COIS 1082 procumbens; 1325 sarcophylla; 1522, 1605 procumbens. WAGENER. 146 sarcophylla. WASSERFALL. 1037 sarcophylla; 1143 africana.v. secundata. WEBB. _ 5, 11 sarcophylla. WEBB & BRANDWYCK. 1479 africana. WELWITSCH. 1092, 1093 africana v. pentandra. WILMAN. 427 namaensis; 3417 exigua. WOLLEY-DOD. 2062, 2168 herniariaefolia; 3127, crystallina v. maritima. WISS. 965 sarcophylla. ZEYHER. 3, 188 pubescens; 189 secunda; 719 sarcophylla; 722 collina; 723 fruticosa ; 724 affinis; 1049 pubescens v. pallens; 2132 secunda; 2135 pubescens ; 2134 secunda; 2632 secunda; 2633 secunda; 2635 pubescens; 2636 pubescens v. pallens; 2637 filiformis ; 2638 africana & pubescens. THE IDENTITY OF ALOH PARVULA BERGER, WITH A NEW SYNONYM, A NEW NAME, AND A NEW SPECIES FROM MADAGASCAR. By G. W. RryNo.ps. (With Plates X VII—X XII). Aloe parvula Berger was described in Engler’s Pflanzenreich Liliac.- Aloin. 172 (1908), from Madagascar, the type being “Ambato-Loha, Grandidier”’. This species was described as having about 20 leaves, 5— 6 cm. long, 5 mm. broad, pedicels 10 mm. long, and cylindric perianth about 23 mm. long. In Perrier de la Bathie’s monograph Lomatophyllum et les Aloes de Madagascar in Mémoires de la Linnéenne Société de Normandie |,1:27 (1926), and again in Flore de Madagascar, Liliacees, 90 (1838), Perrier described “A. parvula” as having only 5—7 leaves (against 20) 8—15 cm. long (against 5—6 cm.), pedicels 8—l10 mm., perianth 14—15 mm. long (against 23 mm.), and widening from base to mouth, i.e. short-campanu- late against long-cylindrical. Notwithstanding this, Perrier 10995! from Mont Amboloandro, near Zazafotsy, South Betsileo, included by Perrier under A. parvula Berger, has leaves no less than 30 em. long, pedicels being 15 mm., and perianth 15 mm. long. It is now clear that Perrier was mistaken in his concept of A. parvula Berger. This involves another species, A. sempervivoides H. Perrier, “in rather large numbers on the Analamamy Mountains west of Itremo, the locality called Ambatoemenaloha by Grandidier. Perrier 12493.” This is the type locality of A. parvula Berger (non Perrier). When a plant collected by Professor J. Millot in the mountains west of Itremo flowered in Johannesburg, it fitted the description of A. parvula Berger, and also of A. sempervivoides Perrier. Professor Millot also collected several plants of another small species in the mountains west of Itremo. Two groups of these plants have flowered in Johannesburg and they differed in too many essential characters to be conspecific with Perrier’s concept of A. parvula Berger from Mont Amboloandro. To clear up the confusion, A. parvula Berger should be more fully described and figured, with A. sempervivoides H. Perrier reduced to synonymy. <A. parvula Perrier non Berger from Mont Amboloandro 129 130 The Journal of South African Botany. should be renamed, and I propose A. Perrieri in commemoration of Mr. Perrier de la Bathie who has contributed such a considerable amount to the knowledge of the Aloes of Madagascar. A small species collected by Professor Millot on the Itremo should be described as new. Professor Millot is in complete agreement with these proposals, and the following is published with his full concurrence. Aloe parvula Berger in Engler Pflanzenr. Liliac.-Aloin. 172 (1908)— non H. Perrier de la Bathie in Lomat. et Aloe, Mem. Soc. Linn. Norm. 1, fase. 1: 27 (1926), et FI. Madag. Liliac. 90 (1938). —A. sempervivoides H. Perr. Mém. Soc. Linn. Norm. 1, fase. 1: 28 (1926), et Fl. Madag. 90 (1938). Plant small, acaulescent, solitary. Leaves about 24, densely rosulate, compactly suberectly spreading, 4—6 cm. long, 8 mm. broad at base, gradually narrowing to an acute apex, rather thick and fleshy; upper surface flat, pale bluish-grey, muricate throughout with rough processes -5—1l mm. long; lower surface convex, similar to upper surface; margins armed with soft to firm (not pungent) narrowly deltoid white cartilaginous teeth 1—2 mm. long low down, 1— 2 mm. distant, the interspaces rounded and the colour of the leaf. Inflorescence simple, sometimes 2, about 35 cm. high. Peduncle basally flattened and 3 mm. broad, terete and slightly narrower upwards, clothed with a few sterile-bracts, the lowest deltoid- acuminate, 12 mm. long, 3 mm. broad, about 7-nerved, smaller and more ovate-acute upwards. Raceme 5—6 cm. long, 4 cm. diam., laxly about 12—15-flowered, the apex a small tuft of dried bracts, youngest buds spreading and greenish- grey tipped, open flowers pendulous. Bracts lowest ovate-deltoid, subscarious, 3—5-nerved, clasping the pedicel, about half as long as the pedicel. Pedicels lowest 10 mm. long, obliquely spreading, recurved near apex. Perianth light coral red, 24 mm. long, cylindric-ventricose, distinctly trigonous and somewhat flat-sided, obtusely tapering at base into the pedicel, shortly stipitate, enlarging to the middle, thence narrowing to the mouth; outer segments free for 7 mm. (tube 17 mm.), white at the margins, obscurely 5-nerved, the nerves greenish at apex, the apices acute, straight, scarcely spreading; inner segments themselves free but dorsally adnate to the outer for 17 mm., broader than the outer, and with broad white marginal border, the apices more cbtuse than the outer. Filaments paie rose, flliform-flattened, the 3 imner narrower and lengthening before the 3 outer with their anthers in turn not or scarcely exserted. Style filiform, with stigma at length scarcely exserted. PLATE XVII. Aloe parvula Berger (non H. Perrier). Photograph x 2/5 of painting No. 63 in the Botanical Dept., Institut de Recherche Scientifique, Tzimbazaza, Tananarive, of a plant collected by Professor J. Millot, west of Itremo, Madagascar. The painting, natural size, depicts leaves 4 cm. long, inflorescence 25 em., pedicels 12 mm. perianth 25 mm. Ini, Il. Hie. 2. PLATE XVIII. Aloe parvula Berger (non H. Perrier). Fra. 1.—Plant (x 1/4 approx.) flowering in Johannesburg, originally collected by Professor J. Millot, west of Itremo, Madagascar. Fic. 2.—Raceme 1/1. The Identity of Aloe Parvula Berger, with a new Synonym, 131 a New Name, and a New Species from Madagascar. Ovary olive-green, 3 mm. long, 1-5 mm. diam., truncate at apex. (Puates XVII and XVIII.) MaTeriAL: Grandidier s.n. Ambatomenaloha (P); Mts. Analamamy, west of Itremo, 2,000 m. “the locality called Ambatomenaloha by Grandi- dier’’, Perrier 12493 (P); mountains west of Itremo, coll. Prof. J.Millot 1949, cult. Johannesburg, fl. 24 Nov. 1951, Reynolds 6017 (PRE). (No material in Kew.) The Analamamy (Ambatomenalcha) mountains in West Betsileo lie south-west of Itremo Village, and to the south-west of Ambatofinandra- hana which is on the main road from Ambositra and Ivato to Morondava on the west coast. A. parvula does not appear to have been collected elsewhere. The specimen collected by Bernier on rocks at Cap Manambato in the north of the Island (Bernier 2nd coll. n. 118) is without leaves and needs further investigation. A. parvula is closely allied to A. haworthioides Baker in leaf characters and small rosettes, but the latter has suberect flowers only 8 mm. long with negligible pedicels. A. parvula is perhaps nearest allied to the South African species A. hwmilis (L) Mill., var. echinata (Willd.) Bak., but the latter has much longer bracts, longer pedicels and longer flowers.—See Aloes S. Afr. 177—8 (1950). A plant of A. parvula from west of Itremo, collected by Professor Millot in 1949, and grown in Johannesburg for five years has shown no signs of suckering. Aloe Perrieri Reynolds. Nom. nov.—A. parvula Perrier (non Berger) in Mém. Soc. Linn. Norm. 1, fase. 1: 27 (1926), Fl. Madag. Liliac. 90 (1938) pro parte. DESCRIPTION (based partly on Perrier’s description of A. parvula Perrier non Berger, and partly on Perrier n. 10995 from Mont Ambolo- andro): Plant succulent, with fusiform roots, acaulescent, suckering at base and forming groups. Leaves about 10, rosulate, linear-attenuate, rather compactly suberect, about 30 cm. long, 10—12 mm. broad at base, gradually narrowing to the apex; wpper surface slightly canaliculate; lower surface convex; both surfaces studded with minute. asperities, maculate, the spots very pale green; margins armed with small narrowly deltoid firm cartilaginous white teeth about 1 mm. long, 1—2 mm. distant. Inflorescence simple, slender, 30—40 em. high. 132 The Journal of South African Botany. Peduncle 2 mm. diam., clothed with a few sterile bracts. Raceme cylindric slightly conical, 12 em. long, rather laxly about 20— 30-flowered. Bracts ovate-acute, scarious, 3-nerved, 4 mm. long in the dried ma- terial. Pedicels 15 mm. long. Perianth cylindric-campanulate, 16 mm. long, basally tapering into the pedicel articulation and shortly stipitate, gradually enlarging to a wide-open mouth; outer segments free to the middle, 3-nerved. Anthers and Stigma not, or very shortly exserted. (PLATES XIX and XX) Madagascar Fianarantsoa Proy., S. Betsileo, Mont Amboloandro, 1,000 m., March 1912, Perrier 10995! holotype (P); M. Decary 5701!with- out locality (P), appears to belong here.—This may be the specimen cited by Perrier from Mont Ivohibory west of Ivohibe, cult. Tananarive. Mont Amboloandro lies to the west of the main road from Zazafotsy to Ihosy at approx. 22° 15’ §., 46° 15’ E., but I was unable to reach it during my travels in Madagascar. I have not seen living plants, and the description is lacking in some characters. A. Perrieri is characterised by having leaves about 30 cm. long and a simple inflorescence less than twice the length of the leaves. This is illu- strated in Perrier’s Fig. XIII, 4, in Fl. Madag. 95 (1938). It is one of three species in Madagascar having short campanulate flowers 13—16 mm. long. One of these species is A. laeta Berger which I have seen on Mont Ibity. <A. laeta is a considerably larger plant with large rosettes of about 24 leaves, the leaves being lanceolate-attenuate, 20 cm. long, 6 cm. broad, pedicels 25 mm., a 10 mm. thick peduncle and a simple or 1- branched inflorescence about 60 cm. high. The other species is described below. I have not seen Perrier 11021 from the Isalo range near Benenitra and do not know whether it belongs to A. Perrieri, or to the following new species: A. bellatula Reynolds. Spec. noy., A. Perrieri affinis, foliis brevioribus, inflorescentia altiore differt. Planta parva, succulenta, acaulescens, caespitosa. Folia ca. 16, dense rosulata, erecto-patula, 1O—13 em. longa, 9—10 mm. lata, 3 mm. crassa; supra leviter canaliculata, subtus convexa, utrinque minute verrucosula, papillis minutis asperata scabrida, maculata; dentibus marginibus deltoideis albidis cartilagineis 1 mm. longis, 1 mm. distantibus munita. Inflorescentia simplex vel interdum l-ramosa, ad 60 cm. alta. Racemus cylindrico-acuminatus, 12—16 em. longus, 4 cm. diametro, laxe Reproduction of Perrier de la Bathie’s Fig. XIII 4, in Flore de Madagascar Liliacees 95 (1938)—A. parvula Perrier (non Berger) « 1/4. Se <=" | | HERB, MUS. PARIS Wlb99 PLATE XX. Aloe Perrieri Reynolds. Photograph of the holotype in Paris—Perrier n. 10995, Mont Amboloandro, near Zazagotsy, S. Betsileo, Fianarantsoa Province, 1,000 m., March 1912. Leaves 30 cm. long, inflorescence 40 cm., perianth 15 mm. PRs. “ PLATE XXI. Aloe bellatula Reynolds. Plant originally collected by Professor J. Millot, west of Itremo, cult. fl. February, 1956. Height 55 em. Johannesburg, ‘sppoudoyy 177N70))99 00)F “TIXX WLVTd ] ‘soBuqs uotruypod-ysod 04 pnq wtody [/] STOMO[Y—§ “OTY] ‘sqgurod oyt-ouypided OIMNUILE GUIMOYS F X OoOVJANS LOMO] JRO] JO UOIWOG—F “YTy] “QOVJANS LOMOT 4ysix foovjans aoddn :qyoy ‘{/] SoAvo] JO UOIWLOg™— | “Ty “I The Identity of Aloe Parvula Berger, with a new Synonym, 133 a New Name, and a New Species from Madagascar. ca. 35-floribus. Bracteae acuminatae, albidae, 1-nervatae, 4—6 mm. longae, 2 mm. latae. Pedicelli 12 mm. longi. Perigonium pallide rubrum, cylindrico-campanulatum, ca. 13 mm. longum, basi obconicum, stipi- tatum, medio 6 mm. diam., fauces versus ampliatum; segmenta exteriora per 7 mm. libera, obscure 3-nervata. Genitalia vix exserta. Ovariwm pallide aurantiacum, 2-5 mm. longum, 1-5 mm. diametro. (PLATES X X and XXII). Fianarantsoa Prov., W. Betsileo, west of Itremo, c. 20° 39’ S., 46°36’ E., c. 1,500 m., leg. Prof. J. Millot, cult. Johannesburg, fl. 29 Jan. 1956, Reynolds 6591, holotype (PRE), second inflorescence 8 Feb. 1956, isotype (TAN), third inflorescence 19 Feb 1956, isotype (P). Our new species was collected b:7 Professor J. Millot in 1949, west of Itremo, and appears to be nearest allied to A. Perrieri described above. A. bellatula is characterised by having leaves only 10—13 em. long, and a very slender inflorescence about five times longer than the leaves, whereas in A. Perrieri the leaves are 30 em. long with the inflorescence less than twice the length of the leaves. Description: Plant small,acaulescent, with cylindric roots, suckering at base and forming dense tufts. Leaves about 16, densely rosulate, compactly ascending, slightly incurved-spreading, 10—13 cm. long, 9—10 mm. broad at base, 3 mm. thick, gradually tapering to an acute apex; wpper surface slightly canalicu- late; lower surface convex; both surfaces minutely verrucose, rough with minute papillae-like points, dark green, copiously white-spotted through- out, the spots pale green, sublenticular, averaging 1 mm. long, -5 mm. broad; margins armed with soft cartilaginous deltoid teeth 1 mm. long, 1 mm. apart low down, smaller to obsolescent upwards. Inflorescence up to 60 cm. high, simple or with 1 branch from about the middle. Peduncle flattened and 3 mm. broad at base, slightly slenderer up- wards, clothed with about 5 sterile-bracts which are narrowly ovate- acuminate, thin, white, subscarious, prominently I-nerved, the lowest 10 mm. long, 3 mm. broad, smaller upwards. Racemes cylindric-acuminate, 12—i6 em. long, 4 em. diam., about 35-flowered, laxer below, slightly denser upwards, youngest buds sub- erect, older buds horizontally disposed, open flowers nutant. Bracts deltoid-acuminate,. thin, white, prominently l-nerved, the lowest 4—6 mm. long, 2 mm. broad, smaller upwards. Pedicels lowest 12 mm. long, nutant at apex, becoming erect in the fruit, the colour of the perianth. 134 The Journal of South African Botany. Perianth light coral-red, ceylindric-campanulate, 13 mm. long, basally obconic and shortly stipitate, 6 mm. diam. at the middle, the mouth wide open; outer segments free for 7 mm., obscurely 3-nerved, the apices sub- acute, slightly spreading; inner segments free but dorsally adnate to the outer to the middle, pale rose-pink with pale coral-red keel, the apices slightly more spreading than the outer. Filaments pale-rose, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn scarcely exserted. Stigma at length scarcely exserted. Ovary pale orange, 2:5 mm. long, 1-5 mm. diam., obtuse at apex. ACKNOWLEDGMENTS. I am grateful to Professor J. Millot, Director of the Institut de Re- cherche Scientifique de Madagascar, and to Dr. R. Paulian, the Deputy Director, for every possible facility and assistance during my travels in Madagascar, and for many plants of Aloe on several occasions. To Professor H. Humbert, Director, Museum National d’ Histoire Naturelle, Paris, I am greatly indebted for sending several sheets of Aloe material to Johannesburg, and for much other assistance. To the South African Council for Scientific and Industrial Research I am indebted for a travelling grant that enabled me to fly to Madagascar to investigate the Aloes. THREE NEW SPECIES AND ONE NEW VARIETY OF ALOE FROM KARAMOJA, UGANDA. By G. W. Rrynorps. (With Plates X XIJI—X XVIII.) Karamoja is that portion of the Northern Province of Uganda lying to the north of the foothills of Mount Elgon and the Suam River, east of Teso and Acholi, south of the Sudan border, and west of the Turkana (Kenya) border. During the last six years Mrs. E. M. Tweedie has sent many Aloe plants from Karamoja to Johannesburg; some she collected herself, while others were collected by Mr. H. Wreford Smith, Mr. T. H. E. Jackson, and Mr. J. T. Wilson, who have all travelled extensively through- out Karamoja. When some plants flowered in Johannesburg they proved to be undescribed species, and I was eager to study them in the wild state. At the invitation of Colonel and Mrs. D. R. Tweedie I flew to Nairobi in October 1955, and travelled by road to Kitale and on to their delightful home ““Mutamayo” on lower slopes of Mount Elgon near Endebess. Colonel and Mrs. Tweedie very generously provided the transport, and together we travelled more than 900 miles in Karamoja. The route led northwards from Kitale, down the spectacular Kacheliba Pass to the plains, rich in Aloe. We visited Amudat, Moroto, Kotido and Loyoro in the far north, and then went to the Labwor Hills south-west of Kotido. On the return journey we went from Moroto down Napau Pass to the border of Turkana. A. Tweediae Christian was found in large numbers almost throughout Karamoja, and proved to be the commonest Aloe of that country. A. marsabitensis Christian was seen near Kotido, and especially a few miles west of Loyoro in the north. A few plants of A. turkanensis Christian were observed near Amudat and Mount Toror, also in Turkana at the type locality. On some mountain slopes (Kacheliba, Kadam, Moroto) a shrubby Aloe was seen, but could not be recognised without flowers. Herbarium material of all Aloe species found in flower has been contri- buted to Pretoria, Kew and Nairobi, together with many photographs. 135 136 The Journal of South African Botany. Three new species and one new variety were investigated and can now be described. Aloe amudatensis Reynolds. Species nova in Sect. Saponariae. Planta succulenta, acaulescens, mox caespitosa. Folia 10—14, dense rosulata, erecto-patentia, anguste lanceolato-attenuata, 22 em. longa, 5 cm. lata, 8 mm. crassa; supra leviter canaliculata, subtus convexa, utrinque viridula, maculis oblongis numerosis transverse irregulariter fasciatim seri- atis picta; ad margines sinuato-dentata, dentibus albidis 2 mm. longis, 5—8 mm. distantibus. Inflorescentia 50—65 cm. alta, simplex vel inter- dum supra medium 1—2 ramosa. Racemz cylindrici, ca. 8 em. longi, 6— 7 cm. diam. Bracteae anguste ovato-deltoideae, 10 mm. longae, 3 mm. latae, scariosae, 3—5-nervatae. Pedicelli ca. 17 mm. longi. Perigonium corallinum, ca. 23 mm. longum, basi inflatum et circa ovarium 9 mm. diametro, supra ovarium valde constrictum (6 mm.) hine leviter decur- vatum et fauces versus paullum ampliatum; segmenta exteriora per 7 mm libera, obscure 3-nervata, interiora latiora. Antherae 0—1 mm. exserta. Stigma demum 1—2 mm. exserta. Ovariuwm 6 mm. longum, 2 mm. dia- metro. (Plate XXIII.) Uganda: Northern Province, Karamoja, Upe County, at Amudat, c. 1° 58’ N., 34° 57’ E., 4,400 ft., fl. 21 Oct. 1955, Reynolds 7996 holotype (PRE), isotype (K, EA); Dodoth County, 7 miles west of Loyoro, c. 3° 13’ N., 34° 12’ E., 4,500 ft., fl. 14 Oct. 1955, Reynolds 7965 (PRE); Matheniko County, at Lokitanyala (37 miles north of Amudat), form with yellow flowers, fl. 20. Oct. 1955, Reynolds 7988 (PRE); Tweedie 670 (K). A. amudatensis was collected by Mrs. E. M. Tweedie several years ago and sent to Kew under her number 670. Our new species differs from its nearest Hast African allies in the Section Saponariae by being a smaller plant, and in having narrow leaves only 5—6 em. broad with white cartilaginous margins armed with small firm almost white teeth only 1—2 mm. long. The inflorescence is mostly simple, averaging 50—65 cm. tall; sometimes it is 1—2-branched in upper third; and flowers vary in length, and in the diameter of the basal swelling. Plants are acaulescent, sucker freely and form small to large dense groups. DistTRIBUTION: A. amudatensis was seen almost throughout Karamoja, from the Suam River to Amudat (91 miles north of Kitale) in the south, to Moroto, Teror, Kotido, Madang, and south and west of Loyoro in the north. Yellow and orange forms occur at Lokitanyala between the Moroto turn-off and the top of Napau Pass which leads down to Turkana in Kenya. A few plants were seen in Kenya Colony between the foot of the Kacheliba Escarpment and the Suam River. Three New Species and One New Variety of Aloe from 137 Karamoja, Uganda. Naturat Hysrip: A. Tweediae Christian occurs in large numbers near Amudat (type locality), and crosses with A. amudatensis were observed among both parents. Description: Plant succulent, acaulescent, freely suckering and form- ing small to large dense groups. Leaves about 12, densely rosulate, erectly spreading, narrowly lanceolate-attenuate, averaging 22 cm. long, 5 cm. broad, 8 mm. thick low down; wpper surface dull green, sometimes reddish-brown tinged, slightly canaliculate, with numerous lenticular whitish spots or blotches throughout; lower surface convex, milky-green, usually with numerous dull very pale greenish lenticular spots or blotches varying from irregu- larly scattered to more or less confiuent into a series of irregular interrupted transverse wavy bands; margins sinuate-dentate, with whitish cartilagi- nous edge armed with small, firm, rather white, deltoid teeth which aver- age 2 mm. long, 5—8 mm. apart, pale reddish-brown at apices only. Inflorescence mostly simple, averaging 50—65 cm. high, sometimes with 1—2 branches high up. Peduncle basally plano-convex and 10 mm. broad, terete upwards, brownish-green with a bloom, sometimes with | or 2 branches in upper third. Racemes cylindric, slightly conical, averaging 8 cm. long, 6—7 cm. diam., sublaxly flowered, the apex a small tuft of dried bracts, youngest buds suberect, grey-green tipped, older buds spreading, open flowers subpendulous. Bracts narrowly ovate-deltoid, 10 mm. long, 3 mm. broad, thin, sub- scarious, 3—5-nerved. Pedicels averaging 17 mm. long, arcuate-spreading, nutant at apex. Perianth rose-pink to coral-red, averaging 23 mm. long, basally inflated and 9 mm. diam. across the ovary, thence abruptly constricted to 6 mm., thence slightly decurved and enlarging to the throat, the mouth trigonous, wide open; outer segments free for 7 mm., paler at the margins, obscurely 3-nerved, the apices subacute, slightly spreading; inner segments broader than the outer, with more obtuse more spreading apices. Filaments lemon, filiform-flattened, the 3 inner narrower and lengthe- ning before the 3 outer with their anthers in turn exserted 0—l mm. Style pale yellow, with stigma at length exserted 2 mm. Ovary yellowish-green, 6 mm. long, 2 mm. diam. Aloe Wilsonii Reynolds. Species nova, forsitan in Sect. Tropicales Berger pertinet; A. Tweediae Christian affinis, sed, inter alia, perigonio basi obconico differt. 138 The Journal of South African Botany. Planta succulenta, breviter caulescens, interdum caudice 50—80 em. Folia ca. 24, dense rosulata, lanceolato-attenuata, patentia, 25 em. longa, 9 cm. lata, 2 cm. crassa; supra planiuscula vel leviter canaliculata, subtus convexa; utrinque saepe immaculata; ad margines sinuato-dentata, dentibus deltoideis, 3—4 mm. longis, 10 mm. distantibus. Inflorescentia paniculata, 50—60 cm. alta. Pedunculus ca. 8-ramosus. Kacemi cylindrico- acuminati, 15—18 cm. longi, 5—6 em. diam. Bracteae ovato-acuminatae, seariosae, 5 mm. longae, 3 mm. latae, l-nervatae. Pedicelli 15 mm. longi. Perigonium coccineum, cylindricum, 2—8 mm. longum, basi obconicum, circa ovarium 7 mm. diametro; segmenta exteriora per 14 mm. libera. Antherae 1—2 mm. exsertae. Stigma demum 2—3 mm. exserta. Ovariwm 6 mm. longum, 3 mm. diametro. (Plates XXIV, X XV.) Uganda: Northern Province, Karamoja, Matheniko County, Mount Moroto near Imagit Peak, 8,500 ft., c. 2° 45’ N., 34° 44’ E., cult. hort. Tweedie, Mutamayo (near Endebess, Kenya) fl. 30 Oct. 1955, Tweedie 1365 (Reynolds 8004) holotype (PRE), isotype (HA). A. Wilsonii is named after Mr. John T. Wilson; Ecologist, Department of Agriculture, Moroto, who first discovered the species, and to whom I am indebted for sending plants to Johannesburg, and for photographs and much data concerning the Aloes of Karamoja. When I visited Karamoja in October 1955, Mr. Wilson and I climbed 3,000 ft. up a spur leading to Imagit Peak, Mount Moroto. The Aloes were found on rocky slopes near the summit in advanced bud, but no flowers were open. A plant was photographed in bud (Plate XXIV) and part of the description was drawn up. A week later I found a plant flowering in the garden of Mrs. E. M. Tweedie, “Mutamayo”’, Endebess near Kitale, Kenya Colony, and com- pleted the description. This plant had originally been collected by Mr. Wilson on Mount Moroto, near Imagit Peak. On Mount Moroto, plants are solitary, acaulescent, or sometimes deve- lop a decumbent smooth stem 50—80 cm. long. Mr. Wilson has also found this species on north-western slopes of Kadam Mountain (Debasien), about 40 miles south of Moroto, at 5,000—8,000 ft., c. 1° 49’ N., 34° 42’ B. At this locality, Mr. Wilson found some plants solitary, and others forming small groups. A striking character of A. Wilsonii is the very acuminate racemes that taper to an acute point. The youngest buds, not hidden by bracts, are round-oval, and only 1—2 mm. long. The perianth is basally rather obconical and tapers into the pedicel articulation, while outer segments are free to the middle. A. Wilsonii is a very distinctive species, and does not appear to be closely allied to any species of Aloe hitherto described. ‘UOTWVIIBA SuLMoYsS [/] StaMopy snore A—: ‘e5v4s uoTyeurod-4sod 04 pnq ulody [/] StaMoyq{— “UID OY WSTOY “GEGT ‘19qG0900 TZ “B “43 OOFF “A LG cbE ON 8e ol ‘9 ‘epuesy ‘elourereyy sppousoy SUSUA]DPNULD 00) F ‘THINN HLV Id —_ eat ig? ‘Aqunoy ody ‘epnury qe squerg—: i 4 Ns is eB § G I DIT OLY “DIT (UOSTIM, “Lf AIN $0204) “GEGT “OQUIOAON 0% SpodopoaAop SUIARY, edUEdDSeELOYUT PUODES OF *OFOAOT “FPN ‘guejd ours oy ,—% “OI ST ‘1040900 6 PUG Ut “43 OOS'S “HU FR .FE NSF oG “BpuRs /) ‘vfoureae sy “OVOIOP, FUNOTY “YROT jlovuly wRou *xoadde 1/1 x 4ue_d— | ‘91 ) u ‘sploudoy Muos {1 207F. “AIXX ChLWId re OU TT “Diy nt PuatE XXV. Aloe Wilsonii Reynolds. Flowers 1/1 from bud to post-pollination stage, from a plant ex Imagit Peak, cult. hort. Tweedie, Mutamayo. “sppousoyy DUDLONGD)] “WEA “MV arypnfuvannyag 30) “IUAXOK (GNI Tel 7 Sl ‘ose4s uorjeurpod-ysod 03 pnq worz [/] SsdoMo]y—'Z “OT “UL (6 qUSIOH “CEG ‘Your LZ ‘Sanqsouueryor 4yno ‘epuvsy) ‘elourerey ‘s[[tFT TOMqey og ULOay QURTG— | “OT yy "2 OE . © n eer ) Three New Species and One New Variety of Aloe from 139 Karamoja, Uganda. It may perhaps fall under Berger’s section T'ropicales, but many of Berger’s sections and series now need drastic revision. A. Tweediae Christian is allied in habit of growth, branching of the inflorescence and cylindric perianth with a definite tube, but the perianth of A. T'weediae is basally rounded, and not obconical as in A. Wilsonit. The apical buds of A. Wilsonii are also considerably smaller. Native Name: EHchuchukua (Karamojong), used throughout Karamoja for Aloe in general—J. T'. Wilson. Description: Plant succulent, solitary or sometimes in small groups. Stem short, or sometimes 50—80 cm. long, 6 cm. diam., decumbent, glabrous, old leaves not persistent. Leaves about 24, densely rosulate, lanceolate-attenuate, erectly spreading to spreading, up to 25 cm. long, 9 cm. broad at base, 2 em. thick low down; upper surface flat to slightly canaliculate, unicoloured deep green with yellowish tinge, slightly glossy, without spots or markings; lower surface convex, of similar colour to upper surface, usually without but sometimes with a few crowded white spots in lower quarter; margins sinuate-dentate, with pale brown horny edge armed with flat-deltoid pungent reddish teeth averaging 3—4 mm. long, 10 mm. distant. Inflorescence a branched panicle 50—60 cm. high, sometimes 2 con- secutively. Peduncle basally plano-convex and 25 mm. broad, brown with a bloom, minutely white-speckled throughout, branched from below the middle with about 8 arcuate-ascending branches, the lowest branch subtended at base by a fleshy, 3-nerved, broadly ovate-acute bract 10 mm. long, 20 mm. broad. Racemes cylindric-acuminate, erect, 15—18 cm. long, 5—6 cm. diam., rather laxly flowered, the lowest flowers 10—15 mm. apart, gradually denser upwards, the youngest buds suberect, round-oval and only 1—2 mm. long, older buds grey-green tipped, erectly spreading, open flowers cernuous to sub-pendulous. Bracts ovate-acuminate, the lowest 5 mm. long, 3 mm. broad, white, thin, scarious, l-nerved. Pedicels reddish-brown, suberect, 15 mm. long, nutant at apex. Perianth dull scarlet with a bloom, cylindric, 28 mm. long, obtusely obconical at base and shortly stipitate, cylindric and 7 mm. diam. across the ovary, thence slightly trigonous to the mouth; outer segments free for half their length (tube 14 mm.), pale lemon at the edges, 3-nerved, apices subacute, spreading; inner segments free but dorsally adnate to the outer to the middle, broader than the outer, with broad pale borders scarlet-keeled, the apices brownish-tipped on the inside, more obtuse and more spreading than the outer. , 140 The Journal of South African Botany. Filaments lemon, filiform-flattened, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn exserted 1—2 mm. Style pale yellow, with stigma at length exserted 2—3 mm. Ovary deep green, 6 mm. long, 3 mm. diam. Aloe Schweinfurthii Bak. var. labworana Reynolds. Varietas nova, a forma typica foliis longioribus, racemis brevioribus et densioribus, et praesertim rache et pedicellis et floribus concoloribus luteis differt. Uganda: Northern Province, Karamoja, Labwor Hills, at Abim, 6 miles S.W. of Rwot, 41 miles S.W. of Kotido, c. 2° 47’ N., 33° 38” E., e. 4,500 ft. (coll. T. H. E. Jackson), fl. 15 Oct. 1955, Reynolds 7980 holotype (PRE), isotype (K, EA); Labwor Hills, Aremo, 58 miles S.W. of Kotido, c. 2° 30’ N., 33° 53” E., (coll. T. H. EK. Jackson), cult. Kitale, fl. 25 Oct. 1955, Reynolds 8006 (PRE, EA, K); Labwor Hills, Loyoroit, 30 miles S.W. of Kotido, c. 2° 46’ N., 33° 41’ E. (coll. H. Wreford Smith), cult. Johannesburg, fl. 27 March 1955, Reynolds 6274 (PRE, K). (Plate XXVI) The typical form of A. Schweinfurthii Bak. on Mount Bangenze, Equatoria, Sudan, was described and figured in this Journal 20: 165—168 (1954). Our new variety comes from the Labwor Hills, south-west of Kotido in Karamoja, Uganda, and is named after the locality which appears to be its headquarters. A few years ago, Mr. H. Wreford Smith collected plants at Loyoroit, about 30 miles south-west of Kotido, and sent them to Mrs. E. M. Tweedie who sent two plants to me. These plants have grown well in Johannesburg and when they flowered they differed sufficiently from typical A. Schwein- furthi to merit at least varietal rank. In October 1955, I visited Loyoroit but found no plants in flower. Mr. T. H. E. Jackson however went further afield and he kindly brought in flowering plants from Abim and Aremo further south-west, from which photographs were secured and herbarium material made. The var. labworana closely resembles typical A. Schweinfurthii in habit of growth, leaves and markings, small 1-nerved bracts, pedicels and shape and size of flowers, but differs in having comparatively longer leaves with larger marginal teeth, and a more branched inflorescence with shorter denser racemes of yellow flowers: The most striking difference is that the rachis, pedicels and flowers are all of a uniform yellow colour. Descrietion: Plant succulent, of medium size, acaulescent, suckering at ground level and forming groups. Leaves 12—16, rosulate, erectly spreading, narrowly lanceolate- attenuate, recurved in upper quarter, 60—80 cm. long, 7—8 em. broad at base, 2 cm. thick low down; upper surface flat to slightly canaliculate, bluish-green with a grey powdery bloom, with numerous very pale- Three New Species and One New Variety of Aloe from 141 Karamoja, Uganda. greenish spots throughout, the spots lenticular, 3—6 mm. long, 1—2 mm. broad at the middle; lower surface convex, greyer-green than upper sur- face, more copiously spotted throughout, the spots smaller and more crowded; margins sinuate-dentate, armed with pungent deltoid reddish- tipped teeth which are 5—6 mm. long, 12—20 mm. distant, with rounded interspaces. Sap dries yellowish. Inflorescence a divaricately many-branched panicle up to 90 cm. high. Peduncle brown, basally plano-convex and 20—25 mm. broad, divari- cately branched from the middle or lower, with 10—12 arcuate-ascending branches, the 1—2 lowest with 3—4 branchlets. Racemes cylindric, slightly conical, 8—9 cm. long, 5 em. diam., youngest buds suberect, denser, open flowers laxer, nutant to subpendu- lous, the whole raceme (axis, pedicels and flowers) uniformly yellow. Bracts ovate-acute, about 2 mm. long and broad, clasping the pedicel, thin, secarious, |-nerved. Pedicels yellow, 10 mm. long. Perianth yellow, 28 mm. long, cylindric, slightly trigonous, basally obtuse, 6—7 mm. diam. across the ovary, thence very slightly narrowed on the underside and enlarging slightly to a wide open mouth; outer segments free for 9 mm. (definite tube of 19 mm.), with 3 greenish nerves throughout, the apices subacute, slightly spreading; inner segments themselves free but dorsally adnate to the outer to beyond the middle, broader than the outer, keeled, the apices more obtuse and more spreading than the outer. Filaments very pale lemon, filiform-flattened, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn exserted 3—4 mm. Style pale yellow. Stigma at length exserted 5 mm. Ovary pale green, 5 mm. long, 2-5 mm. diam. Aloe Wrefordii Reynolds. Species nova (Sect. Grandes-Tropicales), A. abyssinica Lam. affinis, inflorescentia altiore plus ramosa differt. Planta succulenta, saepissime acaulescens. Folia ca. 24, dense rosu- lata, lanceolato-attenuata, patentia, ca. 60 cm. longa, 15 em. lata, 15 mm. crassa; supra planiuscula vel superne leviter canaliculata, obscure lineata; subtus convexa; ad margines sinuato-dentata, dentibus deltoideis 4 mm. longis, 1O—15 mm. distantibus munita. IJnflorescentia paniculata, ca. 120 cm. alta. Racemi subcapitata, 4—5 cm. longi, 6 cm. diam. Bracteae ovato-acutae, 9 mm. longae,.4 mm. latae, albidae scariosae, 5-nervatae. Pedicelli 16 mm. longi. Perigonium coecineum vel aurantiacum, cylin- drico-clavatum, 22 mm. longum, circa ovarium 5—6 mm. diam.; segmenta 142 The Journal of South African Botany. eateriora per 13 mm. libera; tubus 8 mm. Antherae 3—4 mm. exsertae. Stigma demum 5 mm. exserta. Ovariwm 6 mm. longum, 3 mm. diametro. Uganda: Northern Province, Karamoja, Matheniko County, Katumet, 12 miles south-east of Moroto, c. 2° 24’ N., 34° 45’ E., 4,900 ft., cult. hort. Tweedie, Mutamayo, fl. 21 Jan. 1956, Tweedie 659, holotype (PRE), isotype (EA, K), stony foothills, Mt. Moroto, Nov. 1945, Tweedie 659 (K). Our new species js named after Mr. H. Wreford Smith who has travelled extensively throughout Karamoja and who first discovered plants at Katumet, 12 miles south-east of Moroto in 1940. Subsequently Mrs. E. M. Tweedie found the species at another locality nearby, and sent material to Kew under her number 659. In October 1955 I saw plants at these localities but none was in flower. Recently Mr. John T. Wilson, ecologist at Moroto, found several plants at Rupa, 9 miles north-east of Moroto, in stony thorn and bush country. When I visited Rupa, no plants were in flower, but part of the description was drawn up from the material available. A photograph was secured of a large plant in fruit. (Plate X XVII). When one of the plants originally collected by Mrs. Tweedie near Moroto flowered in her garden at Mutamayo, she had it photographed and prepared adequate herbarium material. Mrs. Tweedie also sent several mature racemes in spirit to me by air, from which I was able to complete the description and photograph the flowers natural size. (Plates X XVII, XXVIII). Up to the present, A. Wrefordii has been found only on lower foothills of Mount Moroto. A plant that I collected in 1953 near Yavello, in Borana, Southern Ethiopia, has not flowered in Johannesburg, but judging from the rosette and leaves, it might belong here. A. Wrefordii appears to be nearest related to A. abyssinica Lam., but differs in having different leaves and a much taller much more branched inflorescence. In size, leaves and tall much-branched inflorescence A. Tweediae Christian is a near ally, but A. Tweediae differs in having much longer laxer racemes of very differently shaped flowers. A. calidophila Reynolds, from between Moyale and Mega in Southern Ethiopia, is another near ally in shape and size of flowers, but that species is shrubby and has longer racemes, and longer deeply channelled recurved leaves. A. eru Berger is also closely related to A. Wrefordii in pedicels, bracts, and short cylindric-clavate flowers, but the former is of shrubby growth, and has longer narrower spotted leaves and a lower inflorescence with longer racemes. (‘opeanyy “UdJa[GUIYT “W'S /ojoy_d) “Wd Cy] QUusloy] “OSGT “ATBNLU f* ISTZ “Y ‘elpoeM 7, “q.10y “q[No ‘OJOIOT LeOU Xo quel g—~é “Ivy : - OOS WSO “CEGL “10G0}0Q YAS] ‘pooes puv qmuay url ‘vpuesy ‘efoureaey ‘oyo10;y FO “HON septa G ‘ednyy ye yeqrqey yeanyeu ur yuelq— | ‘splousoy piofory 9077 “ITAXX WLW 1d eee) As | *T (DIT th “OLY =e EMER (Photo lowers 1/1 from bud to post-pollination stage. Ire —~Portion of inflorescence, A. Embleton, Kitale.) Ss. Il. 5) Aloe Wrefordii Reynolds. NXVILT. PLATE Fria. Three New Species and One New Variety of Aloe from 143 Karamoja, Uganda. A. Wrefordi occurs as solitary plants, with little or no stem, and has a tall many-branched inflorescence bearing capitate racemes of cylindric- clavate flowers 22 mm. long. Nothing approaching it in floral characters was seen throughout Karamoja. Description: Plants solitary, not forming groups, acaulescent or with very short stem. Leaves about 24, densely rosulate, lanceolate-attenuate, suberectly spreading and recurved in upper quarter, about 60 em. long, 15 cm. broad at base, 15 mm. thick, the acute apex a reddish-brown spine; wpper surface dull grey-green, sometimes with reddish-brown tinge, mostly without spots or markings, very obscurely lineate, flat low down, canaliculate upwards; lower surface convex, pale grey-green, without markings; margins sinuate-dentate, armed with reddish-brown deltoid pungent teeth 4 mm. long, 10—15 mm. distant, forward-uncinnate, the inter- spaces the colour of the leaf. Sap dries yellow. Inflorescence a compactly branched panicle about 120 cm. high, sometimes 2—3 consecutively. Peduncle basally plano-convex and 20—25 mm. broad, branched well below the middle, with up to 16 ascending branches, the lowest 1—2 branches with 2—5 branchlets producing a total of up to 25 racemes. Racemes sub-capitate, the pedicellate portion of terminal racemes 4—5 cm. long, 6 cm. diam., laterals slightly shorter, densely flowered, the apex a small tuft of dried bracts, youngest buds green-tipped, sub- erect, older buds spreading, open flowers nutant to subpendulous. Bracts ovate-acute, 9 mm. long, 4 mm. broad, very thin, white, scarious, about 5-nerved. Pedicels lowest 16 mm. long, slightly shorter upwards, 20 mm. in the fruit. Perianth scarlet or orange, cylindric-clavate, 22 mm. long, obtusely tapering at base to the pedicel articulation, shortly stipitate,cylindric and 5—6 mm. diam. across the ovary, thence narrowly clavate, the mouth slightly up-turned; outer segments free for 13 mm. (tube about 8 mm.), 3-nerved, apices subacute; inner segments broader than the outer, with broad white border and 3 crowded nerves forming a keel, the apices more obtuse and more spreading than the outer. Filaments pale lemon, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn exserted 3—4 mm. Stigma at length exserted 5 mm. Ovary olive, 6 mm. long, 3 mm. diam., obtuse at apex. 144 The Journal of South African Botany. ACKNOWLEDGMENTS. _ To Colonel and Mrs. D. R. Tweedie I am particularly grateful for providing transport, camping equipment and provisions, and for con- ducting me for over 900 miles through Karamoja, and to Mrs. Tweedie for preparing herbarium material, sending photographs and spirit material, for many plants sent by air to Johannesburg, and for much more. To Mr. T. H. E. Jackson for bringing in flowering plants of A. Schwein- furthii var. labworana from two localities in Labwor, and for much other assistance. To Mr. Wreford Smith for searching for Aloes throughout Karamoja for several years, and for sending valuable material to me through Mrs. Tweedie. ‘To Mr. John T. Wilson, Ecologist, Moroto, for plants, photographs, much data concerning the Aloes of Karamoja, and especially for leading me 3,000 feet up Mt. Moroto to the locality of a new species. I am indebted to the South African Council for Scientific and Indus- trial Research for a travelling grant which enabled me to visit Karamoja to investigate the Aloes. BOOK REVIEW. METHODS OF SURVEYING AND MEASURING VEGETATION: By Dorothy Brown, with a chapter on the Theory of Sampling by G. M. Jolly. Commonwealth Bureau of Pastures and Field Crops, Bulletin 42, 1954. Pp. xv and 223. 29 half-tone plates, 22 figs., 46 tables. 35s. When a survey of vegetation is carried out it is important that the results of such work should be concise and accurate. Accounts drawn up from casual observation are usually subjective, unreliable and subject to personal bias. An objective approach is required, involving counting and measuring. In carrying out a quantitative vegetational survey the selec- tion of a suitable method often presents a problem and the work might easily become a study of various methods in a particular type of vegeta- tion rather than a study of the vegetation itself. This book should help considerably in selecting an appropriate method for a certain type of vegetation. Miss Brown has very successfully brought together in one volume the many and varied techniques which have been applied in quantitative ecological studies in all parts of the world. There is an extensive review of the literature so that a research worker may refer to the original publica- tions if necessary. The various methods are described but there is no critical assessment of their respective merits. Most of the methods described apply to grassland and the book there- fore should be of special appeal to the Pasture Research Officer and Grass- land Ecologist. Although many of the basic principles of the methods could be used or adapted to the study of forest or other vegetation the book has not been written for the worker engaged in the study of these other types of vegetation. The title is well chosen because the book covers little more than “Methods of Surveying and Measuring Vegetation’”’, and this is done very adequately. It is not a handbook of statistical methods and the analysis of results is not dealt with. It shows how to obtain the quantitative results but for their further treatment to assess their significance one must turn to a book on statistical analysis. The publication of a companion volume to supplement the information in this book would fill a long felt want. H. B. Rycrort. 145 JOURNAL OF SOUTH AFRICAN BOTANY VOL. XXII. Published: 21st Novemper, 1956 THE FLORAL STRUCTURE OF NAGELOCARPUS SHRRATUS (THUNB.) BULLOCK. By Marcaret R. Levyns Nagelocarpus serratus Bullock. (Kew Bull. p. 533, 1954), known hitherto as Lagenocarpus imbricatus Klotzsch, shows features of floral structure in the fully developed flower which at first sight make it appear unique in the plant kingdom. Strangely enough J. F. Klotzsch who founded the genus Lagenocarpus in 1838, failed to notice the outstanding peculiarities in the flower to which attention was drawn in the Flora Capensis many years later by that acute observer, the late Dr. N. E. Brown. Two peculiarities noted by Brown were (1) an ovary half inferior to the corolla but lying wholly above the calyx and (2) ovules pendulous from near the apex of a pillar-like central placenta, the placenta itself being quite free from the side walls of the ovary. A study of the development of the floral parts in young flower buds was made in order to see if there might be a simple explanation of the unusual features shown in the open flower. The buds were fixed in formalin-acetic-alcohol, embedded in paraffin wax and sectioned both transversely and longitudinally. Examination of these sections soon revealed that the peculiarities mentioned above were due to the breaking down of delicate spongy tissue in the course of development. The general features of the open flower are shown in Figs. 1—3. Fig. 1 is an external view of a single flower which has just opened. Fig. 2 shows the same flower after removal of the calyx and corolla, the line of attachment of the calyx being at the base (A) and that of the corolla just below the anthers (B). Fig. 3 is the gynaeceum showing the line of attachment of the corolla rather more than half way up the ovary. Fig. 4 is a diagram of a longitudinal section of a flower just about to open. In order to show the attachment of two of the four ovules, the ovary is 147 148 The Journal of South African Botany. depicted slightly to one side of the median plane. For the sake of clarity the anthers which are large and fill the space between corolla and gynae- ceum, are not shown in the diagram. At this stage the flower has a normal structure with both calyx and corolla arising at the same level, about half way up the ovary. It is true that a half-inferior ovary is not GaP. Bea ars pa "1 \ 1 ) t -) . q pi---B | Fig. 1.—Side view of a single flower. x 10. Fig. 2.—The same flower with the calyx and corolla removed. x 10. Fig: 3—The gynaeceum with all the other floral parts removed. x 10. found in other South African members of the Ericaceae but such ovaries are not unusual in the plant kingdom, whereas the ovary described in the Flora Capensis is unique. Below the level of insertion of the calyx and separating it from the wall of the ovary, is a zone of loose spongy tissue which breaks down as soon as the flower opens. By the time the flower is fully open the only indication that such tissue existed, lies in the slight roughness inside the lower part of the calyx and on the ovary. The calyx then has the appearance accurately described by N. E. Brown. Fig. 4.—Diagram of a longitudinal section through a flower which has just opened. The spongy tissue which breaks down is shaded more lightly than the rest. Fig. 5.—Diagram of a cross section through a flower at the same stage of develop- ment as that shown in Fig. 4. The line X - - - - - - - X indicates the level at which the section was cut. The lightly shaded parts eventually break down. The Floral Structure of Nagelocarpus Serratus 149 (Thunb.) Bullock. A diagram of a transverse section taken through a similar flower at the level X -------- X in Fig. 4, is shown in Fig. 5. It depicts a normal 2-chambered ovary with two ovules in each chamber. The central region of the dividing wall is firm and well developed, but the parts connecting it with the outer ovary wall are delicate and before the ovary is mature, break down leaving a pillar-like central column with the pendulous ovules arising near the top. Here too the unusual feature of the mature ovary is due to the breakdown of delicate tissue in the course of de- velopment. The chief difference in floral structure between Nagelocarpus and the genera Salaxis and Coccosperma lies in the half inferior position of the ovary, a feature which is probably sufficient to justify the retention of generic rank for this member of the Ericaceae. THE IDENTITY OF ALOH ABYSSINICA LAMARCK, A. ELEGANS TODARO AND A. CAMPERI SCHWEINFURTH. By G. W. REYNOLDs. (With Plates XXIX—XXXYV.) Aloe abyssinica Lamarck Aloe abyssinica Lamarck was described in Encyclopédie Méthodique 1: 86 (1783) Paris. The description includes, (from the French): ‘Leaves 24 feet long, 4 ins. broad at base, thick, with red marginal spines, convex on the outside, channelled on the inner face. Peduncle simple, up to 3 feet or more, with yellowish-green flowers disposed in a long somewhat lax spike. Flowers deeply cut into 6 parts, stamens protruding a little from the corolla and grouped together against one side.” Lamarck states, “This plant was brought back from Africa by Bruce and is growing in the King’s Garden.” (Paris). The intrepid traveller James Bruce (b. 1730, d. 1794) landed at Massawa, the Red Sea Port of Eritrea, on 19 September 1769. He travelled southwards through Arkiko, and crossed the Taranta Mountain to Dixan. After crossing the Belesa River—a tributary of the Mareb— he reached Adowa on 5 December. Bruce then went to Axum and continued south-westwards to Gondar. After discovering the source of the Blue Nile, he left Gondar on 26 September 1771 and journeyed west- wards to Sennar in the Sudan, then up to Shendi, and reached Cairo on 27 December 1772—three years and three months after landing at Massawa. Bruce had spent two years between Massawa and Gondar, which is country very rich in Aloe, and he took fifteen months from Gondar to Cairo. No Aloe has hitherto been recorded from the Sudan part of his journeys. It seems therefore, that Bruce possibly collected his plant somewhere between Massawa and Gondar, between September 1769 and September 1771. Bruce might have sent the plant to Paris from somewhere in Eritrea or Tigre, but if he literally “brought back” the plant from Africa, as Lamarck clearly states, then he must have carried it with him for at least more than a year—which seems unlikely—unless he picked up a cultivated plant in Cairo or Alexandria, which now seems a possibility. During my own recent and extensive travels in Eritrea and Tigre (Northern Ethiopia), I traversed several areas that Bruce had travelled 151 152 The Journal of South African Botany. through, but I sought in vain for an Aloe fitting Lamarck’s description of A. abyssinica. I had borne in mind Salm Dyck’s figure of “A. abys- sinica’”’ in his Monogr. gen. Aloe sect. 18, fig. 1 (1842), and also Berger’s figures and account in Engler Pflanzenr. Liliac.-Aloin 237 (1908). These accounts were of two different species, and neither agreed with Lamarck’s description. Two species that I found in considerable numbers in Eritrea and Tigre were A. Camperi Schweinfurth (=A. eru Berger—see below), and A. elegans Todaro (=A. aethiopica (Schweinf.) Berger—see below), with A. trichosantha Berger, and A. macrocarpa Tod. in lesser numbers. Bruce must surely have seen these species, but not one of them fits Lamarck’s description. A. Camperi agrees in channelled recurved leaves, but the inflorescence is a many-branched panicle and never bears racemes with “yellow flowers arranged in a long somewhat lax spike.” On reaching Addis Ababa I received a photograph from Professor J. Millot of the specimen named A. abyssinica by Lamarck himself in the Lamarck Herbarium, housed in the Natural History Museum, Paris. This photograph is reproduced herein (Plate X XIX.) and from it it will be seen that A. abyssinica has unusually large sterile- and flori- ferous-bracts, apical buds crowded, suberect, and almost hidden by large imbricate bracts, the open flowers being much more laxly disposed. Professor Millot states that the longest flower is 32 mm. (dry), and the longest pedicel on right is 45 mm. long. I am assured by Professor Millot that the note in the lower left corner of the sheet is in Lamarck’s own handwriting, which eliminates any possibility of the material having been wrongly identified by someone else. The note reads: “Fleurs de l’Aloes d’Abyssinie. Elles sont jaunes. La hampe qui les porte est quel- quefois 4 3 rameaux.”’ This authentic material proves conclusively that Salm Dyck was entirely mistaken in his concept of A. abyssinica. His figure is of a form of A. Camperi Schweinf. and differs considerably from the Lamarck specimen. Salm Dyck (1842) stated: “The distinguished Bruce in about the year 1777, gave the seeds of this plant (which seeds had been collected in Abyssinia) to the Royal Garden in Paris where it has been grown since that time.” Against this, Lamarck stated, “This plant was brought back . from Africa by Bruce.” Salm Dyck’s figure was the first published of “A. abyssinica” and it has misled all subsequent workers up to the present day, including Berger (1908). Berger’s figure of “A. abyssinica Lam.” depicts a form of A. elegans Tod. (see below), and not one of Berger’s citations and synonyms can now be included under A. abyssinica Lam. Heide MES ATS ‘i ee SY Sa ee plone de Cates Ia ey/inie : | PLATE XXIX. Aloe abyssinica Lam. Photograph of the specimen in the Lamarck Herbarium, Paris. Note in lower left corner, in Lamarck’s handwriting, reads; ““Fleurs de | Aloes d’Abyssinie. Elles sont jaunes. La hampe qui les porte est: quelquefois a 3 rameaux.” — By courtesy Professor J. Millot. PLATE XXX. A. elegans Todaro. From Hort. Bot. Panorm. 2: 25, tab. 29 (1882). —By courtesy Mr. J. P. M. Brenan, Kew. ee i Ey r) » > qs ~ a hee te ee 1 4 ‘%. Loe ts Ne, whee we hha! ae 4 OOP SS GS aN ie PLATE XXXII. A. elegans Tod. (=A. aethiopica (Schweinf.) Berger) Photograph of the type plant of A, aethiopica (Schweinf.) Berger, cult. La Mortola, fl. 20 Feb, 1899, The Identity of Aloe Abyssinica Lamarck, A. Elegans 153 Todaro and A. Camperi Schweinfurth. A description lacking in several characters, and based on Lamarck’s description and on obvious characters of the Lamarck specimen would read as follows: Aloe abyssinica Lam. in Hncycl. Méthod. 1: 86 (1783).—Non Salm Dyck, non Roem. et Schult., non Kunth, non Baker, non Schweinfurth, non Hooker, non Berger. Habit of growth unknown. Leaves in a rather open rosette, 75 cm. long, 10 em. broad at base, thick, clear green; wpper surface channelled; lower surface convex; whether unicoloured or spotted not known; margins armed with red teeth, size and distance apart unknown; sap orange- yellow, resinous, “‘quickly thickening into a dry reddish matter.” Inflorescence 1 met. or more tall, simple or sometimes three-branched, with flowers disposed in a long somewhat lax spike; raceme cylindric- conical, 20 cm. long; buds suberect, crowded and partly obscured by large bracts, open flowers laxer, nutant to subpendulous. Bracts evi- dently large and fleshy, narrowly deltoid, about 20 mm. long. Pedicels up to 45 mm. long. Perianth yellow, 32 mm. or more long, apparently cylindric, “deeply cut into 6 parts” (outer segments free to the middle?). Stamens shortly exserted and grouped along one side of the perianth. (Plate X XIX.) Until such time as plants agreeing with Lamarck’s description and material can be re-discovered, it seems that A. abyssinica Lam. will remain a Species non satis cognita. A. elegans Todaro A. elegans Tod. was described and beautifully figured in Hort. Bot. Panorm. 2: 25, t. 29 (1882), but it does not appear to have been recognised again since then. Berger in Engler Pflanzenr. Liliac.-Aloin. 237 (1908) not only wrongly identified A. abyssinica Lam. but he also wrongly placed A. elegans Tod. in synonymy under A. abyssinica Lam. Todaro’s t. 29 is reproduced herein (Plate XXX.), and from this figure it will be seen that one of the most striking characters of A. elegans is the racemes with the young buds spreading horizontally to slightly downwards. Schweinfurth also did not recognise A. elegans, and he described this species as A. vera var. aethiopica in Bull. Herb. Boiss. 2: app. 2, 61 (1894). Berger in Engler’s Bot. Jahrb. 36: 60 (1905), and in Pflanzenr. 243, Fig. 94, (1908) described A. aethiopica (Schweinf.) Berger from a plant that flowered in La Mortola in February 1899. Material from Berger’s type plant is in Kew, together with a photograph of the flowering plant. (Plate XX X1). 154 The Journal of South African Botany. This material is the type of Bot. Mag. t. 7667 (1899) and that plate was prepared from Berger’s photograph—but the species was wrongly named A. Schweinfurthi Bak. by Hooker. This photograph was also used for the figure accompanying the text in Gard. Chron. 23: ser. 3, 197, Fig. 76 (1898) but the species was also wrongly named A. Schweinfurthw Bak. I have photographed and collected the true A. Schweinfurthii Bak. at the type locality, Mt. Bangenze, Equatoria Province, Sudan, near the Congo border—see this Journal 20: 165 (1954). That species differs very considerably from Berger’s A. aethiopica. Berger gives no locality of origin for his plant, but if his photograph is compared with Todaro’s ¢. 29 it will be seen that the plants are very similar in rosettes and kind of inflorescence while in both the buds are spreading horizontally to slightly downwards—an unusual character in Aloe. In my view Berger’s species is clearly conspecific with A. elegans Tod., and the latter being the earlier name has priority. Todaro states that his plants were raised from seeds collected by Schimper in Abyssinia, (which then included the present Eritrea‘and Ethiopia). A. elegans Tod. occurs in considerable numbers over a wide area in Eritrea, and Tigre, Northern Ethiopia. It is abundant near Asmara (7500 ft); for 80 km. along the road towards Keren; southwards to Adi Ugri (6,200 ft.), Adi Kwale (6,500 ft.); not seen in the Mareb valley—which is the Eritrea-Tigre border; southwards to Adowa (6,100 ft.), Axum (6,300 ft.) and east of Adowa. East of Asmara the species abounds south of Nefasit in the broad Piana Dalla valley (5,300 ft.), thence southwards to Decamere (6,400 ft.), Saganeiti (6,900 ft.), Addingfom hill near the Gorge of Gua (6,700 ft.), Adi Caieh (7,700 ft.), on the Kohaito Plateau with A. Camperi Schweinf. and A. percrassa Tod. (8,200 ft.), and near Senafe (7,700 ft.). Im Tigre, A. elegans occurs abundantly to Adigrat (7,700 ft.), southwards to Mekele (6,700 ft.) and to the foot of Amba Alaji (6,600 ft.), 55 km. south of Mekele and 355 km. south of Asmara, which is the furthest south that I saw it. The species can now be more fully described from personal observa- tions in Eritrea and Northern Ethiopia. A. elegans Todaro in Hort. Bot. Panorm. 2: 25, t. 29 (1882). A. Schweinfurthii Hook. f. in Bot. Mag. t. 7667 (1899) non Baker; Bak. ex Berger in Gard. Chron. 23: ser 3, 197, Fig. 76 (1898), non Baker in Journ. Iinn. Soc. 18: 175 (1880). A. aethiopica (Schweinf.) Berger in Engler Bot. Jahrb. 36: 60 (1905), in Pflanzenr. Liliac.-Aloin. 243 (1908). The Identity of Aloe Abyssinica Lamarck, A. Elegans 155 Todaro and A. Camperi Schweinfurth. A. vera var. aethiopica Schweinf. in Bull. Herb. Boiss. 2: app. 2, 61 (1894); Baker in Th. Dyer Fl. Trop. Afr. 7: 465 (1898). A. abyssinica var. Peacocki Baker in Journ. Linn. Soc. 18: 175 (1880), in Bot. Mag. t. 6620 (1882)—fide Todaro; Baker in Th. Dyer Fl. Trop. Afr. 7: 468 (1898). Descrretion. Plant usually solitary, rarely dividing into 2—3 rosettes. Stem none, or sometimes up to 20—30 em. long in old specimens, decumbent. Leaves 16—20, densely rosulate, (rosettes compact, averaging 60 cm. diam.) lanceolate-attenuate, 15—18 cm. broad at base, about 60 cm. long, 25 mm. thick, gradually narrowing to an obtuse shortly-dentate apex; upper surface grey-green, flat low down, slightly canaliculate upwards, without spots or markings; lower surface convex, similar in colour to upper surface; margins with reddish edge, armed with brownish- red teeth that are pungent, deltoid, 3—4 mm. long, 15—25 mm. distant, larger low down, smaller upwards. Sap dries reddish-brown. Inflorescence a many-branched panicle | met. high, 2—4 simul- taneously. Peduncle basally plano-convex and 25 mm. broad, branched from about the middle with about 8 arcuate-ascending branches, the lowest sometimes with 1—2 branchlets. Racemes conical when young becoming corymbose when flowers open, broadly cylindric, subcapitate, the terminal 8 cm. long 7 cm. diam. lateral a little shorter, densely flowered, youngest buds spreading hori- zontally to slightly downwards, open flowers nutant to subpendulous. Bracts narrowly ovate-acuminate, about 8 mm. long, 2—3 mm. broad, thin, subscarious, 3 Pedicels averaging 15 mm. long, compactly arcuate-ascending, nutant at apex. Perianth yellow, orange or scarlet, cylindric, very slightly clavate, 25—30 mm. long, basally flat to very slightly rounded and shortly stipitate, 6 mm. diam. across the ovary, thence very slightly narrowed and trigonously indented, enlarging slightly from the middle to a wide open trigonous mouth; outer segments free for half their length, thinner at the margins, obscurely 3-nerved, the apices subacute, slightly spreading; mner segments free but dorsally adnate to the outer to the middle, broader than the outer, with 3 crowded nerves forming a keel the colour of the perianth, and with their apices more obtuse and more spreading than the outer. 5-nerved. 156 The Journal of South African Botany. y Filaments pale lemon, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn exserted 3 mm. Stigma at length exserted 4 mm. Ovary light green, 5 mm. long, 3 mm. diam. (Plates XX X—XX XIII) Materia: Cult. La Mortola, fl. 20 Feb. 1899! (type of Bot. Mag. t. 7667) (K). Eritrea: Between Adi Ugri and Adi Kwale, 6,000 ft., c. 15° 22’N., 38° 50’ E., 18 Apr. 1956, Reynolds 8092 (PRE, EA, K). Northern Ethiopia, Tigre: 55 km. 8. of Mekele at foot of Amba Alaji, 6,600 ft., c. 13° 08’ N., 39° 29’ E., 27 Apr. 1956, yellow flowers, Reynolds 8110 (PRE, K); red flowers, Reynolds 8111, (PRE). Although occasional plants were found in flower in April, the main flowering time is August—September, after the rains. With buds spreading horizontally to slightly downwards, A. elegans is allied to A. Chabaudii Schonl. (Nyasaland, Northern and Southern Rhodesia, Transvaal), and A. mzimbana Christian (Northern Rhodesia, Nyasaland). A. Camperi Schweinfurth A. Camperi Schweinfurth was described in Bull. Herb. Bowss. 2: app. 2, 67 (1894) from the Great Valley above Ghinda, Eritrea, 1,400 met., Schweinfurth 514a, coll. 1891. Schweinfurth also included: Asmara, 2.400 met. n. 605, coll. 1891: Arbaschiko, 1,600 met., n. 668, coll. 1891; Gheleb, 1,700 met., n. 1074, 1234, coll. 1891; Acrur, 1,900 met., 1342, 1306, coll. 1892, all in HKritrea. Schweinfurth stated: ‘‘Grows at all (above) mentioned places in association with A. abyssinica Lam., but more singly and not in com- munities. It cannot be distinguished by its flowers from A. abyssinica Lam., but differs in its broader shorter leaves not so recurved at the apex...” Schweinfurth had obviously followed Salm Dyck’s description and > which, as has been shown above, was entirely incorrect. He and others have endeavoured to make two species from forms of one shrubby species of Aloe found in considerable numbers from the valley above Ghinda to Nefasit, up the mountains to Asmara, and elsewhere in Eritrea. Another species now becomes involved, namely A. erw Berger. Berger had wrongly identified A. abyssinica Lam.,—his concept being a form of A. elegans Tod. He also wrongly placed A. Camperi Schweinf. in synonymy under his concept of A. abyssinica Lam. (i.e. A. elegans Tod.) which added still more to the confusion. figure of “A. abyssinica Lam.,’ PLATE XXXII. A. elegans Tod. Plant fl. 26 April 1956, at northern foot of Amba Alaji, 55 km. south of Mekele, Tigre, Northern Ethiopia. Alt. 2,000 met. Height 90 cm. IDE, - Dann ee ‘@ PLATE XXXIII. A. elegans Tod. Fic. 1.—Flowers 1/1 from a plant at northern foot of Amba Alaji, 55 km. south of Mekele, Tigre, Ethiopia. Fic, 2,—Flowers 1/1 from a plant between Adi Kwale and Adi Ugri, Eritrea. a (FG ie, 2s PLATE XXXIV. A. Campert Schweinfurth (=A. erw Berger). Plant in the great valley above Ghinda, near Nefasit, Eritrea. c. 15° 23’ N., 39° 04’ E. Alt. 1,700 met., fl. 13 April 1956. Hie, Il, \ Fic. 2. PLATE XXXV. 4. Camperi Schweinfurth (=A. erw Berger) Fie, 1.—Flowers 1/1 from a plant near Nefasit, Eritrea. Fic. 2.—Flowers 1/1 from a plant 6 km. east of Asmara, Hritrea. The Identity of Aloe Abyssinica Lamarck. A. Elegans 157 Todaro and A. Camperi Schweinfurth. Berger founded his A. eru on Schweinfurth’s n. 514 from the Great Valley above Ghinda, and he also cited the identical material from Asmara, Gheleb and Acrur that Schweinfurth had cited for his A. Camperi. In April 1956, when travelling in Eritrea, I specially visited as many of Schweinfurth’s localities as I could reach. I found the shrubby species hitherto so well known as A. erw Berger, in considerable numbers in the valley between Ghinda and Nefasit (the first place where travellers coming up from the coast would see it) and especially on the mountains up to Asmara at 7,600 ft. It is now clear that A. abyssinica Schweinf. non Lam., A. Camperi Schweinf., and A. eru Berger are merely forms of one species from the same locality. A.Camperi being the earlier name has priority over A. erw which should now be reduced to synonymy. A. Camperi occurs in considerable numbers in many parts of Eritrea, from Meschillit Pass north-east of Keren in the north to the Ethiopian border in the south. I found it in the broad Piana Dalla valley south of Nefasit (5,300 ft.), near Decamere (6,400 ft.), Saganeiti (6,900 ft.), Addingfom Hill near the Gua Gorge (6,800 ft.), Adi Caieh (7,700 ft.), on the Kohaito Plateau with A. percrassa Tod., and A. elegans Tod. (8,300 ft.), and near Senafe (7,700 ft.). In Ethiopia, A. Camperi was found in numbers at the northern foot of Amba Alaji (6,700 ft.), near Alamata (4,800 ft.) and Cobbo (4,700 ft.), in the Rift Valley, Tigre Province. In Wollo Province, the species occurs abundantly near Waldia (6,000 ft.), and as far south as near Dessie. A. Camperi is a variable species, weaker forms occurring at lower drier altitudes (4,600 ft.), and more robust forms from 6,000—8,300 feet. Stems vary in length. Leaves vary in length and width and occur with or without spots. Racemes also vary from shorter and denser to longer and laxer. Flowers vary in length and shape, but are invariably clavate, and average 20 mm. in length. The following description is based on personal observations in the valley above Ghinda, near Asmara, and elsewhere in Eritrea: A. Camperi Schweinfurth in Bull. Herb. Boiss. 2: app. 2, 67 (1894). A. abyssinica (non Lamarck) Salm Dyck Cat. rais. 25 (1817), Monogr. gen. Al. ser. 18, fig. 1, (1842); Roem. et Schult. Syst. veg. 7: 695 (1829); Kunth Enum. pl. 521 (1843); Baker in Journ. Linn. Soc. 18: 174 (1880), in Th. Dyer Fl. Trop. Afr. 7: 467 (1898); Schwein- furth in Bull. Herb. Boiss 2: app. 2, 66 et 110 (1894); Hooker in Bot. Mag. t. 7712 (1900). (Note: A. abyssinica Berger in Engler Pflanzenr. Liliac.-Aloin 237 (1908) excl. part syn., is a form of A. elegans Tod.). 158 The Journal of South African Botany. A. eru Berger in Engler Pfianzenr. Liliac.-Aloin 249 (1908). Description. Plant shrubby, branching from base and sometimes forming groups 1—2 met. across. Stem usually 50 cm. long, 6 cm. diam., sometimes up to 1 met. long, 9—10 em. diam., erect, divergent or decumbent, the apical 10—20 cm. densely foliate. Leaves 12—16, rosulate, averaging 50—60 cm. long, 8—12 cm. broad at base, tapering to the apex, basally sheathing, recurved and forming open rosettes; wpper surface dark green, canaliculate, unspotted or sometimes with a few dull white lenticular spots; lower surface convex, deep green, with or without numerous elongated lenticular white spots; margins with reddish edge, armed with deltoid pungent brownish-red teeth averaging 3—5 mm. long, 10—20 mm. distant, more crowded near base, more distant upwards. Inflorescence a branched panicle 70—100 cm. high, usually 2—3 simultaneously. Peduncle basally plano-convex and 15—20 mm. broad, brown with a bloom, 6—8-branched from about the middle, lowest branches sometimes with 1—2 branchlets. Racemes densely flowered, the pedicellate portion 6—9 cm. long, 6—7 cm. diam., youngest buds suberect, scarlet, sometimes greyish- green tipped, older buds spreading, open flowers turning orange to yellow, subpendulous. Bracts very small, deltoid, 2 mm. long, 2 mm. broad at base, dirty white, 5-nerved. Pedicels 12—18 mm. long, 14 mm. the average. Perianth orange to yellow, cylindric-clavate, averaging 20—22 mm. long, basally obconic and shortly stipitate, 5 mm. diam. across the ovary, clavate upwards, the mouth wide open. trigonal; owter segments free for one-third their length (about 7 mm.), obscurely 3-nerved, apices subacute, slightly spreading; inner segments free but dorsally adnate to the outer for two-thirds their length, thinner at the margins, the apices more obtuse and more spreading than the outer. Filaments lemon, the 3 inner narrower and lengthening before the 3 outer with their anthers in turn exserted 2—3 mm. Stigma at length exserted 3—4 mm. Ovary light green, averaging 5 mm. long, 2 mm. diam. (Plates XX XIV—XX XV) Native Name: £rreh in Eritrea, [ret in Ethiopia. MatTeErRiAu. Eritrea: Gheleb, 1,900 met., 3 April 1891, Schweinfurth 1234! 1430! 1796! (K); near Acrur, 1,900 met., 1—6 April 1892, Schwein- The Identity of Aloe Abyssinica Lamarck, A. Elegans 159 Todaro and A. Camperi Schweinfurth. furth 1350! (K), 25—30 April 1892, Schweinfurth 1797! (K). 6 km. east of Asmara, 2,250 met., 12 April 1956, Reynolds 8036! (PRE, K). In valley above Ghinda near Nefasit, 1,700 met., c. 15° 23’ N., 39° 04’ E., 13 April 1956, Reynolds 8051! (PRE, K, EA). On the Kohaito Plateau 13 km. south of Adi Caieh, 2,730 met., 15 April 1956, Reynolds 8071! (PRE, K). Ethiopia, Wollo Prov.: 20 km. south of Waldia, 100 km. north of Dessie, 1,710 met., 25 April 1956, Reynolds 8105! (PRE, K, EA). ACKNOWLEDGMENTS. I am indebted to: The South African Council for Scientific and Industrial Research for a travelling grant that enabled me to investigate the Aloes of Eritrea and Ethiopia. Professor J. Millot for data and a photograph of the Lamarck speci- men of A. abyssinica Lam. in the Lamarck Herbarium, Paris. The Director of Kew Gardens for sending herbarium specimens to Pretoria. Mr. E. Milne-Redhead and Mr. J. P. M. Brenan of Kew for photo- graphs, and much other assistance. MARINE ALGAE OF INHACA ISLAND AND OF THE INHACA PENINSULA. I. By Wo. Epwyn Isaac. (Department of Botany, Cape Town University) (With Plates XX XVI—XLIII) INTRODUCTION The algae listed in this paper were collected at the end of June and in July 1954 and in July 1956 while the author stayed at the Estacao de Biologia Maritima on the island of Inhaca in Delagoa Bay. Algae were collected on the coasts of the island, which lies about twenty miles east of Lourengo Marques, and also at Santa Maria and Ponta Abril on the east coast of the Inhaca Peninsula—a peninsula of the mainland of Portuguese East Africa which extends northwards on the eastern side of Delagoa Bay (Fig. 1). This peninsula is separated from the southern end of Inhaca Island by a narrow strait which at its narrowest at low water spring tides, is about a hundred yards across (Kalk, 1954). The author’s chief immediate interest in the seaweeds of Portuguese Kast Africa is to obtain data relating to the distribution of South African seaweeds to the north of the Union of South Africa on the east coast. There are no published lists or monographs of the seaweeds of Portuguese Kast Africa and so these records should prove of general interest to students of seaweed distribution. Most of the species dealt with in this paper occur also on the Union coasts. A general account of the island and a preliminary account of its marine animal communities has been published by Kalk (1954) and her paper has been much used in writing the general comments which follow. Inhaca Island and Peninsula provide a wide range of marine habitats —rocky coast, sand, mud flats, old coral reef material scattered over sand, and patches of living coral reef off the West Coast of the island. The patches of reef are the most southerly on the east coast of Africa. They are rarely completely uncovered by the sea. The most varied and densest seaweed growth occurs on rocky coasts although individuals of certain species, such as Padina commersonw and Udotea orientalis, attain their maximum luxuriance attached in sand in quiet waters. In places on rocky coasts the algae form mixed communities, the plants sometimes forming turfs in the sense of a dense growth of small plants, the upper 161 162 The Journal of South African Botany. ILHA DA INHACA. Pir IGHTHOUSE ILHA DO , wm e4. ' 6U=Ses (8 [CABO INHAC LOURE NGO ~ MARQUES a N SACO) DA INHACA~Y ¥Z ESTACAOYs BIOLOGIAx MARITIMA: iy Z| Zz, 77 CLIFFS SCALE f MANGROVE SWAMPS KXX CORAL REEF ROCK SHORELINE W, FRESH-WATER SWAMPS 7, IIDAS FEATS — SAND SHORELINE Fig. 1.—Map of Inhaca Island and of the Inhaca Peninsula. Based on amended version by H. B. S. Cooke of map prepared by the Missio Hidrografica da Colonia de Mogambi Marine Algae of Inhaca Island and of the 163 Inhaca Peninsula I. surface of the community being of a more or less even height. Locally, certain seaweeds form more or less pure communities of greater or lesser extent. At the lowest intertidal levels Sargassum forms such communi- ties; on rock platforms communities dominated by Padina commersonii occur; and on vertical rock faces at low intertidal levels very local dense growths of Chamaedoris delphinii and Pseudocodium de-vriesei may be observed. There is reason to believe that there is a more luxuriant growth of marine algae during the summer months. Marine angiosperms are a prominent feature of the region. Extensive mangrove swamps occur, especially at the northern end of the island and in the deep bay or Saco at the southern end. The four dominant mangrove species are Avicennia officinalis Linn., Ceriops candolleana Arn., Bruguiera cylindrica Blum, and Rhizophora mucronata Lam. (Mogg, 1956). Avicennia is the most promi- nent plant along the shoreward edge where its presence is made very evident by the numerous branch roots growing upright through the mud. Ceriops and Bruguiera are more characteristic of the muddy interior while Rhizophora, spectacular on account of its prominent stilt roots, occurs along the banks of the channels running into the mangroves. These mangroves are not estuarine in the sense of being at the mouths of rivers as no rivers flow out to sea from Inhaca. As in true estuaries, however, marked salinity variations occur in these mangroves due to the drainage of water from high lands (especially after rain) and to evaporation following spring tides (Kalk, 1954). A limited number of algae occur in the mangroves including Bostrychia binderi. To seaward of the southern mangrove an abundance of Acetabularia calyculus occurs attached to the empty half shells of the bivalve Loriphes clausus which in places cover a sand and mud substratum in shallow water. There is also on the island coasts an abundance of “gea grasses”: Cymodocea ciliata (Forsk.) Ehrenb., C. isoetifolia Aschers., O. rotundata Aschers and Schweinf., C. serrulata (R. Br.) Aschers and Agnus, Diplanthera uninervis (Forsk.) F. N. Williams, Halophila ovalis (R. Br.) Hook, Halophila sp., and Thalassia hemprichii (Ehrenb.) Aschers. (Moss, 1937, Cohen, 1939, Mogg, 1956). Diplanthera uninervis is uncovered at low water neaps or is left by the sea in very shallow water. Cymodocea rotundata and C. serrulata are uncovered or left in shallow water at low water of springs. Cymodocea ciliata is mostly permanently submerged but it is sometimes uncovered during spring tides at the lowest intertidal levels as at the north east end of the island. Typically all these species, as is reported to be the case with “sea grasses” in general, occur rooted in sand in rela- tively quiet waters. On the east coast of Inhaca island Cymodocea ciliata oceurs attached to rock (including rock pools) in situations where it is subjected to a fair amount of buffeting by the waves but it is more 164 The Journal of South African Botany. abundant in the waters off the west coast of the island where it grows rooted in sand. The extreme tidal range for the coasts of this region is 3-3 m., taking the mean low water level of spring tides as zero while the range for mean high water of spring tides is 3-0 m. (Kalk, 1954). This is wider than for the Union coasts which have a maximum range of less than 2-0 m.— 5-6 ft. (Bauer, 1933). From latitude 11° S. the coastal waters off the east coast of Southern Africa are subjected to the influence of the warm Agulhas Current (Isaac, 1937). Temperatures in general increase northwards along the east coast and although sea temperature data are meagre it is clear that the coastal waters have a higher temperature at Inhaca than off the Natal coast since the number of tropical and sub-tropical genera of marine algae is greater and “‘sea grasses” are prominent. The occurrence on the west coast of Inhaca of patches of living coral reef is further corroborative evidence. The east coast of the island is exposed directly to the waters of the Indian Ocean while the waters off the west coast of Inhaca are shallow and protected by Delagoa Bay. These waters are further protected by sandy shoals to the south and to the north, the northern shoal surrounding the smaller Ilha dos Portuguéses (Fig. 1). Although the effects of relative isolation of the waters off the west coast of Inhaca are modified by currents of water which flow into this region from the open sea it is not surprising that higher sea temperatures should prevail in the waters off the west coast than in the waters off the east coast. The east coast is subject to heavier wave action and the salinity of the open sea here and elsewhere is 35-47 °/o9. The tidal waters of the west coast have a salinity of 30 to 33-9 °/o9. (Kalk, 1954). At least three interacting factors affect the salinity of west coast waters: (a) The inflow of fresh water from large rivers flowing into Delagoa Bay. (b) The drain- age of fresh water into the intertidal region just below the mean high water level of neap tides. (c) An inflow off the west coast of water from the open sea. The position of the drainage line of fresh water on to the shores of Inhaca Island and on to the east shore of Inhaca Peninsula is marked by the flourishing appearing and disappearing green colonies of semi-microscopic turbellarian worms of the Convoluta roscoffensis type, the green colour of which is due to the large numbers of symbiotic micro- scopic algae living in their tissues. (For a note on the identity of this organism see Marcus and Macnae, 1954). The drainage of fresh water into the deeper, more turbulent sea off the east coast is of less significance than drainage into the shallow, quiet waters off the west coast. LIST OF SPECIES In arranging Chlorophyceae, Fritsch’s classification (1935) has been adopted except in the case of siphonous Chlorophyceae where the classi- Marine Algae of Inhaca Island and of the 165 Inhaca Peninsula I. fication of Egerod (1952) has been followed. Kylin’s classification has been used for the Phaeophyta (1933) and for the Florideae (1937). Abbreviated citations to literature are given for each species. Full details are listed under literature cited. Citations included within brackets have not been seen. Where, in the majority of cases, no authority is quoted for the dis- tribution of a species on the coasts of the Union of South Africa, the author has himself collected the species for the range indicated. CHLOROPHYCEAE ULOTRICHALES Ulva rigida C. Ag. J.G. Agardh: Till Alg. Syst. II, p. 168. Levring: “Chlorophyceen u. Phaeophyceen von Siidafrika’’, p. 3., Fig. 1 (sections and surface view). Feldmann: “Les Algues Marines de la Cote des Albéres’’, p. 56. Sub-nom, U. lactuca L. var. rigida Le Jol. Le Jolis: “Algues marines de Cherbourg’’, p. 38. Santa Maria, Ponta Abril, Lighthouse Rocks of Inhaca Island!. This species has somewhat elongated cells as seen in transverse section (Fig. 2) and also a short fleshy stipe. The plants collected were about 14—2 cms. high and up to about 2 cms. broad. The frond is rigid when the plant is taken out of the water. Ulva rigida occurs at the lowest intertidal levels and especially at lower mid-tidal levels. It is frequently fairly common on the tube colonies of Idanthyrsus pennatus, a Sabellarid worm. The species is widely distributed on the east and south coasts of the Union. SIPHONALES Caulerpa cupressoides (Vahl) C. Ag., Weber van Bosse emend. Sub-nom. Fucus cupressoides (Vahl: Endeel Cryptogamiske Planter fra St. Croix). Caulerpa cupresscides. C. A. Agardh: Spec. Algarum, Vol. 1. p. 441. Weber van Bosse: Des Caulerpes, p. 323, Pl. 27. (Figs. 1—14), Pl. 28 (Figs. 1—12) Ponta Abril, Santa Maria, Lighthouse Rocks, west coast of Inhaca Island. As is evident from Weber van Bosse’s monograph this is a very variable species. Inhaca material from Ponta Abril (B. 229) is illustrated 1 Subsequently referred to as Lighthouse Rocks. 166 The Journal of South African Botany. | div =10-~ Fig. 2.—Ulwa rigida C. Ag. (Santa Maria: Isaac 624). Surface view and transverse section. in Fig. 3. It will be observed that the branching fronds bear numerous small pinnae. These pinnae are closely crowded together over the distal parts of the fronds where they are regularly arranged in 3 or 4 rows. The material examined agrees reasonably well with C. cupressoides (Vahl) C. Ag. var. lycopodium (Ag), W. v. B., f. amicorum (Harv.), W. v. B. Caulerpa racemosa (Forssk) Weber van Bosse yar, clavifera (Turn) W. v. B. f. macrophysa. (Kitz) W. v. B., Sub-nom. Fucus racemosus (Forsskal: Flora Agyptiaco-Arabica, p. 191.) Weber van Bosse: Des Caulerpes, p. 357, Plate 33, Fig. 4. Lighthouse Rocks, Santa Maria, Ponta Abril. Plants of f. macrophysa of var. clavifera of this variable species are illustrated in Plate XXXVI A. A common variant of this form is shown in Fig. 4: the pinnae are more or less pyriform and characteristically Marine Algae of Inhaca Island and of the 167 Inhaca Peninsula I. Fig. 3.—Caulerpa cupressoides (Vahl) C. Ag. (Ponta Abril: B. 229) Lem Fig. 4.—Variant of Caulerpa racemosa (Forssk) W. v. B., var. clavifera (Turn.) W. v. B., f. macrophysa (Lighthouse Rocks). 168 The Journal of South African Botany. occur in small clumps, the bases of the pinnae arising very close to one another giving the clumps a characteristic congested candelabra-like appearance. In these plants each frond resembles the right hand frond depicted by Weber van Bosse in Fig. 4, Plate 33 of her monograph on the Caulerpas. These plants grow at the lowest intertidal levels. I have identified var. clavifera f. macrophysa of the type illustrated in Fig. 4, from material collected by Mr. G. M. Thomson at Isipingo where it occurs as a rare plant. In the Inhaca region it is a common plant. Caulerpa scalpelliformis (R. Br.) C. Ag. Sub-nom: Fucus scalpelliformis R. Brown in Turner: Hist. Fuc. Vol. 3, p. 96, Plate 174. Caulerpa scalpelliformis C. A. Agardh: Spec. Algarum, Vol. 1, p. 437. Weber van Bosse: Des Caulerpes, p. 286, Pl. 22 (Fig. 11), Pl. 23, (Figs. 1—10). C. denticulata. (Decaisne, Arch. du Mus. II, pl. 6, Fig. 1). Ponta Abril, Santa Maria, Lighthouse Rocks. The plants examined were of var. denticulata (Decaisne) W. v. B. and var. intermedia W. v. B. (Fig. 5). C. scalpelliformis is a common and widely distributed species of the Inhaca region. Small plants occur as a turf constituent on rocky shores and large sized plants are found growing on sand in quiet waters. C. scalpelliformis var. intermedia has been recorded for the north coast of Natal at St. Lucia rocks in St. Lucia Bay by Papenfuss who also calls attention to material of the plant collected in Delagoa Bay by Moe in 1903 (Papenfuss, 1948). Caulerpa selago (Turn.), C. Ag. Sub-nom: Fucus selago Turner: Fuci 1, p. 122, pl. 55. Caulerpa selago C. A. Agardh: Species Algarum, Vol. I, p. 442. Weber van Bosse: Des Caulerpes, p. 296. Ponta Abril, Santa Maria, Lighthouse Rocks. A common plant at Inhaca (Fig. 6). A constituent species of algal turf. The species was recorded for the Union coast by Pocock and Papen- fuss who collected it in northern Natal near the mouth of the St. Lucia estuary (Papenfuss, 1951). Marine Algae of Inhaca Island and of the 169 Inhaca Peninsula I. : = Fig. 5.—Caulerpa scalpelliformis (R. Br.) C. Ag. A. var. intermedia W. v. B. (Young plant) B. var. denticulata (Decaisne) W. v. B. (B. 232). 170 The Journal of South African Botany. Halimeda cuneata Hering. (Hering: Flora, vol. 29, p. 214). Kiitzing: Species Algarum, p. 505. Kiitzing: Tab. Phyc., vol. 7, p. 8, pl. 21, Fig. 3. E.S. Barton: Halimeda, p. 15, pl. 1, Fig. 7. Santa Maria, Ponta Abril, Lighthouse Rocks. This is a widespread species (Plate XXXVI B) at Inhaca. It shows much variation in size. On vertical rock faces at low water level it is sometimes very abundant. Halimeda cuneata is widely distributed on the eastern and south eastern coasts of the Union. Pseudocodium de-vriesei W. v. B. Weber van Bosse: J. Linn. Soc. Bot., vol. 32, p. 209. ~ B. 0. 5 mm. 0-5 cm. Fig. 6.—Below, Caulerpa selago (Turn.) C. Ag. (B. 279) S-rhizoids covered with sand. A. Portion of frond magnified to show arrangement of pinnules. B. Single pinnule. Marine Algae of Inhaca Island and of the 171 Inhaca Peninsula TI. Levring: “Chlorophyceen u. Phzophyceen von Siidafrika’, p. 14, pl. 4, Fig. 10. Ponta Abril, Santa Maria, Lighthouse Rocks. This is a widespread species. In places it is a very common to domi- nant species on vertical rock faces at low intertidal levels. The type locality of this species is in the vicinity of Durban (rocks near Isipingo! and the Bluff near Durban—Weber van Bosse, 1896). It is widely distributed on the east coast of the Union where it may be a prominent species locally. The Union plants are frequently larger than those of Inhaca Island and Peninsula. As far as I am aware this is the first record of the species outside of the Union of South Africa. Udotea orientalis A. and E. S. Gepp. A. and E.S. Gepp: T'he Codiaceae of the Siboga Expedition p. 119. Sand Bank off Ponta Rasa, Ponta Punduini, Santa Maria, Ponta Abril, Lighthouse Rocks, west coast near Marine Station. $44 Inhaca material of this species agrees ni well with the description of A. and E. 8. Gepp (1911). As Plate XX XVII shows the frond is well developed and broad. The frond filaments (Fig. 7) show constrictions at markedly unequal distances above the points of dichotomy, which feature is characteristic of this species. The frond varies in breadth, and the plant in size, according to habitat condi- tions. The plants growing in quiet sandy pools and in sand in quiet waters behind the old coral reef at Ponta Punduini are large (up to 5 em. and more in height) and have broad (up to 6 em.) fronds. In shallow sandy pools Udotea orientalis was found growing associated with Padinacommersonii and with the Angiosperm Halophila ovalis. Plants growing mn rougher seas on sand and on vertical rock faces are smaller and often a Beets ene, ae are have narrower fronds. - and E. 8. Gepp. (B. 93). ' Recorded as “‘Isipinga”’ by Weber van Bosse. 172 The Journal of South African Botany. Udotea orientalis occurs on the coast of Natal. The South African plants taken as a whole are smaller than those of Inhaca and have narrower fronds. The former not infrequently have an external appear- ance which recalls U. indica A. and E. 8. Gepp but the frond filaments lack the simple appendages characteristic of that species. SIPHONOCLADALES Chamaedoris delphinii (Hariot) Feldm. et Borg. Sub-nom: Siphonocladus delphini. (Hariot: Quelques Algues de Madagascar, p. 470.) Chamaedoris delphinit. (Feldmann: Rev. Général de Bot. v. 50, p. 18) Borgesen: Some Marine Algae from Mauritius I. p. 16, Plate 1. Santa Maria, Ponta Abril, Lighthouse Rocks. A very common plant (Plate XX XVIII) at Inhaca. It forms societies and clans in rock pools and on vertical rock faces at low intertidal levels. This species occurs on the coast of Natal and on the east coast of the Cape Province. It is locally common in some places as at Isipingo (about eleven miles south of Durban) and Chaka’s Rock about 30 miles north of Durban. It was nowhere observed to be as common on the Union coast as on the coasts of the Inhaca region. Microdictyon krausii J. EK. Gray. Gray: Jour. Bot. (Seemann’s) v. 4, p. 69 Setchell: The Genus Microdictyon, p. 519. Ponta Abril, Lighthouse Rocks. The outline of a plant collected at Ponta Abril is shown in Fig. 8a, and in b. a portion of the network towards the base is illustrated showing the long basal segments. The material studied agrees well with the description of the species given by Setchell (1929) and in structure it agrees with material of Microdictyon collected on the coasts of Natal. It grows in somewhat shaded situations and is common on vertical sides of deep pools especially pools at low intertidal levels. At a quick glance it appears in situ somewhat like a large Ulva rigida due to its general form and bright green colour but a more careful observation will detect a certain opacity to its colour. Although when fresh the plant is a bright green colour it soon darkens to a grey-green in preservative and on a herbarium sheet. Microdictyon krausii is a rare plant of the intertidal region on the coast of Natal. The type locality is near Durban. It has not previously been recorded outside the Union. Marine Algae of Inhaca Island and of the 173 Inhaca Peninsula I. Fig. 8.— Microdictyon krausti J. E. Gray, (Ponta Abril B. 250). Below, frond—outline. Above, detail of portion of basal part of frond. Valonia macrophysa Kiitz. Kiitz: Phycologia Generalis, p. 307. Kitz: Tab. Phyc., vol. 6, p. 30, pl. 87, Fig. 3. Sand beach and coral reef on west coast of island, Lighthouse Rocks, Santa Maria, Ponta Abril. Valonia macrophysa is a widespread species in the Inhaca region. It occurs in rock pools with zoanthids; as a constituent of algal turf; and is common on vertical rock faces and underneath low level overhangs at low intertidal levels. This species occurs on the east coast of the Union where in places at Isipingo it dominates rock faces at low intertidal levels. DASYCLADALES Acetabularia calyculus Quoy et Gaimard. (Quoy et P. Gaimard: Voyage autour du Monde, p. 621) Harvey: Phycologia Australica, vol. 5, Plate 249. 174 The Journal of South African Botany. Solms-Laubach: Monograph of the Acetabularieae, p. 26. Boérgesen: Marine Algae of the Danish West Indies 1, p. 75. (The specific epithet is spelt “caliculus” by Boérgesen.) At southern end of Inhaca Island: Saco. The mature plants collected were about 2—3 cms. high. Each stalk terminates in a basin-like “disc”, the whole plant re- sembling somewhat a long stalked glass. The larger “discs” are 4 mm. in diameter. The number of rays in a disc varied from 18 to 32 in the plants examined but the number was usually about 20. The rings of sterile filaments were occasionally present in the material examined but mostly they had fallen off leaving scars. The rays of the “disc” are strongly upcurved, the apices being broadly rounded, or having a central indentation or apiculum (Fig. 9). The rays are either free from one another or laterally closely adherent. Thus the material named A. calyculus here corresponds to the more inclusive description by Bérgesen rather than to the A. calyculus which is demarcated from A. suhru by Solms-Laubach. Bérgesen does not think it possible to regard A. suhrii as a species. (1913, p. 80). The superior and inferior coronal segments are free from one another (Fig. 10). The Bs: _ sterile hair scars on the superior coronal Pe ae segments were not easily visible but there et Gaimard. (Saco: B. 276A) appeared to be two or three when the segment was seen in side view. The fully mature rays contained numerous globular swarmer cysts. The size of the swarmer cysts varied much even sometimes in different rays of the same disc. Although the plants of Acetabularia examined varied in regard to number of rays in a disc, the form of the apex of the rays, the lateral adherence or non-adherence of the rays, and the size and number of swarmer cysts, they all agreed in regard to the following characters:— . Order of size: 2—3 ems. . Strongly upcurved rays. . Relatively small number of rays per disc. . Presence of both superior and inferior coronal segments. . Coronal segments free from one another. S Ol Rm Ww be . Spherical swarmer cysts. Marine Algae of Inhaca Island and of the 175 Inhaca Peninsula I. AL Fig. 10.—Portion of dise of Acetabularia calyculus Quoy et Gaimard (B. 276 A.) S = Axis. C.I. = Corona inferior. C.S. = Corona superior. H = Sterile hair. F. = Fertile ray. C. = Spore Cyst. In rays where the swarmer cysts had not been formed, chloroplasts occurred only along the upper and lower surfaces and at the apex of the rays; they were absent from the sides. As stated in the introductory section, Acetabularia calyculus was found to seaward of the southern mangrove growing attached to shells of Loriphes clausus. It was of local occurrence but it was abundant where it occurred. Neomeris dumetosa Lamour. (Lamouroux: Histoire des Polypiers coralligénes flewibles, p. 243). Howe: Bull. Torrey. Bot. Club., vol. 36, p. 77. Lighthouse Rocks, Santa Maria, Ponta Abril. The material examined agrees with Howe’s description of the species. This is a common and widely spread species in rock pools, at the lower intertidal levels on vertical rock surfaces, and as an algal turf constituent. The Journal of South African Botany. PHAEOPHYTA ECTOCARPALES 176 Ralfsia expansa J. Ag. 7 J. G. Agardh: Spec. Gen. et Ord. Algarum, vol. 1, p. 63. Borgesen: “Two crustaceous algae from the Danish West Indies” p. 123. Ponta Abril, Lighthouse Rocks. Ralfsia is widely distributed on the rocky shores of the Inhaca region. In section, the material examined showed marked symmetry (Fig. 11) in places, but symmetry in the disposition of filaments was barely ota Fig. 11.—Vertical section of Ralfsia expansa J. Ag. (Ponta Abril). E. = erect a s. B. = basal filaments. R. = rhizoid. Marine Algae of Inhaca Island and of the i Inhaca Peninsula I. discernible in other parts. This is charac- teristic of R. expansa while in R. deusta bilateral symmetry is always clearly seen (Weber van Bosse, 1913, p. 146). Mature reproductive organs were not S seen in the sections examined but an early stage in the development of a sporangial sorus was observed (Fig. 12). R. expansa is generally distributed over South African coasts being a species of both warm and cold South African waters. In earlier publications dealing with the biota of South African coasts this species is referred to as Lepadoderma africanum Dickinson. This name, however, was not ,~WH4 validated. 30m. DICTYOTALES Fig. 12.—Early developmental stage of a sporangial sorus of : c aes i Ralfsi J. Ag. Dictyopteris longifolia Papenfuss ined. Rae Crea oe Ponta Abril, Lighthouse Rocks. Plants of Dictyopteris were collected on Inhaca Island and Peninsula which closely resemble plants of a species of Dictyopteris found growing on the east and south eastern coasts of the Union (Plate XX XIX). This species corresponds to Haliseris dichotoma (Suhr) Kiitz depicted by Kiitzing in Tabulae Phycologicae, 1X, Tab. 57 (i) and with material in the Botany Department Cape Town University named Dictyopteris dichotoma (Suhr) Levring by Papenfuss. The change in nomenclature from Haliseris dichotoma Suhr to Dictyopteris dichotoma (Suhr) Levring was made by Levring (1938). Stephenson (1947) makes the following comment in regard to Dictyopteris dichotoma (Suhr) Levring: ‘The common plant which has long been known as dichotoma is not in fact Suhr’s species, one of whose specimens has been examined by Papenfuss; who is accordingly re-naming the common form longifolia’. (p. 298) Dictyopteris longifolia in the Inhaca region grows at low intertidal levels in rock pools and shallow water. Padina commersonii Bory. (Bory: Voyage de la Coquille, Botanique, p. 144). J.G. Agardh: Till Algernes Syst., II p. 119. Weber van Bosse: Des Algues du Siboga., I p. 178. Taylor: Plants of Bikini, p. 100, Plate 54, Fig. 1. Santa Maria, Sandbank off Ponta Punduini, Ponta Abril, Lighthouse Rocks. 178 The Journal of South African Botany. Fertile material was collected both at Santa Maria and Ponta Pun- duini. The relative arrangement of zones of hairs and of organs of fructi- fication is as illustrated by Weber van Bosse (Fig. 51, p. 179, 1913) and is shown in Fig. 13. Fig. 13.—Padina commersonui Bory. (Sand bank off Ponta Punduini: Isaac 185). Tetrasporangial plant. I. = imrolled edge. L. = line of hairs. T. = tetrasporangia. This plant was found growing in a variety of habitats: on sand to landward of a dense growth of Cymodocea ciliata; on old broken coral scattered over the sandbank near Ponta Punduini; rock pools; widely distributed on rocky shores especially on rocky platforms in small, water-filled, shallow depressions and permanently submerged in shallow water. On rocky platforms the plant occurs as a dominant species forming almost pure communities. Here, however, and on rocky shores in general the plant is much smaller than on sand and on broken coral scattered on sand in quiet waters (Plate XL). Also the smaller plants have a denser incrustation of lime. The species is widely distributed on the coast of Natal and on the © east coast of the Cape Province north of East London. Nowhere on the Union coast, however, has it been found in such profusion as on the rocky platforms of the Inhaca region. Pocockiella variegata (Lamour) Papenfuss. (Lamouroux: Observations sur la physiologie des algues marines, p. 331) Marine Algae of Inhaca Island and of the 179 Inhaca Peninsula I. Papenfuss: Amer. Jour. Bot., vol. 30, p. 467, Fig. 1—14. Ponta Abril. The specimens examined agree essentially with Papenfuss’ descrip- tion of the species. Some of the plants collected were somewhat larger than the plants illustrated by Papenfuss and the number of cortical cells (Fig. 14a and b) somewhat greater. The plants, however, are smaller and the number of cortical cells decidedly fewer than in P. papenfussu Taylor (1950). No fertile material was examined. Pocockiella variegata is recorded over the entire length of the Natal coast (Stephenson, 1947). Fig. 14.—Pocockiella variegata (Lamour.) Papenfuss. (Isaac 536). A. Longitudinal section of Apex x 160. B. Longitudinal section of mature thallus x 850. (A. reduced to half and B., reduced to one third.) 180 The Journal of South African Botany. Stypopodium zonale (Lamour) Papenfuss. Sub-nom. Fucus zonalis (Lamouroux: Dissertations sur Plusieurs Espéces de Fucus). Papenfuss: Bot. Notiser, 1940, p. 205. Lighthouse Rocks, Ponta Abril. This is a very variable species in regard to form, colour and thickness. The range of forms collected at Inhaca (Plate XLI) falls within the wide range of forms at the British Museum of Natural History, determined as Stypopodium zonale (Lamour) Papenfuss by Papenfuss. The Museum material had been previously named Stypopodiuwm lobatum (Ag.) Kiitz. In the Inhaca region this species occurs in rock pools and in the upper sub-littoral. It is a fairly common species on the east coast of the island on rocks in the neighbourhood of the lighthouse. Stypopodium zonale is distributed on the east and south coasts of the Union. PUNCTARIALES Chnoospora minima (Hering) Papenfuss, Sub-nom. Fucus minimus Hering: Ann. Mag. Nat. Hist., vol. 8, p. 92 Chnoospora fastigiata var. pacifica. J. G. Agardh: Spec., Gen. et Ord. Algarum vol. 1, p. 172. Chnoospora pacifica Kiitz: Tab. Phyc., vol. 9, p. 36, pl. 86, Fig. 1. Levring: “Chlorophyceen u. Phaeophyceen von Stdafrika”’, p. 21, Fig. 10 (sections), Plate 4, Fig. 12. Santa Maria, Ponta Abril, Lighthouse Rocks. A comparison of the form and structure of Inhaca and Natal plants indicates that they belong to the same species. Material from Isipingo near Durban has been described by Levring (1938). This species has previously been known in South Africa as Chnoospora pacifica J. Ag. Papenfuss (1956, p. 69) points out that since the plant was first described by Hering under the name of Fucus minimus the later name C. pacifica J. Ag. should be changed to C. minima (Her.) Papenfuss. Chnoospora minima occurs on the coast of Natal and on the east coast of the Cape Province. Although not a common plant, it was found widespread on the coasts of the Inhaca region. Colpomenia sinuosa (Roth) Derb., et Sol. Sub-nom. Ulva sinuosa (Roth: Catal. Bot. vol. 3, p. 327, Pl. 12) Marine Algae of Inhaca Island and of the 18] Inhaca Peninsula I. Colpomemia sinuosa (Derbes et Solier: Mémoire sur quelques points de la physiologie des Algues, p. 11) G. M. Smith: Marine Algae of the Monterey Peninsula, p. 128, Plate 20, Fig. 1. Lighthouse Rocks, Santa Maria, Ponta Abril. This is a widespread, occasional species of the Inhaca region. It occurs at low intertidal levels and in rock pools. This plant is widespread on the south and east coasts of the Union. It also occurs at Camps Bay on the west coast of the Cape Peninsula as a locally common plant and in the Saldanha-Langebaan Lagoon region about 70 miles north of Cape Town. In the latter locality the plant was dredged up from a depth of about 12 feet. Both the Camps Bay and Langebaan records are new. Hydroclathrus clathratus (Bory ex C. Ag.) Howe. Sub-nom. Fucus clathratus Bory: Msct. in C. Agardh, Spec. Alg., vol. 1, p. 412. Hydroclathrus clathratus (Howe: In Britton and Millspaugh, The Bahama Flora, p. 590) Taylor: Plants of Bikini, p. 96. Santa Maria, Ponta Abril, Lighthouse Rocks. Plants collected in the Inhaca region measured up to about | dm. in diameter. Recorded for Umpangazi on the northern coast of Natal. (Stephen- son, 1947). FUCALES Hormophysa triquetra (L,) Kiitz. Kiitzing: Phycologia generalis, p. 359. Santa Maria, Ponta Abril, Lighthouse Rocks. The plants of Hormophysa collected correspond with H. latifrons as figured by Kiitzing (Tabulae, 10, Pl. 60). H. latifrons as figured by him is broader than his delineation of H. triquetra (Ibid.) which agrees with Turner’s illustration of that species (Turner: Fuci, vol. 1, Plate 34). H. latifrons is also lacking in bladders which are represented as present in H. triquetra. Borgesen has dealt with the synonomy of this species (Boérgesen, 1937) and maintains that H. latifrons is a form of Hormophysa triquetra which view is accepted in this paper. 182 The Journal of South African Botany. Hormophysa triquetra is a common to frequent plant in deep low level pools in the Inhaca region and where found it is usually present in large numbers. Also it occurs at low intertidal levels on rock where it may be completely uncovered at low water of spring tides. The plant has been collected by me in rock pools at Isipingo on the Natal coast. It has not, however, been observed as a common plant on the east coast of the Union. The Isipingo material collected agrees morphologically with the Inhaca plants (Plate XLII). It may be noted that Papenfuss (1943) suggested that Hormophysa triquetra should be excluded from the South African flora as it had not been collected there in recent years. The previous inclusion of this species in the South African seaweed flora was chiefly based on Linnaeus’ type specimen. Turbinaria ornata J. Ag. J. G. Agardh: Spec. Gen. et Ord. Algarum, vol. 1, p. 266. Kitz: Tab. Phyc., vol. 10, p. 24, Pl. 66 (1). Taylor: Plants of Bikini, p. 101. Lighthouse Rocks, Ponta Abril. This species occurs in pools and also emersed at low intertidal levels. At Ponta Abril and the Lighthouse Rocks it was observed to be locally very common in shallow water in small pools on rocky platforms (Plate XLII). Turbinaria ornata occurs on the coast of Natal. Taylor (1950) states that this species attains heights of 3 dm. on Bikini. The plants seen on the coasts of Natal and at Inhaca are smaller and have not been found to attain heights exceeding one dm. FLORIDEAE NEMALIONALES Galaxaura tenera Kjellm. (Kjellman: Galaxaura, p. 77, Pl. 14 and 20) Kylin: “Rhodophyceen von Siidafrika”’, p. 5, Pl. 1, Fig. 2. Lighthouse Rocks, Ponta Abril. This plant was found as an occasional species in rock pools at low intertidal levels. It is widely distributed on the coast of Natal and on the east coast of the Cape Province. GELIDIALES Gelidiopsis rigida (Vahl) Web. v. Bosse. Sub-nom. Fucus rigidus (Vahl: Naturh. Sellsk. Skrift, 1802) Marine Algae of Inhaca Island and of the 183 Inhaca Peninsula I. Echinocaulon spinellum, E. ramelliferum, HE. rigidum. Kitz: Tab. Phyc. vol. 18, Pl. 38, 39 and 40 respectively. Gelidiopsis rigida (Weber van Bosse: Recueil de trav. bot. Neerl., vol. 1, p. 104). Weber van Bosse: Algues du Siboga, IV, p. 427. Ponta Abril, Santa Maria, Lighthouse Rocks. \e NX 0 Sen Fig. 15.—Gelidiopsis rigida (Vahl) W. v. B. (Santa Maria: B. 291) The specimens of this plant (Fig. 15) collected in the Inhaca region agree well with material of the species in the British Museum of Natural History and with the material in the herbarium at Kew. In the British Museum there are specimens of this species from Dar- es-Salaam, the Persian Gulf and Mauritius. Gelidium caespitosum Kylin. Kylin: “Rhodophyceen von Siidafrika’’, p. 8, Fig. 2, E—H. Ponta Punduini. 184 The Journal of South African Botany. Plants from Inhaca are depicted in Fig. 16 which on comparison with Kylin’s figures will be seen to come within the range of variation of the species. Fig. 16.—Gelidium caespitosum Kylin (Ponta Punduini: B. 189). The type locality of the species is Isipingo but it has been collected by me also at other localities in the Durban area: at Reunion Rocks about 9 miles south, and at Salt Rock about 35 miles north, of Durban. Gelidium reptans (Suhr) Kylin. Sub-nom. Phyllophora reptans von Suhr: “‘Beitrage zur Algenkunde”’, p. 13, Plate 3, Fig. 10. Gelidium reptans Kylin: ‘““Rhodophyceen von Siidafrika’’, p. 6, Fig. 2, A—C. Ponta Abril, Lighthouse Rocks. Some of the plants collected show a certain amount of irregularity of the margin while other plants correspond very closely with Kylin’s drawings (Fig. 17). The former agree more in this respect with von Suhr’s figure. 1 In regard to Suhr’s figure see Borgesen, 1943, p. 6. Marine Algae of Inhaca Island and of the 185 Inhaca Peninsula I. Fig. 17.—Gelidium reptans (Suhr) Kylin. (Ponta Abril: B. 248). At Inhaca this alga is chiefly found at mid-tide levels and occurs commonly on the shells of animals—Patella barbara and Tetraclita serrata. Gelidium reptans occurs on the coast of Natal and on the east coast of the Cape Province. In a northward direction on the east coast of the Union it progressively replaces Gelidium pristoides as the commonest species of Gelidiwm at mid-tide levels. GIGARTINALES Plocamium glomeratum J. Ag. J. Agardh: Spec. Gen. et Ord. Algarum, vol. Il, Part 2, p. 397. Kylin: “Rhodophyceen von Sudafrika”’, Pl. 3, Fig. 7. Ponta Abril, Lighthouse Rocks. This is a common and widespread species at low intertidal levels. It was found in both sandy and rocky habitats. Stephenson (1947) records the species as extending as far south west as Port Elizabeth but adds: “probably farther west also” (p. 307). The present author has collected well developed plants as far west as Tergniet, near Mossel Bay. RHODYMENIALES Champia compressa Harv. Harvey: Genera of South African Plants, 1st Ed., p. 402. Harvey: Nereis Australis, p. 78, Pl. 30. 186 The Journal of South African Botany. Santa Maria, Ponta Abril, coral reef on west coast north of laboratory. Only a few specimens of this beautiful alga were found. The species is widely distributed on the coasts of the Union. Stephen- son (1947) gives its range as from Port St. Johns to Oudekraal, which is near Cape Town. The present writer has dredged up the species from a depth of about 15 feet about 70 miles north of Cape Town in the Lange- baan Lagoon. He has also collected it at Isipingo and Salt Rock on the Natal coast thus extending its recorded distribution on both the east and west coasts. Corallopsis opuntia J. Ag. J. G. Agardh: Spec. Gen. et Ord. Algarum, vol. 3, Part 1, p. 409. De-Toni: Sylloge Algarum, vol. 4, sec. 2, p. 459. Ponta Abril, Santa Maria, Lighthouse Rocks. Material of this species (Fig. 18) was identified as Corallopsis opuntia by Miss C. Dickinson of Kew. Common in rock pools (especially in the presence of sand) and at low intertidal levels. 0-5 cm Fonts Abid dard Susty 1956 Fig. 18.—Corallopsis opuntia J. Ag. (Ponta Abril). Rhodymenia natalensis Kylin Kylin: ‘““Rhodophyceen von Siidafrika”’, p. 13, Pl. 5, Fig. 13. Ponta Abril, Lighthouse Rocks. Marine Algae of Inhaca Island and of the 187 Inhaca Peninsula I. The plants examined consisted of a horizontal creeping portion attached to the substratum at intervals and of erect fronds, 3—6 cms. high, slender, showing fairly regular dichotomous branching and of a somewhat crinkled appearance. The plants in their living condition were of a greenish colour. The plants were tetrasporic, the tetrasporangia occurring in expanded, spoon-like branch tips. The internal structure and tetraspores agreed exactly with the corresponding features of specimens of the species from Durban named by Kylin. As Kylin does not figure a section of the tetrasporangial region in his paper an illustration is included here (Fig. 19). The outer + 34 + Fig. 19.— Rhodymenia natalensis Kylin. Sections showing young and more mature tetrasporangia. 188 The Journal of South African Botany. region of the thallus in which the tetrasporangia occur has cells with thick mucilaginous walls and this region bounds an inner region of large cells with relatively thin cell walls. , On the coasts of Inhaca this species occurs at low intertidal levels. Rhodymema natalensis is a widespread species on the east and south coasts of the Union. On the east coast it is locally a prominent plant on vertical rock faces at about low water level of spring tides. CERAMIALES Amansia glomerata C. Ag. Agardh, C. A: Systema Alg., p. 247. Agardh, J. G: Spec. Gen. et Ord. Algarum, vol. 2, Part 3, p. 1111. Falkenberg, P.: Rhodomelaceen. p. 416, Pl. 1, Figs. 20—21, Pl. 6, Figs. 14—29. Sub-nom. A. fasciculata. Kiitz: Tab. Phyc., vol. 15, Pl. 4, Fig. a—d. Ponta Abril, Santa Maria, Lighthouse Rocks. Plants collected at Inhaca (Fig. 20) agree well with most of the specimens of this species in the British Museum. That collection includes plants from Dar-es-Salaam and Mombasa on the east coast of Africa as well as plants from Mauritius from which island it has also been recorded by Borgesen (1945, p. 43). Bostrychia binderi Harv. Harvey: Nereis Australis, p. 68, Pl. 28. Joly: “Bostrychia in Southern Brazil’, p. 61. tem Fig. 20.—One rosette of Amansia glomerata C. Ag. (Lighthouse rocks. B. 266). I. = inrolled apex. Marine Algae of Inhaca Island and of the 189 Inhaca Peninsula I. Ponta Punduini, Saco, Lighthouse Rocks, Ponta Abril, west coast of island. This plant grows abundantly in sheltered, shaded, rocky habitats at high intertidal levels. It also occurs on the lowest parts of Rhizophora. Bostrychia binder grows on the Natal coast. Digenea simplex (Wulf.) C. Ag. Sub-nom. Conferva simplex. (Wulfen: Cryptogama Aquatica, p. 17) Digenea simplex. C. A. Agardh: Spec. Alg., p. 389. J. G. Agardh: Spec. Gen. et Ord. Algarum, vol.2, Part 3, p. 845. Falkenberg: Rhodomelaceen, p. 159, Pl. 9, Figs. 25—29. Borgesen: Marine Algae Danish West Indies, Part 4, p. 281. Ponta Abril, Lighthouse Rocks, Santa Maria. This species (Fig. 21) was found in low level rock pools. ca Fig. 21..—Digenea simplex (Wulf) C. Ag. (Ponta Abril: B. 272). 190 The Journal of South African Botany. Kiitzingia natalensis J. Ag. J. Agardh: Spec. Gen. et Ord. Algarum, vol. 2, Part 3, p. 1099. Lighthouse Rocks, Ponta Abril. The characters of this species including the inrolled axis apices are shown in Fig. 22. Kiitzingia natalensis occurs in low intertidal level rock pools and at low water level of spring tides. It is a widespread species. It occurs on the coast of Natal and on the east coast of the Cape Province. A 3mm. fe paw Ppaneeacvne Fig. 22.—Kiitzingia natalensis J. Ag. (Lighthouse rocks B. 295). a = Apex of branch. B. = Transverse section of axis: c. = central cell. p. = Peri- central cells. ACKNOWLEDGEMENTS Thanks are due to the Portuguese authorities of Mozambique for affording me the facilities of the Estagao de Biologia Maritima on Inhaca Island. Especial thanks are due to Mrs. Margaret Kalk of the Zoology De- partment of Witwatersrand University for her co-operation and help. Marine Algae of Inhaca Island and of the 191 Inhaca Peninsula I. To the Council of Scientific and Industrial Research for South Africa I am indebted for financial aid which made possible the collecting of material in the Inhaca region and much of the subsequent study of this material. For providing me with facilities to work in the British Museum of Natural History at South Kensington, I wish to thank the Keeper of Botany, Dr. G. Taylor. To Miss Yvonne M. Chamberlain I am indebted for her help and collaboration both at the Natural History Museum and subsequently as my research assistant during 1956 at Cape Town University: most of the drawings in this paper were made by her. Thanks are due also to Miss Carola Dickinson of Kew for collaboration and some identifications (acknowledged in the text); to Miss Judith M. Graves for drawing some of the figures; and to Mr. R. H. Simons for taking most of the photographs included in this paper. REFERENCES Items marked with an asterisk * have not been seen. AGaRpDH., C. A., Species Algarum, vol. 1, Greifswald, 1821—3. Systema Algarum Lund, 1824. AGARDH., J. G. Species Genera et Ordines Algarum, vol. 1, 1848, vol. 2, Part 2, 1852, Part 3, 1863, vol. 3, Part 1, 1876, Lund (Leipzig). Till Algernes Systematik, II Lunds Univ. Arsskrift., vol. 17 (4) III Ibid., vol. 19 (2). (Scripta collectanea, Lund, 1890.) Barton, E.S., The Genus Halimeda, Siboga Expedition Monograph 60, Leiden, 1901. Bauer, H. A., “A World Map of Tides”, Geogr. Rev., vol. 23, p. 259, 1933. Boérersen, F., “Two crustaceous brown algae from the West Indies’, Nuova Notarisia, Serie 23, p. 123, 1912. “The Marine Algae of the Danish West Indies Part 1. Chlorophyceae’’, Dansk Bot. Ark., vol. 1, No. 4, 1913. Part IV. Rhodophyceae (4) Jbid., vol. 3, No. 1, 1918. “Contributions to a South Indian Marine Algal Flora—II.,”’ J. Ind. Bot. Soc., vol. 16, p. 311, 1937. Some Marine Algae from Mauritius. I. Chlorophyceae” K. Danske Vidensk. Selsk. Biol. Medd., vol. 15 (4), 1940. III Rhodophyceae, Part 2., Ibid., vol. 19 (1), 1943, Part 4., [bid., vol. 19 (10), 1945. *Bory DE St. Vincent, J. B. G. M., Voyage de la Coquille, Botanique; Cryptoganie Paris, 1828. *BritTon and Mitispauan, The Bahama Flora 1920. Couen, E., “An Account of the Marine Angiosperms of Inhaca, P.E.A.”, S. Afr. Jour. Sci., vol. 36, p. 246, 1939. *DECAISNE, J., ‘“Plantes de L’Arabie heureuse”’, Arch. du Mus. II, Pl. 6, Fig. 1, 1939. *Derpes, A. ET Souier, A. J. J., Mémoire sur quelques points de la physiologie des Algues, Paris 1856. Der-Tont, J. B., Sylloge Algarum, vol. 4, Sec. 2, Padua, 1900. Earrop, L. E., “An analysis of the Siphonous Chlorophycophyta”’, Univ. Calif. Publ. Bot. vol. 52, p. 325, 1952. FAaLKENBERG, P., Die Rhodomelaceen des Golfes von Neapel, Fauna und Flora des Golfes von Neapel Monograph 26, Berlin, 1901 FELDMANN, J., ‘“Les Algues marines de la Cote des Albéres. 1—111. Cyanophycées, Chlorophycées, Pheophycées”’, Rev. Algol. vol. 9, 1937. *_______ “‘Sur la classification de |’ordre des Siphonocladales”’, Rev. Générale de Bot., vol. 50, 1938. 192 The Journal of South African Botany. *ForssKAL, P., Flora Hgyptiaco—Arabica Hauniae, 1775. Fritsceu, F. E., The Structure and Reproduction of the Algae, vol.1, Cambridge, 1935. Gepp, A. and H. 8., The Codiaceae of the Siboga Expedition, Siboga Expedition Monograph 62, Leiden, 1911. Gray, J. E., “On Anadyomene and Micredictyon, with the description of three new allied genera-discovered by Menzies in the Gulf of Mexico”, Jour. Bot. (Seemann’s), vol. 4, p. 41, 1866. *Hariort, P., “Quelques algues de Madagascar”, Bull. du Mus. d’Hist. nat., No. 6, 1902. Harvey, W. H., The Genera of South African Plants, First Edition, Cape Town, 1838. —_—_— Nereis Australis, London, 1847. Phycologia Australica, vol. 5, London, 1863. Herine, D., “Diagnoses algarum novarum a cl. Dre Ferdinand Krauss in Africa Australi lectarum”, Ann. Mag. Nat. Hist., vol. 8, p. 90, 1841. *HeERING, D., Flora, vol. 29, 1846. Hows, M. A., “The Genus Neomeris and Notes on other Siphonales”’, Bull. Torrey Bot. ‘Club., vol. 36, p. 75, 1909. Isaac, WM. Epwyn, “South African Coastal Waters in relation to Ocean currents”’, Geographic Rev., vol. 27, p. 651, 1937. Jory, A. B., “The genus Bostrychia Montagne, 1838 in Southern Brazil. Taxonomic and ecological data,’ Boletius de Faculdade de Filosofia, Ciéncias e Letras da Uni. Sao Paulo, No. 173—Botanica No. 11, p. 55, 1954. Kak, MarGaret, “Marine Biological Research at Inhaca Island, Mozambique: An interim Report”, S. Afr. Jour. Sci., vol. 51, p. 107, 1954. *KJELLMAN, F’. R., “Om florideslagtet Galaxaura dess organografi och Systematik”’, Vet. Akad. Handl., vol. 33, 1900. Kurzine, F. T., Phycologia Generalis Leipzig, 1843. — Species Algarum, Leipzig, 1849. 5 Tabulae Phycologicae, vol. 6, (1856). vol. 7, (1857) vol. 9, (1859) vol. 10, (1860). vol. 15, (1865) and vol. 18, (1868). Nordhausen. Kyun, H., ‘Uber die Entwicklungsgeschichte der Phaeophyceen”’, Lunds Univ. Arsskr. N. F. Avd 2, vol. 29, (7), 1933. ———— Anatomie der Rhodophyceen, Berlin, 1937. —_____— “Verzeichnis einiger Rhodophyceen von Siidafrika’”’, Lunds Univ. Arsskr. N. F. Avd. 2, vol. 34, (8), 1938. *LamourRoux, J. V. F., Dissertations sur Plusieurs Hspéces de Fucus peu connues ou nouvelles: avec leur description en latin et en francais. Paris, 1805. x_______ “Observations sur la Physiologie des Algues Marines, et Description de cing Nouveaux genres de cette Famille’, Nouv. Bull. Sci. Soc., Philom., vol. 1, p. 330, 1809. *_______ Histoire des Polypiers Coralligenes Needles, vulgairement nommés Zoophy- tes. Caen., 1816. Le Jotis, A., “Liste des Algues Marines de Cherbourg”, Mémoires Soc. Impér. Sct. : Nat. de Cherbourg, vol. 10, 1864. (Reprinted as a small book in 1880.) LevyrineG, T., “Verzeichnis einiger Chlorophyceen und Phaecophyceen von Stid- Afrika” , Lunds Univ. Arsskr. N. F. Avd. 2, vol. 34, (9), 1938. Marcus, E. and Macnas, W., ‘“‘Architomy in a species of Convoluta’”’, Nature, vol. 173, p. 130, 1954. Moege, A. O. D., Systematic List of the Flora of Inhaca Island, P. E. A. Typescript, revised July, 1956. Moss, M., “A Preliminary account of the Sea-grasses of Delagoa Bay”’, S. Afr. J Sci. vol. 33, p. 344, 1937. Pargenrfuss, G. F., “Notes on South African Marine Algae I’, Bot. Notiser, p. 200, 1940. “Notes on South African Marine Algae II”, Jour. S. Afr. Bot. vol. 9, p. 79, 1943. “Notes on algal nomenclature II, Gymnosorus J. Agardh,’”’ Amer. Jour. Bot., vol. 30, p. 463, 1943. “Notes on South African Marine Algae. III’, Jour. S. Afr. Bot., vol. .17, p. 167, 1951. — “Notes on South African Marine Algae IV”, Ibid., vol. 22, p. 65, 1956. *Quoy, J. R. C. pt Gatmarp, J. P., Zoologie, Voyage autour du Monde exécuté sur les Corvettes L’ Uranie et la Physicienne, Paris, 1824. Marine Algae of Inhaca Island and of the 195 Inhaca Peninsula I. *Rotu, A. W., Catalecta botanica, Leipzig. Fase. 3, 1806. SercHetL, W. A., “The Genus Microdictyon”, Univ. Calif. Publ. Bot., vol. 14, p. 453, 1929. SmiruH, G. M., Marine Algae of the Monterey Peninsula Stanford Univ. Press, 1944. Sotms-LausBacu, H., “Monograph of the Acetabularieae”’, Trans. Linn. Soc. Lond., 2nd series, Botany, vol. 5, Pt. 1, 1895. STEPHENSON, T. A., “The constitution of the intertidal fauna and flora of South Africa, Part III’, Ann. Natal Mus., vol. XI, p. 207, 1947. von Suur, J. N., “Beitrage zur Algenkunde”’, Acta Acad. Caes. Leop. Carol. Nat. Cur., vol. 18, Suppl. 1840. Taytor, W. R., Plants of Bikini, Univ. Michigan Press. 1950. TURNER, Dawson, Fuci, vol. 1, London, 1808. Vol. III London, 1811. *VauL, Beskrivelse over endeel cryptog. Planter fra St. Croix. Naturh. Sellsk. Skrift 1802. WEBER VAN Bosse, A., “On a new genus of Siphonean Algae—Pseudocodium’’, Jour. Linn. Soc. Lond., Bot., vol. 32, p. 209, 1896. “Monographie des Caulerpes”’, Ann. Jard. Bot. Buitenzorg, vol. 15, p. 248, 1898. *________ ‘Note sur deux algues de l’Archipel Malaisien’’, Recueil de trav. bot. Neerl., vol. 1, p. 104, 1904. Liste des Algues du Siboga, I Siboga Expedition Monograph 59a—Myxophy- ceae, Chlorophyceae, Phaeophyceae, Leiden, 1913. Ibid., 1V Monograph 59d—Rhodophyceae (3) Leiden, 1928. *WouLFEN, F., Cryptogama Aquatica Leipzig, 1803. Prats XXXVI. Above (A): Caulerpa racemosa (Forssk) W. v. B., var. clavifera (Turn.) W.v. B. f. macrophysa. Below (B): Halimeda cuneata Hering. ee Prath XX XVII. Udotea orientalis A. and KE. 8S. Gepp. Range of form and size. BIog 49 ‘ploq (qolrey) wureydjap sisopenwumyy) “TITAXXX FLVIg & : a , nfm, od if ee Nee -| f S&, J Py, £4, GY, ~ hae = &, i oN 7 & 3 Pratt XXXIX. Dictyopteris longifolia Papenfuss ined. PuatE XL. Padina commersonw Bory. Above, from sand in quiet waters; Ponta Punduini. Below, attached to rock and exposed to rougher seas: Santa Maria, ‘yeyeN (uequnq “u) osurdisy urosy ‘yYSt4 uC- "RIIB], BIUBY UO “Jo, UC “Zany (T) vuanbiuy vshydousoH “TTTX ALvig Prats XLII. Turbinaria ornata J. Ag. Rocky platform, Ponta Abril. ae INDEX OF PLANT NAMES—VOL. XXII, 1956. PAGE Acetabularia calyculus ob GS 3 ealyeulus Quoy et Gaimard.. 173,174 59 suhrti .. at 174 Actiniopteris australis oe so Uh © Adiantum caudatum ie oa Ups Aizoon contaminatum oo IOS) » erystallnum E & Yi, on LUE) a5 elongatum E. & Z. lod » fruticosum .. ‘ mee LG » Herniariae Reichb. ao HOI > microphyllum ae so UO! » papulosum EK. & Z. ae Le » propinquum EH. & Z. meLOD, » pubescens EH. & Z. .. ao OY » secundum .. so IO eae (E. & Z. Viva motes! glee, Aletris F at 42, 43 Aloe abyssinica Lam. 33, 141, 151 », abyssinica v. Peacockii Baker. 155 ;, acutissima H. Perrier v. antanimorensis Reynolds 27, 28 ,, aethiopica (Schweinf.) Berger 153 ,, amudatensis Reynolds .. 186 » antandroy H. Perrier 23, 28 » Bakeri Scott Eliot .. 23, 24 » bellatula Reynolds .. .. 182 ,, bulbillifera H. Perrier v. Paulianae Reynolds -. 26 », Calidophila Reynolds . . so (WEY , Camperi Schweinfurth 152, 156 ,, Chabaudii Schonl. ais oo EG » Decaryi Guillaumin .. 27, 28 , elegans Todaro A 152, 153 , eru Berger aie ae 142, 158 ., itremensis Reynolds .. ae eo) , kniphofioides Bak. .. So , laeta Berger .. ne so ey » macrocarpa Tod. = .. 152 » linearifolia Berger .. co ha) ., marsabitensis Christian oo, Weis » Millotii Reynolds o6 co 2B » minima Bak. on ao (sia) », modesta Reynolds .. .. 85 ,, mzimbana Christian .. DO 5, parvula Berger a oo 2) 5, percrassa Tod. ee .. 154 ,, Perrieri Reynolds .. 180,181 , schweinfurthii Bak. .. 140, 154 ;, schweinfurthii Hook. eA: 5, sempervivoides Perrier sa WAY) », trichosantha Berger emo 2 » Tweediae Christian .. 135, 139 », ukambensis Reynolds co 63} » verav.aethiopica .. ao lbs) », Wilsonii Reynolds .. .. 137 » Wrefordii Reynolds .. .. 141 195 PAGE Amansia fasciculata ess) es glomerata C. A. Ag. .. 188 Anigosanthus : 41, 42 Anisocladus congestus Reinke. 67, 68 Antanimorensis 0 te eS Aristea angolensis .. a: so weal Avicennia officinalis Linn. .. lS. Azolla pinnata ie o'0 so UO Barberetta .. ays a go ne '8} Blancoa : ome af .. 43 Bostrychia binderi .. : 163, 188 Bruguiera cylindrica Blum. 66) G33 Camptosorus rhizophyllus Se eel Carpomitra filiformis (Suhr.) Papenfuss ee so AO) Caulerpa crassifolia » (C. Agardh.) J. Agardh .. 65 99 cupressoides (Vahl.) 165, 167 55 Harveyana Kiitzing OD % mexicana Sonder Yo. GH 5p pinnata sensu W. v. B. 65, 66 aS racemosa (Forssk.) 166, 167 x3 sealpelliformis (R. Br.) Ob AVG) og oA 168, 169 , selago (Turn.) C.Ag. | 168,170 5p taxifolia (Vahl.) C. Agardh. ints oo GSS Ceriops candolleana Arn. .. sa Gs Ceterach cordatum a: Se aD) Chamaedoris delphinii fe 163, 172 Champia compressa Harv. ae) U85 Cheilanthes dinteri x Sate Mig 9 96 gracillima 10 56 hirta .. oe 7, 12 multifida 15 U2 parviloba oo 7, 12 Chnoospora fastigiata o% 69, 180 m6 minima (Hering) Papen- fuss bi rae 69, 180 5 pacifica J. Agardh. 69, 180 Colpomenia sinuosa (Roth.) .. 180 Conanthera oe oC as Conferva simplex .. a6 SO, Conostylis. .. 46 on) 4a Convoluta roscoffensis 38 .. 164 Corallopsis dichotoma Suhr. ee) 9 opuntia J. G. Ag. .. 186 Crinum bulbispermum Pe (Burmets) Sion hel es graminicola V ondlorran eS 2, », macowanii Bkr. .. ae!) », moorei Hook.f. .. oo teh) Cyanastrum 20 a0 42, 43 Cyanella 06 ae bio oa 453 196 PAGE Cymodocea ciliata (Forsk.) ae Ehrenb. 163 - isoetifolia Aschers 163 00 rotundata Aschers and Schwein ae so UGS} serrulata (R. Br.) Aschers and Agnus 5:6 yel63 Dictyopteris dichotoma (Suhr.) Leyring . o WaZ 5 longifolia Papenfuss . 177 Digenea simplex (Wulf.) C.A.Ag. .. 189 Dilatris Js . 43 90 pillansii Barker 4] Diplanthera uninervis (Forsk.) F. N. Williams 163 Doryopteris concolor v. kirkii 7,9 Dryopteris thelypteris Vv. squamigera .. U5 9 Echinocaulon ramelliferum 183 53 rigidum 183 6 spinellum 183 Eeklonia maxima 74 Ketoclinium dentatum 72 kowiense Holmes 71 Encephalartos altensteinil .. 2,3 5 arenarius R. A. Dyer 1,3 9p ferox 3 5 horridus 3 Fp kosiensis 3 55 latifrons 5 25 83 %) lehmannii .. 3 tridentatus .. : 2 Erica Bakeri Salter 37, 39 » cyrillaeflora Salisb. . 38 » empetrifolia L. 38 , heleogena Salter 38 » lmosa L. Bol. 38 s, mollis Andr. 38 », paludicola L. Bol. 38 Salteri L. Bol. 38 Eugalenia Fenzl. 89 Fucus clathratus 181 >, cupressoides é 165 » minimus Hering Be 69, 180 > pygmaeus 69 5, racemosus 166 » vigidus 182 » selago 168 zonalis 180 Galaxaura Diesingiana 71 50 magna Kjellman 70 .; tenera Kjellm. 182 Galenia acutifolia Adamson 112 3 affinis Sond. 104 3 africana 87 Se africana x namaensis 95 o Gollbwae, (IB, &5 A) oo oo OY » crystallina (E. & Z.) Fenzl. 119 Index of Plant Names. PAGE Galenia cymosa Adamson 122 ;, Dinteri Schellenb. 103 5 Dregeana Fenzl. 123 5 Ecklonis 97 » exigua Adamson 101 Ap fallax Pax... as 120 ka filiformis (Thunb.) 100 s, fruticosa (L. ») Sond. 7, * glauca 97 » glauca Walp. 30 co HIG) ,, hemisphaerica Adamson .. 122 A herniariaefolia Eee Fenzl. 90, 101 » hispidissima Fenzl. 93 » humifusa Fenzl. 102 , leucoclada Schellenb. & Schltr. 119, 121 sie leucodermis a5 UII >», linearis Thunb. 90 - Meziana K. Mull. 103 oe namaensis Schinz. 94 >» hamaensis x procumbens 97 >, namaquensis Schellenb. 119 » pallens (EK. & Z.) oo LUG) 5 papulosa (EK. & Z.) Sond. 112 s portulacea Fenzl. .. 116 ; procumbens L.f. .. 96 5 prostrata Schellenb. plat Bs pruinosa Sond. ate 121 3 pubescens (HE. & Z.) 107 , vigida Adamson 94 = sarcophylla Fenzl. 114 » secunda (L.f.) Sond. 105 » spathulata Fenzl. 107 >» Squamulosum (E. & Z.) Fenzl. 117 subearnosa Adamson -- 108 Gelidiopsis rigida (Vahl.) 182, 183 Gelidium caespitosum Kylin. 183, 184 a pristoides We 72,185 2p reptans (Suhr.) Kylin. 184, 185 Haemodorum 43 Hagenbachlia cut ye 43 Halimeda cuneata Hering. co 7D) Haliseris dichotoma (Suhr.) Kutz 177 Halophila ovalis (R. Br.) Hook. 163 Halophila sp. 2 as 163 Halopteris congesta EO) D funicularis (Montagne) Sauvageau. é 66 59 Scoparia 67 Halymenia reptans Suhr. al Herposiphonia. monocarpa. (Montagne) De Toni. 73 Hormophysa latifrons 181 triquetra (L) Kiitz 181 Hydroclathrus clathratus (Bory. ex C. Ag.) Howe. 181 Index of Plant Names. 197 PAGE Idanthyrsus pennatus O10 1) 165 Kolleria Presl Symb. Bot. .. .. 88 ” colina KE. & Z. .. oo OY herniariaefolia Pres. co Oil pallens E. & Z. ... so HIG) » glauceaak. &Z. .. so LIC Kiitzingia natalensis J. Ag. .. 190 Lachnanthes : 43 Lagenocarpus imbricatus Klotzsch. 147 Lanaria : 43 Lepadoderma africanum Dickinson W702 Liriope as BSS ae sa 43 Lophiola O0 ae 20 50), 4B) Loriphes clausus_.. a 56 IK} Macropidia .. oi aca ao | 48) Marsilea spp. a6 06 eo of macrocarpa se sos) F nubica ae ; 9,16 0 profundelaciniata Dinter 9 oD pubescens Se as 9 0 subterranaea of so | villifolia ‘ 9 Mesembryanthemum filiforme (Thunb) aeLOO Microdictyon krausii ¥. E. Gray 172, 173 Microlepia speluncae sis seeds, @. Nagelocarpus serratus Bullock .. 147 Neomeris dumetosa Lamour .. 175 Nienburgia Andersoniana (J. Agardh.) Kylin. .. 73 =p borealis (Kylin.) Kylin. 73 30 japonica (Yamada) Kylin 73 0 Kylin. 73 serrata (Suhr.) Papenfuss 73 Nitophyllum serratum Suhr. so Ue Notholaena eckloniana Des 7,12 o) marlothii ie ome) < rawsoni dys Be TAS Odontostomum : 42, 43 Ophioglossum polyphyllum o4 WY) Dp reticulatum Ue Wal Ophiopogon Be a 1. 43 Padina commersonii ae oo IG >» commersonii Bory 177, 178 Patella barbara eG ‘its ao JY Pauridia ae Le oe 42, 43 Peliosanthes . . ot fe so | C8} Pellaea atropurpurea ai eo lO » calomelanos : Aree Coit!) » deltoidea .. ane Powetaie: » goudotii Me 7,8 » robusta 7,8 PAGE Pellaea viridis wie ie seeds 8 Phlebocarya . . : ; -. 43 Phloiocaulon squamulosum. so) OW Phyllophora reptans 00 184, 185 Placophora Eckloniae Pocock oo US monocarpa (Montagne) Papenfuss 60 Ke so 783 Plocamium Ase oe 56) Ua 3 glomeratum J.Ag. ao Its Pocockiella variegata (Lamour) Papenfuss : an 178, 179 Polysiphonia Binderi ae arma Pseudocodium de-vriesei .. 163, 170 Pyrrosia schimperana ie eels Ralfsia deusta ae og ay » ©xpansa J. AG. ae .. 176 Rhizophora mucronata Lam. 63 Rhodophyllis capensis Kiitzing. .. 171 y reptans (Suhr.) Papenfuss 71 Rhodymenia natalensis Kylin. 186, 187 oD nicricans Harvey yer anvil Sanseviera .. a ifs -. 43 Sargassum .. ae GS Sialodes Eck. & Zeyh at 5 ths